TUBERCULOSIS

Springer
Berlin
Heidelberg
New York
Hong Kong
London
Milan
Paris
Tokyo
M. Monir Madkour (Editor)
A. Al Saif . M. Al Shahed . K. R. Al Moutaery
A. Al Kudwah (Associate Editors)

Tuberculosis
Foreword by
David A. Warrell

With 444 Figures in 846 Separate Illustrations, 139 in Color and 89 Tables

Springer
Editor:
M. MONIR MADKOUR
Department of Medicine
Armed Forces Hospital
Riyadh 11159
Saudi Arabia

Associate Editors:
ABDULAZIZ AL SAIF MONA AL SHAHED
Department of Medicine Department of Radiology
Armed Forces Hospital Armed Forces Hospital
Riyadh 11159 Riyadh 11159
Saudi Arabia Saudi Arabia

KALAF AL MOUTAERY AIDA AL KUDWAH

Department of Neurosurgery Department of Dermatology
Armed Forces Hospital Armed Forces Hospital
Riyadh 11159 Riyadh 11159
Saudi Arabia Saudi Arabia

ISBN 3-540-01441-1 Springer-Verlag Berlin Heidelberg

Library of Congress Cataloging-in-Publication Data

Tuberculosis I M. Monir Madkour, ed

p.cm.
Includes bibliographical references and index.
ISBN 3-540-01441-1 (alk. paper)
1. Tuberculosis. 1. Madkour, M. Monir.
RC311.T4172003
616.9'95--dc22 2003059067

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned,
specifically the rights of translation, reprinting, reuse of illustrations, recitations, broadcasting, reproduction on
microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is per-
mitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and
permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the
German Copyright Law.
Springer-Verlag is part of Springer Science+Business Media
httpllwww.springeronline.com
© Springer-Verlag Berlin Heidelberg 2004
Printed in Germany
The use of general descriptive names, trademarks, etc. in this publication does not imply, even in the absence of a
specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free
for general use.
Product liability: The publishers cannot guarantee the accuracy of any information about dosage and application
contained in this book. In every case the user must check such information by consulting the relevant literature.
Cover-Design: Erich Kirchner, Heidelberg
Typesetting: Verlagsservice Teichmann, Mauer
2113150xq - 5432 1 0 - Printed on acid-free paper
Foreword

It was through his superb monograph on brucellosis, first published in 1989, that I came
to know Dr. M. Monir Madkour and later to meet him in Riyadh. From this introduction,
I was quick to enlist his help as an author of the Oxford Textbook ofMedicine. He has now
turned his considerable talents as physician, author and editor to a disease at least as
interesting but even more important, tuberculosis. Considered against the background
of a large literature on this subject, Dr. Madkour's monumental monograph is distinc-
tive in several respects. As he emphasises in his preface, his book, unlike most others
published on tuberculosis in English, has not been written by western experts whose
experience of the disease is based in developed countries. The Kingdom of Saudi Arabia
has afforded the unusual combination of an abundance of clinical material with state-of-
the-art facilities for laboratory diagnosis, investigation, imaging and medical and surgi-
cal management. Dr. Madkour and his four associate editors, all from the prestigious
Riyadh Armed Forces Hospital, have assembled a distinguished international cast of 64
authors; 29 from the Kingdom of Saudi Arabia, 16 from Europe, 12 from North America,
five from Asia and two from Africa. The result is an exciting, comprehensive and origi-
nal work incorporating a wide range of clinical and scientific skills and experience and
carrying enormous authority.
Tuberculosis is the most prevalent of all bacterial infections, existing actively or in
latent form in a third of the world's population. In the developing world, almost every-
one is infected and the tubercle bacillus might almost be regarded as part of the normal
human condition. It is not surprising that the disease has played such an important role
in the political, social, industrial, artistic and scientific history of mankind. This com-
manding historical perspective is clearly identified in the two opening chapters. I was
fascinated by Dr. Saleh A. Bedeir's review of tuberculosis in Ancient Egypt, although the
examples of Pott's disease in remains from the predynastic period may be antedated
by neolithic finds near Heidelberg, Germany (5000 B.C.) and Liguria, Italy (4000 B.C.)
(Formicola et al. 1987). So was it Europe that gave the world tuberculosis?
In Chap. 2 ("Historical aspects of tuberculosis"), Dr. Madkour and colleagues include
an important section on the achievements of Arabian medicine during the period
500-1500 A.D., usually regarded as the "Dark Ages" in Europe. The roles of the Baghdad
and Cairo Schools during this era, and notably the contributions of Ibn an-Nafis, Rhazes,
Avicenna and Albucasis, have certainly been neglected. Understandably, this account
does not dwell on the enormous impact of tuberculosis on the arts, literature and sci-
ence in Europe, especially during the nineteenth century, when this disease determined
the creative life of many outstanding authors, artists and composers and supplied many
fictional heroes, heroines and story lines.
In Chap. 3 ("Global epidemiology of tuberculosis"), Marcos Espinal and Mario Ravi-
glione of WHO Geneva, provide compelling evidence of the current worldwide crisis in
tuberculosis control. An annual incidence of some 8 million new cases, with 2 million
deaths and a case notification rate of around 50 per 100,000, shows no evidence of decline
 despite the availability of effective anti-tuberculosis chemotherapy. The most chilling
evidence for a failure of implementation of chemotherapy, the impact of the HIV-AIDS
epidemic and emerging multi-drug resistant disease, is given in their Fig. 3.5, which pre-
dicts an increase to 10 million new cases in 2010. The epidemiological situation in the
Kingdom of Saudi Arabia (Chap. 4) is complicated by the large influx of pilgrims from all
over the world during the Hajj.
Molecular fingerprinting of Mycobacterium tuberculosis (Chap. 5) has provided
fascinating insights into the epidemiology, history and geographical origins of the
pathogen, for example into the Cape coloured population of Cape Town (van HeIden
2002). Tuberculosis has become a major occupational health concern for medical per-
sonnel and others who work in high-risk communities. This important area is covered
by Jaime Esteban in Chap. 7 ("Tuberculosis in Special Groups and Occupational Haz-
ards"). There is a strong section on the microbiology, immunology, pathogenesis and
pathology of tuberculosis, including a substantial review of immunology and pathogen-
esis by Graham Rook (Chap. 9). Pathological changes in the central nervous system are
often dominated by secondary effects of obstructive or communicating hydrocephalus
or the involvement of major blood vessels or cranial nerves at the base of the brain
(Chap. 10).
Environmental and other non-tuberculous mycobacteria, many of which have been
regarded as virtually non-pathogenic in immunocompetent people, have caused life-
threatening infections in those with subtle immune deficits, for example in the IL-12-
dependent interferon-gamma pathway (Lammas et al. 2000; Fieschi et al. 2003) (Chap. 11).
These rare deficiency syndromes provide insights into normal immune defence mecha-
nisms, as does reactivation of latent tuberculosis by corticsteroids, anti-TNF antibodies
and other agents. The section on diagnosis (Chaps. 12-15) covers conventional and newer
methods, including PCR.
The last decade has witnessed increasing investment in basic science research into the
cell biology, genetics and immunology of tuberculosis. A landmark achievement has been
the sequencing of the M. tuberculosis genome. However, these advances in the understand-
ing of the biology of the bacillus and the host response have not yet resulted in any tangible
improvements in diagnosis and management of tuberculosis patients and their contacts.
One area in which recent scientific advances are beginning to make an impact on clinical
practice is the diagnosis of latent tuberculosis infection. Until very recently, this relied
upon the century-old tuberculin skin test, introduced by Robert Koch (Chap. 15). Blood
tests measuring interferon-gamma secretion in response to purified protein derivative
(PPD) may offer a more rapid and convenient approach to detecting latent infection. Cel-
lular immune responses to PPD may, however, cross-react with BCG vaccination, just as
with the delayed type hypersensitivity response to PPD detected by skin testing, resulting
in poor specificity. Comparative mycobacterial genomics has identified a stretch of DNA
(region of difference I, RD1) that is present in M. tuberculosis but absent from all strains
of BCG vaccine. The antigens encoded by RD 1 genes thus offer the opportunity to develop
more specific tests for M. tuberculosis infection. Two of these antigens, ESAT-6 and CFP10,
have now been incorporated into a rapid, sensitive blood test that enumerates individual
interferon-gamma-secreting T cells, the enzyme-linked immunospot (ELISPOT) assay. In
active tuberculosis, EllSPOT appears to have a higher sensitivity and specificity than the
tuberculin skin test (Lalvani et al. 2001). In latent infection, the absence of a gold standard
reference test makes evaluation of any new test difficult, but recent studies indicate that
ELISPOT successfully distinguishes latent infection from BCG vaccination and thus has a
higher specificity than skin testing, while the observation that ELISPOT correlates more
closely with M. tuberculosis exposure than skin testing suggests that it may also have a
higher sensitivity (Lalvani et al. 2001; Ewer et al. 2003).
From the clinical point of view, tuberculosis is the most protean of all diseases,
demanding inclusion in the differential diagnosis of almost every symptom, sign and
abnormal result of investigation. The organ-related chapters of the book provide a
thorough coverage of disease manifestations in the different age groups (Chaps. 16-18),
during pregnancy (Chap. 19), in the face of immunosuppression (Chap. 29) and in the
various sytems, organs and tissues (Chaps. 20-43). A great strength of the book is its
inclusion of a rich variety of high-resolution pictures of clinical cases (see, for example,
Chap. 36 on skin manifestations, and Chap. 40 on the eye). The clinical suspicion that
allows the crucial early diagnosis of tuberculous meningitis (Chaps. 32 and 33) is partic-
ularly subtle, but the diagnosis may be difficult to confirm in the developing world. I have
seen too many patients undergoing inappropriate therapeutic trials of antituberculosis
chemotherapy for this condition. Tuberculous meningitis is a prevalent and commonly
misdiagnosed condition that deserves a far more detailed and lengthy discussion.
Compliance to the necessarily prolonged treatment regimens has been improved by
the "DOTS" (directly observed treatment) strategy (Chap. 45). Once the correct treat-
ment has been implemented, its efficacy must be monitored (Chap. 47), as the great-
est threat to successful cure is drug resistance. Peter Davies provides an authoritative
review of this, the most challenging obstacle to cure and control (Chap. 46). Of the many
suggested ancillary treatments, corticosteroids are becoming less controversial for peri-
cardial and pleural effusions and for tuberculous meningitis. It is to be hoped that the
current large randomised controlled trial of corticosteroids in tuberculous meningitis,
being conducted by The Centre for Tropical Diseases in Ho Chi Minh City, Vietnam, will
finally provide decisive evidence in this long-running debate.
I congratulate Dr. Madkour and his expert team of contributors for producing a timely,
thorough, scholarly and truly fascinating account of one of the world's most important
and complex diseases. Their book deserves to be consulted by all medical practitioners
and to be read in its entirety by all infectious diseases specialists. It will surely become a
classic in the literature of this disease.

Oxford, UK DAVID A. WARRELL

References

Ewer K, Deeks J, Alvarez L et al (2003) Comparison of T-cell-based assay with tuberculin skin test for
diagnosis of Mycobacterium tuberculosis infection in a school tuberculosis outbreak. Lancet 361:
1168-1173
Fieschi C, Dupuis S, Catherinot E et al (2003) Low penetrance, broad resistance, and favorable outcome
of interleukin-12 receptor beta-l deficiency: medical and immunological implications. J Exp Med
197:527-535
Formicola V, Milanesi Q, Scarsini C (1987) Evidence of spinal tuberculosis at the beginning of the
fourth millennium B.C. from Arene Candide Cave (Liguria, Italy). American J Physical Anthropol-
ogy 72:1-6
Van HeIden PD (2002) Molecular epidemiology of TB: challenging dogmas and asking new questions.
IUBMB Life 53:219-223
Lalvani A, Pathan AA, McShane H et al (2001) Rapid detection of Mycobacterium tuberculosis infec-
tion by enumeration of antigen-specific T cells. Am J Respir Crit Care Medicine 163:824-828
Lammas DA, Casanova J-L, Kumararatne DS (2000) Immunodeficiency review. Clinical consequences
of defects in the IL-12-dependent interferon-gamma (IFN-gamma) pathway. Clin Exp Immunol
121:417-425
Preface

In 1993, the World Health Organization declared tuberculosis a global emergency. Yet,
the incidence of tuberculosis continues to rise. It is not clear, however, whether the rising
incidence can be attributed to failure of control strategies, the recent impact of the HIV
pandemic or the increasing problem of multidrug resistance. At present, it is estimated
that one-third of the world population is infected with Mycobacterium tuberculosis,
with over 8 million new cases of active disease detected every year and almost 2 million
annual deaths. The impact of the present situation and the current inadequate efforts
necessitates the development of a novel and alternative approach to the control and
therapy measures.
It is interesting to note that while the incidence of tuberculosis is generally low in
developed countries, most of the authoritative textbooks on tuberculosis in the Eng-
lish language have been written by western experts. As an expatriate, working in Saudi
Arabia, I found a rare opportunity to work in an endemic area yet enjoy the diagnostic
and therapeutic facilities of developed countries. Such a combination of circumstances
is rarely present in developing or developed countries. The rich clinical material is
clearly reflected in this book in the form of "evidence-based" clinical practice enriched
with vast numbers of images of various body organs and systems as well as by a thor-
ough survey of all the relevant literature. The numerous figures that are present in a lot
of chapters, many in colour, give a vivid picture of what is seen in clinical practice and
comprise a special feature of this textbook. Separate imaging chapters were compiled
whenever the inclusion with the text would have rendered the corresponding chapter
unwieldy.
Some of the world's most distinguished medical scientists from North America,
South-East Asia, the Far East, South Africa and many European countries have con-
tributed chapters on global epidemiology, molecular epidemiology, the molecular
epidemiology of Mycobacterium bovis, polymerase chain reaction (PCR), serological
methods, latent tuberculosis, prevention and control, directly observed therapy (DOT),
multidrug-resistant tuberculosis, HIV and tuberculosis, thoracic surgery, new vaccines
and many other clinical topics. Animal tuberculosis, which may contribute to human
disease, is included in a separate chapter.
The main goal of this book is to present a clear account of all aspects of tuberculosis
as seen and practiced both in developing and developed countries. All authors took this
into consideration while writing their chapters. Recent advances in diagnosis, preven-
tion and control measures, and treatment are also addressed.
I am humbled by the honour that was given to me by Professor David A. Warrell,
Founding Director of the Centre for Tropical Medicine (Emeritus), University of Oxford
and editor of the Oxford Textbook of Medicine, in agreeing to write the foreword in this
book, and I am most grateful to him.

Riyadh M. MONIR MADKouR

This work is dedicated to:

The renaissance of the ancient glorious Library of Alexandria, Egypt, and to those who
contributed to its resurrection over the original site. Alexandria was the chief cultural
and commercial center of its time, and when the Library was built in approximately 332
Be it attracted manuscripts and scholars from the entire Hellenistic world. It contained
over 700,000 rolls of the work of Socrates, Plato, Aristotle, Hippocrates and all the writ-
ing of the known world, but was destroyed by a great fire. Virtual images of the ancient
Library are available at UNESCO. The inauguration ceremony of the "new--ancient"
Library of Alexandria in September 2002 was attended by heads of state, scientists, art-
ists and dignitaries from all over the world.
The World Health Organization, for its enormous, determined and unyielding efforts
to control tuberculosis, particularly in endemic, poorly resourced, and developing coun-
tries.
Acknowledgements:

I am indebted to Gen. Dr. S.M. Al Deeb and the Hospital Management Board for provid-
ing the facilities I required for this project. I am grateful to our medical library staff:
R. Ahtashemuddi, S.Z. Khan, Y.S. AI-Zharani, M.A. Khushi, S.M. AI-Rakaf, S. Al-Salafi,
R. AI-Sous, and A. AI-Sultan; our medical record staff: J. Al Anazi, A. Al Humaid and
T. Al Shuriti; and our medical illustration staff: S. Nally, T. Michalak, D. Hargreaves, K.
Jefferson, N. Morgan, R. Ponmbath, M. Horaib.
Support from our colleagues from the department of microbiology, Dr. S. Bakhesh-
wain and Dr. N. EI-Khizzi, is highly appreciated. I would like to thank Dr. M. Sofi and Dr.
S.A. AI-Mohaimeed for their support. My colleagues from the department of medicine
helped and supported this project, including Drs. T. AI-Tassan, B. AI-Dosary, S. AI-Faraj,
A. AI-Zaid, N. AI-Qahtani and many others, and I am grateful.
I would like to thank our secretaries for typing the manuscripts of my chapters and
for their secretarial help: I. Andries, J. Misquith, E. Sabbagh.
I would like to thank my son Amr Madkour, pre-medical student at Austin, Texas for
helping me with his experience in the Internet and other computer facilities for this
project during his summer vacations. My other children, Rasha, Monir and Reem, helped
with typing and sending correspondence bye-mail and with MedLine searches. Without
their help, my handicap in computer technology would have been insurmountable.
I am most grateful to Dr. Julia Heidelmann, Dr. Ute Heilmann, Ms. Wilma McHugh
and the staff at Springer-Verlag for their meticulous care and efforts in the production
of this book.
Despite the help I have received from others, I alone stand responsible for any short-
comings this book may contain.

M. MONIR MADKouR
Contents

Historical Background .

Tuberculosis in Ancient Egypt

SALEH A. BEDEIR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Historical Aspects of Tuberculosis

M. MONIR MADKOUR, KITAB E. AL-OTAIBI, R. AL SWAILEM 15

Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3 Global Epidemiology of Tuberculosis

MARCOS A. ESPINAL and MARIO C. RAVIGLIONE 33

4 Epidemiology of Tuberculosis in Saudi Arabia

ABDULRAHMAN A.ALRAJHI and ALI M.AL-BARRAK 45

5 The Molecular Epidemiology of Tuberculosis

KATHRYN DERIEMER and CHARLES L DALEY 57

6 Molecular Epidemiology of Mycobacterium bovis

ROBIN A SKUCE and SYDNEY D NEILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

7 Tuberculosis in Special Groups and Occupational Hazards

JAIME ESTEBAN 93

Microbiology, Immunology, Pathogenesis and Pathology 113

8 Microbiology of Tuberculosis
A OLU OSOBA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 115

9 The Immunology and Pathogenesis of Tuberculosis

GRAHAM A. W. ROOK 133

10 Pathology of Tuberculosis
MOHAMMED AKHTAR and HADEEL AL MANA 153

11 Nontuberculous Mycobacteria
JAE-JOON YIM and STEVEN M. HOLLAND 163

Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 185

12 Serologic Testing for Tuberculosis

CHAKHRADHAR KOTARU and EDWARD D. CHAN. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 187
XIV Contents

13 PCR and Diagnosis of Tuberculosis

DIANA LWILLIAMS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 199

14 Hematologic Findings in Mycobacterial Infections Among Immunosuppressed

and Immunocompetent Patients
GORGON AKPEK 213

15 Immunodiagnostics for Latent Tuberculosis Infection

ROHIT K. KATIAL 231

Organ-related Chapters 241

General

16 Childhood Tuberculosis
PETER R. DONALD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 243

17 Primary Tuberculosis in Adults

M. MONIR MADKOUR 265

18 Miliary/Disseminated Tuberculosis
M. MONIR MADKOUR 273

19 Tuberculosis and Pregnancy

M. MONIR MADKOUR 301

Thorax

20 Post-primary Pulmonary Tuberculosis

M. MONIR MADKOUR, Y. ABUSABAAH, ALI BEN MOUSA, ALI AL MASOUD . . . . . . . .. 313

21 Endobronchial Tuberculosis
HEE SOON CHUNG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 329

22 Pleural Tuberculosis
M. MONIR MADKOUR, MAJDY IDREEs, M. AL SHAHED 349

23 Radiology of Pulmonary Tuberculosis

MONA AL SHAHED, MOHAMMED ABD EL BAGI, M. MONIR MADKOUR . . . . . . . . . . .. 359

24 Pulmonary Function Test and Tuberculosis

MAJDY M. IDREES, SIRAJ O. WALl, ABDULLA AL-AMOUDI 385

25 Thoracic Surgery for Tuberculosis

YUJI SHIRAISHI 395

26 Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis

DAVID HAMILTON and JAWDA AL-NABULSI 411

27 Tuberculosis of the Heart and Pericardium

ERNESTO E SALCEDO and AHMAD S OMRAN 431
Contents xv

Immune System

28 Mycobacterial Lymphadenitis
M. MONIR MADKOUR and RASHID AL KUHAYMI 445

29 Tuberculosis and Co-infection with the Human Immunodeficiency Virus

ALIMUDDIN ZUMLA and JOHN M GRANGE 455

eNS

30 Spinal Tuberculosis
M. MONIR MADKOUR, M. WASEF AL SEBAI, KHALAF R. AL MOUTAERY .. . . . . . . . .. 481

31 Surgical Management of Spinal Tuberculosis

M. WASEF AL SEBAI, M. MONIR MADKOUR, KAHLAF R. AL MOUTAERY . . . . . . . . . .. 493

32 Tuberculosis of the Central Nervous System

M. ZUHEIR AL-KAWI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 535

33 Imaging of Brain and Spinal Cord Tuberculosis

FRANCIS MCGUINNESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 547

Musculo-Skeletal and Soft Tissue

34 Extraspinal Musculoskeletal Tuberculosis

M. MONIR MADKOUR, AIDA J. AL-KUDWAH, MOHAMMED ABD EL BAGI 587

35 Imaging of Musculoskeletal Tuberculosis

MOHAMMED ABD EL BAGI, MONA AL SHAHED, M. MONIR MADKOUR . . . . . . . . . . .. 605

36 Tuberculosis of the Skin

AIDA J. AL KUDWAH 627

Gastrointestinal

37 Abdominal Tuberculosis
MOHAMMAD SULTAN KHUROO and NAIRA SULTAN KHUROO 659

38 Imaging of Gastrointestinal Tuberculosis

MONA AL SHAHED and MOHAMMED ABD EL BAGI 679

Genitourinary

39 Genitourinary Tuberculosis
M. MONIR MADKOUR 699

Head

40 Ocular Manifestations of Tuberculosis

SHWU-JIUAN SHEU 731

41 Otorhinolaryngeal Aspects of Tuberculosis

SUJATA MURANJAN 741
XVI Contents

Endocrine

42 Endocrine and Metabolic Manifestations of Tuberculosis

SOHAIL INAM and MONA AL-SHAHED 751

Vascular

43 Tuberculous Vasculitis and Mycotic Aneurysms

M. MONIR MADKOUR 771

Treatment 779

44 Treatment of Tuberculosis
DEAN E SCHRAUFNAGEL 781

45 Directly Observed Treatment for Tuberculosis Control

ZHANG LI-XING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 801

46 Multi-Drug-Resistant Tuberculosis
PETER D. O. DAVIES 809

47 Monitoring Treatment Efficiacy

P. H. LAGRANGE, N. SIMONNEY, A. O. SOUSA, A. WARGNIER, J. L. HERRMANN 839

Control and Prevention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 853

48 Tuberculosis Control in Developing and Developed Countries

KEVIN SCHWARTZMAN .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 855

49 BCG and New Tuberculosis Vaccines

ZHOU XING 881

50 Tuberculosis in Animals
P. L. NICOLETTI 893

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 903

About the Editor 930

List of Contributors

ABD EL BAGI MOHAMMED, MB, BCh, DMRD, FSRRCSI AL-KuDWAH AIDA J, MD, DD, FRCP (Ed)
Department of Radiology Consultant Dermatologist, Department of Dermatology
Riyadh Armed Forces Hospital Riyadh Armed Forces Hospital
P.O. Box 7897 P.O. Box 7897, C-117
Riyadh 11159 Riyadh 11159
Saudi Arabia Saudi Arabia

ABUSABAAH Y, ABIM
AL KUHAYMI RASHID, FRCS
Fellow, Respiratory Medicine
Department of Medicine Director of Surgery
Riyadh Armed Forces Hospital Riyadh Armed Forces Hospital
P.O. Box 7897 P.O. Box 7897
Riyadh 11159 Riyadh 11159
Saudi Arabia Saudi Arabia

AKHTAR MOHAMMED, MD, FCAP, FRCPA, FRCPath AL MANA HADEEL, MD

Chairman, Department of Pathology Fellow, Department of Pathology ad Laboratory Medicine
and Laboratory Medicine King Faisal Specialist Hospital and Research Centre
King Faisal Specialist Hospital and Research Centre P.O. Box 3354
P.O. Box 3354, MBCI0 Riyadh 11211
Riyadh 11211 Saudi Arabia
Saudi Arabia

AKPEK GORGON, MD AL MAsouD ALI, MD

Assistant Professor of Medicine Department of Medicine
Greenebaum Cancer Center at University of Maryland Riyadh Armed Forces Hospital
School of Medicine P.O. Box 7897
22 South Greene Street Riyadh 11159
Baltimore MD 21201 Saudi Arabia
USA

AL-AMOUDI ABDULLA, MD, FRCP (C) AL MOUTAERY KHALAF R, MD, FRCS (Ed), FACS
Deputy Head, Department of Medicine Head of Neurosurgery
King Faisal Specialist Hospital & Research Center Riyadh Armed Forces Hospital
Jeddah P.O. Box 7897
Saudi Arabia Riyadh 11159
Saudi Arabia
AL-BARRAK ALI M, AmBIM, DTM & H, FRCP (C)
Consultant Infectious Diseases Unit AL-NABULsI JAWDA, DMRD, MSc NM
Department of Medicine Nuclear Medicine
Riyadh Armed Forces Hospital Riyadh Armed Forces Hospital
P.O. Box 7897 P.O. Box 7897
Riyadh 11159 Riyadh 11159
Saudi Arabia Saudi Arabia
AL-KAwI M. ZUHEIR, MD, FACP
Senior Consultant Neurologist & Deputy Chairman AL-OTAIBI KITAB E, Facharzt (Urology), MD, FRCS (Edin)
Department of Neurosciences Director General of Medical Services Department
King Faisal Specialist Hospital and Research Center Riyadh Armed Forces Hospital
P.O. Box 3354 P.O. Box 7897
Riyadh 11211 Riyadh 11159
Saudi Arabia Saudi Arabia
XVIII List of Contributors

ALRAJHI ABDULRAHMAN A, MD, MPH, FIDSA CHUNG HEE SOON, MD, FCCP
Consultant and Head, Section of Infectious Diseases, Associate Professor of Medicine
Chairman, Department of Medicine (MBC #46), Seoul National University College of Medicine and Seoul
King Faisal Specialist Hospital and Research Centre, Municipal Boramae Hospital affiliated to Seoul National
P.O. Box 3354, University Hospital # 395
Riyadh 11211, Shindaebang - 2 Dong
Saudi Arabia Dongjak-Gu, Seoul, 156-707
Korea
AL SAIF A, MD, PhD, FCCP, FRCP (Lon)
Consultant Respiratory Medicine DALEY CHARLES L, MD
Director of Medicine Associate Professor of Medicine
Riyadh Armed Forces Hospital Division of Pulmonary and Critical Care Medicine
P.O. Box 7897 San Francisco General Hospital
Riyadh 1159 Room 5K-l
Saudi Arabia 1001 Potrero Ave
San Francisco, CA 94110
AL SEBAI M WASEF, FRCS. FFCS USA
Consultant Spinal Surgeon
Riyadh Armed Forces Hospital
DAVIES PETER D 0, DM, FRCP
P.O. Box 7897
Director, Tuberculosis Research Unit
Riyadh 11159
Cardiothoracic Centre
Saudi Arabia
Liverpool Ll4 3PE
UK
AL SHAHED MONA, MBBS, FRCR
Consultant Radiologist DERIEMER KATHRYN, PhD
Department of Radiology Senior Research Scientist
Riyadh Armed Forces Hospital Division of Infectious Diseases
P.O. Box 7897 and Geographic Medicine
Riyadh 11159 Stanford Medical Center, Stanford
Saudi Arabia California, USA

AL SWAILEM R, MRCP (UK), ABIM DONALD PETER R, MD

Consultant Rheumatologist Department of Paediatrics and Child Health
Department of Medicine Faculty of Health Sciences
Director of Postgraduate and Academic Affairs University of Stellenbosch
Riyadh Armed Forces Hospital P.O. Box 19063
P.O. Box 7897 7505, Tygerberg, South Africa
Riyadh 11159
Saudi Arabia
ESPINAL MARCOS A, MD
World Health Organization
BEDEIR SALEH A, MD Stop TB Department
Dean, Faculty of Medicine Tuberculosis Strategy & Operations
Cairo University Ave Appia # 22
Kasr El-Aini Hospitals 1211 Geneva
Cairo Switzerland
Egypt
ESTEBAN JAIME, MD
BEN MOUSA ALI, MD Department of Medical Microbiology
Department of Medicine, Fundaci6n Jimenez Diaz
Riyadh Armed Forces Hospital Av. Reyes Cat6licos 2
P.O. Box 7897 28040 Madrid
Riyadh 11159 Spain
Saudi Arabia
GRANGE JOHN M, MSc, MD (Lon)
CHAN EDWARD D, MD Visiting Prof, Centre for Infectious Diseases
Associate Professor of Medicine & International Health
K613e Goodman Bldg. Royal Free and University College Medical School
National Jewish Medical and Research Center, Denver, CO Windeyer Institute of Medical Sciences
1400 Jackson St 46 Cleveland St
Denver, CO 80206 London WIP 6DB
USA UK
List of Contributors XIX

HAMILTON DAVID, PhD, FIPEM KOTARU CHAKRADHAR, MD

Department of Nuclear Medicine Department of Medicine and Program in Cell Biology
Riyadh Armed Forces Hospital National Jewish Medical and Research Center
P.O. Box 7897 Division of Pulmonary Sciences and Critical Care Medicine,
Riyadh 11159 University of Colorado Health Sciences Center, and Denver
Saudi Arabia Veteran Administration Medical Center
K613e, Goodman Building
1400 Jackson Street
HERRMANN J L, MD
Denver, CO 80206
Service de Microbiologie
USA
H6pital Saint Louis
Assistance Publique-H6pitaux de Paris
LAGRANGE P H, MD
Universite Denis Diderot
Service de Microbiologie
1, avenue Claude Vellefaux
H6pital Saint Louis
75475 Paris
Assistance Publique-H6pitaux de Paris
France
Universite Denis Diderot
1, avenue Claude Vellefaux
HOLLAND STEVEN M, MD 75475 Paris
Immunopathogenesis Unit France
Building 10, Room llN103
10 Center Dr. MSC 1886 LI-XING ZHANG, MD
Bethesda, MD 20892 1886 Vice President & Secretary General
USA Chinese Anti-Tuberculosis Association
5, Dong Gang Hu-Tong
IDREES MAJDY M, MD, FRCP (C), FCCP Beijing 100035
Head, Pulmonary Function Laboratory China
Division of Pulmonary Medicine
Department of Medicine MADKOUR M MONIR, MD, DM, FRCP (London)
Riyadh Armed Forces Hospital Consultant
ClIO, P.O. Box 7897 Department of Medicine
Riyadh 11159 Riyadh Armed Forces Hospital
Saudi Arabia P.O. Box 7897, C-119
Riyadh 11159
Saudi Arabia
INAM SOHAIL, MBBS, FRCP (Edin), FRCP
Head of Endocrinology Division
MCGUINESS FRANCIS, MD
Department of Medicine
Apt. 169, Al Haurin EI Grande
Riyadh Armed Forces Hospital
2912 Malaga
P.O. Box 7897
Spain
Riyadh 11159
Saudi Arabia
MURANJAN SUJATA, MS (ENT), DNB, DORL
Consultant ENT
KATIAL ROHIT K, MD Suman Apptartment, 3rd Floor
Associate Professor of Medicine 16B Naushir Bharucha Road
National Jewish Medical and Research Center Denver Tardeo
1400 Jackson St. Mumbai 400007
Denver, CO 80206 India
USA
NEILL SYDNEY D., BSc, PhD
KHUROO MOHAMMAD SULTAN, MD, DM, FRCP (Edin), MACP Veterinary Sciences Division
Professor, Consultant Gastroenterologist Department of Agriculture and Rural Development
Department of Medicine - MBC 46 Stoney Road
King Faisal Specialist Hospital and Research Centre Stormont,
P.O. Box 3354 Bellfast BT4 3SD
Riyadh 11211 UK
Saudi Arabia
NICOLETTI P L, DVM, MS
KHUROO NAIRA SULTAN, MBBS College of Veterinary Medicine
Department of Radiology Department of Veterinary Pathology
King Faisal Specialist Hospital and Research Centre P.O. Box 110880
P.O. Box 3354 Gainseville
Riyadh 11211 Florida 32611 0880
Saudi Arabia USA
xx List of Contributors

Ow OSOBA A, MD, FRCPath (UK), FFPath (Ireland), FWACP SHIRAISHI YUJI, MD

Consultant & Head of Microbiology Head, Section of Chest Surgery
King Khalid National Guard Hospital Fukujuji Hospital
P.O. Box 9515 3-1-24 Matsuyama Kiyose
Jeddah 2143 Tokyo 204 8522
Saudi Arabia Japan

OMRAN AHMAD S, MD
Consultant Cardiologist SIMONNEY N, MD
King Abdulaziz Cardiac Center, Service de Microbiologie
King Abdulaziz Medical City, National Guard, H6pital Saint Louis
Riyadh, Saudi Arabia Assistance Publique-H6pitaux de Paris
P.O. Box 22490 Universite Denis Diderot
Riyadh 11426 1, avenue Claude Vellefaux,
Saudi Arabia 75475 Paris
France
RAVIGLIONE MARIO C, MD
World Health Organization SKUCE ROBIN A, BSc, PhD
Stop TB Department Veterinary Sciences Division
Ave Appia # 22 Stoney Road
1211 Geneva Stormont
Switzerland Belfast, BT4 3SD
Northern Ireland
ROOK GRAHAM A.W, BA, MB Bchir, MD
Professor, Department Medical Microbiology
Royal Free and University College Medical School SOUSA A O,MD
Windeyer Institute of Medical Sciences Service de Microbiologie
46 Cleveland Street H6pital Saint Louis
London WIT 4JF Assistance Publique-H6pitaux de Paris
UK Universite Denis Diderot
1, avenue Claude Vellefaux
SALCEDO ERNESTO E, MD 75475 Paris
Department of Cardiology France
King Abdulaziz Cardiac Center
National Guard Hospital
P.O. Box 22490 WALl SIRAJ 0, MD, FRCP (C), FCCP
Riyadh 11426 Head, Division of Pulmonary Medicine
Saudi Arabia Department of Medicine
King Khalid Hospital
National Guard
SCHRAUFNAGEL DEAN E, MD
Jeddah
Department of Medicine M/C 787
Saudi Arabia
University of Illinois at Chicago
840 S. Wood St
Chicago, IL 60612 7323
USA WARGNIER A, MD
Service de Microbiologie
H6pital Saint Louis
SCHWARTZMAN KEVIN, MD, MPH, FRCPC Assistance Publique-H6pitaux de Paris
Assistant Professor Universite Denis Diderot
Department of Medicine and Epidemiology And Biostatistics 1, avenue Claude Vellefaux
Respiratory & Epidemiology Unit 75475 Paris
McGill University France
1110 Pine Avenue West
Montreal, Quebec
Canada H3A lA3 WILLIAMS DIANA L, PhD
Molecular Biology Research Department
SHEU SHWU-JIUAN, MD Laboratory Research Branch
Department of Ophthalmology National Hansen's Disease Programs
Kaohsiung Veterans General Hospital At LSU-SVM, Rm 3517W
386 Ta-Chung Skip Bertman Dr.,
1st Road, Kaohsiung Baton Rouge, LA 70803
Taiwan 813 USA
List of Contributors XXI

XING ZHOU, MD ZUMLA ALIMUDDIN, BSc, MBChB, MSc, PhD, FRCP (Lon),
Associate Prof. Pathology & Molecular Medidine FRCP (Edin)
Head, Division of Infectious Diseases Professor of Infectious Diseases and International Health
Centre of Gene Therapeutics Director, Centre for Infectious Diseases
McMaster University Health Sciences and International Health
1200 Main Street W Royal Free & University College Medical School
Hamilton, Ontario L8N 3Z5 Windeyer Institute of Medical Sciences
Canada Room G41
46 Cleveland St
YIM JAE-JOON, MD London WIP 6DB
Respiratory and Critical Care Medicine UK
Seoul Natinal University College of Medicine
Seoul
South Korea 110744
Historical Background
1 Tuberculosis in Ancient Egypt
SALEH A. BEDEIR

CONTENTS 1.1
Introduction
1.1 Introduction 3
1.2 Predynastic Period 3
1.2.1 The Foundation of Egypt 4 In all probability, tuberculosis is older than the
1.2.1.1 Predynastic Period 4 human race, and different species in the genus Myco-
1.2.1.2 Archaic Period 4 bacterium may have caused the disease in mammals,
1.2.1.3 Old Kingdom 4 birds, reptiles and fish (Breed et al. 1957). Tubercu-
1.2.2 Development of Egyptian Language 4
losis in man and other animals, both warm-blooded
1.2.3 A Chronology of Literature 5
1.2.3.1 From Alexander to Napoleon 5 and cold-blooded, may have arisen from a common
1.3 Ancient Egyptian Art 5 Mycobacterium ancestor many millions of years
1.4 Paleopathology 6 ago (Cockburn and Cockburn 1980). Other authors
1.5 Differential Diagnosis 7 believe that tuberculosis in man may be of much
1.6 The Author's Case 7
more recent origin. The earliest report on presump-
1.7 Pulmonary and Osseous Tuberculosis
in Ancient Egyptian Mummies 8 tive tuberculosis in ancient Old World human remains
1.8 PCR and the Diagnosis of Tuberculosis is an article by Bartels (1907) on a Neolithic skeleton
in Ancient Egyptian Population 8 found near Heidelberg, Germany. The fourth and fifth
1.8.1 Environment 9 thoracic vertebrae have collapsed and fused with the
1.8.1.1 The Nile 9
sixth vertebra, creating an angulation often seen in
1.8.1.2 Climate 9
1.8.1.3 Housing 9 spinal tuberculosis (Ortner and Putschar 1981).
1.8.1.4 Diet 9
1.8.1.5 Religion 10
1.8.1.6 Sports and Recreation 10
1.8.1.7 Physique 10
1.9 The Medical Papyri 10
1.2
1.10 Sanatorial Treatment 11 Predynastic Period
References 12
Tuberculosis has existed for thousands of years in
Egypt and Nubia, as shown by art forms and skeletal
material. The ancient Egyptian obtained his first
tuberculosis bacillus from the close contact with
livestock that occurred after the Neolithic revolu-
tion, as early as 3300 B.C., during the initial herding
period of Fayoum (Filer 1995). One of the earliest
examples of spinal tuberculosis has been found at
excavations at the predynastic site of Adaima, 8 km
south of Esna.
The chronological order of Ancient Egyptian
Dynasties are given here according to:

a. Foundation of Egypt
S. A. BEDEIR, MD b. Development of Egyptian Language
Dean, Faculty of Medicine, Cairo University, Kasr El-Aini c. A Chronology of Literature
Hospital, Cairo, Egypt d. From Alexander to Napoleon

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
4 S.A. Bedeir

1.2.1
The Foundation of Egypt

1.2.1.1
Predynastic Period

Badarian Period Naqada I Period Naqada II Period Naqada III Period

5000-4000 B.C. 4000-3500 B.C. 3500-3150 B.C. 3150-3000 B.C.

First evidence for Development of First major towns First hieroglyphs
pottery in Egypt culture

1.2.1.2
Archaic Period

Dynasty I Dynasty II

3050-2815 B.C. 2815-2600 B.C.

Unification of Egypt; Royal tombs at Saqqara
royal tombs at Abydos

1.2.1.3
Old Kingdom

Dynasty III Dynasty IV

2660-2600 B.C. 2600-2470 B.C.

First pyramids Great pyramids at Giza

1.2.2
Development of Egyptian Language

Archaic Period Old Kingdom First Intermediate Period

Dynasties I-II Dynasties III-IV Dynasties VII-XIa

3050-2660 B.C. 2660-2200 B.C. 2200-2070 B.C.
Earliest script Old Egyptian Transition to Middle Egyptian,
which lasts through the Middle Kingdom,
dynasties Xlb-XIII

Second Intermediate New Kingdom Third Intermediate Period

Period

Dynasties XIV-XVII Dynasties XVIII-XX Dynasties XXI-XXV

1650-1550 B.C. 1550-1070 B.C. 1070-664 B.C.
Middle Egyptian Middle Egyptian transition Late Egyptian Demotic
to Late Egyptian

Late Period Graeco-Roman Period Coptic Period

Dynasties XXVII-XXXI Dynasties XVIII-XX Romans Byzantines
664-332 B.C. Ptolemies 332-30 B.C. 30 B.C.-A.D. 395 A.D. 395-640
Late Egyptian Demotic Demotic Coptic Demotic
Tuberculosis in Ancient Egypt 5

1.2.3
A Chronology of Literature

Archaic Period Old Kingdom First Intermediate Middle Kingdom

Period

Dynasties I-II Dynasties III-IV Dynasties VII-XIa Dynasties XIb-XIII

3050-2660 B.C. 2660-2200 B.C. 2200-2070 B.C. 2070-1650 B.C.
Labels; short Pyramid texts, Decrees, biographies Coffin texts, stories,
inscriptions only first biographies wisdom texts

New Kingdom Third Intermediate Period Graeco-Roman Period

Dynasties XVIII-XX Dynasties XXI-XXV Ptolemies Romans

1550-1070 B.C. 1070-664 B.C. 332-30 B.C. 30 B.C.-A.D. 395
Book of the Dead; Major mythological texts: Increasing Greek influence
"Historical" texts continue through Later Period,
Dynasties XXVI-XXXI

1.2.3.1
From Alexander to Napoleon

Ptolemaic Period Roman Period Coptic Period Arab Period Ottoman Period

332-30 B.C. 30 B.C.-A.D. 395 A.D. 395-640 A.D. 640-1517 A.D. 1517-1805
Egypt ruled by Egypt made a Egypt part of GeneralAmr The Turkish Sultan
heirs of Alexander Roman province the Eastern conquers Egypt Selim I invades
the Great Empite, based on for the Caliphs and incorporates the
Constantinople country into his empire

Spinal tuberculosis, or Pott's disease, is more likely to
be found in an archaeological context. Evidence for pul-
monary tuberculosis is less tenable, as the bacilli disap-
pear soon after the death of the victim. The discovery of
lung collapse or pleuritic adhesions in some mummies
has been mentioned as evidence of its existence, but
other conditions may cause these complications.
1.3
Ancient Egyptian Art
Pott's disease is well presented in both ancient Egyp-
tian art and human remains. Several clay statuettes
of men, dating to the Predynastic period, have been
considered to be cases of Pott's disease. The men
are depicted with humped spines and an emaci-
ated look that suggests tuberculosis. Yet some of the
figures are in large clay bowls, which may reflect a
style of burial (Filer 1995). A wooden figurine in
the Brussels Museum, probably Predynastic period,
shows an angular deformity of the spine and chest,
strongly suggesting Pott's disease. A gardening scene
from the tomb of Ipwy of the 19th dynasty shows a
gardener with a humped back denoting Pott's disease
(Fig. 1.1). In the Cairo Museum, there is an Old King-
dom statue with a humped spine and a concomitant
deformity of the chest (Fig. 1.2). A false door from a
mastaba tomb of Ankh-my-was of the Old Kingdom
(4th to 5th dynasty), now preserved in Glyptotek Ny
Carlsberg, Copenhagen, represents the deceased with
a deformed upper part of the torso, indicating Pott's
disease (Cave 1939). The tomb of Seshen Nufer I from
the 4th dynasty near the Giza pyramids compromises
the scene of a humped-back serving maid clearly suf-
fering from Pott's disease (Mahmoud 1998). Ruffer
(1921) described two drawings denoting tuberculosis
of the spine. The first, in Beny Hassan, shows affec-
tion of the lower cervical and upper dorsal spine.
The other, in Tel EI-Amarna, dates back to the 18th
dynasty and shows the kyphosis is in the lower dorsal
and upper lumbar region.
There are many representations of humped-back
servants in the tomb chapels, but it is difficult to dis-
tinguish between Pott's disease, Porter's hump, anky-
losing spondylitis and simply poor posture. In Pott's
disease, the deformity is localized and the angulation
is usually in the form of a gibbus. The chest wall may
show a corresponding deformity.
6 S.A. Bedeir

Fig. 1.1. A scene from the tomb of Ipwy, 19th dynasty, Deir Fig. 1.2. An Old Kingdom statue with a humped back and chest
el-Medinah. A gardener irrigating a garden with a "shadouf". deformity denoting Pott's disease
His spine appears to have Pott's disease

1.4 A 4- to 6-year-old girl from the Ptolemaic period,

Paleopathology whose mummy is in the Bolton Museum, appears to
have Pott's disease. Lines of growth arrest (Harris line)
There are several reports of lesions attributed to tuber- are shown at the ends of the long bones due to the long
culosis in Egyptian skeletal materials and mummies. period of disease and stress (Filer 1995). When the
The classic, and perhaps the most convincing case of tomb of Iurudef at Saqqara was excavated in 1991, an
skeletal tuberculosis, comes from the 21st dynasty, or intrusive burial of a non-royal female was found to have
about 1000 B.C. It is the mummy of Nespaheran, a 25
to 30-year-old Egyptian priest of Amun, near Thebes.
It was found among a cache of 44 priests of Amun.
The spine is acutely angulated due to destruction of
the lower thoracic and upper lumbar vertebrae. A huge
abscess cavity is present in the sheath of the right psoas
muscle (Ruffer 1910, 1921; Cave 1939; Reyad 1960;
Kamal 1964; Filer 1995; Nunn 1997). The drawing, by
Mrs. Cecil M. Firth, was first introduced by Ruffer in
1910 and is now a classic (Fig. 1.3). Most of the known
possible cases from the Predynastic to the 21st dynasty
were reviewed by Morse et al. (1964). Buikstra et al.
(1993) added new cases. Flinders Petrie and Quibell
collected 13 specimens with tuberculosis affection
from 2000 graves in Naqada (Nunn 1997). Derry (1938)
reported nine cases. The Nubian collection of the Royal
College of Surgeons of England contained many cases
of skeletal tuberculosis, but many were lost together
with their records during an air raid in 1941.
Fig. 1.3. Nespaheran, a priest of Amun from the 21st dynasty.
Morse et al. (1964) described a single case of a The side view shows marked kyphosis of the spine and, in the
mummy with a collapsed lung and pleural adhesions, front view, a large cold abscess cavity is seen in the right psoas
which are common complications of tuberculosis. major muscle. The anterior abdominal wall is removed
Tuberculosis in Ancient Egypt 7

severe degeneration of the spine with marked angula-

tion and a possible paravertebral abscess. Iurudef was
a high Memphite official in the household of Princess
Tia, sister of Rameses II of the 19th dynasty. It is
believed that the woman, who died at approximately
the age of 20 years, had lived her life as an invalid and
only survived through family care (Walker 1991).
It is interesting to note that one of the possible
causes of the early death of Tutankhamen is tubercu-
losis (Reyad 1960).

1.5
Differential Diagnosis Fig. 1.4. Tuberculosis of the right iliac bone and right sacro-
iliac joint, of a male aged 35-39 years old at time of death. Giza
Although there is very little doubt that tuberculosis necropolis. There is ankylosis of the sacroiliac joint, erosion
of the iliac bone and the sacrum and a round defect in the
was the cause of the pathological findings in most of
iliac bone
the reported cases, it is sometimes difficult to exclude
compression fractures of the spine, osteomyelitis,
spinal metastasis degenerative spondylosis, fungal
infection and bone cysts.

1.6
The Author's Case

I was privileged to have the opportunity to examine

the Giza collection of disarticulated skeletons, which
should be considered one of the most important
materials in the history of archaeology and archaeo-
pathology. The material belongs to a very important
period, the Old Kingdom, which was when the great
pyramids were built. The United Nations Educational,
Scientific and Cultural Organization (UNESCO) had
added the entire archaeological area of Giza to its list
of world heritage, as it is the oldest and most famous
necropolis in the world. The material includes the
skeletal remains of two different classes that lived
in the same country during the same period of the
third millennium B.C. The high officials and the dig-
nitaries were buried in the western cemetery and the
workmen and artisans were buried in the southeast
cemetery. The excavations were headed by Reisner in
1942 and Hawass in 1994.
Skeleton number 1932 (male) is one of the high
officials buried in the western cemetery. His age at the
time of death was between 35 years and 39 years. The
right iliac bone and the adjoining right side of the
sacrum are markedly affected by a chronic infective
condition (Fig. 1.4). There is complete bony ankylosis
of the right sacroiliac bone (Fig. 1.5). The outer part
Fig. 1.5. Radiograph of the specimen in Fig. 1.4
of the bone is markedly eroded, with bone buildup at
the iliac crest and parts of the iliac bone and sacrum.
The greater trochanter of right femur shows exuber-
ant bone formation (Fig. 1.6). It seems that the tuber-
culous cold abscess trickled down along the gluteal
muscle to involve the trochanter in the tuberculous
process. The sacroiliac joint is the most common site
of the pelvis that is affected by tuberculosis (Ortner
and Putschar 1981) and chronic tuberculous affection
of the greater trochanter is a characteristic lesion that
affects young adults more than children. The spine
and the hip joints do not show obvious pathological
or radiological changes.
The differential diagnosis of this case is pyogenic
osteomyelitis, which is very rare in ancient Egyptian
8 S.A. Bedeir
Fig. 1.6. Chronic tuberculosis of the greater trochanter of the
right femur. Same case as in Fig. 1.4. There is erosion of the
bone and exuberant bone buildup denoting a good healing
reaction. There is no involvement of the femoral head
skeletal remains (Wood-Jones 1910). Patients with
hematogenous osteomyelitis were expected to die
from toxemia and pyemia before the inflammation
produced manifested bone changes.
The skeletal response to the tubercle bacilli is
dependent on the virulence of the organism and the
host response. If the host had little ability to defend
himself, then tuberculosis could cause a progressive
type of acute disease and early death. If the host's
resistance was considerable, it is presumed that the
disease then would be chronic, with the possibility
of extensive healing and exuberant bone reaction. In
our case, we believe that the patient's resistance was
good, as manifested by the complete bony fusion of
the right sacroiliac joint, the marked bone healing
reaction of the iliac bone and the excessive bony
build-up of the greater trochanter. The patient was
one of the high dignitaries of the king. He most likely
enjoyed eating a good, nutritive diet and lived in an
appropriate house. It is interesting to comment on the
presence of three healed fractures of both ulnae and
left fibula. Fracture of the ulna is usually considered
a "parry" injury, in which the patient tried to fend off
a blow using the forearm. These injuries, which were
in different stages of healing, indicate that our patient
was active and aggressive during his life.
1.7
Pulmonary and Osseous Tuberculosis
in Ancient Egyptian Mummies
Pulmonary tuberculosis has been suspected by many
authors to have occurred in ancient Egyptian popula-
tions (Cockburn et al.1975). The first to prove its occur-
rence, with microscopic confirmation of the presence
of Mycobacterium tuberculosis bacilli, was Zimmerman
(1979) from Michigan, USA. This author reported his
anatomical and microscopic finding on the mummy of
an ancient Egyptian child of five years of age belonging
to Dra-Abu-el-Naga, found on the road to the Valley of
the Kings, the tombs of the officials of 19th and 20th
dynasties (1314-1085 B.C.). The author described the
anatomical features that suggested pulmonary and
osseous tuberculosis with recurrent and finally fatal
pulmonary hemorrhage. Ziehl-Neelsen stain of a bone
specimen showed scattered acid-fast bacilli. In a more
recent report, pulmonary tuberculosis in an ancient
Egyptian mummy was diagnosed and confirmed by
polymerase chain reaction (PCR) (Nerlich et al. 1997).
These German authors described their excavation
findings at the tombs of the Nobles at Thebes-West in
Upper Egypt. They reported a mummy of 35-year-old
male dated from the New Kingdom (1550-1080 B.C.).
These authors noted residue of the affected right lung
with extensive pleural adhesions while the pleura of
the left lung were normal. They also found two lumbar
vertebral bodies severely affected anteriorly with lytic
lesions and new bone formation. Specimens from both
lungs were examined using PCR-amplification ofmyco-
bacterial DNA. A specific positive band was identified
from right lung specimens but not from the left unaf-
fected lung. This was probably the first confirmed case
of pulmonary tuberculosis using PCR in an ancient
Egyptian mummy (Nerlich et al. 1997).
1.8
peR and the Diagnosis of Tuberculosis
in Ancient Egyptian Population
Many authors have tried to use the relatively recently
acquired knowledge and development of DNA
molecular technology for various purposes (Leek
1979; Harrison and Connolly 1969, David 1997).
Svante Piiiibo (1985) used this technology, aiming
at studying the descent and relationship of the ancient
pharaonic families, individual relationships between
members of these families and population in the Nile
Valley. The author sampled 23 mummies from various
Tuberculosis in Ancient Egypt
dynasties between the 6th dynasty (2370-2160 B.C.) up
to the Roman times. The ages ofthese mummies ranged
between one-year-old boys to older ages, as estimated
by the radiocarbon age technique of accelerator mass
spectroscopy. Blunt-ended extract material was mixed
with E. coli DNA polymerase with labeled radioactive
nucleotides. The author concluded that recombinant
DNA techniques could allow the study of the descent
of the ancient Egyptian pharaonic families and their
individual relationships. However, this author failed to
recognize what would become the much wider use of
this technique-detecting diseases in ancient Egyptian
pharaonic population and families.
The molecular approaches to DNA probe analysis
using a nucleic acid-base methodology was originally
described in early 1980s. It has been used extensively
and has begun to have a major influence on current
clinical management of various diseases. Several meth-
ods of PCR assays have been developed and IS6110
PCR can be done with a turnaround time of 24-36 h.
It is a rapid and reliable molecular approach that can
detect Mycobacterium tuberculosis DNA directly from
clinical specimens. Such advances in investigative
techniques have been exploited by many scientists and
researchers who aim to trace diseases to an ancient time.
In a large series of ancient Egyptian cadavers, Zink and
colleagues (2001) from Munich, Germany, collected 41
skeletal specimens. Four were from Abydos (3000 B.C.)
and thirty-seven were from Thebes-West, Upper Egypt
(2120-500 B.C.). Three had typical macromorpho-
logical evidence of skeletal tuberculosis, 17 showed
probable disease and 21 were without morphological
skeletal changes. DNA was extracted from bone speci-
mens and amplified by PCR with a primer pair that
recognized the Mycobacterium tuberculosis complex
insertion sequence IS611O. Nine of these cases were
positive for Mycobacterium tuberculosis DNA. Two of
the three cadavers with typical macromorphology of
osseous tuberculosis had positive results for Mycobac-
terium tuberculosis DNA. Five of those with probable
and two of those without morphological changes had
positive results for Mycobacterium tuberculosis DNA.
The authors concluded that their findings confirmed
the relatively frequent occurrence of tuberculosis in
ancient Egypt, dating back to 3000 B.C.
1.8.1
Environment
The lives of ancient Egyptians were shaped by their
environment. The geography of Egypt was, and is
still, like that of no other country. The Nile valley
9
was enclosed by deserts that protected Egypt against
invasion, which resulted in long periods of stability.
1.8.1.1
The Nile
The Nile played a central role in daily life. Between
July and October of each year, the Nile flooded its
banks, depositing a rich and fertile layer of alluvial
slit on the land. Inundation saved an enormous labor
in the transportation of water and renewed the fertil-
ity of the land.
1.8.1.2
Climate
Herodotus stated that the good health of the ancient
Egyptians was due to the lack of variability in the
weather. The Egyptian sky was blue almost all year
round. The sun's rays nourished the Egyptians'bodies
with vitamin D, and decreased the incidence of rick-
ets. The dry, hot climate of Egypt is better suited to
the preservation of human remains than that of any
other country in the world. However, heat encourages
flies, which were responsible for the spread of infec-
tions, especially eye infections and diarrhea.
The Sahara (Khamasin) winds blew up dust
storms and sand particles, resulting in chest infec-
tion, pneumoconiosis, laryngitis and sinusitis.
1.8.1.3
Housing
The hot and dry climate influenced the design of
ancient Egyptian housing. Houses varied from the
large, villa-like houses of the top members of society
to the small, crowded houses, with limited amount
of light and space, of the poor people. In Amarna,
smoke-blackened roof timbers indicated a heavy pas-
sage of soot. Tuberculosis thrived in such conditions
of ill-housing and bad ventilation.
1.8.1.4
Diet
Egyptians were already enjoying reasonably healthy
and varied diets. Food production was not a major
problem in Egypt, provided that the inundation of the
Nile occurred every year. The main cereal crops were
barley and emmer, with wheat being introduced only
in the late period. Barley was used to make loaves and
beer. Vegetables and fruits were grown on a large scale.
Many of the familiar varieties were known: grapes, figs,
10
dates,pomegranates, melons, onions, leeks, cucumbers,
lettuce, garlic, lentils and chickpeas. These provided
many of the vitamins essential for good health. The
ancient Egyptians consumed a wide variety of animal
products, including cattle, sheep, goats and pigs. They
kept sheep for producing wool, milk, cheese and meat.
They ate ducks, geese, pigeons and quails. Hens were
known only during the Ptolemaic period. The ancient
Egyptians consumed fish that was fresh or dried
and salted. Pure salt has been found from the sixth
dynasty, but the method of preparation is unknown.
Dates or honey was used for sweetening. Bee keeping
was actively practiced, and honey was used to prepare
medicines. The funerary offerings for King Unas of the
fifth dynasty included:
• Milk, three kinds of beer, five kinds of wine, ten
loaves, four of bread, ten cakes, fruit cakes, four
meats, different cuts, joints, roast, spleen, limb,
breast, tail, goose, pigeon, figs, ten other fruits,
three kinds of corn, barley, spelt, five kinds of oil
and fresh plants (Nunn 1997).
1.8.1.5
Religion
Religion had a beneficial effect on the general health of
the ancient Egyptians. Sacred holidays allowed them
to rest and go to the temples to see the King. The poor
people took their shares from the offerings. Vacations
and ceremonial activities had the effect of raising the
morale of the people. They had to take baths using
running water and natron to clean their bodies. Men
would shave their hair and women removed their
unwanted hair. The ancient Egyptians believed that
diseases were caused by excessive food, thus they
used drugs and suppositories to evacuate their stom-
achs and bowels three times every month.
1.8.1.6
Sports and Recreation
Athletic sports were an ancient Egyptian invention.
The ancient Egyptians were skilled in wrestling, stick
combat, gymnastics, ball games, hunting, fishing and
bull fighting. Families played cup games and board
games (senet). They also enjoyed music, singing, and
dancing.
1.8.1.7
Physique
The ancient Egyptians were small; the mean height
of the males was 1.67 cm and the females 1.53 cm.
S.A. Bedeir
The arithmetic mean age at death was 36 years in the
dynastic period. There are very few burials of people
over the age of 60 years (Sarry EI-Din 1995). Pepi II
and Rameses II survived to well in their eighties.
1.9
The Medical Papyri
Although there can be no doubt of the existence
of tuberculosis in ancient Egypt, many authors
believe that nothing in the medical papyri can be
directly related to the disease. The medical papyri
do not contain any clear descriptions of tubercu-
losis (Ortner and Putschar 1981; Steinbock 1976).
However, Ebbell (1937) and Kamal (1964) believe
that the Ebers Papyrus (Ebers 1875) contains the
description of two cases with tuberculosis of the
cervical lymph nodes. The first case is described in
section 860, column 195; Ebbell entitled the section
"Soft Tuberculosis Gland". The section is translated
as follows:
\'Jl "Information about enlarged cystic node in his
neck"
- Examination: if you examine a man having an
enlarged cystic node in his neck, and you find it like
the thymus(?) gland because of its softness at pal-
pation and its white discharge ... (empty space).
- Diagnosis: you shall then say concerning him he
has enlargement of the cystic node in his neck.
An ailment I will treat by surgery to protect the
vessels.
- Treatment: you shall prepare medicines to treat
him by a bandage that makes the cyst open
through the skin: acacia seyal, pea, fruit, animal
blood, insect's blood, honey, common salt (and
other things). Grind. Mix together and bandage.
«I The second case, section 861, reads as follows:
- Title: information about enlarged infected node
in the neck of a man.
- Examination: if you examine an enlarged infected
node in the neck of a man. After it is enlarged and
removed the skin that covered it, and caused sup-
purating granulation for many years and months.
It discharges a secretion like the seminal fluid of
synodontis (a type of fish).
- Diagnosis: you shall then say concerning him he
has an enlarged infected node. An ailment I will
treat.
- Treatment: you shall prepare the medicines for
him: wax or fat, As(?), ink, salt, goose fat, fruit(?)
Galena. Heat and bandage the neck (Fig. 1.8),
Tuberculosis in Ancient Egypt 11

Fig.I.7. Healed fractures of the right

ulna and left fibula of the same case as
in Fig. 1.4

o -'- --
'h 5' JC
0
III
\
(-~
t-"o. M ~ .. fI n
00
'
.~""')JOIft': " I
II--. 2 n-~ <::Yl Q
II
~l ........

J! C\1d -
I "0......... 0
d•... -
"1: Q A
E 11/\ 0
tJ tJ. nf'rl))
J)nf\ c:>~....,
oC)
t2
1f' A·
~~(.) I~.a'-

t ~ "s..J l>,,:1I~UI ~ ;,,,......., "-t-- ci' J

Fig.1.8. Ebers papyrus, part of para-

graph 861, describing a chronic tuber-
culous node in the neck. Arabic transla-
tion by Kamal (1964)

The Ebers Papyrus was purchased in Luxor by
Edwin Smith in 1862. George Ebers referred to the
Papyrus in 1873, and it is now in the University
Library of Leipzig. It was translated into English
by Dr. Ebbell in 1836. It comprises 110 pages and
is by far the longest of the medical papyri. It is in
very good condition and the handwriting is clear.
It was written in 1534 B.C. It is claimed that the
Ebers' Papyrus (or at least parts of it) date back to
an ancient origin.
1.10
Sanatorial Treatment
In 1938, Derry excavated four possible cases of
spinal tuberculosis from two neighboring graves. In
one grave, Derry found the remains of a man and a
woman and, in the other grave, he found the skeletons
of two adult males and a nine-year-old boy. This clus-
ter of cases was regarded as evidence of a cemetery
for a tuberculosis sanatorium. The other possibility
is that the individuals were related and acquired the
disease from a common source of infection within
the family.
The ancient Egyptians knew sanatorial treatment.
The sanatorium was a building in which patients could
be totally or partially immersed in healing holy water.
In certain temples, incubation or "temple sleep" was
practiced; the patients were allowed to dream in the
hope of receiving a cure from a deity (Quirke 1992).
Sanatorial treatment was practiced in the great tem-
ples of Hathor and Dendera. Certain basins were sup-
plied with water from the sacred lake, which could be
made more curative by allowing it to ft.ow over healing
statues and certain texts. The effect of the sacred water
was obtained by drinking it (Dunn 1997).
12
Parts of Deir el-Bahri temple (Fig. 1.9) are consid-
ered by certain authorities to be the first sanatorium in
history. The temple was established by Queen Hatshep-
sut in the 18th dynasty, 1500 B.C. Patients and invalids
from different countries visited the sanatorium and
stayed beside its columns, hoping for recovery. During
the Ptolemaic period, Imhotep (Fig.UO) and Amen-
hotep, the son of Hapu, were worshipped on the upper
terrace of the temple at Deir el-Bahri, where the faithful
and sick gathered to request grace from the two gods of
medicine. In 1923, Sir William Osler said that Imhotep
was "the first figure of a physician to stand out clearly
Fig. 1.9. Deir el-Bahri. Is it the first sanatorium in the world?
Fig.1.10. Statue of Imhotep. He was the
patron of science, medicine and writing-one
of the "wise men" of the past. Graeco-Roman
period, Cairo Museum
S.A. Bedeir
from the mists of antiquity" (Nunn 1997). Milne and
Joseph (1914) concluded that the temple was used as a
sanatorium, and currently visitors can see the writings
of patients who wrote while sitting; the levels of these
writings corresponded with the heights of the patients.
Frequently, the writings include the name and occupa-
tion of the patient as well as the times of his arrival and
departure. One of those paragraphs reads as follows:
I am Andropachus from Macedonia. I came to visit
Amenotis. I was very ill, but now I am cured thanks to
the mighty God. I ask you merciful God to give us all
good health. Goodbye (Reyad 1960).
References
Bartels P (1907) Tuberkulose in der Jiingeren Steinzeit. Arch
AnthropoI6:243-250
Breed RS, Murray EGD, Smith NB (eds) (1957) Bergey's
manual of determinative bacteriology. Williams and
Wilkins, Baltimore
Buikstra AE, Baker BJ, Cook DC (1993) What diseases plagued
the ancient Egyptians? A century of controversy consid-
ered. In: Davies WV, Walker R (eds) Biological anthropol-
ogy and the study of Ancient Egypt. British Museum Press,
London, pp 231-250
Cave AJE (1939) The evidence forthe incidence of tuberculosis
in ancient Egypt. Br J TubercoI33:142-152
Cockburn A, Cockburn E (eds) (1980) Mummies, diseases and
Ancient cultures. Cambridge University Press, Cambridge
Cockburn A, Barraco RA et al (1975) Autopsy of an Egyptian
mummy. Science 187:1155-1160
David AR (1997) Disease in Egyptian mummies: the contribu-
tion of new technologies. Lancet 349:1760-1763
Derry DE (1938) Pott's disease in ancient Egypt. Med Press
Circ 197:196-199
Ebbell B (1973) The Papyrus Ebers. Oxford University Press,
London
Ebers GM (1875) Papyros Ebers (two vols). Englemann, Leipzig
Filer J (1995) Disease. British Museum Press, London
Harrison RG, Connolly RC (1969) Microdetermination of blood
group substances in ancient human tissue. Nature 224:326
Hawass Z (1996) The workmen's community at Giza plateau.
Sonderdruck aus Haus und Palast in alien Agypten. Inter-
nationales Symposium, 8 bis 11 April 1992 in Kairo.
Kamal H (1964) Ancient Egyptian medicine (two vols). Egyp-
tian Organization for editing, translation and distribution
(in Arabic)
Leek FF (1979) Teeth and bread in ancient Egypt. J Egyptian
ArchaeoI58:126-132
Mahmood MMR (1998) Occupational rehabilitation of the
disabled in Ancient Egypt. Thesis, Master's degree, Helwan
University
Milne M, Joseph R (1914) The sanatorium of Der el-Bahri. J
Egypt Archaei 1:96-99
Morse D, Brothwell DR, Ucho PJ (1964) Tuberculosis in
Ancient Egypt. Am Rev Respir Dis 90:524-541
Nerlich AG, Haas CJ et al (1997) Molecular evidence for tuber-
culosis in an ancient Egyptian mummy. Lancet 350:1404
Tuberculosis in Ancient Egypt
Nunn JF (1977) Ancient Egyptian medicine. British Museum
Press, London
Ortner DJ, Putschar WGJ (1981) Identification of pathological
conditions in human skeletal remains. Smithsonian Insti-
tution Press, Washington DC
Quirke SGT (1992) Ancient Egyptian religion. British Museum
Press, London
Reisner GA (1942) History of Giza necropolis, vol I. Cambridge
Harvard University Press, London
Reyad N (1960) Ancient Egyptian Medicine. Karnak Press,
Cairo (in Arabic)
Ruffer MA (1910) Pott'sche Krankheit an einer agyptischer
Mumie aus der zeit der 21 Dynastie. Zur historischen Biolo-
gie der Krankheitserreger, 3. Heft, Giessen
Ruffer MA (1921) Studies in palaeopathology of Egypt. Uni-
versity of Chicago Press, Chicago
Sarry El-Din AM (1995) Comparative study of skeletal mate-
rial from Giza old Kingdom with Sudanese Nubia. Thesis,
PhD degree, Cairo University
13
Steinbock RT (1976) Paleopathological diagnosis and inter-
pretation. Bone diseases in ancient human populations.
Thomas, Springfield, Illinois
Svante Piiiibo (1985) Molecular cloning of ancient Egyptian
mummy DNA. Nature 314:644-645
Walker R (1991) Skeletal remains. In: Raven MJ (ed) The tomb
of Iurudef, a Memphite official of Rameses II. Egyptiam
Exploration Society, London, pp 55-76
Wood-Jones F (1910) General pathology (including diseases
of the teeth). In: Elliot-Smith G, Wood-Jones F (eds) The
Archaeological Survey of Nubia Report for 1907-1908, vol
II. Report on human remains. National Printing Depart-
ment, Cairo, pp 263-292
Zimmerman MR (1979) Pulmonary and osseous tubercu-
losis in an Egyptian mummy. Bull NY Acad Med 55:
604-608
Zink A et al (2001) Molecular analysis of skeletal tuberculosis
in an ancient Egyptian population. J Med Microbiol 50:
355-366
2 Historical Aspects of Tuberculosis
M. MONIR MADKouR, KITAB E. AL-OTAIBI, R. AL SWAILEM

CONTENTS 2.25 Para Aminosalicyclic Acid (PAS) 28

2.26 Isoniazid 28
2.1 Islamic Medicine During Europe's Dark Ages 16 2.27 Rifampicin (Rifampin) 28
2.2 The Pulmonary Circulation and Ibn an-Nafis 16 2.28 Drug Therapy for Cure 28
2.3 Islamic Medicine and Its Scientific Contribution 2.29 Conclusion 29
to Tuberculosis 17 References 29
2.4 The Germ Theory 18
2.4.1 Miasma or Contagion 18
2.5 Percussion by an Amateur Musician 19
Tuberculosis was first described by ancient Egyptian
2.6 Laennec 19
2.6.1 Laennec's Work on Tuberculosis 19 doctors who described clearly the clinical features
2.6.2 Laennec's Invention of the Stethoscope 19 and medical and surgical treatment of scrofula (Fig.8,
2.6.3 Laennec's Tubercles 19 Chap. 1). The Ancient Egyptian artists contributed
2.7 Gaspard Laurent Bayle (1774-1816) 20 to the description through their exquisite paintings,
2.8 "Tubercle" and "Tuberculosis" 20
which showed spinal tuberculosis and its gibbous
2.9 The White Plague 20
2.10 Robert Koch (December 11, 1843-May 27,1910) 21 deformity (Fig. 1, Chap. 1). The first known sanato-
2.10.1 The Discovery of the Tubercle Bacillus 21 rium in the history of mankind was established by
2.10.2 Koch's Secret Cure (Tuberculin) and the Storm the Ancient Egyptians and is now open to visitors
That Followed 22 (Fig. 9, Chap. 1). Archaeological and anthropological
2.10.3 The Ziehl-Neelsen Stain, or Koch-Ehrlich-
studies indicated that tuberculosis infected the Chi-
Rindfleisch-Ziehl-Neelsen Method? 23
2.11 Franz Ziehl (1857-1926) 23 nese, Indo-European and pre-Columbian American
2.12 Friedrich Karl Adolph Neelsen (1854-1894) 24 tribes (Buikstra 1983; Powell 1992; Verano and Ube-
2.13 Paul Ehrlich 24 laker 1992; Paulsen 1987).
2.14 The Evolution of Tuberculosis Treatment 25 Hippocrates (460-361 B.C.) described tuberculo-
2.15 Climatological and Sanatorium Therapy 25
sis as "phthisis" and considered it to be an inherited
2.16 Sanatorium-Bimaristan or Deir el-Bahri 26
2.17 Artificial Lung Collapse Therapy 26 and non-contagious disease (Chadwick and Mann
2.18 Thoracoplasty 26 1978). Galen of Pergamon (130-200 A.D.) believed
2.19 Plombage 27 that tuberculosis was contagious. From Galen's time
2.20 Pulmonary Resection 27 until the 8th century, nothing of significant impor-
2.21 The History of BCG 27
tance in the history of tuberculosis occurred. The
2.22 The History of Chemotherapy
for Tuberculosis 27 Middle Ages in European history, also known as the
2.23 Streptomycin and Modern Antituberculosis "Dark Ages" (approximately 500-1500 A.D.) roughly
Therapy 27 corresponded to the "Golden Age of Islam." Scholars
2.24 Animal and Human Trials on the Efficacy in Islamic medicine, such as Rhazes, Avicenna and
of Streptomycin 28
Albucasis, described symptomatology, pathology
and predisposing factors for the development of
M. M. MADKouR, MD, DM, FRCP tuberculosis. Avicenna, suggested the communi-
Consultant, Department of Medicine, Riyadh Armed Forces
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
cable nature of tuberculosis. Avicenna's "Canon
K. E. AL-OTAlBI, MD, FRCS of Medicine" was translated to Latin and served
Director General of Medical Services Department, Riyadh as foundation for university courses in medicine
Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi between 1250 and 1600 in Europe (Cruse 1999).
Arabia Bimaristan (house of the sick) was first established
R. AL SWAILEM, MRCP (UK)
Consultant Rheumatologist, Department of Medicine, Riyadh
in Baghdad by the famous Harun AI-Rashid in 803.
Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi Ibn an-Nafis (AI-Quraishi, d. 1288) discovered pul-
Arabia monary circulation (Magner 1992).

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
16
During the early development of the industrial
revolution in Europe (approximately 1780), the
incidence of tuberculosis increased and reached epi-
demic proportions. It was the most common cause of
death in the 18th century. In Europe, from the later
part of the 18th century throughout the 19th century,
progress and advances have been achieved in the
diagnosis of tuberculosis. France was the champion,
with many French scholars such as Corvisart, Bayle,
Laennec and Villemin, greatly advancing our under-
standing of tuberculosis. The breakthrough in the
field of the diagnosis of tuberculosis was made by
the Germans. Robert Koch discovered the causative
bacillus in 1882 and Ziehl-Neelsen reported the stain-
ing method to identify the bacillus in 1882-1888. In
1895, Roentgen discovered Roentgen's X-ray, an
advanced investigative method.
Despite the identification of the bacillus, the dis-
covery of antibiotics and their use for tuberculosis
treatment lagged behind for over 60 years. However,
surgical interventions for the treatment of tubercu-
losis started much earlier than chemotherapy. The
human immunodeficiency virus (HIV) epidemic
started in 1981 and its association with the resur-
gence of tuberculosis in areas where the disease had
been controlled is a major setback in the fight against
tuberculosis. As there is no cure for HIV at present,
its management is reminiscent of tuberculosis prior
to the antibiotics era.
2.1
Islamic Medicine
During Europe's Dark Ages
The Middle Ages of European history, from approxi-
mately 500 to 1500 A.D., an era also popularly known
as the Dark Ages, roughly corresponds to the Golden
Age of Islam. The great sages of Islamic medicine were
established and studied in their own right and not just
in terms of their role in preserving classical Greek
medicine. Latin translation of the works of many
brilliant Islamic scholars, including Rhazes, Avicenna
and Albucasis, were most influential in Europe. Other
Arabic scholars held prominent status in medical
knowledge and practice in the Muslim world that had
no parallel in the west (Magner 1992).
During the European Dark Ages, western scholars
were merely interested in recovering Greek medi-
cal knowledge and had less interest in the practical
aspects of Islamic medicine, such as the development
of hospitals and clinical training. In 750, Baghdad was
M. M. Madkour et aI.
established as the capital of the Islamic Empire and by
the end of the 10th century, Baghdad and Cairo had
developed into independent centers of scholarship
(Magner 1992). The library in Cairo had over 100,000
volumes of manuscripts. Baghdad had 36 public
libraries by 1258, the year when the Moguls conquered
the city and its great libraries were destroyed.
2.2
The Pulmonary Circulation
and Ibn an-Nafis
Western scholars long maintained that Arabic
medical manuscripts were a Greek translation and
lacked originality. However, the strange story of Ibn
an-Nafis (AI-Qurashi, d. 1288) and his discovery of
the pulmonary circulation demonstrates that such
previous assumptions were unsound. The discovery
and the writings of Ibn an-Nafis on pulmonary cir-
culation were essentially ignored until 1924, when
an Egyptian physician (Dr Muhyi ad-Din at-Tatawi)
presented his doctoral thesis to the Medical Faculty
of Freiburg, Germany and brought it to the atten-
tion of Prof. Max Meyerhof (Magner 1992). Dr. Lois
N. Magner (1992), from the Department of History,
Indiana, USA, in his book, A History of Medicine;
wrote: "In the midst of a fairly conventional discus-
sion of the structure and function of the heart, Ibn
an-Nafis departed from the accepted explanation of
the movement of the blood. His description of the
two ventricles of the heart accepts the Galenic doc-
trine that the right ventricle is filled with blood and
the left ventricle with vital spirit. His next statement,
however, boldly contradicted Galen's teachings on
the pores in the septum. An-Nafis insisted that there
were no visible or invisible passages between the two
ventricles and argued that the septum between the
ventricles was thicker than other parts of the heart in
order to prevent the harmful and inappropriate pas-
sage of blood or spirit between them. Thus, to explain
the path taken by the blood, he reasoned that after
the blood had been refined in the right ventricle, it
was transmitted to the lung where it was rarefied
and mixed with air. The finest part of this blood
was then clarified and transmitted from the lung
to left ventricle. That is, the blood can only get into
the left ventricle by way of the lungs. Perhaps some
still obscure Arabic, Persian, or Hebrew manuscript
contains a commentary on the curious doctrines of
Ibn an-Nafis, but there is as yet no evidence that later
Arab authors or European anatomists were interested
Historical Aspects of Tuberculosis
in these anti-Galenic speculations. Thus, although
Ibn an-Nafis did not influence later writers, the fact
that his concept was so boldly stated in the thirteenth
century should lead us to question our assumptions
about progress and originality in the history of sci-
ence. Since only a small percentage of the pertinent
manuscripts have been studied, edited, translated,
and printed, the question may go unanswered for
quite some time:'
2.3
Islamic Medicine and Its Scientific
Contribution to Tuberculosis
Nothing of significant importance in the history of
tuberculosis occurred until the 9th and 10th centu-
ries, when the Islamic physicians Rhazes, Avicenna
and Albucasis wrote excellent descriptions of the
symptomatology, pathology and predisposing fac-
tors for the development of the disease. The com-
municable nature of tuberculosis was also suggested
by Avicenna, who believed in the curability of the
disease (Lyons and Petrucelli 1978).
Rhazes (Abu Bakr Muhammad Ibn Zakariya al-
Razi) (850-925) was a brilliant physician who prac-
ticed medicine in Baghdad and at various locations
in Iran. He has been known as the greatest physician
of the Islamic world. He is considered one of the
most scientifically minded physicians of the Middle
Ages and one of the greatest physicians of all time
(Magner 1992). He was the author of over 200 medi-
cal and philosophical treatises, including his massive
masterpiece the Continens, or Comprehensive Book
of Medicine, which was translated into Latin by the
Jewish physician Faraj Ibn Salim (known in Latin as
Farragut) for King Charles of Anjou (Magner 1992).
Rhazes' manual of the healing arts Kitab al-Mansori
(Almansour) was very influential in the West into the
16th century. He was chosen as director of one of
the first great hospitals in Baghdad and selected the
most healthful location for it. Rhazes provided clear
insight into the relationship between patients and
physicians, the signs and symptoms that the physi-
cian thought significant, the type of treatment given,
the family background and occupation of the patient.
Rhazes' book on smallpox and measles highlighted
an extraordinary ability of his clinical acuteness,
diagnosis, therapy and concepts of diseases.
Mesue (Yuhanna Ibn-Masawayh) was the court's
physician to four Abbasid Caliphs during the late 8th
and first half of the 9th centuries. He was a brilliant
17
physician and renowned clinician and teacher. He
wrote influential texts on nosology and therapeutics.
One of the 9th century physicians was Joannitius
(Hunay Ibn-Ishaq AI-Ibadi), who was proficient in
Greek, Syriac and Arabic and translated many Greek
medical texts. He wrote about many diseases, includ-
ing a monograph on ophthalmology.
Haly Abbas (Ali Ibn AI-Abbas AI-Majusi), who died
in 994 and who wrote the famous Kitab AI-Malaki
(also known as Kamil AI-Sinaa Filtib), one of the most
concise and well-recognized expositions of Greco-
Islamic medicine, was a physician at the Buwayhid
court (Johnston 1993; Gallagher 1993; Risse 1993,
quoted by Bishop and Neuma 1970).
Avicenna (Abu-Ali AI-Husayn Ibn-Sina) (980-
1037) was also known as a "Prince of Physicians"
(Cruse 1999). He was the first scholar to create a com-
plete philosophical system in the Arabic language
and his influence in medicine cannot be underes-
timated. Some critics indicated that his influence
could not be challenged by other physicians, which
therefore inhibited further development in medicine.
At the age of ten, Avicenna memorized the Qur'an
and turned to the natural sciences and to learning
medicine. He soon realized that some things could
only be learned by experience rather than solely
from books. Early in the 11 th century, he wrote his
leading medical encyclopedia, Canon of Medicine, a
five-volume study dealing with physiology, nosology,
etiology, symptomatology and therapy, pathology
and the preparation of compound remedies. His
enormous knowledge influenced the Islamic world
and was translated into Latin. The Latin translation
served as foundation for university courses in medi-
cine between 1250 and 1600 in Europe (Cruse 1999).
Significant advances made by Arab physicians served
to teach European physicians during the 8th through
the 11 th centuries. A well-known, brilliant Jewish
physician in Arabic medicine was Maimonides
(l135-1204), who translated "Canon of Medicine"
into Hebrew and served as a liaison for medical
knowledge between East and West (Cruse 1999).
Rhazes, Avicenna and Albucasis wrote about the
virtues of dry air and consuming fresh milk as cli-
matological therapy. They linked the ulcers of the
lungs to those of the extremities. They advised on the
values of using camphor in sugar roses and believed
in the curability of the disease. Avicenna suggested
the communicable nature of tuberculosis. Avicenna
gave an excellent description of the clinical features
and pathology of tuberculosis in Arabic script. Avi-
cenna, Rhazes and Haly Abbas described the people
who were at great risk of developing tuberculosis as
18 M. M. Madkour et al.

having narrow chests, thin bodies and ages between 2.4

18 years and 30 years. They described the symp- The Germ Theory
toms of tuberculosis (known in Arabic as; As'sul
or Ad'daran) as fever (more so at night), sweating, 2.4.1
malaise, anorexia, weight loss, cough, expectoration Miasma or Contagion
of sputum, hemoptysis and swelling of the legs. Al
Majousi also described clubbing of the fingers in The idea that tuberculosis can be transmitted by
patients with tuberculosis (Al-Amoud 1993). contact is an ancient folk belief, but it was generally
ignored in Hippocratic doctrine (Magner 1992). Physi-
Albucasis or Abu'l-Qasim (936-1013): Abu'l-Qasim cians and scientists have argued since antiquity about
KalafIbn-'Abbas al-Zahrawi (Albucasis), an extremely the terms contagion theory and miasma theory as the
ascetic man who devoted much of his time to work- cause of diseases. The miasma theory of disease was
ing among the poor, wrote a more specialized guide known as disease-inducing noxious and contaminated
to Arab Surgery. His book on Surgery and Instruments air and the theory was applied until the 19th century.
is one of the first comprehensive illustrated treatises The contagion theory of disease was suspected by
on this subject. He practiced in Cordoba, Spain and Fracastoro who published On Contagion in 1546.
wrote an encyclopedic study that contained an influ- Girolamo Fracastoro (1478-1553) was an Italian phy-
ential section on surgery based on Greek sources and sician, scientist, mathematician, astronomer, geologist
his own findings (Magner 1992). and poet. He created the story of the shepherd named
Pharmacology, alchemy and optics were other Syphilis, who brought about the first outbreak of
areas of great interest to Arab scientists. Arabic trea- syphilis because he cursed the sun and became the first
tises on medical pharmacology, including drug prod- victim of the new pestilence. Fracastoro's contagion
ucts and their nomenclature and aromatic medicinal theory of disease interested and inspired the microbi-
plants, were written in the 8th to the 11th centuries. ologists of the 19th century, and he was regarded as the
first to give an exposition of germ theory. Although the
AI-Kindi (800-870): Yaaqub Ibn-Ishaq-Al-Kindi wrote microscope was invented at the end of the 16th century,
a medical formulary and served as a model for Arabic its use to establish the existence of tiny "animalcules"
treatises on pharmacology, botany, zoology and min- was not applied until the 17th century by Antoni Van
eralogy. Persian and Indian drugs that were unknown Leeuwenhoek (1632-1723) who described "little ani-
to Hippocrates and Galen appeared in such formular- mals" (Magner 1992).
ies, as did new kinds of drug preparations. Professor Jacob Henle of Zurich (1809-1895)
The eleventh century chemist and author Abu AI- wrote, "physicians blamed disease on miasma, which
Biruni composed a treaties on pharmacy in which they defined as something that mixed with and poi-
he listed 720 drugs, including metallic compounds, soned the air, but no one had ever demonstrated the
for the treatment of diseases. Such expansions led to existence of miasma with scientific instruments".
the establishment of a medical pharmacy (Johnston Henle's hypothesis was that "Contagia Animata"
1993; Gallagher 1993; Risse 1993). (living organisms) caused contagious diseases
Islam devoted secular institutions to care for (Magner 1992).
the sick, named bimaristan, a Persian term mean- Giovanni Cosimo Bonomo (d.1697) was the first to
ing "house for the sick." The first bimaristan was provide convincing evidence that scabies was caused
established by Ibn Barmak in Baghdad before the by a tortoise-like mite. He also provided the contigu-
year 803, during the rule of the famous "Harun ous nature of scabies, "the itch", from one person to
AI-Rashid". Inmates came from all sectors of the another by means of bedding or clothing. Yet, the
population, including psychiatric patients (Galla- itch mite observation was not used as an example
gher 1993). Bimaristans were operated by a direc- of other infectious diseases (Magner 1992). Agostino
tor, medical staff, pharmacist, servants (as nurses) Bassi (1773-1857) also showed that silkworm disease
and had their own libraries, which contained many could be transmitted to healthy silkworms and sug-
medical texts. They also offered educational experi- gested its contiguous nature. In 1839, Lucas Schon-
ences for medical students serving apprenticeships. lein (1793-1864) a German professor of medicine
Lectures and discussions by famous physicians were reported his finding of the fungus that caused ring-
offered for students who were issued licenses for worm. In the second half of the 19th century, Louis
practicing medicine if they had successfully com- Pasteur and Robert Koch developed medical micro-
pleted theoretical and clinical courses. biology as a new science in the late 19th century.
Historical Aspects of Tuberculosis
2.5
Percussion by an Amateur Musician
Leopold Joseph Avenbrugger (1722-1809) was an
Austrian physician who worked in Vienna, he was
also a gifted amateur musician and composer. Using
his sensitive ear and musical talent, he invented chest
percussion as a diagnostic procedure. His published
work, entitled Invention Novum, became a landmark
in the history of medicine. He gave an account of a
new diagnostic method called "chest percussion",
also known as "chest thumping", which could give
the physician insight into the internal status of the
chest cavity by carefully evaluating the sounds pro-
duced by chest percussion (Magner 1992). Only few
physicians paid attention to Avenbrugger's invention
until Jean Nicolas Covisart (1755-1821) published a
translation and commentary in 1808.
2.6
Laennec
Rene Theophile Hyacinthe Laennec (1781-1826) was
a leading French physician, known for his work on
tuberculosis, a disease that he certainly died from
at the early age of 45. He postulated that all tuber-
cular phenomena (phthisis, scrofula and miliary)
constituted a single disease, which therefore raised
the theories of etiology. His invention of the initial
primitive stethoscope in 1816 made some of his
critics scoff at the perceived value of his instrument.
However, in 1825 his instrument was referred to as
"Laennec's immortal work".
Laennec was born in 1781 at Quimper in Brittany,
northwest France. At the age of 14 1/2 years, he stud-
ied medicine at Nantes under his uncle, Guillaume
Francois Laennec, rector of Nantes University. He
later went on to study in Paris in 1801 and qualified
in 1804. He was awarded many prizes and was asked
to withdraw from some competitive examinations so
as not to discourage his fellow-pupils.
2.6.1
Laennec's Work on Tuberculosis
The extensive pathological anatomy made by
Laennec was recorded in his 800-page manuscript,
still in existence in Nantes and Paris, but was never
published. He identified the various stages of pulmo-
nary consumption and defined the tubercle that was
19
found in many other body organs. During his work,
Laennec was subjected to many cuts to his fingers
while cutting through the cadavers, which possibly
contributed to his contracting and dying from the
disease (Keers 1981; Sakula 1981).
2.6.2
Laennec's Invention of the Stethoscope
While examining a stout female patient in 1816 at
Necker hospital, he used tightly rolled sheets of paper
attached to a wooden model in order to better hear
her heart sounds. This primitive stethoscope of his
went through many variations of form and material
throughout the 19th century. He published his first
book on auscultation in 1819, titled Del' Auscultation
Mediate etc. This book introduced a diagnostic tool
into medicine for the first time. He also wrote the
most clear, accurate and almost complete account of
chest diseases.
While working in Cornwall, UK in 1821, Dr. John
Forbes (later Sir John), translated Laennec's book
into English, abridged and condensed. The first edi-
tion of Forbes' translation seriously underestimated
the role and influence of Laennec's stethoscope, but
in the second edition he admitted that he had been
wrong in this estimation.
2.6.3
Laennec's Tubercles
Laennec wrote on "phthisis Pulmonalis" and "of the
essential, or anatomical character of tubercles in the
lung" and stated:
"The existence, in the lungs, of those peculiar
productions to which the name of Tubercles has
been restricted by modern anatomists, is the cause,
and constitutes the true anatomical character, of
Consumption". Laennec then went on to describe
the appearances of tubercles, tracing them from
their first, observable state, in which he wrote that
they vary from the size of a millet seed to that of a
hemp seed (and may be called miliary tubercles),
to the ultimate fibro-cavernous destruction of lung
substance (Bishop 1981). Later, he wrote: "There is
perhaps no organ free from the attack of tubercles,
and wherein we do not, occasionally discover them in
our examination of phthisical subjects. The following
are the parts in which I have met with these degen-
erations, and I enumerate them in the order of their
frequency: the bronchial, mediastinal, the cervical,
20
and the mesenteric glands; the other glands through-
out the body; the liver - in which they attain large
size, but come rarely to maturation; the prostate - in
which they are often found completely softened, and
leave, after their evacuation by the urethra, cavities
of different sizes; the surface of the peritoneum and
pleura, in which situations they are found small and
very numerous,... the epididymis, the vasa deferentia,
the testicle, spleen, heart, uterus, the brain and cere-
bellum, the bodies of the cranial bones, the substance
of the vertebra or the point of union between these
and the ligaments, the ribs, and lastly, tumours of the
kind usually denominated scirrus or cancer, in which
the tuberculous matter is either intimately combined
with, or separated in distinct patches from, the other
kinds of morbid degeneration existing in them." (p.
11; quoted from Bishop 1981).
Laennec restored the hope and confidence for a
cure or suspension of symptoms for patients suffer-
ing from tuberculosis. He discussed remedies that
has been used since antiquity, such as blood-letting,
mercurial salivation, emetics, charcoal, mushrooms,
red cabbage, the conserve of roses, crabs, oysters,
frogs, chocolate, alcohol, opium, cinchona, prepara-
tion of lead, hydrocyanic acid and the use of a swing
for the cure of phthisis (Bishop 1981). Laennec never
embraced these remedies, and it was said that his
approach to medicine was Hippocratic.
2.7
Gaspard Laurent Bayle (1774-1816)
A French physician, Laennec's friend and one-time
co-worker, published the result of his research on
tuberculosis. He carried out extensive pathological
studies of 900 post-mortem specimens. He recog-
nized and reported six types of pulmonary consump-
tion including: tubercular phthisis, glandular phthisis,
phthisis with melanosis, ulcerous phthisis, calculous
phthisis and cancerous phthisis. He noted "Four of his
ex-named varieties of phthisis represented one dis-
ease, tuberculosis in varying phases of developmene'
He contributed ideas to the knowledge of the pathol-
ogy and the unity of tuberculosis although he failed
to grasp the etiological identity of one phase with
another (Bishop 1981). Thomas Young (1773-1829),
a well-known London physician who initially studied
medicine at St. Bartholomew's Hospital in London
(Bishop 1973) wrote A Practical and Historical Trea-
tise on Consumptive Diseases published in London in
1815. He devoted nearly 12 pages to Bayle and wrote:
M. M. Madkour et al.
"Mr. Bayle's elaborate work has very lately been intro-
duced to the English public, with no small parade, by
his translator Dr. Barrow; and it must not be denied
that it contains much valuable matter, although the
practical importance of anatomical observations in
general may have been somewhat overrated by the
author. The translator talks of the propriety of some
legislative measures for enforcing the keeping of reg-
isters of cases in all public hospitals, observing that,
if such a regulation were adopted, France would not
'long exult as at present in her claim to preeminence,
either in pathological or practical knowledge'" (p. 447;
quoted by Bishop 1973).
2.8
"Tubercle" and "Tuberculosis"
The term "tubercle" was first used by Franciscus
(Francois) de la Boe, also known as Sylvius of
Leyden (1416-1672), an outstanding Dutch anatomist
and iatrochemist. He attempted to classify catarrhal
diseases into groups distinguished by the chemical
qualities of the humor produced-either alkaline or
acidic in nature. He established the first university
chemical laboratory at Leyden. He was the first to
use the term "tubercle" and stated that tubercles
were often seen in the lungs of consumptives (John-
ston 1993). The term "tuberculosis" was first used by
Laennec and his friend and colleague, Bayle. Laennec
postulated that all tubercular phenomena (phthisis,
scrofula and miliary) constitute a single disease and
raised the theory of etiology.
2.9
The White Plague
"I decided to entitle this presentation ...by avoiding
the unfashionable word 'tuberculosis' and substitut-
ing a more romantic appellation:' noted Professor
John F. Murray, University of California, San Fran-
cisco, when he gave "The 1989 J. Burns Amberson
Lecture", entitled: "The White Plague". He hoped to
stimulate an interest in and recruit an audience for
his lecture, and certainly he did. Prof. Murray went
on to say, "I borrowed my title from that of a book by
Rene and Jean Dubos, who borrowed it, in turn, from
Oliver Wendell Holmes, Sr., who first used 'the white
plague' in 1861 to describe the enormity of the prob-
lems caused by tuberculosis at the time:' The phrase
Historical Aspects of Tuberculosis
"the white plague" has intrigued many historians,
such as Dr. Allen B. Weisse (1995) from the University
of New Jersey Medical School who wrote: "I wondered
if the 'white' in 'white plague' related to race, pathol-
ogy, the appearance of the patient or perhaps some
other aspect of the disease". Although the population
of Western Europe and the USA is predominantly
white, non-whites are more susceptible to the disease
than whites. Pathological appearance of the lesion in
early post-mortem descriptions included terms such
as "tumor albus, or white swelling" as in Richard M.
Burke's book, An Historical Chronology of Tuberculo-
sis (Myers 1977). Others described hepatic lesions as
"the white tubercle of the liver", while Matthew Bail-
lie (in 1703) described pulmonary tubercle as gray, or,
as noted by the great Laennec, grayish or, following
caseation, yellowish (quoted by Weisse 1995).
If "white" reflected the patients' appearance, Dr.
Weisse explained, clinicians have often commented
on flushed skin rather than pallor reflecting the
febrile state. The term "plague", from the Latin "to
blow or strike", is an epidemic disease causing a high
rate of mortality: pestilence. Early prehistoric ancient
Greeks knew tuberculosis when over 130,000 persons
died over a 5-year period (430-426 B.C.). Hippocrates
(460-370 B.C.) was alive in Greece during that plague
but he used the term "phthisis" to describe the rapidly
progressive pulmonary form of the disease, which
later became known as "galloping consumption".
In the USA and Western Europe during the 18th
and 19th centuries, tuberculosis was one of the most
common causes of death (the other was cholera,
the black death), and the usual chronic course of
the illness was well recognized. Therefore the term
"plague" was obviously a misnomer when referring
to tuberculosis (Weisse 1995).
2.10
Robert Koch
(December 11, 1843-May 27,1910)
While working on anthrax in 1877, Robert Koch, a
great German physician and bacteriologist, was the
first to demonstrate that a specific pathogen caused
a specific disease. He discovered tubercle bacillus on
24 March 1882 as the cause of tuberculosis. On 17
September 1883, while working in Alexandria, Egypt,
he discovered Vibrio cholerae, but was unsuccessful in
transmitting the infection to an experimental animal.
This made his critics doubt his "postulates", which
were not fulfilled; a criterion that he determined in
21
order to relate an organism as a causative agent to a
disease. He also discovered the method of transmis-
sion of bubonic plague and sleeping sickness as well as
other tropical diseases. One day after his birthday, on
12 December 1905, Koch received the Nobel Prize.
Heinrich Herrmann Robert Koch was born on
11 December 1843 in Clausthal, a small mining city
located between Hannover and Gottingen. He was the
3rd in a family with 11 (surviving) children (9 boys
and 2 girls) and was closely attached to his mother.
He was interested in photography, mathematics and
natural sciences as a child. He studied English and
French and was fluent. He joined Gottingen Univer-
sity at age 19 years to become a teacher but decided
to change to studying medicine. He graduated at the
age of 23 years and wanted to become a ship's doctor
but gave up his idea in favor of becoming engaged to
his sweetheart, Emmy Fraatz. He worked as medical
assistant in Hamburg General Hospital as a source of
income, but the income was not enough to provide for
a family. He married Emmy on 16 July 1867. However,
he had financial difficulties as he moved to other jobs.
He volunteered in the Franco-German war in 1870
after having been rejected for army service because he
was nearsighted. In 1872, he took an exam for District
Medical Officer. His work on anthrax began in 1875.
At the end of that year, he discovered the causative
organism after inoculating a rabbit through the ear
and showing the bacteria on microscopic examination
(Brock 1999). The discovery of the etiology of anthrax
made Koch famous among the intellectuals.
2.10.1
The Discovery of the Tubercle Bacillus
Tuberculosis was strongly suspected to be a conta-
gious disease before the discovery of the bacillus.
In 1865, it was shown by a French physician, Jean
Antoine Villemin (1827-1892), that tuberculosis
was transmissible to experimental animals; how-
ever, the suspected causal organism had never been
seen (Brock 1999). In August, 1881, Koch started
his experimental studies, aiming at identifying the
causative organism. On 24 March 1882, Koch gave
his historic lecture in Berlin on the discovery of
the bacillus. He was aware that the presence of the
organism did not indicate that it was the cause of
the disease. He proposed a criteria in order to make
such an association, which became known as "Koch's
Postulates". The postulates are a series of steps that
should be followed to prove that a specific organ-
ism is the cause of a specific disease. The organism
22
must be isolated and grown in a pure culture and the
pure culture must induce the disease when injected
in experimental animal. Koch himself failed to fulfill
his postulate when he discovered Vibrio cholerae as
the cause of cholera while working in Alexandria,
Egypt on 17 September 1883. He was not successful
in establishing an animal model at that time and his
postulates were thrown back at him by his critics in a
proceeding of the first cholera conference. An edito-
rial that appeared in the «British Medical Journal" in
1884 (quoted by Brock 1999) noted, «Of course, the
whole point turns on whether Dr. Koch has made out
that the comma-bacillus is really the cause of the dis-
ease. In order to demonstrate that a given bacterium
is the cause of a disease, it must be proved:
1) that a special bacterium with definite characteris-
tics marking it out from other forms of bacteria,
is constantly present in the parts affected
2) that his bacterium is present in sufficient numbers
to account for the disease
3) that it is not similarly associated with other dis-
eases
4) that this bacterium can be cultivated apart from
the body, and that its introduction into lower
animals is followed by the same effects as the
introduction of the infective material itself"
In response to his critics, Koch wrote,«The comma
bacillus is a specific bacterium, found exclusively in
association with Asiatic Cholera. As long as this
statement is not contradicted, all of my conclusions
regarding the diagnostic utility of this bacterium and
its relation to the pathology of cholera remain valid.
However, my opponents state that a causal relation-
ship between the comma bacillus and cholera has
not been established because it has not been possible
to induce cholera artificially in experimental ani-
mals. This objection is not valid because Professors
Rietsch and Nicati succeeded, during the last cholera
epidemic in Marseille, to bring about cholera-like
symptoms in dogs and guinea pigs, provided the
bile duct had been tied off and a pure culture of
the comma bacillus injected directly into the duo-
denum... These studies have been recently repeated
here in my laboratory, using a pure culture diluted
so much that less than one-hundredth of a drop was
injected... With few exceptions, the treated animals
died in 1.5-3 days... These animal studies have been
extended in other directions and have shown without
a doubt that the comma bacillus is pathogenic. Under
these circumstances, it would be very advisable to
avoid completely any inoculation studies in humans
(as has been recently discussed), and restrict all inoc-
M. M. Madkour et al.
ulation studies to guinea pigs and other experimental
animals" (quoted by Brock 1999).
2.10.2
Koch's Secret Cure (Tuberculin) and the Storm
That Followed
In August, 1890, Koch dramatically reported secret
research in a speech given at the 10th International
Congress of Medicine in Berlin. He implied that he
found a cure for tuberculosis and said: «[after long
study of many chemicals 1I have at last found sub-
stances which both in the test-tube and in the living
body prevent the growth of the tubercle bacilli. All
such investigations... are very exhausting and slow;
and my experiments with these substances, though
lasting more than a year, are not yet concluded, so
that all I can say at present is that if guinea pigs are
treated they cannot be inoculated with tuberculosis,
and guinea pigs which already are in the late stages
of the disease are completely cured, although the
body suffers no ill effects from the treatment. From
these experiments I will draw no other conclusion at
present than that it is possible to render pathogenic
bacteria within the body harmless without ill effect
on the body itself" (quoted from Brock 1999).
News spread rapidly and thousands of patients
traveled from all over Europe to Berlin, seeking the
injections of «Koch's lymph" hoping to get cured. His
new «remedy" was tried on a human at the Charite
Hospital, but the validity of the cure were doubted.
Tuberculous patients commonly showed a strong reac-
tion to injection with tuberculin as a hypersensitivity
reaction and became known as «Koch Phenomenon;'
Later «tuberculin" was used as an important diagnos-
tic test. Koch injected himself with the tuberculin and
reported fever, rigor, joint pains, nausea and vomiting
which improved after 12 h to few days. He refused to
describe the method of tuberculin preparation and
was severely criticized for not revealing the nature
of his remedy. In 189011891, it became clear that
Koch's tuberculin had no therapeutic value but was
of great diagnostic use. The compound was known as
«remedy" or «Koch's lymph:'but in 1891 Koch and Lib-
bertz labeled it «tuberculin". The stress of the tubercu-
lin storm created by his cynical critics was intense and
Koch arranged for a long vacation to Egypt in 1891,
leaving his son-in-law in charge of tuberculin and the
Institute for Infectious Diseases (Magner 1992).
The «tuberculin" crisis had severely affected
Koch's scientific life. He became acquainted with
Hedwig Frieberg, a pretty, 17-year-old art student.
Historical Aspects of Tuberculosis
His relationship with his wife Emmy had been dete-
riorating and they eventually divorced in 1891. He
married Hedwig on 13 September 1893, when he
was 50 years and she was 20 years of age. They lived
together until he died in 1910. Hedwig Koch died on
16 June 1945. "In his lifetime, Robert Koch did more
to single handedly advance the world's understand-
ing of microbes as causes of disease than any other
man, with the exception only of his great French rival
Louis Pasteur" (James Strick, who wrote the foreword
to Thomas D. Brock's book, Robert Koch: A Life in
Medicine and Bacteriology, 1999).
2.10.3
The Ziehl-Neelsen Stain, or Koch-Ehrlich-
Rindfleisch-Ziehl-Neelsen Method?
The history of staining the tubercle bacilli started when
Koch's made his discovery of the tubercle bacillus on
24 March 1882. The translation of Koch's first publica-
tion on the subject on 10 April 1882 Die Aetiologie der
Tuberculose was translated by William de Rouville in
1938 and quoted in the Bishop and Neumann (1970)
paper on the history of the Ziehl-Neelsen Stain.
Koch's method of staining the tubercle bacillus
included the three most important factors: the use
of a chemical substance that made staining possible
(mordant), heat, and decolorizing. He wrote: "The
material to be examined is prepared in the usual
manner for examining for pathogenic bacteria, and
either spread on the cover slip, dried and heated or
cut into sections after fixation in alcohol. The cover
slips or sections are placed in a staining solution of
the following constitution: 200 cc of distilled water
are mixed with 1 cc of a concentrated alcoholic solu-
tion of methylene blue, shaken up, and then 0.2 cc of
a 10% solution of potassium hydroxide is added with
repeated shaking. This mixture must show no precip-
itate after standing for several days. The materials to
be stained remain in this solution for 20 to 24 hours.
By heating the solution to 400C in a water bath this
time can be shortened to a half to one hour. Following
this the cover slips are covered with a concentrated
aqueous solution of vesuvin (Bismarck brown) which
is filtered each time before using, and one to two min-
utes rinsed with distilled water. When the cover slips
come from the methylene blue, the attached layer
appears dark blue and markedly (p. 222) overstained.
During the treatment with vesuvin this blue colour is
lost and it appears stained a faint brown. Under the
microscope all the constituents of animal tissue, that
is, the cell nuclei and their products of disintegration
23
appear brown, while the tubercle bacilli on the other
hand, stain a beautiful blue. Moreover, all other bacte-
ria which I have investigated to date, with the excep-
tion of the lepra bacilli, take on a brown colour with
this staining method. The color contrast between the
brown stained tissue and the blue tubercle bacilli is
so striking that the latter, which are present often
only in very small number, nevertheless, are to be
found and identified with the greatest certainty" (pp.
227-228 in the original: Koch 1882; the translation is
that of William de Rouville, 1938, pp. 854-55).
2.11
Franz Ziehl (1857-1926)
A neurologist at Lubeck, Franz Ziehl was born in
Germany at Wismar in Mecklenburg-Schwerin. He
qualified and worked in Heidelberg as assistant in
a medical clinic from 1881 until 1886. He moved to
Lubeck in 1887 to work as neurologist until his death
on 7 April 1926. Ziehl published his first paper on
staining the tubercle bacillus on 12 August 1882 and
described how he had modified Ehrlich's method by
changing the mordant from aniline to carbolic acid.
He wrote: "Up to the present we have not been able
to obtain the pure constituents of aniline oil. As it
was necessary to use a substance belonging to the
aromatic series we tested such a one, the reaction of
which was known to us from that second aniline oil,
namely carbolic acid" (Ziehl 1882 p. 451, quoted by
Bishop and Neuman 1970).
Description of the acid-fastness of the bacilli by
Ziehl was published in another paper on 25 April
1883, entitled "On the staining of tubercle bacilli"
(Ziehl 1883b, quoted by Bishop and Neumann 1970).
"The importance of finding tubercle bacilli
especially with regard to diagnosis and prognosis"
was the title of another paper made by Ziehl that
appeared on 31 January 1883 and reported the find-
ing of the bacilli in 73 cases of pulmonary tuberculo-
sis and concluded:
"1.In most cases of pulmonary tuberculosis tubercle
bacilli may be found; exceptions do occur.
2. The finding of tubercle bacilli may confirm the
diagnosis of tuberculosis. In some cases the dif-
ferential diagnosis is possible by this finding.
3. If tubercle bacilli cannot be found, it is not always
possible to exclude tuberculosis.
4. The number and kind of tubercle bacilli do not
allow any prognostic conclusions." (Ziehl 1883a,
quoted by Bishop and Neuman 1970).
24
2.12
Friedrich Karl Adolph Neelsen
(1854-1894)
A German pathologist born at Utersen, Holstein,
Neelsen studied, qualified and worked as assistant at
the Pathology Institute at Leipzig in 1876 until 1878.
Between 1878 and 1885 he had a similar position at
Rostock and became a professor in 1884. He moved
to Dresden in 1885 and worked as Prosector at the
Staedtisches Krankenhaus. He died at the age of 40 years
on 11 April 1894. His death was almost certainly due
to tuberculosis, as he suffered from emaciation, cough,
progressive hoarseness and general deterioration.
These symptoms started in the autumn of 1893. In 1883,
Neelsen combined ZieW's mordant with Ehrlich's red
stain fuchsin in staining the tubercle bacillus. Neelsen
published the first article on the subject on 14 July 1883
in a German journal "Centralblattfur die medicinischen
Wissenschaften". The title of the article was "A casuistic
note on the theory of tuberculosis" and noted:
"In this case I used first as a staining fluid a three-
quarter per cent solution of fuchsin in 5% carbolic acid
with admixture of some alcohol and decolorized with
25% sulphuric acid. This method, which I have used
almost exclusively for a long time, gives in my hands
better results than the other methods (perhaps only
because I am most accustomed to it)" (Neelsen 1883
p. 500, quoted by Bishop and Neuman 1970).
2.13
Paul Ehrlich
A German bacteriologist and pathologist born in 1863,
Ehrlich made an improvement of Koch's method of
staining the tubercle bacillus. His first communication
on the subject was made to the Berlin Society ofInternal
Medicine on 1 May 1882. He used decolorization with
a mineral acid, used a red stain, fuchsin, and changed
the mordant to aniline oil instead of alkaline methylene
blue. He wrote: "After these remarks I wish to briefly
report the method. I have worked almost exclusively
with dry preparations, Le. of sputa, but I have done con-
trolled trials which have proved that the method is also
applicable to section preparations. I proceed by taking
out a small particle with a dissecting needle from the
sputum and squeezing it flat between two cover glasses,
and I have found that those 0.10 to 0.12 mm are best.
Under these conditions it is easy to obtain uniformly
thin layers from the small plug of sputum. Then both
glasses are torn from each other and one then gets two
M. M. Madkour et aI.
thin layers, which can easily be air-dried. These prepa-
rations are not yet suitable. Usually I have done this by
keeping the preparations for one hour at 100 to 110oe;
however there exists a more convenient method which
I have seen used in the Kaiserliches Gesundheitsamt
and which consists in passing the air-dried preparations
three times through the flame of a Bunsen burner. For
staining I use water, saturated by aniline oil, which may
be prepared in a few minutes by shaking water with sur-
plus aniline oil and filtering it through a moistened filter.
To the water-clear fluid obtained in this way one adds
drop by drop a saturated alcoholic solution of fuchsin
or methylene violet until a distinct opalescence indi-
cates saturation with the dyestuff. Upon this fluid the
preparations are made to float and staining is adequate
within a quarter to half and hour. The colour of the
tubercle bacilli does not disappear under the influence
of vesuvin (Bismarck brown) or only very slowly. It is
necessary to use an acid and I soon turned to the use of
strong, I may even say heroic ones. A mixture consisting
of 1volume of official nitric acid and 2 volumes ofwater
can be recommended. Under its influence one sees the
preparation grow pale within a few seconds, yellow
clouds rise and the preparation becomes white:'
"If one examines the preparation at this stage, it
is seen to be decolorized and only the bacillus has
maintained its intensive staining. It might be possible
to examine such a preparation, but the technical dif-
ficulties of focusing the bacillus are extremely great.
It is recommended to stain the background yellow if
the bacilli are violet, and blue if they are red."
"I have succeeded in staining the bacillus with
all basic aniline dyes, even with Bismarck brown,
and hence it follows that the substance of the bacil-
lus itself does not differ from that of other bacilli
in its staining properties. But the discrimination
of the tubercle bacillus by staining depends on the
presence of a surface layer with characteristic and
specific properties. The first of these to which Koch's
technique is pointing is that the covering layer is per-
meable only to dyes under the influence of alkalis'."
(Ehrlich 1882 pp. 269-270)
In 1882, Edvard Georg Rindfleisch, a Wiirzburg
pathologist, made an improvement on Koch's method
of staining tubercle bacillus by heating the slide
instead of putting it in hot water.
Historians of tuberculosis made good accounts in
the literature of the history of the staining methods
of the tubercle bacillus since its first discovery by
Robert Koch in 1882. The time between 1882 and
1888 was noted by Predohl, where all the methods
of staining Koch's bacillus have been described. He
was the first to mention the words, the Ziehl-Neelsen
Historical Aspects of Tuberculosis
method, and noted,"I must describe the modification
of Ziehl's method given by Neelsen".
The phrase Ziehl-Neelsen method of staining
tubercle bacilli was not mentioned in the first edition
of Edgar March Crookshank's book of bacteriology
in 1886. Although his list included both Ziehl's and
Neelsen's methods, the phrase"Ziehl-Neelsen method"
was not used. In the fourth edition in 1896, Crookshank
noted: "The Ziehl-Neelsen method is preferred by the
author for sections and cover-glass preparations;'
The phrase and label of the Ziehl-Neelsen method
may have become known sometime between 1890 and
1893 (Bishop and Neumann 1970). These two German
doctors, Ziehl and Neelsen, described their methods
nearly 100 years ago. These methods remain in use at
present and will always be in the literature of tuber-
culosis.
2.14
The Evolution of Tuberculosis Treatment
From the time of the Ancient Egyptians, Hippocrates,
Galen, and the Middle Ages, up until the early 19th cen-
tury, little has been added to the therapy of tuberculo-
sis. The ancient Egyptians treated scrofula with surgery
and dressings (see Chap. 1). Hippocrates' treatment was
simple and related to confinement into the temples of
health, resting, praying, drinking milk, dieting, exercis-
ing and avoidance of extreme weather. Drug treatment
was not emphasized. Five centuries later, Galen added
little to the classic Hippocratic treatment for pulmonary
tuberculosis. Galen recommended bed rest, gargles,
application of plasters to the chest and head and the diet
and dry air ofhealth resorts. Such concepts of treatment
for tuberculosis remained for more than 1000 years. In
medieval Europe, scrofula was more common than
the pulmonary form of the disease. It was customary
to believe in the "King's touch", in which the kings of
France and England were believed to have the power
to cure scrofula simply by touching its victims. This
custom originated in the 12th century and persisted
until the 18th century (Risse 1993; Gallagher 1993;
Johnston 1993). The Oxfordshire village of Stanton St.
John was burdened with scrofula, "The King's Evil". Its
residents desired the cure by the touch. The touching was
performed by English sovereigns from Edward the Con-
fessor to Queen Anne. Those who did not respond chose
a range of equally efficacious modes of therapy accom-
panied by prayer. Records at Stanton St. John indicated
that certificates attested to the eligibility of parishioners
to practice "The King's Touch" (Crawfurd 1911).
25
In North America, early in the 17th century, the
Reverend Cotton Mather, famous for his persecution
of witchcraft in Salem, wrote a medical work entitled
The Angel of Bethesda, which quoted Galen on ther-
apy and directed the patient to repent his sins. He
recommended a concoction obtained by strangling
two roosters, beating them to a pulp, boiling them in
a mixture of wine and water, straining the mixture
through a cloth and adding raisins, hartshorn, maid-
enhair and saffron (Rosenblatt 1973).
2.15
Climatological and Sanatorium Therapy
During the early industrial revolution, around 1780,
the recorded mortality rate due to tuberculosis in
England was 1,120 per 100,000 population (Rogers
1969). Early in the 19th century the mortality rate
started to decline and the 19th century was called the
century of tuberculosis. Treatment of tuberculosis at
the beginning of the 19th century was summarized
in a thesis of Edward Delafield published in 1816
(quoted by Rogers 1969). Delafield recommended
blood letting, emetics, mercury, opiates, digitalis
and Peruvian bark. He recommended sea voyages
and mild climate, particularly in the south of France.
He speculated that the nausea of seasickness was
the principle that explained the benefits of sailing.
In 1834, Samuel Morton, who studied in Paris and
was lectured by Laennec, Bayle and Louis, published
the first American textbook on tuberculosis (Rogers
1969). Morton deplored blood letting for the treat-
ment of tuberculosis and was in favor of the effect
of climate, particularly cold, dry atmosphere. Many
other authors in the 19th century were interested in
climatotherapy and recommended other areas, such
as the Rocky Mountains and Pike's Peak. Rogers
(1969) quoted from Disturnell (1867) who quoted
reports of surgeon G.K. Wood, of the US Army, sta-
tioned at Fort Laramie, Wyoming: "The climate of
these broad and elevated table-lands which skirt the
base of the Rocky Mountains on the east, is especially
beneficial to persons suffering from pulmonary dis-
eases, or with a scrofulous diathesis... It is of great
importance that the climate of this region should be
generally known..."
Daniel Dark (1850), quoted by Rogers (1969),
described the Missouri river and its environs around
Fort Leavenworth in a work entitled, "Journeys of
health on the great plains;' He wrote: 1. The patient
escapes from malaria, or that condition of the atmo-
26
sphere, which, in the larger portions of the Great
Valley, gives origin to intermittent and remittent
fevers, too often followed by infirmities for which a
change of locality is the only effectual remedy. 2. He
is constantly immersed in a dryer air.... 4. He lives on
a reduced, solid, and simple diet. ..6. He takes regular
saddle exercise...8. He is divested of his old cares.....
9. He is redeemed from the dominion of empiricism
and polypharmacy."
In the second half of the 19th century, climatogra-
phy was well established. The American Climatologi-
cal Association was formed in 1884 and held the first
meeting in Washington, D.C. in 1908. The Sixth Inter-
national Congress on Tuberculosis was held, and W.
Jarvis Barlow of Los Angeles said, "The high altitude
treatment received great impetus from this congress:'
The transactions of the congress of 1908 showed
something quite different: "The apogee of high alti-
tude therapy have been reached and its advocacy
would slowly subside thereafter" (Rogers 1969).
The name of the "American Climatological Asso-
ciation" was changed in 1914 and again in 1933 as
the problem of the efficacy of altitude therapy was a
classic case of the problem of multiple variables. In
1960, Spriggs (quoted by Rogers 1969) published an
article and said, "What statistical evidence there was
for the value of rest treatment does not nowadays
convince.....". By the time of World War II, climato-
therapy in the United States was a dead issue (Rogers
1969; Wilson 1979).
2.16
Sanatorium-Bimaristan or Deir el-Bahri
"I am Andropachus from Macedonia. I came to visit
Amenotis. I was very ill, but now I am cured thanks
to the mighty God. I ask you merciful God to give us
all good health. Goodbye:' These were the writings in
hieroglyphic by an ancient sick Greek patient on the
wall of Deir el-Bahri temple, which was believed by
Milne and Joseph (1914) to be the first Sanatorium in
history (see Chap. I). Nowadays visitors can see the
writing of patients who sought cure in this sanatorium
"Temple:' Bimaristan (houses of the sick in Persian)
were first established in 803 by the famous "Harun AI-
Rashid" in Baghdad.
Sanatorium, is a word that comes from the Latin
Sanatorius meaning "an establishment providing
therapy" (Wright 1988). In 1836, George Bodington,
a practitioner in Birmingham, UK, was the first to
propose housing tuberculosis patients in one build-
M. M. Madkour et aI.
ing (Rosenblatt 1973). His aim was to provide these
patients with diet, physical activities and fresh air, but
the proposal failed for lack of support. The first sana-
torium for the treatment of patients with pulmonary
tuberculosis was opened in 1859 in Gorbersdorf, Ger-
many by Dr. Hermann Breemer (Rosenblatt 1973). He
believed that mountain air and exercise were benefi-
cial. Many other sanatoria were opened, and perhaps
one of the most famous was Trudeau Sanatorium in
Saranac Lake, N.Y. It was established by Edward L.
Trudeau in 1885, a young New Yorker who decided on
a sudden impulse to become a doctor as he passed by
the College of Physicians and Surgeons while walking
along Fourth Avenue in 1868. Shortly after gradua-
tion, he was struck by tuberculosis.
Fifty years later, the number of sanatoria in the
USA had increased to 600 with a total bed capacity
of 95,000.
2.17
Artificial Lung Collapse Therapy
Bourru (1774) in Paris was the first to think of inducing
artificial lung collapse,aiming at the arrest ofthe disease,
but the concept was not given a clinical trial (Rosenb-
latt 1973). An attempted trial to induce pneumothorax
in two tuberculous patients by Carson had failed, but
despite these failures its concept was intriguing. In 1835,
McRuer from the USA and Houghton and Strokes from
the UK were successful and their patients improved.
Others had similar encouraging experiences, including
Cayley (1885) who used the therapy in a patient who
had had a pulmonary hemorrhage (Rosenblatt 1973).
The most significant work on artificial pneumothorax
therapy was reported by Carlo Forlanini (b. 1847)
in 1882. He used an apparatus to force nitrogen gas
through a tube and a needle inserted into the patient's
pleural cavity and was successful (Rosenblatt 1973).
Pneumothorax remained popular despite its therapeu-
tic limitations and complications, such as development
ofpleural effusion, emphysema and increased mortality,
and remained in use until the 1940s.
2.18
Thoracoplasty
Rib resection to collapse and close tuberculosis
cavities (thoracoplasty) was a popular operation in
the late 19th and early 20th centuries. The first tho-
Historical Aspects of Tuberculosis
racoplasty was performed by de Cerenville in 1885,
and in 1909, Braver introduced a two-stage technique
to reduce the surgical shock that was noted in the
one-stage thoracoplasty (Rosenblatt 1973). Phrenic
nerve crush was used to paralyze the diaphragm
and compress the affected lung. Pneumoperitoneum
with phrenectomy were popular in the 1930s and
early 1940s (Godfrey and Norman 1969).
2.19
Plombage
Plombe is a word introduced by Professor Custodis of
Dusseldorf, Germany to describe the implants he used
to produce infolding of the sclera in cases of retinal
detachment. Plombage (a French word meaning stop-
ping) is the procedure of packing the pleural cavity
with inert material. Extrapleural pneumothorax by
instillation of oil to maintain lung collapse was first
suggested by Mayer 1913 (quoted in Magner 1992),
but was not used until two decades later. Insertion of
plombe from fatty tissue, paraffin or Lucite balls into
the extrapleural space was called plombage. The pro-
cedure was used to provide permanent compression of
the underlying diseased lung, particularly with large
peripheral cavities.
2.20
Pulmonary Resection
In 1883, Block (quoted in Magner 1992) was the first to
perform pulmonary resection for tuberculosis in man.
The patient died and at autopsy there was no evidence
of tuberculosis. Block shot himself. In 1884 Kronlein
(quoted in Magner 1992) resected apical tuberculosis
in two patients and both died. Later, despite advances
in surgery in the pre-antibiotics era, postoperative
mortality was high.
2.21
The History of BCG
Leon Charles Albert Calmette (b.l863) and Camille
Guerin (b.l872) were French bacteriologists working in
the Pasteur Institute in Lille. Their work on the vaccine
began in 1908 on a strain of Mycobacterium tuberculo-
sis isolated by Nocard from a cow that had tuberculous
27
mastitis. An attenuated bovine bacillus was first isolated
in 1921 by Calmette and Guerin after years of laboratory
cultivations. The attenuated strain of tubercle bacillus
was incapable of producing tuberculosis in any labora-
tory animals (Calmette 1923). In 1930, a tragic incident
occurred whereby administration of a virulent tubercle
bacilli stored in the same refrigerator as the BCG led
to the death of 67 German babies in Lubeck. BCG was
initially given to many infants in France as an oral vac-
cine but the Scandinavians pioneered its administration
intradermally in 1950.
2.22
The History of Chemotherapy
for Tuberculosis
The antibiotics era began in 1928 when Alexander
Fleming (1881-1955), a Scottish bacteriologist and
immunologist in London, discovered the effect of
the mold Penicillium notatum on bacteria. His dis-
covery remained a laboratory curiosity until World
War II. "Fleming, recalled that neither bacteriologists
nor physicians paid any attention to penicillin until
the introduction of sulphonamide" (Quoted from
Magner 1992 p. 353).
A sulfur containing red dye called Prontosil,
which protected mice from staphylococcal and strep-
tococcal infections was discovered by Domagk in
1932. Gerhard Domagk (1895-1964) was a German
chemist and pathologist. In 1935, Domagk reported
"A contribution to the chemotherapy of bacterial
infections;' and with that report he initiated the age
of chemotherapy. Prontosil and other sulfonamides,
including promizole and diasone, were hailed as
"miracle drugs" in the 1930s. In 1939 Domagk was
awarded the Nobel Prize for Physiology and Medi-
cine "For the discovery of the antimicrobial effect of
Prontosil;' but Nazi officials would not allow him to
accept it. He finally received the Nobel Medal in 1947.
Sulfonamides were tried on patients with tuberculo-
sis but were found to be of little value (Hinshaw et al.
1944,1946; Hinshaw 1969).
2.23
Streptomycin and Modern
Antituberculosis Therapy
In 1943, Selman A. Waksman (1888-1973) and co-
workers isolated streptomycin from two strains of
28
an actinomycete species. Much of the initial work
was done by his graduate student, Albert Schatz.
Waksman was a Russian immigrant to the USA who,
in 1915, graduated from the Agricultural College of
Rutgers University in New Jersey (Ryan 1992). He
published a book titled, Principles of Soil Microbiol-
ogy.
2.24
Animal and Human Trials on the Efficacy
of Streptomycin
Experimental trials of streptomycin on animals
infected with tuberculosis and on tuberculous
patients demonstrated its efficacy (Feldman et al.
1947; Hinshaw et al. 1946; Pfuetze et al. 1955) and by
the late 1940s streptomycin's effectiveness was clearly
established. Schatz shared royalties with Waks-
man but did not share the Nobel Prize, which was
awarded to Waksman alone. The use of streptomycin
as a single agent for the treatment of tuberculosis was
difficult (Cooke et al. 1946).
2.25
Para-Aminosalicyclic Acid (PAS)
Dr. Jorgen Lehmann was born in Denmark and
graduated from a Swedish medical school. He went
to New York as a research fellow then returned back
to work as Chief of Chemical Pathology in Goteborg,
Sweden. He developed PAS in 1943, after synthesizing
analogues and various derivatives of salicylic acid in
Swedish laboratories. PAS was first used as a topi-
cal agent for the treatment of tuberculous fistulous
tract caused by cervical scrofula or by intrapleural
installation in patients with tuberculous empyema.
The first oral tablets of PAS were used in October
1944 and in 1946, Lehmann reported a preliminary
laboratory and clinical report on the drug (Lehm-
ann 1946, 1964). Feldman and colleagues (1947)
from Mayo clinic tried PAS on animals infected with
tuberculosis and found it to be effective (Feldman
et al. 1947). The British Medical Research Council
(MRC) study on using PAS alone, streptomycin (SM)
alone or PAS plus SM found SM alone was superior
to PAS alone and that their combination showed
clinical and radiological improvement with reduced
emergence of resistant organisms (British Medical
Research Council 1950).
M. M. Madkour et al.
2.26
Isoniazid
Isoniazid (isonicotinic acid hydrazide, INH) was first
synthesized by Meyer and Malley in Prague (Meyer
and Malley 1912). However, it was not until 1942
that its antituberculous effect was noted by Chorine
(1942). In 1951-1952, Hoffman-La Roche, Squibb in
the USA, and Bayer in Germany independently and
simultaneously developed INH. The first clinical
trial of INH alone for the treatment of tuberculous
patients was in Seaview Sanatorium, New York in
1951 by Robitzek and Selikoff (Robitzek and Selikoff
1952; McDermott 1969).
2.27
Rifampicin (Rifampin)
The initial microbiologically active components of the
rifamycin group were obtained from organisms of the
species Amycolatopsis mediterranei found in the soil of
a French pine forest in 1957 by Lepetit Research Labo-
ratory in Milan. The microbiologically active compo-
nents were nicknamed "Rififi" after a popular French
movie. Rifamycin B had no intrinsic microbiological
activity, but the spontaneous breakdown product,
rifamycin SV, was active but was not absorbed by the
gastrointestinal tract. Collaboration between Lepetit
(Milan) and Ciba-Geigy (Basel), laboratories synthe-
sized a product that was better absorbed orally, and
more active for a prolonged period (Maggi et al. 1965;
Hobby and Lenert 1968; Sensi 1983).
2.28
Drug Therapy for Cure
The principle of multi-drug therapy of tuberculosis
was found in the 1960s to be the cornerstone of any
effective treatment regimen, as it also prevented
the development of drug resistance (Canetti 1962;
Mitchison 1965).
Canetti (1962) proposed the principle of two-
phase chemotherapy: in the first phase, with a high
cavitary load of bacilli, intensive combined chemo-
therapy is required, and in the second phase, with
a small load of bacilli, continued and prolonged
therapy with isoniazid alone is sufficient. Mitchison
(1965) recommended initial therapy with a three-
drug combination.
Historical Aspects of Tuberculosis
The British Medical Research Council trial recom-
mended three drugs for the first 6 weeks and two
drugs for a total of 1 year (British Medical Research
Council 1962). Long-term therapy, though effective,
was expensive and compliance of patients was dif-
ficult. Short-course therapy on large-scale trials,
first introduced in East Africa by the British Medical
Research Council using several 6-month regimens
(East African/British Research Council 1973, 1974),
clearly established their effectiveness.
2.29
Conclusion
Tuberculosis has been known and recorded since
antiquity by the Ancient Egyptians and by Greek, Arab,
European and other scholars. Despite the development
of knowledge on the pathology and clinical features, the
cause remained unknown till the discovery of the bacil-
lus late in the 19th century. The discovery of antibiotics
lagged behind the identification of the causative organ-
isms. In the 1930s and 1940s, the use of antibiotics in the
treatment oftuberculosis marked a new and distinct era
in the history of tuberculosis. Despite the application
of antibiotics and subsequent developments of other
new anti-tuberculous drugs over the following years,
tuberculosis remained a common disease in developing
countries. The discovery of HIV in 1981 and its effect on
the resurgence of tuberculosis both in developing and
developed countries together with the development of
drug resistance, remains a challenge.
References
Al-Amoud M (1993) The history of tuberculosis. Saudi Med
J 14:515-520
Bishop PJ (1973) Thomas Young and his 'A practical and his-
torical treatise on consumptive diseases', 1815. Tubercle
54:159-164
Bishop PJ (1981) Laennec: a great student of tuberculosis.
Tubercle 62: 129-134, 147-148
Bishop PJ, Neumann G (1970) The history of the Ziehl-Neelsen
stain. Tubercle 51:196-206
British Medical Research Council (1950) Treatment of pulmo-
nary tuberculosis with streptomycin and para-aminosali-
cyclic acid: a Medical Research Council investigation. BMJ
2:1073-1085
British Medical Research Council (1962) Long-term chemo-
therapy in the treatment of chronic pulmonary tuberculo-
sis with cavitation: a report to the Medical Research Coun-
cil by their Tuberculosis Chemotherapy Trials Committee.
Tubercle 43:201-219
29
Brock TD (1999) Robert Koch - a life in medicine and bacte-
riology. ASM Press, Washington DC
Buikstra JE (1983) Prehistoric tuberculosis in the Americas
(Evanston, Ill.: Northwestern University Archeological
Program, 1981) (reviewed by William D Sharpe). Trans
Stud Coll Phys Phil 5(5):278-281
Calmette A (1923) Tubercle Bacillus Infection and tuberculo-
sis in Man and Animals (trans WB Soper and GH Smith).
Williams and Wilkins, Baltimore MD (quoted by Evans CC,
1998, Clinical tuberculosis, 2nd edn. Davies, Chapman and
Hall Medical)
Canetti G (1962) The eradication of tuberculosis: theoretical
problems and practical solutions. Tubercle 43:301-321
Chadwick J, Mann WN (1978) Hippocratic writings. Penguin,
NewYork,pp 117-118
Chorine V (1942) Action de l'amide nicotinique sur les bac-
illes du genre Mycobacterium. CR Acad Sci (Paris) 220:
150-151
Cooke RE, Dunphy DL, Blake FG (1946) Streptomycin in tuber-
culous meningitis. Yale J Bio Med 18:221-226
Crawfurd R (1911) The king's evil. Oxford
Cruse JM (1999) History of medicine: the metamorphosis of
scientific medicine in the ever-present past. Am J Med Sci
318:171-180
East African/British Medical Research Council (1973)
Controlled clinical trial of four short-course (6-month)
regimens of chemotherapy for treatment of pulmonary
tuberculosis. Second report. Lancet 1:1331-1338
East-African/British Medical Research Council (1974)
Controlled clinical trial of four short-course (6-month)
regimens of chemotherapy for treatment of pulmonary
tuberculosis. Third report. Lancet 2:237-240
Feldman WH, Karlson AG, Hinshaw HC (1947) Para-amino-
salicyclic acid in experimental tuberculosis in guinea pigs.
Proc Sraff Meet Mayo Clin 22:473-479
Gallagher NE (1993) Islamic and Indian Medicine. In: Kiple
KF (ed) The Cambridge world history of human disease.
Cambridge University Press, Cambridge, pp 27-34
Godfrey LG, Norman CD (1969) Surgical history of pulmonary
tuberculosis: the rise and fall of various technical proce-
dures. Can J Surg 12:381-388
Hinshaw HC (1969) Tuberculosis chemotherapy: reminiscences
of early clinical trials. Scand J Respir Dis so: 197- 203
Hinshaw HC, Pfuetzek and Feldman WH (1944) Chemother-
apy of clinical tuberculosis with promin (Promizole). Am
Rev Tuber 50:52-57
Hinshaw HC, Feldman WH, Pfuetze KH (1946) Treatment of
tuberculosis with streptomycin. A summary of observa-
tions on one hundred cases. JAMA 132:778-782
Hobby GL, Lenert TF (1968) The antimycobacterial activity of
Rifampin. Am Rev Respir Dis 97:713
Johnston WD (1993) Tuberculosis. In: Kiple KF (ed) The
Cambridge world history of human disease. Cambridge
University Press, Cambridge, pp 1059-1068
Keers RY (1981) Laennec: his medical history. Thorax 36:91-94
Lehmann J (1946) Para-aminosalicyclic acid in the treatment
of tuberculosis. Lancet 1: 15-16
Lehmann J (1964) Twenty years afterward: historical notes on
the discovery of the antituberculous effect of para-amino-
salicyclic acid (PAS) and the first clinical trials. Am Rev
Respir Dis 90:953-956
Lyons A, Petrucelli R (1978) Medicine, an illustrated history.
Abrams, New York
30
Maggi N, Pallanza R, Sensi P (1965) New derivatived of Rifa-
mycin Sv. Antimic. Agents Chemother 5:765
Magner LN (ed) (1992) A history of medicine. Dekker, New
York
Magner LN (1992a) Islamic medicine. In: Magner LN (ed) A
history of medicine. Dekker, New York, pp 133-152
Magner LN (1992b) The scientific revolution and the circula-
tion of the blood. In: Magner LN (ed) A history of medi-
cine. Dekker, New York, pp 189-216
McDermott W (1969) The story of INH. J Infect Dis 119:
678-683
Meyer H, Malley J (1912) Hydrazine derivatives of pyridin-
ecarboxyclic acids. Monatsh 33:393-414
Mitchison DA (1965) Chemotherapy of tuberculosis: a bacte-
riologist's viewpoint. BMJ 1:1333-1340
Murray JF (1989) The white plague: down and out, or up and
coming? Am Rev Respir Dis 140:1788-1795
Myers JA (1977) Captain of all these men of death. Tuberculo-
sis historical highlights. Green, St Louis
Paulsen HJ (1987) Tuberculosis in the native American: indig-
enous or introduced? Rev Infect Dis 9:1180-1186
Pfuetze KH, Pyle MM, Hinshaw HC et al (1955) The first clini-
cal trial of streptomycin in human tuberculosis. Am Rev
TubercoI71:752-754
Powell ML (1992) Health and disease in the late prehistoric
southeast. In: Verano JW, Ubelaker DH (eds) Disease and
demography in the Americas. Smithsonian Institution Press,
Washington DC, pp 41-53
M. M. Madkour et al.
Risse GB (1993) History of western medicine from Hip-
pocrates to Germ theory. In: Kiple KF (ed) The Cambridge
world history of human disease. Cambridge University
Press, Cambridge, pp 11-27
Robitzek EH, Selikoff IJ (1952) Hydrazine derivatives of
isonicotinic acid (Rimifon, Marsilid) in the treatment of
active progressive caseous-penumonic tuberculosis. A pre-
liminary report. Am Rev Tubercol 65:402-428
Rogers FB (1969) The rise and decline of the altitude therapy
of tuberculosis. Bull Hist Med 43:1-16
Rosenblatt MB (1973) Pulmonary tuberculosis: evolution of
modern therapy. Bull NY Acad Med 49: 163-196
Ryan F (1992) The Forgotten Plague: how the battle against
tuberculosis was won and lost. Little Brown, Boston
Sakula A (1981) RTH Laennec 1781-1826 his life and work: a
bicentenary appreciation. Thorax 36:81-90
Sensi PG, Greco AM, Ballotta R (1959/1960) Rifomycin I.
isolation and properties of Rifomycin Band Rifomycin
complex. Antibiot Annu 262
Verano JW, Ubelaker DH (eds) (1992) Disease and demog-
raphy in the Americas. Smithsonian Institution Press,
Washington DC
Weisse AB (1995) Tuberculosis: why "The white plague"? Per-
spect Bioi Med 39:132-138
Wilson JL (1979) History of the American Thoracic Society.
Am Rev Respir Dis 119:177-184
Wright KW (1988) A history of the Onondaga Sanatorium for
the treatment of tuberculosis. NY State J Med 137-145
Epidemiology
3 Global Epidemiology of Tuberculosis
MARCOS A. ESPINAL and MARIO C. RAVIGLIONE

CONTENTS every year and almost two million deaths annually

(Corbett et al. 2003). Despite the existence of a proven
3.1 Introduction 33 comprehensive cost-effective strategy-DOTS-aimed
3.2 Exposure and Infection with M. tuberculosis 33
3.2.1 Exposure 33
to reduce mortality, morbidity, and transmission of
3.2.2 From Exposure to Infection 34 the disease, only 32% of the estimated new smear-
3.2.3 Magnitude of Infection Worldwide 34 positive tuberculosis cases worldwide were managed
3.3 Active Tuberculosis 34 under this strategy in 2001. The emergence of the
3.3.1 Risk of Disease 34 human immunodeficiency virus (HIV)/acquired
3.3.2 Morbidity 35
3.3.3 Mortality 36
immunodeficiency syndrome (AIDS) and of multi-
3.4 The Impact of HIV/AIDS on the Epidemiology drug-resistant tuberculosis (MDR-TB-resistance to
of Tuberculosis 37 at least isoniazid and rifampicin) have posed addi-
3.5 The Impact of Multidrug-Resistant tional challenges to tuberculosis control. In order to
Tuberculosis 38 accelerate control efforts and overcome these threats,
3.6 Future Trends of Tuberculosis 38
3.7 Concluding Remarks 40 partnership efforts by the international community
References 41 and a revised framework for tuberculosis control
have been launched at the beginning of this new
century. This chapter will review the epidemiology
of infection and disease with M. tuberculosis glob-
ally, present morbidity and mortality data, discuss
the impact of HIV and MDR-TB, and examine future
3.1 trends.
Introduction

Tuberculosis has been present in the lives of human

beings for many centuries. Evidence of bone lesions
suggestive of tuberculosis in mummies of North
America, Peru, and Egypt confirms the ancient impact
of this disease on early civilizations (Rothschild et al.
2001; Nerlich et al. 2000; Salo et al. 1994). More than
50 years after the introduction of anti-tuberculosis
chemotherapy, one-third of the world population
is estimated to be infected with Mycobacterium
tuberculosis and active disease remains a worldwide
pandemic, with more than eight million new cases
M. A. ESPINAL
World Health Organization, Stop TB Department, Tuber-
culosis Strategy & Operations, Ave Appia 22, 1211 Geneva,
Switzerland
M. C. RAVIGLIONE
World Health Organization, Stop TB Department, Ave Appia 22,
1211 Geneva, Switzerland
3.2
Exposure and Infection with M. tuberculosis
3.2.1
Exposure
Tuberculosis is caused by bacilli belonging to the M.
tuberculosis complex, especially the human variety.
Tuberculosis caused by milk contamination with M.
bovis is not an issue any longer. Likewise, tubercu-
losis caused by M. africanum is rarely reported. The
natural history of tuberculosis can be described in
four different stages: exposure, infection, disease, and
death. Exposure to M. tuberculosis is influenced by
the number of infectious cases and the duration of
their infectiousness. The intensity of exposure will
depend on various factors, including population den-
sity, family size, climatic changes, and age of sources
of infection (Rieder 1999).

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
34
3.2.2
From Exposure to Infection
M. tuberculosis infection is usually airborne. The prob-
ability of infection depends on the number of infectious
droplet nuclei that remain suspended in the air and the
duration of exposure of a susceptible individual. Drop-
lets ofless than 5 f.lIll diameter transmit M. tuberculosis
(Sonkin 1951). Coughing is the main source of air drop-
lets, but they can also be generated by sneezing and talk-
ing (Wells 1955; Riley and O'Grady 1961). The number
of bacilli present in the source case is estimated to be
between 102 and 104 in nodular lesions and between
8
107 and 10 in cavitary lesions (Canetti 1965). Smear-
positive tuberculosis cases constitute by far the most
important source of infection (Styblo 1991).
The presence of infection is usually measured
with the tuberculin intradermal skin test through the
assessment of the size of induration generated by the
test. Incidence of infection is usually estimated through
the risk of infection (also called annual risk of tuber-
culosis infection; or ART!) from tuberculin prevalence
surveys. The risk of infection also suggests the extent of
transmission of M. tuberculosis in the past. The ART!
has been defined as the probability that any individual
will be infected with M. tuberculosis in 1 year. It has
been the basis of several exercises to estimate the
global magnitude of tuberculosis (Sudre et al. 1992;
Dolin et al. 1994; Murray and Lopez 1997). Styblo esti-
mated that in the absence of tuberculosis control, an
annual risk of infection of 1% corresponded roughly to
an incidence of new smear-positive tuberculosis cases
of 50 per 100,000 population (Styblo 1985).
The disadvantages of tuberculin prevalence surveys
include high costs and technical problems inherent to
the application and reading of the tuberculin test. Fur-
thermore, HIV infection has made the interpretation
of tuberculin prevalence surveys very difficult and
the previously established fixed relationship between
ART! and incidence of smear-positive cases is no
longer valid. Thus, in many settings it is not possible to
obtain accurate and reliable data of the levels of infec-
tion and estimates need to be constructed.
3.2.3
Magnitude of Infection Worldwide
Prevalence of infection with M. tuberculosis varies
according to demographic (age, sex, place) and socio-
economic (crowding, poverty, confinement) deter-
minants and the degree of contact and intensity of
exposure. Empirical data available and mathematical
M. A. Espinal and M. C. Raviglione
models suggest various prevalence patterns in differ-
ent regions of the world. In high-income countries the
population most infected is that above 65 years of age.
In Europe with the decline of the risk of infection by
an estimated 10-12% annually, the young generations
are basically free of tuberculosis infection (Styblo et
al. 1969; Sutherland et al. 1971; Styblo 1991). In the
Netherlands, data from military recruits suggest that
between 1910 and 1939, the annual risk of infection
was already declining by 5.4% annually, largely due to
isolation of cases and better socio-economic condi-
tions. Such a declining trend increased further up to
12.9% annually between 1940 and 1969, concomitant
to the introduction of milk pasteurization and anti-
tuberculosis drugs (Styblo et al. 1969). Likewise, in the
United States the risk of infection has been declining
for several decades. Data from Navy recruits show that
the proportion of positive reactions to the tuberculin
test fell from 6.6% in 1949-1951 to 3.1% in 1967-1968
and to 1.5% in 1980-1986 (Comstock 2000).
The scenario is completely different in resource-
limited countries, where young generations and eco-
nomically productive populations are still currently
subject to a high risk of infection with M. tuberculo-
sis. This is due to a slower decline of the annual risk of
infection that was estimated at 1-6% per year in the
1980s (Cauthen et aI.1988).
Estimates released by the World Health Organization
(WHO) in 1991 suggested that one-third of the world
population, or 1.7 billion people, were infected with M.
tuberculosis (Kochi 1991). Revised estimates suggested
that in 19971.86 billion people [32% of the world popu-
lation] were infected with the tuberculous bacillus (Dye
et al.1999). These numbers are not much different from
newly revised estimates, which suggest that in 2000 1.8
billion people were infected with M. tuberculosis, again
one third of the world population (Corbett et al. 2003).
South-East Asia and Africa, with an estimated 46% and
31% of their populations infected with M. tuberculosis,
were the most affected regions of the world.
3.3
Active Tuberculosis
3.3.1
Risk of Disease
Tuberculosis develops as a result of progression of pri-
mary infection, exogenous reinfection, or endogenous
reactivation. The highest risk of disease is posed imme-
diately (first 24 months) after infection is acquired. With
Global Epidemiology of Tuberculosis
time, the risk declines. In the United States, 82,269 chil-
dren between 1to 18 years of age who reacted positively
to the tuberculin skin test were followed for 20 years
(Comstock et al. 1974). Two incidence peaks of active
disease where identified: at ages 1-4 years, and during
late adolescence and early adulthood. Another cohort of
children followed for 20 years in Great Britain showed
that almost half of the tuberculosis cases developed
in the first 2 years of age and 68% in the first 5 years
after infection (D'Arcy Hart and Sutherland 1977).
The accepted principle is that the risk of disease is
greatest within the first 2 years following infection and
after about 7 years starts to level off, remaining low
and unchanged over the following decade and, usu-
ally, during lifetime. The lifetime risk of progression
from infection to disease is estimated to be about 10%
if the infection has been acquired during childhood
(Comstock et al. 1974). Half of the risk is during the
12-24 months post-infection.
Several factors have been identified that are respon-
sible for progression to active disease. These include
HIV infection, spontaneously healed fibrotic lesions,
malnutrition, age, sex, use of immunosuppression
agents, medical conditions including silicosis, a vari-
ety of neoplastic disorders, and diabetes. The contri-
bution of many of these factors to the pandemic of
tuberculosis is probably not very extensive, because
their prevalence is low or they are weakly associated
with an increase risk of disease. HIV infection is the
most potent risk factor for tuberculosis yet identified.
The risk of tuberculosis in people co-infected with M.
tuberculosis and HIV has been shown in several stud-
ies to range from 5% to 15% per year (Hopewell and
Chaisson 2000). Differences in the quantification of the
risk of tuberculosis among HIV-infected individuals
are especially related to the degree of immunosuppres-
sion as measured by the CD4+ lymphocyte counts.
Studies conducted in the last 10 years using DNA
fingerprinting techniques have suggested that half of
tuberculosis cases are attributable to recent or ongo-
ing transmission in some settings (Wilkinson et al.
1997; Godfrey-Faussett et al. 2000). Even in low-prev-
alence settings, where previously most disease was
though to be due to reactivation, one third of cases
have been attributed to ongoing transmission (Small
et al.1994; Alland et al. 1994; Borgdorff et al. 1997).
3.3.2
Morbidity
Tuberculosis started to decline in Europe by the
19th century, a trend that mirrored the increase in
35
industrialization, better sanitation, and improvement
of living conditions (Evans 1998). This trend was
accelerated after the introduction of chemotherapy
by mid 20th century. As of today, tuberculosis in
Western Europe, Australia, Canada, Japan, and the
United States is only a problem of selected groups
as the general population of these countries is
minimally affected. Only 5% of the world burden of
tuberculosis is estimated to be in the industrialized
world (Corbett et al. 2003). The highest toll of tuber-
culosis in industrialized countries is predominantly
among immigrants from high-prevalence settings,
the homeless, intravenous drug addicts, and HIV-
infected people.
Since 1980 yearly case notifications to WHO have
allowed monitoring and modeling of the future
impact of the tuberculosis pandemic (Raviglione et
al. 1997a; World Health Organization 2003). Overall,
tuberculosis notifications to WHO have remained at
around 60 per 100000 people since 1980, with little
fluctuations (Fig. 3.1). In industrialized countries
data show that tuberculosis continues its declin-
ing trend, after a few years of stagnation in the late
1980s and early 1990s. Tuberculosis notifications in
Asia and Latin America have remained fairly stable
throughout the years, without significant changes.
On the other hand, tuberculosis notifications in
Africa and Eastern Europe have shown very differ-
ent epidemiological trends. In sub-Saharan Africa,
notifications have been increasing on average 10%
per year since 1980, largely due to the HIV/AIDS
epidemic. In the former Union of Socialists Soviet
Republics (USSR) tuberculosis was declining at 5%
per year up to 1990; thereafter, it started to increase at
8% per year due to deteriorating socioeconomic con-
ditions leading to poverty, congestion, malnutrition,
and to the collapse of the public health system and of
Case notiticafion rate (pe, 100,000)
80
60
40
20
o
1980 1985 1990 1995 2000
Years
Fig. 3.1. Global trend in the tuberculosis case notification rate.
Source: Global Tuberculosis Control. World Health Organiza-
tion Report 2003. WHO/CDS/TB2003316
 36 M. A. Espinal and M. C. Raviglione

tuberculosis control efforts. In the recent years rates
of increase have started to slow down.
The most recent data show that, in the year 2001,
3,813,109 (62 per 100,000 people) cases were noti-
fied worldwide, of which 59% were notified in Asia
and 21 % in Africa (Fig. 3.2). Of the two million cases
notified in Asia, India, China, Bangladesh, the Philip-
pines, and Indonesia accounted for 87%. In Africa, of
the 811,172 cases notified, Nigeria, Ethiopia, South
Africa, the Democratic Republic of Congo, and Kenya
accounted for 53%. Of the 22 high-burden countries
of the world responsible for 80% of the global case-
load, 18 are located in Asia and Africa. The global
notification of smear-positive cases in 2001 was
1,602,153 (26 per 100,000 people), of which 59% were
notified in Asia and 23% in Africa.
Notification data have been pivotal to track the
tuberculosis pandemic; however, they do not reflect the
true picture of the global situation. Lack of access to
health services, poor capacity to suspect and diagnose
cases, and under-reporting are some of the issues that
limit the significance of notification data. Therefore,
estimates are needed in order to have a clearer picture
of the burden of disease. Several figures have been pub-
lished in the past 25 years, ranging from 4 million new
tuberculosis cases in 1977 to 8.3 million cases in 2000
(Bulla 1977; Sudre et al.1992; Murray et al. 1993; Dolin
et al. 1994; Raviglione et al. 1995; Murray and Lopez
1997; Dye et al. 1999; Corbett et al. 2003). These esti-
mates have been useful in portraying the global situ-
ation of tuberculosis; however, the earlier studies only
provided data for regions of the world and not for indi-
vidual countries. Using several sources of data, expert
opinion and mathematical methods, WHO estimated
the burden of disease for 212 countries individually for
1997 (Dye et al.1999). There were 16 million prevalent
tuberculosis cases worldwide and 8 million (range: 6.3-
11.1 million) new cases for a case rate of 136 per 100,000
people. Of these, 3.5 million were smear-positive for a
case rate of 60 per 100,000 people. Eighty percent of all
new cases were in 22 countries. Newly revised estimates
for 2000 report 8.3 million (range: 7.3-9.2 million) new
tuberculosis cases (Fig. 3.2) for a case rate of 137 per
100,000 people, of which 3.7 million were smear-posi-
tive for a case rate of 61 per 100,000 people (Corbett et
al. 2003). Incidence in adults 15-49 years was estimated
to be 5.4 million for a case rate of 172 per 100,000
people. The greatest incidence rate was in sub-Saharan
Africa (290 per 100,000 people). However, the greatest
number of cases (more than five million) is estimated
to be in Asia including 3.2 million cases in India and
China. Between 1997 and 2000, the increase of new
tuberculosis cases and incidence rates were at 1.8%
and 0.4% per year respectively. The greatest rise was
observed in Eastern Europe and sub-Saharan Africa
due to reasons explained earlier.
3.3.3
Mortality
Today, HIV/AIDS is the most important condition influ-
encing tuberculosis mortality. Other factors include
lack of access to chemotherapy, missed diagnosis, late
diagnosis, treatment default, and multidrug-resistant
tuberculosis.

Cc:nual and Eastc:tn

Europe
410000 c.... cslimalCd
(87 per 100 000 pop)
326 739 cases notified
(70 per 100 000 pop)

LatIn America

.....
360 000 ...... CIIiD*Od
(70 per 100000 pop)
' 215842 _ _lIIed
(42 per 100 000 pop)

I 900 000 cues Cllilllllled Fig. 3.2. Estimated and

(290 per 100000 pop) notified tuberculosis
811 In caseSDOlified cases in the world,
(115 per 100 000 pop)
2000
Global Epidemiology of Tuberculosis
Tuberculosis mortality has decreased to a mini-
mum in the industrialized world. Mortality in Euro-
pean countries was highest between the late 18th
century and early 19th century, reaching 50-60%
of all tuberculosis cases, after 5 years of diagnosis
(Drolet 1938; Thompson 1943). Thereafter, mortality
declined with the introduction of chemotherapy and
currently is not a major issue.
A different pattern, however, is observed in
resource-limited countries, regardless of the avail-
ability of effective chemotherapy for more than
50 years. In the 1980s and 1990s, various studies esti-
mated that almost three million deaths were due to
tuberculosis globally (Styblo 1983; Dolin et al. 1994).
The general impression today was that these num-
bers were overestimating the magnitude of mortality
from tuberculosis because of the very high case-fatal-
ity ratio, which in the past was at around 30% (Enar-
son and Rouillon 1998). In the absence of treatment
the case-fatality ratio is estimated to be around 50%.
However, such a high estimate may not be completely
accurate, as already many countries were providing
chemotherapy. In 1997 Murray & Lopez new esti-
mates reported two million deaths from tuberculo-
sis worldwide (Murray and Lopez 1997). However,
these numbers excluded deaths due to HIV. Recent
estimates suggest that mortality from tuberculosis
may be lower than previously estimated (Corbett et
al. 2003) and that 1.82 million tuberculosis deaths
(30 per 100,000 people) occurred in 2000. Ninety-
eight percent of such deaths were in resource-limited
countries and 226,000 were attributable to HIV.
3.4
The Impact of HIV/AIDS
on the Epidemiology of Tuberculosis
In a very short period of time, spanning over 20 years,
the HIV/AIDS pandemic has rampaged all regions
of the world and today more than 60 million people
may have been infected with HIY. The impact of
HIV/AIDS on the tuberculosis pandemic has been
devastating. The Joint United Nations Program
on HIV/AIDS (UNAIDS) estimates that 42 million
individuals are currently living infected with HIV
(UNAIDS 2002). An estimated five million became
infected in 2002 and 800,000 of them were children.
Three-quarters of the world population infected with
HIV live in sub-Saharan Africa and a further 17% in
Asia. HIV/AIDS has claimed more than 22 million
deaths in the past two decades.
37
Worldwide tuberculosis is the most common
HIV-associated disease of public health importance
occurring in AIDS patients. As mentioned already,
HIV is the strongest risk factor for reactivation of
tuberculosis infection. However, the impact of HIV
infection on the tuberculosis pandemic is not only
limited to reactivation of tuberculosis infection.
Consistent evidence shows that people infected with
HIY, who later become infected or reinfected with M.
tuberculosis, progress faster to full-blown tuberculo-
sis than people who are infected with M. tuberculosis
but not with HIV (Di Perri et al. 1989; Daley et al.
1992; Girardi et al. 2000a; Sonnenberg et al. 2001).
Fortunately, HIV-infected tuberculosis patients do
not appear to be potent transmitters of M. tuberculo-
sis. In fact, household contact studies have consistently
shown that HIV-infected people with tuberculosis are
less likely to transmit the tubercle bacilli to their close
counterparts than HIV-uninfected people with tuber-
culosis (Klausner et al.1993; Elliot et al.1993; Cauthen
et al. 1996; Espinal et al. 2000a). Therefore, the ART!
may not be affected in spite of the increasing tubercu-
losis incidence in high HIV-prevalence settings.
Eleven million adults aged 15-49 years (0.36%
prevalence) were estimated to be co-infected with
HIV and M. tuberculosis in 2000 (Fig. 3.3), of which
9.5 million were in sub-Saharan Africa and 2.3 mil-
lion in South-East Asia. The overall global prevalence
of HIV among new adult tuberculosis cases has been
estimated to be 11% (Corbett et al. 2003). However,
in sub-Saharan Africa, this figure is 38%, the highest
worldwide, compared with only 5.9% in the Americas
and 3.2% in South-East Asia. Furthermore, in some
African countries such as Cote d'Ivoire, Malawi,
South Africa, Uganda, and Zambia prevalence in
new tuberculosis cases can be as high as 40-78%
(Raviglione et al. 1997b). Also, several sub-Saharan
African countries have reported continued increases
in tuberculosis notifications (Fig. 3.4), regardless of
Fig. 3.3. Estimated distribution of adults infected with HIV
and tuberculosis, 2000. Source: World Health Organization
38
No~flca~on Rates (xl00.000)
500
450
400
350
300
250
••
200 •• •, / UR T nzanle
150
0--------
1980 1985 1990 1995 2000
Years
Fig. 3.4. Tuberculosis trends in selected African countries
1980-2001. Source: World Health Organization
the existence of successful national tuberculosis con-
trol programs (Narain et al. 1992). The proportion of
new tuberculosis cases attributable to HIV/AIDS in
sub-Saharan Africa in 2000 was 31 % or 421,000 cases.
This proportion is very well above the global estimate
of 9% (511,000 cases). This clearly shows the negative
impact that HIV/AIDS is posing over tuberculosis in
this region of the world.
3.5
The Impact of Multidrug-Resistant
Tuberculosis
Drug resistance to at least isoniazid and rifampicin, a
combination known as MDR-TB, is, after HIV/AIDS, the
most important threat to tuberculosis control. Surveil-
lance efforts launched in the 1990s by WHO and the
International Union Against Tuberculosis and Lung
Disease (IUATLD) have shown that the magnitude of
the tuberculosis problem is grossly of drug-susceptible
cases (World Health Organization 1997,2000; Pablos-
Mendez et al. 1998; Espinal et al. 2001). Data from 67
countries/sites show a global prevalence of MDR-TB
of 1% in new cases of tuberculosis and 9.1% in previ-
ously treated cases. However, the same data also show
that MDR-TB is a severe problem in several countries
of Eastern Europe including Estonia, Latvia, and Russia.
Areas of India and China also show concerning MDR-
TB prevalence. The problem appears to be less severe
in Africa and Latin America. Several of these countries
did not follow internationally recommended guidelines
for tuberculosis control-DOTS-until just recently. For
instance, in several former Soviet Union countries,
interventions that are customarily used for tuberculo-
M. A. Espinal and M. C. Raviglione
sis control include long hospitalization of tuberculosis
cases, individualized treatment regimen, and unsu-
ZImbabwe pervised administration of therapy (Perelman 2000).
Hospitalization of tuberculosis cases in the absence of
infection control measures can be a major vehicle for
the rapid dissemination of MDR-TB. Lack of standard-
Kenye ization of treatment and use of inadequate regimens
! Melewl
without proper supervision are clearly conducive to
creation and spread of drug-resistant bacilli (World
Health Organization 1997) On the other hand, countries
implementing sound tuberculosis control measures for
many years such as Benin, Botswana, Chile, Cuba, Kenya,
and Uruguay have reported a very low prevalence of
MDR-TB (World Health Organization 2000).
Since the available data are limited and more
than half of the world countries have not yet been
surveyed, a mathematical model has estimated the
magnitude of MDR-TB globally (Dye et al. 2001).
This model suggests that in 2000 3% (273,000, 95%
confidence intervals: 185,000 and 414,000) of all new
estimated tuberculosis cases were MDR-TB. Eight of
the twenty-two high-burden countries of the world,
from which data are not available yet, were estimated
to have more than 4000 cases of MDR-TB each. These
countries need to be surveyed urgently in order to
have a clear picture of the magnitude of the problem.
If current estimates prove accurate, the impact of
MDR-TB on tuberculosis control in these countries
may be overwhelming, as the management of MDR-
TB is more costly, complex, and lengthy than that of
drug-susceptible tuberculosis.
MDR-TB is a man-made problem and most coun-
tries with a high prevalence have a history of poor
tuberculosis control. The spread and evolution of
MDR-TB will depend on the efforts these countries
are willing to undertake to accelerate tuberculosis
control according to recognized and tested interna-
tional guidelines. The only effective way of preventing
MDR-TB from becoming a problem of the magnitude
of drug-susceptible tuberculosis is by ensuring the
cure of all tuberculosis cases susceptible to first-line
drugs under proper management conditions (Dye et
al. 2002a). In addition, careful management of MDR-
TB with second-line drugs will be also needed in set-
tings where the problem is at epidemic levels.
3.6
Future Trends of Tuberculosis
Tuberculosis is expected to continue increasing in
the first decade of the XXI century. About ten million
Global Epidemiology of Tuberculosis 39

new tuberculosis cases are expected in 2010, if current tional130,OOO extra new smear-positive tuberculosis
trends continue and control efforts are not accelerated cases on average every year have been recruited
(Fig.3.5) (Dye 2000). Several factors are expected to under DOTS since 1996. In order to achieve the
influence such trends. These include lack of political global targets by 2005, an annual average of 360,000
will to effectively expand tuberculosis control, the extra new smear-positive tuberculosis cases needs to
HIV/AIDS pandemic, drug resistance, lack of involve- be detected and treated under DOTS. Detection and
ment of the private sector in tuberculosis control, provision of short-course chemotherapy to these
increasing poverty, demographic changes, the current cases may halve tuberculosis incidence in low HIV-
burden of M. tuberculosis infection worldwide, and prevalence areas in about 10 years at a rate of decline
lack of better tools to speed-up prevention, diagnosis, of 5-10% annually (Dye et al.1998).At the moment,
and cure of tuberculosis cases. this is not happening; thus, tuberculosis will continue
Strong focus on the expansion of effective tuber- to spread uncontrolled in many settings.
culosis control is the key to reverse the current trends HIV/AIDS is another factor that will continue
of tuberculosis worldwide. In 1991 WHO adopted impacting the tuberculosis pandemic for several
the DOTS strategy for tuberculosis control (World years to come, unless effective measures beyond
Health Organization 1991), which was later labeled DOTS and specific to HIV-associated tuberculo-
as one of the most cost-effective health interventions sis are implemented to reverse the current trends.
available (World Bank 1993). The effectiveness of WHO has estimated that there will be 3.5 million
DOTS has been successfully tested in the field with new tuberculosis cases in sub-Saharan Africa in
impressive results (Suarez et al. 2001; Dye et al. 2000). 2005 (World Health Organization 2001). It is now
As of 2001, 155 countries have implemented DOTS widely accepted that in spite of the existence of good
, which is increased from only 10 countries in 1991 tuberculosis control programs in this continent,
(World Health Organization 2003). However, in many DOTS alone cannot reverse the increasing toll of
of these countries population coverage by DOTS is HIV-associated tuberculosis (de Cock and Chaisson
low. Today, only half of the world's population lives 1999; Elzinga and Nunn 2002). Therefore, intensive
in countries implementing DOTS. Furthermore, only collaboration between tuberculosis and AIDS con-
32% of the estimated new smear-positive tuberculo- trol programs, exploring novel options designed to
sis cases in the world were detected and managed enhance control of the two diseases should be imple-
under DOTS conditions in 2001, which means that mented. Useful strategies to test in field conditions
68% of the world infectious cases were managed with include active tuberculosis case-finding and making
unclear or inappropriate strategies. voluntary counseling and HIV testing widely avail-
Mathematical modeling suggests that if tubercu- able not only to tuberculosis patients but also to the
losis control is not expanded effectively and quickly, general population in order to facilitate entry into
the global targets of 70% case-detection and 85% preventive and treatment strategies. In particular,
cure rate among new tuberculosis cases will not be administration of preventive therapy with isoniazid
achieved until 2013 (Dye et al. 2002b). Only an addi- and highly active antiretroviral therapy (HAART) to
people with HIV/AIDS needs strong and urgent con-
New TB cases ('OOOs)lyear
sideration. The current evidence shows that therapy

--
with isoniazid prevents reactivation of tuberculosis
12000
10000
r among HIV-infected people who are tuberculin skin
test positive (WHO/UNAIDS 1999). There is also
increasing evidence that HAART can reduce the
8000 _ _ -
risk of tuberculosis by more than 80% by restoring
immunocompetence (Girardi et al. 2000b; Jones et al.
6000
2000; Badri et al. 2002; Santoro-Lopes et al. 2002).
4000 The threat of MDR-TB on tuberculosis control
efforts cannot be undermined. Current estimates
suggest that at least 80% cure rate will be needed to
2~L prevent or control an epidemic of MDR-TB, assum-
ing 70% of new MDR-TB cases are detected annually
1995 2000 2005 2010
Years and that the reproductive fitness of MDR strains
Fig. 3.5. Global tuberculosis pandemic. New cases annually to is similar to that of drug-susceptible strains (Dye
2010. Source: World Health Organization and Williams 2000). Such a cure rate may be only
40 M. A. Espinal and M. C. Raviglione

achieved by the careful introduction of second-line to control tuberculosis worldwide suggest that US
drugs, as short-course chemotherapy cannot achieve $1 billion per year is needed to achieve the global
an 80% cure rate among MDR-TB cases (Espinal et targets in the 22 high-burden countries (Floyd et al.
al. 2000b). However, most resource-limited countries 2002). A further US $200 million per year are needed
lack experience in dealing with longer, more expen- for low- and middle-income countries outside the 22
sive and more toxic treatment regimens employing high-burden countries. Of the estimated total costs
second-line drugs (Gupta et al. 200l). Thus, feasible needed for the 22 high-burden countries, 69% is
and cost-effective strategies are needed to address currently contributed by the government of these
MDR-TB in countries with high MDR-TB prevalence countries and only a minor 4% is contributed by
within the current framework of tuberculosis control the international community. The remaining 27% or
(Espinal et al.1999). Operational research in this field US $300 million per year is the financial gap that the
is ongoing and preliminary evidence suggests that 22 high-burden countries will need to fill to control
management of MDR-TB with second-line drugs in tuberculosis. However, new financial resources will
resource-limited settings is feasible and cost-effec- not be all that is needed. Innovative approaches to
tive when a strong tuberculosis control program is in increase case finding and cure rates are necessary, as
place (Suarez et al. 2002). Fortunately, it appears that is a sound strategy to human resource development,
MDR-TB is more of a focal problem, limited to some if new funds are to be spent effectively.
settings or countries, rather than a global problem of Demographic changes will also play their role in
the magnitude of drug-susceptible tuberculosis (Dye keeping tuberculosis incidence at high rates in the
and Espina1200l). current decade. Changes in the age structure of the
Involving the private sector in tuberculosis control population of the developing world will certainly
needs to be strongly pursued. A large proportion of contribute to increasing rates of tuberculosis. Most
tuberculosis patients in high prevalence countries of the population of resource-limited countries is
including India, Pakistan, Philippines, Vietnam, and young, contrary to the industrialized world where
Uganda first approach a private practitioner (Uplekar the population is increasingly aging. Children of
et al. 200l). A study in India suggested that 88% of poor countries infected with M. tuberculosis that are
rural and 85% of urban patients with tuberculosis entering the economically productive age will defi-
first went to a private practitioner (Uplekar et al. nitely be at risk of developing active disease. Since
1998). Strategies to involve the private sector are one third of the world population is infected with
currently under testing. The challenge will be their M. tuberculosis infection, the vast majority of which
implementation on a wider scale. Nevertheless, with- are in resource-limited countries, the joint effects of
out involving private practitioners the international demographic changes and the high burden of M.
community will not achieve tuberculosis control in tuberculosis infection will impact negatively on the
many countries. morbidity and mortality of tuberculosis.
Growing poverty is another factor that will pre- Finally, the lack of new tools to speed-up control
vent tuberculosis incidence from declining in the and elimination of tuberculosis is another major
next several years. Poverty has been always associ- obstacle to overcome. New vaccines, drugs and diag-
ated with tuberculosis (Moore-Gillon 1998). The nostics are not expected to be available for at least the
decreasing incidence of tuberculosis in Europe in next 10 years (Freire and Roscigno 2002; Perkins and
the 19th century was parallel to the reduction of Kritski 2002). For instance, the neglect in the field
poverty. Recently, the most clear impact of poverty of new drugs has been so immense that, in the last
on the incidence of tuberculosis has been observed in 25 years, only 16 new chemical entities out of 1393
Russia. The economic downfall in Russia, due to the were marketed for tropical diseases, including tuber-
transition from the socialist economy model to the culosis (Trouiller et al. 2002).
market economy one, has been linked to the growing
incidence of tuberculosis observed in the last decade
(Dye 2000).
Tuberculosis is one of the three major diseases of 3.7
poverty along with HIV/AIDS and Malaria. The 22 Concluding Remarks
tuberculosis high-burden countries, where increas-
ing poverty is making control of infectious diseases The tuberculosis pandemic will be affecting human
an enormous task, are all very limited in resources. lives for many more decades. While much progress
Recent estimates regarding the resources needed has been made in the last few years, control of tuber-
Global Epidemiology of Tuberculosis
culosis is only slowly becoming a priority in many
countries. Clearly strong financial, political and
managerial actions are needed in order to accelerate
DOTS coverage, reach the WHO global targets, and
incorporate new tools in disease control. To move
forward, a Stop TB partnership was created in 1998
to ensure that endemic countries are adequately
supported technically and financially to control
tuberculosis (Raviglione and Pio 2002). The original
framework for tuberculosis control (World Health
Organization 1994) has been also recently revised to
widen the scope of DOTS, and make it more com-
prehensive and more patient-centered in order to
address some of the current threats to tuberculosis
control (WHO 2002).
Nevertheless, current efforts are not sufficient. To
make tuberculosis control more efficient and effica-
cious, novel interventions need to be added urgently
in high HIV/AIDS and MDR prevalence settings
or where private practitioners are not involved in
national tuberculosis control efforts, or where health
system changes are not addressing tuberculosis con-
trol as a top priority. The recent creation of the Global
Fund to Fight AIDS, Tuberculosis, and Malaria offers
a tremendous opportunity to change the future of the
tuberculosis pandemic if resources are properly used.
The world should not miss it; otherwise, tuberculosis
will continue to be a major scourge to individuals and
communities.
Acknowledgements. Thanks to C. Dye and C. Watt for
providing some of the maps and figures.
References
Alland D et al (1994) Transmission of tuberculosis in New York
City - an analysis by DNA fingerprinting and conventional
methods. N Engl J Med 330:1710-1716
Badri M, Wilson D, Wood R (2002) Effect of highly active anti-
retroviral therapy on incidence of tuberculosis in South
Africa: a cohort study. Lancet 359:2059-2064
Borgdorff MW et al (1997) Analysis of tuberculosis trans-
mission between nationalities in the Netherlands in the
period 1993-5 using DNA fingerprinting. Am J Epidemiol
147:187-195
Bulla A (1977) Global review of tuberculosis morbidity and
mortality in the world (1961-1971). World Health Statistics
Rep 30:2-38
Canetti G (1965) Am Rev Respir Dis 92:687
Cauthen GM et al (1996) Transmission of Mycobacterium tuber-
culosis from tuberculosis patients with human immunodefi-
ciency virus infection or AIDS. Am J EpidemioI144:69-77
Cauthen GM, Pio A, ten Dam HG (1998) Annual risk of tuber-
culosis infection (WHO/TB/88.154). WHO, Geneva
41
Comstock GW, Livesay VT, Woolpert SF (1974) The prognosis
of a positive tuberculin reaction in childhood and adoles-
cence. Am J Epidemiol 99: 131-138
Comstock GW (2000) Epidemiology of tuberculosis. In: Reich-
man LB, Hershfield ES (eds) Tuberculosis: a comprehensive
international approach. Dekker, New York, pp 129-156
Corbett et al (2003) The growing burden of tuberculosis:
global trends and interactions with the HIV epidemic.
Arch Intern Med 163:1009-1021
Daley C et al (1992) An outbreak of tuberculosis with acceler-
ated progression among persons with the human immu-
nodeficiency virus. An analysis using restriction-fragment-
length polymorphism. N Engl J Med 326:231-235
D'Arcy Hart P, Sutherland I (1977) BCG and vole bacillus
vaccines in the prevention of tuberculosis in adolescence
and early adult life. Final report to the Medical Research
Council. Br Med J 2:293-295
De Cock KM, Chaisson RE (1999) Will DOTS do it? A reap-
praisal of tuberculosis control in countries with high rates
of HIV infection. Int J Tuberculosis Lung Dis 3:457-465
Di Perri et al (1989) Nosocomial epidemic of active tuberculo-
sis among HIV-infected patients. Lancet 21:502-504
Dolin PJ, Raviglione MC, Kochi A (1994) Tuberculosis: a global
overview of the situation today. Bull WHO 72:213-220
Drolet GJ (1938) Present trend in case fatality rates in tuber-
culosis. Am Rev Tuberculosis 37:125-151
Dye C (2000) Tuberculosis 2000-2010: control, but not elimi-
nation. Int J Tuberculosis Lung Dis 4:S146-S152
Dye C, Espinal M (2001) Will tuberculosis become resistant to
all antibiotics? Proc R Soc Lond Bioi Sci 268:45-52
Dye C, Williams BG (2000) Criteria for the control of
drug-resistant tuberculosis. Proc Nat! Acad Sci USA 97:
8180-8185
Dye C et al (1998) Prospects for worldwide tuberculosis control
under the WHO DOTS strategy. Lancet 352:1886-1891
Dye C et al (1999) Global burden of tuberculosis: estimated
incidence, prevalence, and mortality by country. J Am Med
Assoc 282:677-686
Dye C et al (2000) Evaluating the impact of tuberculosis con-
trol: number of deaths prevented by short-course chemo-
therapy in China. Int J Epidemiol 29:558-564
Dye C et al (2001) Worldwide incidence of multidrug-resistant
tuberculosis. J Infect Dis 185:1197-1202
Dye C et al (2002a) Erasing the world's slow stain: strategies
to beat multidrug-resistant tuberculosis. Science 295:
2042-2046
Dye C et al (2002b) Low access to a highly effective therapy: a
challenge for international tuberculosis control. Bull WHO
80:437-444
Elliot A et al (1993) The impact of HIV-l on infectiousness
of pulmonary tuberculosis: a community study in Zambia.
AIDS 7:981-987
Elzinga G, Nunn P (2002) TB and HIV: joint problems, joint
solutions? Bull WHO 80:469-470
Enarson DA, Rouillon A (1998) The epidemiological basis of
tuberculosis control. In: Davies PDO (ed) Clinical tuber-
culosis, 2nd edn. Cambridge University Press, Cambridge,
pp 35-52
Espinal MA et al (1999) Rational DOTS-Plus for the control of
MDR-TB. Int J Tuberculosis Lung Dis 3:561-563
Espinal MA et al (2000a) Infectiousness of Mycobacterium
tuberculosis in HIV-1 infected patients with tuberculosis:
a prospective study. Lancet 355:275-280
42
Espinal MA et al (2000b) Standard short-course chemotherapy
for drug-resistant tuberculosis: treatment outcome in 6
countries. J Am Med Assoc 283:2537-2545
Espinal MA et al (2001) Global trends in resistance to antitu-
berculosis drugs. N Engl J Med 344:1294-1303
Evans CC (1998) Historical background. In: Davies PDO (ed)
Clinical tuberculosis, 2nd edn. University Press, Cam-
bridge, pp 1-33
Floyd K et al (2002) Resources required for global tuberculosis
control. Science 295:2040-2041
Freire M, Roscigno G (2002) Joining forces to develop new
weapons against TB: together we must. Bull WHO 80:429
Girardi E et al (2000a) Impact of the HIV epidemic on the spread
of other diseases: the case of tuberculosis. AIDS 14:S47-S56
Girardi E et al (2000b) Impact of combination antiretroviral
therapy on the risk of tuberculosis among persons with
HIV infection. AIDS 14:1985-1991
Godfrey-Faussett P et al (2000) Tuberculosis control and
molecular epidemiology in a South African gold-mining
community. Lancet 356:1066-1071
Gupta R et al (2001) Responding to market failures in tuber-
culosis control. Science 293:1049-1051
Hopewell PC, Chaisson RE (2000) Tuberculosis and human
immunodeficieny virus infection. In: Reichman LB, Hersh-
field ES (eds) Tuberculosis: a comprehensive international
approach, 2nd edn. Dekker, New York, pp 525-527
Jones JL, Hanson MS, de Cock KM (2000) HIV-associated
tuberculosis in the era of highly active antiretroviral
therapy. Int J Tuberculosis Lung Dis 4:1026-1031
Klausner JD et al (1993) Mycobacterium tuberculosis in house-
hold contacts of human immunodeficiency virus type-l-
seropositive patients with active pulmonary tuberculosis
in Kinshasa, Zaire. J Infect Dis 168: 106-111
Kochi A (1991) The global tuberculosis situation and the
new control strategy of the World Health Organization.
Tubercle 72:1-6
Moore-Gillon JC (1998) Tuberculosis and poverty in the devel-
oped world. In: Davies PDO (ed) Clinical tuberculosis, 2nd
edn. Cambridge University Press, Cambridge, pp 383-393
Murray CJ, Lopez AD (1997) Mortality by cause of eight
regions of the world: global burden of disease study. Lancet
349:1269-1276
Murray q, Styblo K, Rouillon A (1993) Tuberculosis. In:
Jamison JT, Mosley WH, Measham AR, Bobadilla JL (eds)
Disease control priorities in developing countries. Oxford
University Press, New York, pp 233-259
Narain JP, Raviglione MC, Kochi A (1992) HIV-associated
tuberculosis in developing countries: epidemiology and
strategies for prevention. Tubercle Lung Dis 73:311-321
Nerlich AG et al (2000) Ancient Egyptian prosthesis of the big
toe. Lancet 356:2176-2179
Pablos-Mendez A et al (1998) Global surveillance for antitu-
berculosis-drug resistance, 1994-1997. N Engl J Med 338:
1641-1649
Perelman MI (2000) Tuberculosis in Russia. Int J Tuberculosis
Lung Dis 4:1097-1103
Perkins M, Kritski AL(2002) Diagnostic testing in the control
of tuberculosis. Bull WHO 80:512-513
Raviglione MC et al (1995) Global epidemiology of tubercu-
losis: morbidity and mortality of a worldwide epidemic. J
Am Med Assoc 273:220-226
Raviglione MC et al (1997a) Assessment of worldwide tuber-
culosis control. Lancet 350:624-629
M. A. Espinal and M. C. Raviglione
Raviglione MC et al (1997b) Tuberculosis and HIV: current
status in Africa. AIDS II:SI15-S123
Raviglione MC, Pio A (2002) Evolution of WHO policies for
tuberculosis control, 1948-2001. Lancet 359:775-780
Rieder HL (1999) Epidemiologic basis of tuberculosis control,
1st edn. International Union Against Tuberculosis and
Lung Disease, Paris
Riley RL, O'Grady F (1961) Airborne infection: transmission
and control. Macmillan, New York
Rothschild BM et al (2001) Mycobacterium tuberculosis com-
plex DNA from an extinct bison dated 17,000 years before
the present. Clin Infect Dis 33:305-311
Salo WL et al (1994) Identification of Mycobacterium tubercu-
losis DNA in a pre-Columbian Peruvian mummy. Proc Nat!
Acad Sci USA 91:2091-2094
Santoro-Lopes G et al (2002) Reduced risk of tuberculosis
among Brazilian patients with advanced human immu-
nodeficiency virus infection treated with highly active
antiretroviral therapy. Clin Infect Dis 34:543-546
Small PM et al (1994) The epidemiology of tuberculosis
in San Francisco. A population-based study using con-
ventional and molecular methods. N Engl J Med 330:
1703-1709
Sonkin LS (1951) The role of particle size in experimental air-
borne infection. Am J Hyg 70:185-196
Sonnenberg P et al (2001) HIV-l and recurrence, relapse, and
reinfection of tuberculosis after cure: a cohort study in
South African mineworkers. Lancet 358:1687-1693
Styblo K (1983) The epidemiological situation of tuberculosis
and the impact of control measures. Bull Int Union Against
Tuberculosis 58: 179-186
Styblo K (1985)The relationship between the risk of tubercu-
lous infection and the risk of developing tuberculosis. Bull
Int Union Against Tuberculosis 60:117-119
Styblo K (1991) Epidemiology of tuberculosis, 2nd edn. Royal
Netherlands Tuberculosis Association, The Hague
Styblo K, Meijer J, Sutherland I (1969) The transmission of
tubercle bacilli - its trend in a human population. Tubercu-
losis Surveillance Research Unit report no 1. Bull Int Union
Against Tuberculosis 42:1-104
Suarez PG et al (2001) Epidemiological impact of tuberculosis
control in Peru. J Infect Dis 184:473-478
Suarez PG et al (2002) Feasibility and cost-effectiveness of
standardised second-line drug treatment for chronic
tuberculosis patients: a national cohort study in Peru.
Lancet 359: 1980-1989
Sudre P, ten Dam G, Kochi A (1992) Tuberculosis: a global
overview of the situation today. Bull WHO 70:149-159
Sutherland I et al (1971) Annual risks of tuberculous infec-
tion in 14 countries, derived from the results of tuberculin
surveys in 1948-1952. Bull Int Union Against Tuberculosis
Lung Dis 45:75-114
Thompson BC (1943) Survival rates in pulmonary tuberculo-
sis. Br Med J 2:721
Trouiller P et al (2002) Drug development for neglected dis-
eases: a deficient market and a public-health policy failure.
Lancet 359:2188-2194
UNAIDS (2002) Joint United Nations Program on HIV/AIDS I
World Health Organization. AIDS epidemic update. 02.46E.
UNAIDS, Geneva
Uplekar M et al (1998) Tuberculosis patients and practitioners
in private clinics in India. Int J Tuberculosis Lung Dis 2:
324-329
Global Epidemiology of Tuberculosis
Uplekar M et al (2001) Private practitioners and public
health: weak links in tuberculosis control. Lancet 358:
912-916
Wells WF (1955) Airborne contagion and air hygiene. Harvard
University Press, Cambridge
WHO/UNAlDS (1999) World Health Organization I Joint
United Nations Program on HIV/AlDS. Preventive therapy
against tuberculosis in people living with HIV. Weekly Epi-
demiol Rec 74:385-398
Wilkinson D et al (1997) Molecular epidemiology and trans-
mission dynamics of Mycobacterium tuberculosis in rural
Africa. Trop Med lnt Health 2:747-753
World Bank (1993) Investing in Health. World Development
Report. Oxford University Press, New York
World Health Organization (199l) Forty-fourth World Health
Assembly. Resolutions and decisions. Resolution WHA
44.8. WHO, Geneva
43
World Health Organization (1994) Global tuberculosis pro-
gramme. Framework for effective tuberculosis control.
WHO/TB/94.179. WHO, Geneva
World Health Organization (1997) Anti-tuberculosis drug
resistance in the world. WHO/TB/97.229. WHO, Geneva
World Health Organization (2000) Anti-tuberculosis drug
resistance in the world. Report no 2: prevalence and trends.
WHO/CDS/TB/2000.278. WHO, Geneva
World Health Organization (2001) Global tuberculosis con-
trol. WHO Report 2001. WHO/CDS/TBI2001.287. WHO,
Geneva
World Health Organization (2002) An expanded DOTS frame-
work for effective tuberculosis control. Int J Tuberculosis
Lung Dis 6:378-388
World Health Organization (2003) Global tuberculosis control:
surveillance, planning, financing. WHO report 2002. WHOI
CDS/TBI2003.316. WHO, Geneva
4 Epidemiology of Tuberculosis in Saudi Arabia
ABDULRAHMAN A. ALRAJHI and ALI M. AL-BARRAK

CONTENTS tion of transmission, number of infected persons

without active disease, distribution of drug-resistant
4.1 General Epidemiology 45 tuberculosis, and tuberculosis in certain populations.
4.1.1 Incidence of Pulmonary Tuberculosis 46
4.1.2 Incidence of Extrapulmonary Tuberculosis 48
There are no data on the impact of tuberculosis on
4.1.3 Prevalence of Infection 49 community, economy or development. Mortality data
4.2 Epidemiology of Drug-Resistant M. tuberculosis 51 are not available either.
4.3 Tuberculosis in Certain Populations 51
4.3.1 The Elderly 52
4.3.2 Diabetics 52
4.3.3 Persons Infected with Human Immunodeficiency
Virus 52 4.1
4.3.4 Hemodialysis Patients 53 General Epidemiology
4.3.5 Cancer and Bone Marrow Transplant Patients 54
4.3.6 Solid Organ Transplant Recipients 54 Measuring the size of tuberculosis in a community
4.4 Mycobacterial Other than M. tuberculosis 54
4.5 Conclusion 55
is not as easy as it may appear initially. The disease
References 55 course and the community it afflicts are constantly
changing. Definitions used for various tuberculosis
cases are not standardized and surveillance appli-
Tuberculosis is older than recorded history. Typical cation in the field has enormous variations. The
lesions of spinal tuberculosis were noted in Neolithic community is also changing its composition, living
human remains and paintings. Pre-Islam Arab poetry standards, and exposure between members.
had described the disease and its association with a There are several measures to indicate the size of
non-cure. The writings of Moslem physicians indi- tuberculosis as a community problem. Before the
cate that the disease was well known along with its era of tuberculosis chemotherapy discovery and its
risk factors, symptoms and complications. Some of use in treatment, an indirect measure, which could
the factors in this part of the world were encourag- also assess the impact of the disease, was mortality.
ing the introduction of the organism on a continu- Tuberculosis chemotherapy and control measures
ous basis through commerce; as the peninsula was make mortality an imprecise measure of disease
involved in commerce roads, and pilgrims were extent in the community. Another measure identified
coming from all corners. early in the 20th century is tuberculin skin test. It is
The purpose of this chapter is to present the a useful measure to identify the prevalence of infec-
available epidemiology about tuberculosis in Saudi tion in a community at a certain point of time and to
Arabia. It includes data on numbers of new cases follow trends. Unfortunately, the value of tuberculin
annually, type of tuberculosis involvement, demo- testing is affected by BeG vaccination, exposure to
graphics of affected patients, geographic distribu- nontuberculous mycobacteria, and waning immunity
(Thompson et al. 1979). A third measure was used in
A. A. ALRA1HI, MD, MPH, FIDSA some areas, which is field surveillance by chest X-ray
Consultant and Head, Section of Infectious Diseases, Chairman, and sputum cultures from community samples. This
Department of Medicine (MBC #46), King Faisal Specialist Hos- method has to be repeated so incidence rates can be
pital and Research Centre, P.O. Box 3354, Riyadh 11211, Saudi measured. Indirectly, incidence can be estimated to
Arabia be half of the prevalence, considering the disease
A. M. AL-BARRAK, AmBIM, DTM & H, FRCP (C)
Consultant Infectious Diseases Unit, Department of Medicine, lasts for 2 years without treatment. Another epidemi-
Riyadh Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, ologic concept was developed; annual risk of tuber-
Saudi Arabia culosis infection reflects the probability of a person

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
46 A. A. Alrajhi and A. M. AI-Barrak

becoming infected within a given year. This measure
requires a prevalence assay on multiple occasions in
order to be precise. BCG vaccination will influence
prevalence rates that are measured using Mantoux
tuberculin test. Lastly, the World Bank has attempted
to estimate the economic impact of tuberculosis in
its 1993 World Development Report though disabil-
ity-adjusted life years (DALYs), the years that are
consumed by various diseases, including tubercu-
losis (World Bank 1993). This chapter will attempt
to present available data on these various measures
pertaining to Saudi Arabia.
Tuberculosis is a reportable disease in Saudi
Arabia. The Ministry of Health (MOH), represented
by Preventive Medicine has been collecting data
on newly diagnosed cases of tuberculosis annually.
These incidence rates are published every year. Some
of the information included was gender, age, nation-
ality, and type of tuberculosis involvement. These
rates are the only available data at population level.
Although they may not include all cases because of
reporting deficiency from care providers, they pro-
vide valuable means to monitor trends. These data
have been considered along with the social, develop-
Case identification and treatment is considered one
of the most important components of tuberculosis
control and prevention. The annual incidence rates
of pulmonary tuberculosis were higher for foreigners
compared with Saudis. In 1996, the rate was 23.9 cases
per 100,000 population for non-Saudis compared with
7.4 cases per 100,000 population for Saudis (Ministry
of Health 1996). The female and male rates for Saudis
were almost similar,6.5 and 8.3 cases per 100,000 pop-
ulation, respectively. For non-Saudis, these rates were
22.7 for males and 26.9 cases per 100,000 population
for non-Saudi females. Certain regions of the coun-
try consistently scored higher rates. Aseer, ]azan, and
Holy Makkah have higher rates than Hail and Gaseem.
Table 4.1 summarizes incidence rates for pulmonary
tuberculosis in various groups and regions over the
period of 1996-1997.
Apart from reported pulmonary tuberculosis,
investigators have attempted to estimate incidence
Cases per 100.000 population
1600
1400
1200

.• • •
mental, and economic changes in the country. 1000
800
600 - I--- -

4.1.1
Incidence of Pulmonary Tuberculosis
400
200
-
-
I---
I---
-
- -
o
70 71 72 73 74 75 76 77 78 79
The first available incidence rate of pulmonary Year

tuberculosis was in 1970. During that year, there were Fig. 4.1. Annual incidence rates of pulmonary tuberculosis in
76,748 new cases of pulmonary tuberculosis, making Saudi Arabia, 1970-1979 (Ministery of Health: Annual Health
a rate of 1298.5 cases per 100,000 population per Report)
year (Ministry of Health 1979). Over the subsequent
years, incidence rate of pulmonary tuberculosis has
decreased steadily to 166.1 cases per 100,000 popula- Reported cases per 100,000 population
140
tion in 1979. Figure 4.1 depicts the trend of incidence
rates over the 10 years of the 1970s. 120
Incidence of reported pulmonary tuberculosis
continued to decrease steadily during the 1980s and 100

the 1990s. For the year 1980, the incidence rate was 80
135 cases per 100,000 population. In 1997, the rate has
dropped to 12 cases per 100,000 population (Ministry 60

of Health 1997). When considering other changes in

40
the health care system availability, advances in diag-
nosis capabilities, improved reporting associated with 20 ••
communications improvement, this steady reduction
becomes more impressive, Fig. 4.2. This reduction o III1IIIIII
80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
was attributed to BCG vaccination program that have Year

been started many years ago, and more importantly, Fig. 4.2. Annual incidence rates of pulmonary tuberculosis in
the active screening program for foreign workers Saudi Arabia, 1980-1997 (Ministery of Health: Annual Health
coming to the country (Ministry of Health 1996). Report)
Epidemiology of Tuberculosis in Saudi Arabia
Table 4.1. Annual incidence rates of pulmonary tuberculosis
in various groups and regions during 1996-1997 (Ministry of
Health 1997,1996)
Group Rates per 100,000 population
Overall 12.0
Nationality
Saudis 7.4
Non-Saudis 23.9
Gender
Saudi, male 8.3
Saudi, female 6.5
Non-Saudi, male 22.7
Non-Saudi, female 26.9
Regions
Jazan 20.5
Makkah 17.0
Riyadh 13.9
Eastern 10.9
Hail 2.6
rates at community level (Zaman 1991a; Milaat et al.
1994; el-Kassimi 1994). However, these efforts have
not provided more reliable data than Ministry of
Health annual reports. Milaat et al reviewed all tuber-
culosis cases in a regional chest hospital in Jeddah
for 1 year, 1989-1990. Population data were obtained
from MOH records to calculate incidence rates. The
overall incidence rate was estimated at 63.4 cases per
100,000 populations. Extrapulmonary tuberculosis
was found in 7.6% of all cases. There were 2.6 cases
of tuberculosis in non-Saudis for each one case of
tuberculosis in Saudis from the same population
number (99.6 non-Saudis and 38.8 Saudi tuberculo-
sis cases per 100,000 population). Reporting tubercu-
losis cases from the private health care facilities has
been grossly deficient. This is noted in Jeddah, where
private medicine provides care for a large proportion
of population in that region. Milaat and colleagues
think that the higher estimated rate noted in Jeddah
relative to the country in general is related to the
crowding and high immigrant population in Jeddah.
In addition, Jeddah serves as the major port of entry
for visitors to the two holy cities. This is suggested
by peaking of cases during the 11th and 12th Hijra
months. In a nationwide survey, positive Mantoux
reaction was noted to be three times higher in the
Western region than the nation as a whole (AI-Kas-
simi et al. 1993). AI-Kassimi has calculated incidence
rates from the annual risk of infection (AI-Kassimi
1994). The result was double the rates from reported
cases to MOH. He correctly points to the problem of
under-notification by health care providers to MOH.
The high rates of tuberculosis as infection and
incidence of active disease are reflected in the pro-
47
portion of patients admitted with pneumonia that
end up diagnosed as having pulmonary tuberculosis.
In a review of 112 patients admitted with pneumonia
over the year 1983 at Riyadh Military hospital, tuber-
culosis was confirmed in 7.1% of them. In fact, it was
only second to Streptococcus pneumoniae (Mohamed
and Evans 1987). Some of the radiological changes
noted in tuberculosis patients could also be seen in
patients with community acquired bacterial pneu-
monia (AI-Dabbagh et aI.1991)."Unusual" chest radi-
ography findings were noted in 13% of 714 patients
with pulmonary tuberculosis in two hospitals in
Jeddah. Cavities were seen in 78% of all patients, sug-
gesting a prolonged and unrecognized disease. The
most common radiological finding, however, was
upper or multilobar lesions, seen in 86% of patients.
However, more severe pneumonia requiring intensive
care had a slightly different etiology (Dahmash et al.
1994; Dahmash and Chowdhury 1994). In a univer-
sity hospital in Riyadh, 113 patients were admitted
to the intensive care unit with pneumonia between
1991 and 1992. Patients were followed prospectively,
50 patients (44%) had hospital-acquired pneumo-
nia. Out of the community acquired pneumonias,
9 (14.3%) were due to M. tuberculosis, which was
second only to S. pneumoniae, isolated from 10
patients but far more fatal than S. pneumoniae-5 of 9
(55.6%) compared with 2 of 10 (20%) (Dahmash and
Chowdhury 1994).
The population at risk for pulmonary tuberculosis
remains large in Saudi Arabia. Cases of pulmonary
tuberculosis most frequently involve young adults
between 15 years and 35 years of age (AI-Dabbagh
et al. 1991; AI-Hajjaj et al. 1991). This seems to be the
case for last thousand years (AI-Amoud 1993). Some
argue that this is related to the number at risk being
higher in the age group because of the immigrant
work force. However, this is not the reason, as the cal-
culated rates are still highest for the young age group
(Ministry of Health 1996). In the large review of 1566
cases of pulmonary tuberculosis at Riyadh Chest
Hospital between 1983 and 1987, 41 % of patients
were aged 21-30 years (AI-Hajjaj et al. 1991). These
patients were admitted for therapy initiation at a rate
of 32.6 cases monthly. Around 52% of patients were
non-Saudis, mostly from Yemen, 20% of all patients.
A concerning phenomenon in this cohort is the high
rate of "old" cases, which represent 29% of the cases
with most of the patients being Saudis. They had more
hemoptysis and were more difficult to cure. Sputum
smear positivity was relatively high at 75%. The
conversion rate was low at two months (80%). These
patients from Riyadh had larger proportion of lower
48 A. A. Alrajhi and A. M. AI-Barrak

lung field tuberculosis compared with Jeddah-lO.3%
compared with 5.3% (AI-Dabbagh et al.199l; Pandya
et al. 1990). These percentages represent the post pri-
mary infection. The patients from Riyadh had higher
rates of sputum positive smears-84% compared with
75% for all pulmonary tuberculosis cases.
Pulmonary tuberculosis from the South Region of
Saudi Arabia has been characterized in a prospective
study managed at the Chest Hospital in Abha between
1989 and 1991 (al-Wabel et al. 1995) One hundred
ninety patients were admitted; 69.5% of them had
no previous diagnosis of tuberculosis. As in other
areas of the country, the group aged 20-40 years
constituted more than 34% of all cases. In this region,
patients older than 60 years were 25% of all patients
with pulmonary tuberculosis. Unlike in the central or
western regions, less than 25% of patients were non-
Saudis. Another peculiar finding in these patients was
the high proportion of diabetics (21 %), reflecting the
higher rates of diabetes in elderly nationals. Sputum
smears were positive in 42% of patients, posing a
considerable risk for disease transmission from
these cases. Radiologically, cavitary lung lesions were
noted in 38% of patients, lower lung field in 3.2%, and
miliary pattern in 1.3%.
Saudi Arabia hosts millions of foreigners as part
of the workforce, visitors for business, or Hajj and
Omrah. Many of the workers come from countries
where tuberculosis is highly endemic. Screening
for these workers is performed before they obtain
their visas and after they arrive. In spite of this, they
represent more than 50% of tuberculosis cases in
a large city like Riyadh (AI-Hajjaj et al. 1991). The
impact of these cases is not limited to economy, but
to the extent of increasing the hazard of tuberculosis
infection. Drug-resistant M. tuberculosis is also of
major concern as these patients come from areas
where drug-resistant rates are very high. Screening
for tuberculosis is not applied to visitors, including
pilgrims for Hajj and Omrah. Alzeer et al. (l998)
reviewed the microbiology of pneumonia in patients
mens other than expectorated sputum (EI-Sheikh et
al.1998). The focus was mainly on bacterial and viral
pathogens. Nevertheless, two patients of eighty-six
(2.3%) with lower respiratory tract infection had
positive smears for acid-fast bacilli.
4.1.2
Incidence of Extrapulmonary Tuberculosis
Extrapulmonary tuberculosis requires tissue sam-
pling to confirm diagnosis. This may not be available
to many health care facilities. Subsequently, diagno-
sis is delayed if attempted. The number of reported
extrapulmonary tuberculosis cases are even smaller
than the number of actual cases. The annual Health
Reports of the MOH started presenting data on extra-
pulmonary tuberculosis in 1989. The annual inci-
dence rate of reported extrapulmonary tuberculosis
was 3.7 cases per 100,000 population nation-wide in
1989 (Ministry of Health 1989). This figure decreases
to 1.7 per 100,000 population in 1993 (Ministry of
Health 1993). This rate has been steadily increasing
to 4.7 cases per 100,000 population in 1997, Fig. 4.3
(Ministry of Health 1997). During the same time
frame, reported pulmonary tuberculosis rates were
decreasing (Ministry of Health 1997). The increasing
rates of reported extrapulmonary tuberculosis may
have resulted from better diagnostic facilities, which
identified more cases, and better reporting by health
providers.
Earlier reports describe their series in their local
hospitals. In 1982, Froude and Kingston reviewed 162
cases diagnosed with extrapulmonary tuberculosis
between 1979 and 1981 from King Faisal Specialist Hos-
pital and Research Centre (Froude and Kingston 1982).
Although this finding may be due to referral pattern,
CNS disease (including spine involvement) was the
New cases per 100 population

admitted during Hajj of 1994 (l414 H). Out of 64 5

I--
4.5
admitted patients, M. tuberculosis was the causative 4 I--

agent in 13 (20%) patients. It was the most common 3.5 - f--- I--
3 - - I--- I--
agent for pneumonia identified, followed by S. pneu- 2.5 f--- - f--- I--
moniae and coliforms in 9% and 11%, respectively. 2 f--- I--- - I--- I--

Radiographic features of tuberculous and non-tuber- 1.5 I--- - - I--- I--- - I--- I--
1 I--- - - I--- I--- - I--- I--
culous pneumonias were the same. This high rate of 0.5 I--- - - I--- f--- - f--- I--
tuberculosis among pilgrims was thought to be the o '--r- ~ '-
1989 1991 1992 1993 1994 1995 1996 1997
reason for the above national annual risk of infec- Year

tion noted in pilgrim-receiving cities (AI-Kassimi Fig. 4.3. Annual incidence rates for extrapulmonary tubercu-
et al. 1993). An earlier report of the 1991-1992 Hajj losis in Saudi Arabia, 1989-1997 (Ministery of Health: Annual
season did not attempt to obtain respiratory speci- Health Report)
Epidemiology of Tuberculosis in Saudi Arabia
most common, comprising 36% of cases, followed by
lymphadenopathy 22%, gastrointestinal disease 16%,
and surprisingly, disseminated disease was diagnosed
in 14% of the extrapulmonary tuberculosis cases. The
morbidity associated with extrapulmonary tubercu-
losis can be very high. Seventeen patients (11 %) had
spinal cord damage related to vertebral involvement.
During the same study period of 27 months, six new
pulmonary tuberculosis cases were diagnosed every
month. Around the same time, a similar report was
published about patients from Jeddah. Mokhtar and
Salman published details of 125 patients with extra-
pulmonary tuberculosis in 1983 (Mokhtar and Salman
1983). Lymph node involvement was identified in 54%
of the patients, followed by abdominal disease and
urogenital disease in 16% each, and bone and joints in
10%. Surprisingly, only one CNS disease was diagnosed
as meningitis. During the same period, 20 new cases of
pulmonary tuberculosis were diagnosed every month.
In their Hospital, Mokhtar and Salman found extrapul-
monary tuberculosis in 15% of all tuberculosis cases
identified. In 1980, incidence of tuberculosis was 120
cases per 100,000 population in Jeddah (Mokhtar and
Salman 1983).
Fig. 4.4. Distribution of involvement in extrapulmonary tuber-
culosis (Froude and Kingston 1982)
10 CNS 0 Lymph node 0 Urogenital • Abdominal 0 Bone & Joint • Othenl
Fig. 4.5. Distribution of involvement in extrapulmonary tuber-
culosis (Mokhtar and Salman 1983)
49
4.1.3
Prevalence of Infection
Human exposure to M. tuberculosis will result in
infection. No disease may be apparent initially. After
a variable period of time and factors related to host
immunity, M. tuberculosis will be reactivated, result-
ing in disease. The prevalence of infection can be
measured by the tuberculin test. In 1908, Mantoux
described his intradermal test, which was developed to
a purified protein derivative (PPD) in 1934 by Seibert
(Schein and Huebner 1995). When this test was applied
to a large population, it was realized that most people
get infected without having active disease. This test,
known as Mantoux, tuberculin, and PPD served as
good tools to measure the proportion of the popula-
tion that had been exposed to M. tuberculosis.
In a rural area in the western region of Saudi
Arabia in 1979, prevalence of pulmonary tuberculo-
sis was as high as 1% of the population (Hammam et
al.1979). In Saudi Arabia, the first reference of testing
of asymptomatic population for tuberculosis expo-
sure was in 1979. Out of81,006 subjects tested, 48,483
(59.9%) subjects had a positive test. During the same
year, 66% of tested males were positive, compared
with 40% of tested females. These percentages were
higher with the increasing age of the subjects (Minis-
try of Health 1979). Although no details are available
on these subjects as to whether or not they are patient
contacts, these percentages give an indication of how
common tuberculosis infection is in the area. Prob-
lems of the test, its administration, or interpretation
may influence the prevalence estimation towards the
underestimation side.
Subsequent to 1980, many field studies published in
peer-reviewed journals have reported the tuberculin
skin testing results. Saudi Arabia has been among the
middle prevalence countries (2-14%) according to the
classification proposed by the International Union
Against Tuberculosis (AI-Kassimi et al. 1993). The
World Health Organization (WHO) expert commit-
tee on tuberculosis recommends continuing early life
vaccination until the percentage of tuberculin reactor
entering school drops to 2%, which is a low prevalence
level (WHO 1964). The protective effect of BCG vac-
cination does not last beyond 10 years. Saudi Arabia
has adopted a BCG vaccine given at birth. In a national
immunization coverage survey in 1991, BCG immu-
nization coverage was high (99%) both in urban and
rural area settings, but the western and southern area
had slightly lower coverage (Farag et al.1995).
In early 1990s, community based studies for
the prevalence of Mantoux reaction were done at
 50 A. A, Alrajhi and A, M. AI-Barra!<

different regions of Saudi Arabia. In a nationwide Bener and Abdullah confirmed the above finding
community survey, about 24% of the children aged in another Saudi Arabia population sample (Bener
5-14 years were not vaccinated (AI-Kassimi et al. and Abdullah 1993). Positive Mantoux reaction was
1993). The prevalence of positive Mantoux reaction significantly higher among BCG recipient subjects
(10 mm induration) in children not vaccinated aged in all age groups, male gender, nationalities, and
5-14 years was 6%, and was 4% in Saudi nationality. occupations. Studies from different regions of Saudi
In this age group, there was marked difference in Arabia confirmed the finding of this prevalence (el-
prevalence between urban and rural dwellers (10% Kassimi et al. 1991; Bener 1990; Abdullah et al. 1991;
and 2%, respectively). This difference was more AI-Kassimi et al. 1991). In the central region, 7% of
in western (20% and 1%) and southern (12% and unvaccinated children aged 4 to 14 years had posi-
1.5%) provinces. The authors have attributed this tive Mantoux reaction, while 15% of those vaccinated
to the presence of settlers after pilgrimage in the were positive. There was increased prevalence of
western province and Yemeni laborers' exemption Mantoux positivity with age, more in the BCG-vacci-
from medical screening in the south, adjacent to nated population. There was a steep upward increase
their country. The prevalence of positive Mantoux in prevalence of positivity between age 5 years and
reaction increased with age steadily to peak at age 14 years and 25 years and 34 years, probably due to
45-64 years at 56% for non-vaccinated and 70% for an increased number of a previously infected expa-
BCG vaccinated subjects, Fig. 4.6. triate population who are workers in the age group

a 100%
Central Region 101W. North Region

- +------------f-----l

....
80% 80%

?' "'~'-"
60%
/ .•... \
60";' +-------------.I--.....--l
40";' + - - - - - - - - - : ;......'--------,_-----l
.. , \
......
40%
~
20% 20";' +-----::.-=---:+---'--'-----------1
' .......
0% 0% +-~...,....-"""T""-__r-__.--..__----l
'5-14 15-24 25·34 35-44 45-64 65+ 5-14 15-24 25-34 35-44 45-64 65+

Age groups Age groups

100% 100"1. South Region

Eastern Region

80% 80% + - - - - - - - - -...----1

60%

40%
. '
60";' +-------+----\;--'---1

40"/. t - - - - - : : f - -.....-'---+----I

20% 20% +------.f---:.-'------4----I

0";'
'5-14
• 15-24 25-34 35-44 45-64 65+
0% +-:....,r--__r-"""T""-...,....-T""'i~

'5·14 15·24 25-34 35-44 45-64 65+

Age groups Age groups

I··... No-BeG - - - BCG·Recipients 1

b 100%
Nation-wide

80%

+-----/---.....=-----=----.-=-.•-..-..-..-...
.'
60% - -1

40% +----/---:;,L,,-..-,..-•-.. 7"

•• --:..:----------1
20%
~
•.... '
....
+--.,,4---::.,....:....-----------1 Fig.4.6. a: Rates of positive Mantoux test in various
regions in BCG-recipient and non-recipient population
0% +------.--..__--.-----.--...,....---1 (el-Kassimi et al. 1991; Bener 1990; Abdullah et aI. 1991;
5-14 15-24 25-34 35-44 45-64 65+ AI-Kassimi et aI. 1991) b: Rates of positive Mantoux tests
Age groups nationwide in BCG-recipient and non-recipient popula-
1.. ···No-BCG --BCG-Recipients I
tion (Al-Kassimi et aI. 1993)
Epidemiology of Tuberculosis in Saudi Arabia
25-44 years, and an increased opportunity to acquire
infection by adolescents as they leave the sheltered
life of young age. Saudi nationality again had lower
prevalence, but higher prevalence in male gender
than in female. Urban dwellers had, in general,
higher prevalence. In the northern area, only 4% of
unvaccinated children aged 5-14 years had a positive
Mantoux reaction. A positive Mantoux reaction was
seen in significantly more BCG-vaccinated subjects
(manifested only in adults) and in older age groups.
The overall prevalence of positive Mantoux reaction
in the unvaccinated group was 19% and reached
its peak (58%) in the 65+ age group, while in those
vaccinated was 15%. The overall rate of positivity in
males was 37% which is significantly higher than in
women (12%). There was no difference in subjects
living status, nationality, level of education, and his-
tory tuberculosis in the family or chest symptoms. In
the Eastern area, 5% of unvaccinated children aged
5-14 years had a positive Mantoux reaction. Again, a
positive Mantoux reaction was seen in significantly
more BCG-vaccinated subjects (manifested only in
adults) and in older age groups. In the Southern area,
4% of unvaccinated children aged 5-14 years had a
positive Mantoux reaction. Again, a positive Mantoux
reaction was seen in significantly more BCG-vacci-
nated subjects and in older age groups. The overall
prevalence of a positive Mantoux reaction in the
unvaccinated group was 26% and reached its peak
(67%) in the 65+ age group, while in those vaccinated
was 17%. The overall rate of positivity in male was
35% as opposed to 42% in female with difference
between the sexes is more marked in the 25-64 year
age group. There was significant difference between
urban and rural residents, and in nationality status.
In 1992, MOH has established the National Tuber-
culosis Control Committee. The field application
of the National Tuberculosis Control Program was
implemented in Riyadh in 1996 according to WHO
guidelines (AI-Hajjaj 2000).
4.2
Epidemiology of Drug-Resistant
M. tuberculosis
The reporting procedures of all tuberculosis cases in
Saudi Arabia did not include the susceptibility results
of isolated organisms. Microbiological diagnosis at
point of care areas did not attempt to grow the organ-
isms. Diagnosis was limited to smears of respiratory
specimens and histopathological findings for extra-
51
pulmonary tuberculosis. Certain "Chest Hospitals"
had the facilities to isolate and test susceptibility
of M. tuberculosis. Testing in these facilities was not
done for all isolates. Health reports did not include
the limited data on drug susceptibility. Thus, patterns
and distribution of drug-resistance are not known at
the community level.
Other health care facilities were performing sus-
ceptibility on their isolates. Many reports of sensitiv-
ity patterns were published (Schiott et al. 1985; AI-
Orainey 1986; AI-Orainey et al. 1989; Zaman 1991b;
Jarallah et al.I992; Ellis et al.1996; AI-Jama et al.1999;
Khan et al. 2001; Alrajhi et al. 2002). These reports
used different methodologies, definitions, and vari-
ous ways of presenting the data. So any epidemio-
logical analysis of these reports lacks the necessary
details to identify patterns, geographic and popu-
lation distribution. Rates varied as high as 43.7%
resistant to at least one of the first-line agents and
as low as 8.7% of isolates being resistant. Table 4.2
summarizes the published data on drug resistance.
Two reports on drug resistance rates appeared from
the same institutions over two periods of time. Khan
et al have shown clearly that resistance rates were
increasing between the period of 1993-1995 and the
period of 1996-1998 This increase was statistically
significant for one or more antituberculous agents
and for all first-line agents singularly except for eth-
ambutol and pyrazinamide (Khan et al. 2001). On the
other hand, from King Faisal Specialist Hospital and
Research Centre, where most of tuberculosis cases
are Saudis, the resistance rates were not statistically
significant between the period 1989-1994 and the
period 1995-1999 (Alrajhi et al. 2002). As most of the
reported resistance rates came from referral centers,
they may not represent the actual rates in the com-
munity as a whole. In a recent report by the WHO,
no data from the Eastern Mediterranean Region,
where the Kingdom of Saudi Arabia is grouped, were
included (WHO 1997).
4.3
Tuberculosis in Certain Populations
As M. tuberculosis activity is closely related to host
immunity, rates of reactivation of a latent infection
will depend mostly on host-related factors. Thus
reactivation rates will vary widely among various
patient and non-patient populations. Over the next
few sections, available data of disease prevalence in
various populations will be reviewed.
52 A. A. Alrajhi and A. M. AI-Barrak

Table 4.2. Drug-resistance rates against isoniazid and rifampin from various reports

Author (Reference year) Year Region Total Isoniazid Rifampin Isoniazid and At least resistant
number (%) number (%) rifampin, no (%) to one agent

Schiott et aI. (1985) Not indicated Jazan 103 42 (40.8%) 21 (20.4%) 20 (19.4%) 45 (43.7%)
AI-Orainey (1986) 1979-1982 Riyadh 1968 82 (4.2%) ** ** 259 (13.2%)
AI-Oraineyet aI. (1989) 1986-1988 Riyadh 432 84 (19.4%) 42 (9.7%) 38 (8.8%) 92 (21.3%)
Zaman (1991b) Not indicated Jeddah 483 86 (17.8%) 113 (23.4%) 70 (14.5%) 145 (30%)
Jarallah et aI. (1992) 1986-1988 Taif 678 44 (6.5%) 104 (15.3%) 26 (3.8%) 153 (22.6%)
Chowdhury and 1987-1991 Riyadh 214 20 (9.3%) 8 (3.7%) 7 (3.3%) 24 (11.2%)
Kambal (1992)
Ellis et aI. (1996) 1989-1994 Mixed 289 22 (7.6%) 9 (3.1 %) 8 (2.8%) 25 (8.7%)
AI-Jama et aI. (1999)* 1989-1998 Eastern 1239 211 (17%) 78 (6.3%) 34 (2.7%) 302 (24.4%)
Khan et aI. (2001) 1996-1998 Western 101 29 (28.7%) 21 (20.7%) 21 (20.7%) 30 (29.7%)
AIrajhi et aI. (2002) 1995-2000 Mixed 320 29 (9.1%) 9 (2.8%) 9 (2.8%) 36 (11.3%)
Total** 5827 649 (11.1 %) 405 (10.5%) 233 (6.0%) 1111 (19.1%)

* Report included mycobacteria other than tuberculosis

** Rifampin and MDR rates are calculated without Ref. no. (AI-Orainey 1986)

4.3.1 insulin requirements (Boucot et al. 1952). Pandya et

The Elderly
The high prevalence of tuberculosis in Saudi Arabia
increases with age. As health care improves and life
expectancy is prolonged, more population is at risk of
tuberculosis reactivation with advanced age. A report
from a teaching hospital in Riyadh has outlined the
diagnostic challenges of tuberculosis in the elderly
(Dahmash et al. 1995). Dahmash et al. reported on
80 patients older than 60 years admitted with tuber-
culosis between 1988-1993. Pulmonary tuberculosis
accounted for 80% and miliary tuberculosis in 20%
of these patients. Diagnosis was delayed in 23% of
patients, and 21 % died during the same admission.
Of the cohort that had follow-up in the same hospi-
tal, 21 % died (9 of 43), seven had miliary tubercu-
losis and two had pulmonary tuberculosis. Fatality
occurred within the first two months of therapy.
Another problem in elderly patients with tuberculo-
sis is intolerance to antituberculous chemotherapy. In
their series, Dahmash et al. reported adverse effects
in 18% of the patients.
4.3.2
Diabetics
Diabetes mellitus is higWy prevalent in Saudi Arabia.
Prevalence in the general population ranges between
7% in males of rural areas to 15% in females from
urban areas. It increases with age, and one of every two
females above 50 years had diabetes (Al-Nuaim 1997).
The association of tuberculosis and diabetes mellitus
is well documented and is directly proportional to
al. (1991) noted that out of 1566 patients with pulmo-
nary tuberculosis at Riyadh Chest Hospital between
1983 and 1987,136 (8.7%) had diabetes mellitus. The
proportion was higher for Saudi patients (13%) com-
pared with non-Saudi patients (5%) with tuberculosis.
However, prevalence of diabetes mellitus in patients
with pulmonary tuberculosis does not seem to be
higher than in the general population (AI-Nuaim
1997). Over a 4-year period starting in 1995, diabetes
mellitus was found in 6.2-15.5% of active tuberculosis
seen in Riyadh Chest Hospital (AI-Hajjaj 2000).
4.3.3
Persons Infected with Human Immunodeficiency
Virus
The Human Immunodeficiency Virus (HIV) and its
Acquired Immunodeficiency Syndrome (AIDS) global
epidemic have had a profound impact on global tuber-
culosis (Narain et al. 1992). Saudi Arabia is no excep-
tion when it comes to interaction between HIV and M.
tuberculosis. In fact, the high prevalence of tuberculosis
infection at an early age makes the impact of HIV/AIDS
on tuberculosis reactivation more pronounced.
Among the HIV/AIDS patients at King Faisal
Specialist Hospital and Research Centre, 11 out of
233 patients had culture-proven tuberculosis, giving
a rate of 4.7% for reactivation. Some patients pre-
sented with tuberculosis before they were known to
have HIV infection. As of 1999,7 of 171 HIV positive
patients (4.1%) had tuberculosis (Aljurf et al. 1999).
Screening of all tuberculosis patients for HIV infec-
tion has been recommended.
Epidemiology of Tuberculosis in Saudi Arabia
4.3.4
Hemodialysis Patients
As of 1999, there were 5706 patients receiving hemo-
dialysis in Saudi Arabia; 2084 patients started hemo-
dialysis in 1999 (Jondeby et al. 2001). These patients
started recently on hemodialysis are older on average,
and 16% of them have diabetes mellitus. In fact, in
1999-initiated patients alone, 40% of them have diabe-
tes mellitus. Uremia associated impaired cell mediated
immunity is well documented (Wilson et al. 1965). In
1974, the first report of tuberculosis in dialysis patients
was published (Pradhan et al. 1974). This increased
incidence continues until the late 1990s (Cengiz 1996;
Murthy and Pereira 1997). In 1990, the first report from
Saudi Arabia on tuberculosis in hemodialysis patients
was published by AI-Khader et al. (1990) Forty-two out
of fifty dialysis centers in the Kingdom responded to a
questionnaire. Tuberculosis was diagnosed in 4.4% of
1450 patients on chronic hemodialysis. Around fifty
percent had lymphadenitis and less than one third
had pulmonary involvement. On average, tuberculosis
patients were older, mean age 42 years compared with
general dialysis patients, mean age of 33 years. The
report did not indicate the period on hemodialysis
that preceded tuberculosis diagnosis. In 1991, Mit-
walli reported on tuberculosis in dialysis patients at
one hospital in Riyadh (Mitwalli 1991). Patients were
diagnosed with tuberculosis after either one or two
months on dialysis. Of 26 dialysis patients, 7 (27%)
were diagnosed with tuberculosis, predominantly,
lymphadenitis (57%), and only one with pulmonary
tuberculosis as pleurisy. From another dialysis center
in Tau, western province that provides hemodialysis to
325 patients, 30 patients (9.2%) had tuberculosis. Again
extrapulmonary tuberculosis was predominant (70%)
mostly lymphadenopathy (57%). Pulmonary tubercu-
losis accounted for 30% of tuberculosis cases (Hussein
et al.1990, 1994) Frequently, pulmonary tuberculosis in
hemodialysis patients is associated with pleurisy. From
Table 4.3. Number and percentage of dialysis patients with tuberculosis in Saudi
Arabia
Author, Reference year Years Region Total on
covered dialysis
Al-Khader*, Mixed 1450
(Al-Khader et aI. 1990)
Mitwalli (1991) 1989-1990 Riyadh 26
Hussein et al. (1994) 1980-1993 Al-Hada 325
Al-Homrany (1997) 1986-1993 Abha 270
Shohaib et al. (1999) 1992-1997 Jeddah 210
Total 2281
* From a survey of 42 dialysis centers, some cases may be reported in other reports
53
a center in Abha, southwest of the country, tuberculosis
was diagnosed in 13 patients (4.8%) out of 270 dialysis
patients after a mean of 19 months on dialysis. Pul-
monary tuberculosis was found in 8 patients (62%) in
contrast to other reports finding higher percentages of
extrapulmonary tuberculosis (al-Homrany 1997). From
Jeddah, two dialysis centers reported 17 tuberculosis
cases among 210 dialysis patients with a prevalence
rate of 8.1% (Shohaib et al. 1999). Again, as in Abha,
ten patients had pulmonary involvement and nine
had lymph node involvement. Table 4.3 summarizes
the numbers of patients on dialysis and tuberculosis
in various published reports. Hemodialysis patients
who developed tuberculosis and completed therapy
did not relapse after kidney transplantation (Hussein
et al. 1996). Reactivation oflatent infection seems to be
the role rather than a primary infection. However, this
depends on the prevalence of disease in the community.
The other co-morbid illnesses in hemodialysis patients,
such as viral hepatitis, diabetes, and hypertension make
chemotherapy in this population complex and in need
of close monitoring. PPD test in patients on hemodialy-
sis with tuberculosis is unreliable. Reports of positive
test rates ranged from zero to 62% of patients. High
level of clinical suspicion for tuberculosis is essential
for prompt diagnosis and therapy institution.
A unique report has prospectively evaluated the risk
of tuberculosis in patients with renal impairment not
requiring dialysis (AI Shohaib 2000). Al Shohaib has
prospectively followed 80 patients with a creatinine
clearance of 40 ml/min or less for 3 years in Jeddah. At
enrollment, all patients had a normal chest X-ray exam-
ination and 50% of patients had a positive PPD test of
more than 10 mm induration. During follow-up, eight
patients (10%) developed tuberculosis, four pulmo-
nary, two renal, one meningitis, and one adenitis. Four
of them had a positive PPD test at enrollment. AIl had
an excellent response to a nine-month course except
the meningitis case, who was treated for one year. This
study has provided valuable data on patients with renal
Patients with Percent
tuberculosis
64 4.41
7 26.92
30 9.23
13 4.81
17 8.1
131 5.74
54
impairment without dialysis and risk for tuberculosis.
However, time of tuberculosis reactivation in relation
to renal impairment course was not indicated, and no
attempt was made to identify risk factors for reactiva-
tion. An indirect estimate for the tuberculosis reactiva-
tion rate in this population can be made at one case for
every 30 persons per year. This is equal to 3300 cases
annually per 100,000 population, a rate more than 51
times the general population in Jeddah (Milaat et al.
1994). This study also provides evidence that PPD test
is unreliable in this population to identify patients at
risk for tuberculosis-reactivation.
4.3.5
Cancer and Bone Marrow Transplant Patients
Neoplasia constitute a major immune compromise
with and without cytotoxic therapy. In hematologic
malignancies, tuberculosis reactivation has been
reported from developed countries. However, data
from developing countries are limited. The King
Faisal Specialist Hospital and Research Centre in
Riyadh has been performing bone marrow transplan-
tation and tuberculosis was diagnosed in 4 (0.6%) of
641 adult recipients of bone marrow transplantation
between 1986 and 1997 (Aljurf et al. 1999).
Diagnosis in these severely immunocompromised
hosts was delayed. Two patients died before tuber-
culosis was diagnosed. Rates of tuberculosis in bone
marrow transplantation patients in developed coun-
tries ranged between 0.1% and 2.0%. For some reason,
tuberculosis reactivation rates in bone marrow trans-
plantation patients in Saudi Arabia are lower than
expected given the prevalence rate in the country and
degree of immune suppression in these patients.
Another group of cancer patients in which tuber-
culosis rates have been poorly evaluated is lym-
phoma patients. They have a considerable degree of
immunosuppression, even prior to chemotherapy.
Lymphoma is one of the most common cancers in
Saudi Arabia (Cancer registry report: KFSH-RC).
There are no reports on lymphoma cohorts who
developed tuberculosis to assess the reactivation
risk. Among a group of 3921 lymphoma patients at
King Faisal Specialist Hospital and Research Centre,
culture-proven tuberculosis was documented in
26 patients (0.7%). Rates were similar between
Hodgkin's and non-Hodgkin's disease. Pulmonary
tuberculosis was noted in ten patients, extrapul-
monary in thirteen patients and three had dissemi-
nated disease. Diagnosis was delayed and mortality
was high at 46%. Three of the twelve mortalities
A. A. Alrajhi and A. M. AI-Barral<
were diagnosed to have tuberculosis post-mortem
(Neamatallah et al. 2000).
4.3.6
Solid Organ Transplant Recipients
The immunosuppressive therapy used in solid organ
transplant patients enhances the chances for reac-
tivation of tuberculosis (Barber and Sugar 1994).
This has been documented from a kidney transplant
series at King Faisal Specialist Hospital and Research
Centre in Riyadh. Qunibi et al. (1990) reviewed 403
kidney-transplanted patients over 9 years. Tubercu-
losis was documented in 14 cases, a rate of 3.5%.
This rate was thought to be 50 times higher than the
annual incidence rate. Pulmonary tuberculosis was
diagnosed in 29% of the cases, disseminated in 64%.
One case acquired infection from the transplanted
kidney. At the same institution, King Faisal Specialist
Hospital and Research Centre, two liver transplant
patients developed tuberculosis out of a cohort of 72
patients. The rate is 2.7%, close to the rate in kidney
transplant recipients (A. Alrajhi, unpublished data).
4.4
Mycobacteria Other than M. tuberculosis
Mycobacterial species other than M. tuberculosis
(MOTT) has not been studied well in Saudi Arabia.
This could be due to non-culture methods of iden-
tifying tuberculosis in some centers, difficulty in
identifying non-tuberculosis mycobacterium, or
considered as contamination. Zaman found 9%
(47 isolates) of mycobacterium isolates are non-
tuberculosis at National Guard Hospital in Jeddah
over 2 years (Zaman 1991a). M. fortuitum and M.
chelonei were the most prevalent. Respiratory and
urogenital tracts were most frequently affected. At
King Faisal Specialist Hospital and Research Centre
in Riyadh, MOTT was found in 396 (64%) of the 620
mycobacteria isolates over a 3-year period between
1987 and 1989 (Brodie and Qadri 1990). M. avium-
intracellulare complex was cultured most frequently.
Respiratory specimens were the most frequent source
of positive source. This high incidence is probably
due to bias in patient population selection in this
tertiary care center, which treats transplanted and
malignancy patients.
There is a need for reliable data collection to be
able to establish the true incidence and prevalence of
Epidemiology of Tuberculosis in Saudi Arabia
MOTT in Saudi Arabia. Progress in health care has
led to an increase of numbers in immunocompro-
mised patients who are predisposed to these organ-
isms. Specimens processed for mycobacteria need to
be cultured and processed even if they are MOTT.
4.5
Conclusion
Saudi Arabia has been in the middle percentage
countries affected by tuberculosis. The incidence
of reported pulmonary tuberculosis continued to
decrease steadily during the 1980s and the 1990s.
However, extra-pulmonary tuberculosis has resurged
late in the 1990s. The incidences around the Kingdom
differ according to area of high traffic of a population
during a short time, e.g., Jeddah area and Hajj or a con-
centration of a certain population. There is alarming
data of increasing multi-drug resistance tuberculosis.
There is an increasing demand for nationwide
studies on incidence and prevalence of tuberculosis
in different populations in the new centuries in order
to asses the effect of the national tuberculosis pro-
gram that was established last decade.
References
Abdullah A et al (1991) Tuberculosis epidemiology survey:
Mantoux test results from Central region of Saudi Arabia.
Saudi Med J 12:107-110
Al-Amoud M (1993) The history of tuberculosis. Saudi Med
J 14:515-520
AI-Dabbagh A, El-Deeb H, AI-Baghdadi T (1991) The radio-
graphic features of pulmonary tuberculosis observed for
the Western region of Saudi Arabia. Ann Saudi Med 11:
194-200
Al-Hajjaj M (2000) The outcome of tuberculosis treatment
after implementation of the national tuberculosis control
program in Saudi Arabia. Ann Saudi Med 20:125-128
AI-Hajjaj M et al (1991) Pulmonary tuberculosis in Saudi Arabia:
a retrospective study of 1566 patients. Ann Saudi Med 11:
443-447
Al-Homrany M (1997) Successful therapy of tuberculosis in
hemodialysis patients. Am J NephroI17:32-35
Al-Jama A, Borgio F, Al-Qatari K (1999) Patterns of resistance
to antituberculous drugs in Eastern Province, Saudi Arabia.
Saudi Med J 20:927-930
Aljurf M et al (1999) Mycobacterium tuberculosis infection
in allogeneic bone marrow transplantation patients. Bone
Marrow Transplant 24:551-554
Al-Kassimi F (1994) Review of tuberculosis in Saudi Arabia.
Saudi Med J 15:192-195
Al-Kassimi F et al (1991) Mantoux Reaction Survey conducted
55
in the Northern region of Saudi Arabia. Ann Saudi Med
11:315-321
Al-Kassimi F et al (1993) Nationwide community survey of
tuberculosis epidemiology in Saudi Arabia. Tubercle Lung
Dis 74:254-260
AI-Khader A et al (1990) Prevalence of tuberculosis in the
dialysis population in Saudi Arabia. Saudi Kidney Dis
Transplant Bull 1:155-157
Al-Orainey I (1986) Resistance to standard antituberculous
drugs in Saudi Arabia. Saudi Med J 7:363-368
Al-Orainey let al (1989) Resistance to antituberculosis drugs
in Riyadh, Saudi Arabia. Tubercle 70:207-210
Alrajhi AA, et al (2002) Risk factors for drug-resistant Myco-
bacterium tuberculosis in Saudi Arabia. Saudi Med J 2002;
23:305-310
Al Shohaib S (2000) Tuberculosis in chronic renal failure in
Jeddah. J Infect 40:150-153
AI-Wabel A et al (1995) Spectrum of pulmonary tuberculosis in
the Asir Region of Saudi Arabia. Saudi Med J 16:105-107
Al-Wabel A et al (1997) Symptomatology and chest roent-
genographic changes of pulmonary tuberculosis among
diabetics. East Afr Med J 74:62-64
Alzeer A et al (1998) Tuberculosis is the commonest cause of
pneumonia requiring hospitalization during Hajj (pilgrim-
age to Makkah). J Infect 36:303-306
Barber T, Sugar MA (1994) Mycobacteriossis and nocardiosis
in the immunocompromised host. In: Robin RH, Young LS
(eds) Clinical approach to infection in the compromised
host. Plenum, New York, pp 251-253
Bener A (1990) Prevalence of tuberculosis infection in ABHA
and BAHA. Eur J EpidemioI6:376-381
Bener A, Abdullah A (1993) Reaction to tuberculin testing in
Saudi Arabia. Indian J Publ Health 37:105-110
Boucot K, Cooper D, Dillon E et al (1952) Tuberculosis among
diabetics: the Philadelphia survey. Rev Tuberculosis 65:127
Braun M et al (1991) A retrospective cohort study of the risk of
tuberculosis among women of childbearing age with HIV
infection in Zaire. Am Rev Respir Dis 143:501-504
Brodie L, Qadri SM (1990) Isolation of mycobacteria other
than that of tuberculosis at a major tertiary care hospital
in Saudi Arabia. Med Sci Res 18:257-258
Cengiz K (1996) Increased incidence of tuberculosis in
patients undergoing hemodialysis. Nephron 73:421-424
Chowdhury M, Kambal A (1992) Mycobacterial resistance in
a general hospital in Riyadh Saudi Arabia. Med Sci Res 20:
21-22
Dahmash N, Chowdhury M (1994) Re-evaluation of pneu-
monia requiring admission to an intensive care unit: a
prospective study. Thorax 49:71-76
Dahmash N et al (1994) Infections in critically ill patients:
experience in MICU at a major teaching hospital. Infec-
tion 22:264-270
Dahmash N et al (1995) Diagnostic challenge of tuberculosis
of the elderly in hospital: experience at a university hospi-
tal in Saudi Arabia. J Infect 31:93-97
El-Kassimi F (1994) The annual risk of tuberculosis infection
used as predictor of the future incidence of smear-positive
cases. Respir Med 88:589-592
El-Kassimi F et al (1991) Tuberculin survey in the Eastern
Province of Saudi Arabia. Respir Med 85:111-116
Ellis M et al (1996) High proportion of multi-drug resistant
Mycobacterium tuberculosis in Saudi Arabia. Scand J Infect
Dis 28:591-595
56
EI-Sheikh S et al (1998) Bacteria and viruses that cause respi-
ratory tract infections during the pilgrimage Haj season in
Makkah, Saudi Arabia. Trop Med Int Health 3:205-209
Farag M et al (1995) National immunization coverage survey
Saudi Arabia, 1991. J Trop Pediatr 41:S59-S67
Froude J, Kingston M (1982) Extrapulmonary tuberculosis
in Saudi Arabia, a review of 162 cases. King Faisal Special
Hosp Med J 2:85-95
Hammam H, Kamel L, Hidayat N (1979) Third Saudi Medical
Conference, Dammam
Hussein M, Bakir N, Roujouleh H (1990) Tuberculosis in
patients undergoing maintainence dialysis. Nephrol Dial
Transplant 5:584-587
Hussein M et al (1994) Observations in a Saudi-Arabian
dialysis population over a 13-year period. Nephrol Dial
Transplant 9: 1072-1076
Hussein M, Mooij J, Roujouleh H (1996) Tuberculosis in hemo-
dialysis patients. Saudi J Kidney Dis Transplant 7:6-9
Jarallah J et al (1992) High rate of rifampicin resistance of
Mycobacterium tuberculosis in the Taif region of Saudi
Arabia. Tubercle Lung Dis 73:113-115
Jondeby M et al (2001) Caring for hemodialysis patients in
Saudi Arabia. Saudi Med J 22:199-204
Khan M et al (2001) Increasing resistance of M. tuberculosis
to anti-TB drugs in Saudi Arabia. Int J Antimicrob Agents
17:415-418
Mackay A, Cole R (1984) The problem of tuberculosis in the
elderly. Quart J Med 212:497-510
Milaat W et al (1994) Epidemiology of tuberculosis in Jeddah
region, Saudi Arabia. Saudi Med J 15:133-137
Ministry of Health (1979) Tuberculosis. First Annual Health
Report, 1979. Ministry of Health, Riyadh, pp 199-208
Ministry of Health (1991) Tuberculosis. Annual Health Report,
1991. Ministry of Health, Riyadh, pp 252-258
Ministry of Health (1996) Tuberculosis. Annual Health Report,
1996. Ministry of Health, Riyadh, pp 35-37
Ministry of Health (1997) Tuberculosis. Annual Health Report,
1997. Ministry of Health, Riyadh, pp 46-49
Mitwalli A (1991) Tuberculosis in patients on maintenance
dialysis. Am J Kidney Dis 18:579-582
Mohamed A, Evans D (1987) The spectrum of pneumonia in
1983 at the Riyadh Armed Forces Hospital. J Infect 14:31-37
Mokhtar A, Salman K (1983) Extrapulmonary tuberculosis.
Saudi Med J 4:317-322
Murthy B, Pereira B (1997) A 1990 s perspective of hepatitis
C, human immunodeficiency virus, and tuberculosis infec-
tions in dialysis patients. Semin NephroI17:346-363
Narain 1, Raviglione M, Kochi A (1992) HIV-associated tuber-
A. A. Alrajhi and A. M. AI-Barrak
culosis in developing countries: epidemiology and starte-
gies for prevention. Tubercle Lung Dis 73:311-321
Neamatallah A, Alrajhi N, Alrajhi A (2000) Mycobacterium
tuberculosis infection during lymphoma chemotherapy.
10th European congress of clinical microbiology and
infectious diseases, Stockholm, Sweden
Pandya L et al (1990) Lower-lung-field tuberculosis in Saudi
Arabia. Ann Saudi Med 10:374-377
Pandya L et al (1991) Pulmonary tuberculosis in diabetic
patients. Ann Saudi Med 11:293-296
Pradhan R et al (1974) Tuberculosis in dialyzed patients. JAMA
229:798-800
Qunibi W et al (1990) Mycobacterial infection after renal
transplantation - report of 14 cases and review of the
literature. Quart J Med 77:1039-1060
Schein M, Huebner R (1995) Tuberculin skin testing. McGrraw-
Hill, New York, pp 73-75
Schiott C et al (1985) Incidence of drug resistance amongst
isolates of Mycobacterium tuberculosis recovered in the
Gizan area, Saudi Arabia. Saudi Med J 6:375-378
Selwyn P et al (1989) A prospective study of the risk of tuber-
culosis among intravenous drug users with human immu-
nodeficiency virus infection. N Engl J Med 320:545-550
Shanks N, Khalifa I, Al Kalai D (1983) Tuberculosis in Saudi
Arabia. Saudi Med J 4:151-156
Shohaib S, Scrimgeour E, Shaerya F (1999) Tuberculosis in
active dialysis patients in Jeddah. Am J NephroI19:34-37
Styblo K (1980) Recent advances in epidemiological research
in tuberculosis. Adv Tuberculosis Res 20:1-63
Thompson N et al (1979) The booster phenomenon in serial
tuberculin testing. Am Rev Respir Dis 119:587-597
WHO (1964) WHO technical report series no 290. World
Health Organization, Geneva
WHO (1997) Anti-tuberculosis drug resistance in the world.
The WHO/IUATLD global project on anti-tuberculosis
drug resistance surveillance 1994-1997. World Health
Organization, Geneva
Wilson W, Kirkpatrick C, Talmage D (1965) Suppression of
immunologic responsiveness in uremia. Ann Intern Med
62:1-8
World Bank W (1993) World development report 1993: invest-
ing in health. Oxford University Press, Oxford
Zaman R (199Ia) Tuberculosis in Saudi Arabia: epidemiology
and incidence of Mycobacterium tuberculosis and other
mycobacterial species. Tubercle 72:43-49
Zaman R (1991b) Tuberculosis in Saudi Arabia: initial and
secondary drug resistance among indigenous and non-
indigenous populations. Tubercle 72:51-55
5 The Molecular Epidemiology of Tuberculosis
KATHRYN DERIEMER and CHARLES 1. DALEY

CONTENTS M. tuberculosis for the purpose of tracking strains

in the community and designing prevention and
5.1 Introduction 57
control strategies to block further transmission of
5.2 Genotyping Methods 57
5.3 Commonly Used Secondary Typing Methods 59 M. tuberculosis are now available.
5.3.1 PGRS Typing 59 Molecular epidemiology combines the strain-
5.3.2 Spoligotyping 59 typing capabilities of genotyping with conventional
5.3.3 Other PCR-Based Methods 59 epidemiologic approaches in order to study the dis-
5.3.4 Comparisons of Different Typing Methods 60
tribution and determinants of disease occurrence.
5.3.5 Methodological Issues in Interpreting
IS6110 RFLP Typing 60 Molecular techniques can help address epidemiolog-
5.3.6 Whole Genome and Microarray Technologies 62 ical problems that cannot be approached or would be
5.4 Lessons Learned Using Molecular more difficult to address by conventional methods.
Epidemiology 62 Epidemiologic investigations that incorporate geno-
5.4.1 Contact and Outbreak Investigations 62
typing of M. tuberculosis have been used to provide
5.4.2 Community Epidemiology and Tuberculosis
Control 65 novel information about the spread of tubercle bacilli
5.4.3 Exogenous Reinfection with M. tuberculosis 66 in mini-epidemics and outbreaks, to determine the
5.4.4 Geographical Distribution and Dissemination risk factors for transmission in a community, to
of M. tuberculosis 67 distinguish exogenous re-infection from relapses,
5.4.5 Evaluation of Laboratory Cross-contamination 68
to track the geographic distribution and spread of
5.5 The Future of Molecular Epidemiology 69
References 70 strains ofM. tuberculosis of public health importance,
and to identify instances of laboratory contamina-
tion (Foxman and Riley 2001). This chapter updates
current knowledge about the accomplishments and
future directions of this relatively new field.
5.1
Introduction

The ability to track specific strains of Mycobacterium 5.2

tuberculosis as they spread through a population is
critical for our understanding of the transmission
and pathogenesis of tuberculosis. However, until
recently, the only phenotypic markers that distin-
guished different strains of M. tuberculosis were
drug resistance patterns and mycobacterial phage
typing (Gruft et al. 1984). Molecular genotyping
techniques that allow us to differentiate isolates of
K. DERIEMER, PhD
Senior Research Scientist, Division of Infectious Diseases
and Geographic Medicine, Stanford Medical Center,
Stanford, California, USA
C. 1. DALEY, MD
Associate Professor of Medicine, Division of Pulmonary
and Critical Care Medicine, San Francisco General Hospital,
Room 5K-1, 1001 Potrero Ave, San Francisco, CA 94110, USA
Genotyping Methods
Until recently, drug susceptibility testing and phage
typing were the only bio-markers available for epide-
miological studies of M. tuberculosis, both of which
have serious limitations. Drug susceptibility test
results can change in a strain as the strain acquires
resistance to one or more antimicrobials during
treatment. Bacteriophage typing of M. tuberculosis is
oflimited value for epidemiologic studies as only a few
phage types can be recognized and thus, it cannot ade-
quately distinguish between different strains (Rado et
al. 1915; Grange et al. 1978; Engel 1978; Crawford and
Bates 1984).
During the past decade, several nucleic acid-based
genotyping methods based on the variations in the

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
58 K. DeRiemer and C. L. Daley

sequence in bacterial genomes were developed.
These new methods allow us to accurately distin-
guish between different strains of M. tuberculosis.
The most widely used method is referred to as
restriction fragment length polymorphisms (RFLP)
analysis (Fig. 5.1). Specific restriction endonucleases
cleave the mycobacterial DNA at the sites of specific
repetitive sequences, producing DNA restriction
fragments of different lengths. The restriction frag-
ments can be separated by size using conventional
agarose gel electrophoresis and the banding patterns
of these restriction fragments can be detected by
staining the gel with ethidium bromide (Collins and
de Lisle 1985, 1987; Swaminathan and Matar 1993).
However, the resulting genotype pattern contains
thousands of fragments and is very difficult to read
and interpret. In order to decrease the number of
fragments to be compared, the chromosomal DNA
restriction fragments are probed with specific, repet-
itive, labeled DNA fragments (Eisenach et al. 1988;
Reddi et al. 1988). Only the genomic DNA restriction
fragments that are complementary to and hybridize
with the probe are visible, resulting in fewer bands
and a more readable band pattern.
An internationally accepted, standardized proto-
col for RFLP typing of the M. tuberculosis complex
was published in 1993 and is still used today (van
Embden et al. 1993). This method takes advantage
of a specific, well-characterized, repetitive element
insertion sequence 6110 OS611O) (Thierry et al.
1990). It was originally thought that IS6110 inser-
tions occur randomly (Zhang et al.1992) but this was
disproved when the H37Rv strain of M. tuberculosis
was sequenced (Philipp et al. 1996; Cole et al. 1998).
Between 0 and 25 copies of IS611 0 are found in
almost all strains of the M. tuberculosis complex (van
Soolingen et al.1993; Agasino et al.1998). The position
and number of copies of the IS6110 elements varies
from strain to strain; this variability in the chromo-
some of M. tuberculosis is exploited to distinguish
between different strains (Cave et al.1991; van Soolin-
gen et al. 1991). The RFLP band patterns of strains may
be compared visually or scanned optically by a com-
puterized reading system and matched to a reference
library of strain profiles (Woelffer et al.1996; Heersma
et al.1998). When used in conjunction with standard-
ized international databases and computer-assisted
analysis, this approach allows comparisons of strains
between different laboratories in widely separated
geographical regions.
However, the standardized method of IS6110 RFLP
typing has several disadvantages. It is slow, labor inten-

Mycobacterium tuberculosis Chromosomal DNA Digested DNA

m© A
-..
0 B
-.. ~i
0c1;D IS6110 site
Extract
DNA
0
Digest
DNA
)/
\52)1
\~

Separate by gel J C
electrophoresis ,

[2] 0
Hybridization performed
with labeled IS611 0

• D
-- --
Fig. 5.1. Method ofI56110-based restriction fragment length polymorphisms (RFLP) analysis. Depicted are two strains of Myco-
bacterium tuberculosis, labeled 1 and 2. The location and number of the insertion sequence, 156110, is noted by the small black
rectangles. A. The first step in the process is the extraction of the chromosomal DNA. B. The extracted DNA is then cleaved
with a restriction endonuclease (PvuII). Note that in reality, thousands of fragments are created. C. After digestion, the DNA
fragments are separated according to size by gel electrophoresis. In reality, this process results in thousands of bands that are
nearly confluent on the gel. D. Hybridization with probe for 156110 results in a gel with bands only containing the 156110 element.
Thus, there are fewer bands and the pattern is easier to read. The two strains can be seen to differ in the number of bands (i.e.,
the number of 156110 copies in the genome) and the location of the bands (i.e., the size of the restriction fragments)
The Molecular Epidemiology of Tuberculosis
sive, and technically demanding. It requires a relatively
large amount (e.g. 2 j.1g) of high-quality DNA from
each strain of M. tuberculosis, an amount that can only
be extracted from a large number of bacteria grown
from clinical material. For isolates with less than six
copies of IS6110 (less than six bands in the RFLP pat-
tern), it has relatively poor discriminatory power and
should be supplemented by analyses with other probes.
Finally, sophisticated computer software and technical
support are required to compare, analyze, and interpret
large numbers of IS6110 RFLP band patterns. Despite
these limitations, IS6110 RFLP typing remains the most
commonly used method in molecular epidemiologic
studies of M. tuberculosis and is considered the gold
standard technique.
S.3
Commonly Used Secondary Typing Methods
Several basic assumptions underlie our interpreta-
tion of genotyping results. For example, we assume
that two strains with an identical IS6110 RFLP pat-
tern have a common origin, albeit, perhaps, remote.
In addition, we assume that strains with identical
IS6110 RFLP band patterns with multiple bands are
more likely to have a common origin and be clonal
than are strains with relatively few bands. Strains with
few bands or low copy numbers in the RFLP pattern
could appear identical by chance alone. A secondary
probe, such as the polymorphic guanine/cytosine-
rich repetitive sequences (PGRS) or spoligotyping
can be used to further differentiate between strains
with low copy numbers (Burman et al. 1997b).
Early comparative studies showed that using mul-
tiple molecular techniques in a single study provided
better discrimination between strains and insight for
phylogenetic groupings (Bifani et al. 1999). Today,
most molecular epidemiologic studies rely on IS611 0
RFLP typing and a secondary typing method, such
as PGRS or spoligotyping, for those isolates with less
than six bands in the IS6110 RFLP band pattern.
5.3.1
PGRSTyping
The most common repetitive DNA sequences in the
genome of M. tuberculosis are the PGRS, a triplet
repeat of GTG, and the major polymorphic tandem
repeat (Hermans et al. 1992; Ross et al. 1992; Poulet
and Cole 1995; Chaves et al. 1996). The PGRS typing
59
system uses the polymorphic GC-rich repeat sequence
contained in the recombinant plasmid pTBN12 as a
probe (Cousins et al. 1992; Ross et al. 1992). The main
disadvantage of the PGRS typing system is that some
of the PGRS regions include many non-perfect repeats
and the patterns that PGRS typing generates are com-
plex and difficult to interpret. In addition, PGRS typing
is less discriminatory than IS6110-based RFLP geno-
typing. Despite these limitations, PGRS genotyping is
a commonly used secondary typing method.
5.3.2
Spoligotyping
Spoligotyping is a polymerase-chain-reaction (PCR)-
based method that interrogates a small direct repeat
(DR) sequence with 36-base pair (bp) repeats inter-
spersed with short, unique, non-repetitive sequences
(Hermans et al. 1991; Groenen et al. 1993; Kamer-
beek et al. 1997). All of the unique, non-repetitive
sequences, or "spacers:' between the direct repeats
can be amplified simultaneously using one set of
primers. Strains are differentiated by the number
and position of the spacers that are missing from
the complete spacer set. The complete spacer set
is defined by sequencing this region from a large
number of isolates of M. tuberculosis.
This spacer oligotyping or "spoligotyping" is more
rapid and easier to perform than IS6110 RFLP typing,
making it an excellent tool in resource-poor settings.
However, it has a lower level of discrimination than
IS6110 RFLP typing (Diaz et al.1998; Goyal et al.1999).
Nevertheless, it can be used to validate the IS6110
RFLP typing results, to further differentiate strains
with low IS6110 copy numbers (less than six bands in
the RFLP pattern) (Wilson et al.1998; Bauer et al.1999;
Goyal et al. 1999; Yang et al. 2000; Soini et al. 2001),
and to show more distant phylogenetic relationships
among isolates. As recommended in several studies
(van Soolingen et al. 1995; Stauffer et al. 2000), spoligo-
typing could be used as a preliminary screening step,
to be followed by another typing method with greater
discriminatory power.
5.3.3
Other peR-Based Methods
Relative to IS6110 RFLP genotyping, PCR-based meth-
ods are easier to perform, require small amounts of
genomic DNA, and can even be performed on nonvi-
able organisms or directly from clinical specimens,
60 K. DeRiemer and C. L. Daley

thereby eliminating the lengthy time required to
culture M. tuberculosis (Driscoll et al. 1999; Qian et
al. 1999; Taylor et al. 1999). Several other PCR-based
typing assays have been developed using IS6110 as the
target. Mixer-linked PCR (Haas et al. 1993) ligation-
mediated PCR (Bonora et al.1999), heminested inverse
PCR, IS6110 inverse PCR, IS6110 ampliprinting, and
double-repetitive (DR) element PCR (Friedman et al.
1995a) are among the techniques developed to date.
None of these techniques have been studied or applied
as extensively as IS611O-based RFLP genotyping.
Analysis by variable number tandem repeat
(VNTR) is a PCR-based method that differentiates
between strains by identifying the number and length
of exact tandem repeats present in an isolate, inde-
pendently of the IS6110 polymorphisms (Skuce et al.
2002). A high-resolution typing method based on the
VNTRs of mycobacterial interspersed repetitive units
(MIRUs) based on 12 specific intergenic regions of the
M. tuberculosis genome has been proposed (Mazars et
al. 2001; Supply et al. 2002). In a study of 72 clinical
isolates, the correlation between genetic relationships
inferred from MIRU-VNTR and IS6110-RFLP typing
was highly significant. Relative to IS611 0 RFLP typing,
MIRU-VNTR profiling is fast, appropriate for strains
regardless of their IS611 0 RFLP copy number, and per-
mits rapid comparison of results from independent
laboratories using a binary data classification system.
5.3.4
Comparisons of Different Typing Methods
Despite the development of these different methods,
IS6110 RFLP remains the most widely used geno-
typing method for the molecular epidemiology of
tuberculosis and is the gold standard to which other
methods are compared. In an inter-laboratory study,
Kremer and colleagues compared most of the cur-
rently available molecular typing methods for M.
tuberculosis (Fig. 5.2) (Kremer et al. 1999). Five RFLP
typing methods including IS611O, IS1081, PGRS, the
DR and the (GTGh repeat as probes were highly
reproducible. Of the PCR-based methods that were
compared, VNTR typing, mixed-linker PCR, and spo-
ligotyping were highly reproducible between differ-
ent laboratories. However, the double repetitive ele-
ment PCR, IS611 0 inverse PCR, IS611 0 ampliprinting
and arbitrarily primed PCR were not very reproduc-
ible. The most discriminatory methods were IS6110
RFLP typing and the mixed-linker PCR.
5.3.5
Methodological Issues in Interpreting 156110
RFLPTyping
While the methods for generating IS6110 RFLP pat-
terns are well-standardized, the interpretation of these
patterns depends on the epidemiological question
being asked and the stability of the IS6110 element
in the mycobacterial genome. Genetic markers that
change rapidly obscure epidemiological links, whereas
those that are too stable suggest direct links and relat-
edness where true links do not exist. In well-defined
outbreak settings, some tuberculosis patients may
have isolates whose RFLP band patterns differ by one
or two insertion elements (Fig. 5.3) (Daleyet al. 1992).
This latter observation suggests that as a strain spreads
through the community, the strain is evolving and pro-
duces progeny with subtle variations in their RFLP
patterns and the rate at which these changes occur is

Slrain 156110 DR mixed· VNTR

PGRS spollgotyping (GTG)S
linked peR

2
,
I
103
108
22
I
1
2
1
1
2
, ,,
I
1
2
• 31432
31432
02_
1
•
1
1
2 2
2 20 2 2 2 02_ 2 2
3 12' 3 II 3 I I 3 22233 3 3
3 lMI II I I
••
3 3 3 22233 3

•• •• •5• ••
3

•5•
Sol II 22232
23 II 22232

,
5 '2 5 5 I I 22432 5
5 78 5 5 I I 5 22432 5 5
8 32 I 5 5 I 8 33C33 8 8
5 58 5 5 I I 8 33C33 5 8
eo
7
7
"59'0
7
7
7
7 II I 7
7
22033
22033
7
7
7
7
,
5 I I I I. II 8 8 8 2'033 8 5
8 I I I I I. II 8 8 I 8 21033 8 8
9 35 9 9 I' 9 31333 9 9
9 75 e e
,
9 9 II 31333 9
'0 89 .0 10 10 32333 10 10
'0 S3 10 10 10 32333 10 ,n

Fig. 5.2. Comparison of different genotyping methods. Depicted above are patterns of ten Mycobacterium tuberculosis complex
duplicate samples obtained by various typing methods. The methods are ordered from most to least reproducible (left to right).
The 156110 RFLP patterns are shown as representatives of the Pvull-based RFLP method. Numbers in vertical direction indicate
duplicate samples. (Adapted from Kremer et al. 1999)
The Molecular Epidemiology of Tuberculosis
~
I~ ~ ir.',~.---=-.-,~.~,~,;'~~.':",.."..-,-,-"~"""'--'1Ci
CAS! tAlER
-22 b
-fA
-u
-4.4
- --22
-13
..
-1.1
-o.t
-- u....
-4.A
--
-u
-1.0
-u
-1.1
-0"
Fig. 5.3. Restriction fragment length polymorphisms (RFLP)
patterns of M. tuberculosis isolates obtained from an outbreak
of tuberculosis in an HIV congregate living site (A) and con-
trols living in San Francisco at the same time (B). The first two
lanes are the RFLP patterns of M. bovis, BCG, and M. intra-
cellulare. Cases 1 and 2 (A) were receiving antituberculosis
therapy when they entered the facility. Tuberculosis developed
in the remaining patients while they lived in the facility. Note
the shift of the band in the upper part of lane 11. (Adapted
from Daley et al. 1992)
important (Bifani et al. 1996). At least two studies have
shown that the 156110 RFLP band patterns of different
strains of M. tuberculosis are relatively stable, with a
half-life of approximately 3-4 years (Yeh et al. 1998; de
Boer et al.1999). Therefore, the 156110 element should
be useful for studying transmission of M. tuberculosis
over relatively short periods of time.
To interpret the results of 15611 0 RFLP strain typing
and other typing methods, we assume that epidemio-
logically related strains will have the same genotype
pattern and epidemiologically unrelated strains will
have different patterns. Ideally, strain typing will pro-
vide a clear, objective basis for identifying an outbreak
strain and distinguishing it from epidemiologically
unrelated isolates. Many studies of tuberculosis have
extended these assumptions to define clusters of
patients in the community whose isolates had identi-
cal or nearly identical genotype patterns. Clustering
has often been equated with recent or ongoing trans-
mission, and the factors associated with clustering
have been sought as a means to identify and target
subpopulations with substantial ongoing transmission
61
(Glynn et al.1999a). Since the first large-scale molecu-
lar epidemiologic studies of tuberculosis transmis-
sion using 156110 RFLP typing, it has generally been
accepted that tuberculosis cases with identical 156110
RFLP band patterns that are isolated from two or more
persons in the same community represent recent or
ongoing exogenous infection and rapid progression
to disease. By contrast, patients whose isolates of M.
tuberculosis have genotype patterns that do not match
any other isolates in the community are considered
to be unique. Unique isolates likely represent disease
caused by reactivation of latent tuberculosis infec-
tion (LTBI) (Alland et al. 1994). Molecular techniques
enable us to distinguish tuberculosis due to recent or
ongoing infection versus reactivation of LTBI and to
estimate the fraction of active tuberculosis transmis-
sion in a community.
However, some strains may be identical for reasons
other than recent transmission and there is not always
an epidemiological link between patients whose iso-
lates have identical genotype patterns. For example,
a large proportion of the isolates from newly diag-
nosed tuberculosis cases in a relatively stable, rural
population in the United States showed high degrees
of clustering based on 156110 RFLP typing, but most
of the patients had no discernible contact or other
epidemiological links among themselves (Braden et
al. 1997). Similar findings were reported from Malawi
and Kenya; some strains from a large rural area had
identical genotype patterns, but there were no discern-
ible epidemiological links between the patients (God-
frey-Faussett and Stoker 1992). Combined, such stud-
ies suggest that the epidemiological interpretation of
molecular epidemiologic studies may be confusing in
areas where it likely that ancestral, endemic, well-con-
served strains with stable genotype patterns occur.
The amount of transmission represented by clus-
tering using molecular typing techniques will depend
on the sampling strategy and duration of the study
(Murray 2002; Murray and Narde1l2002b). Undersam-
pling can bias the estimates of the proportion of tuber-
culosis cases that were likely caused by recent tubercu-
losis transmission and it can bias the estimates of the
risk factors associated with clustering. When an early
hospital-based study of tuberculosis transmission in
New York City, USA, was re-examined and adjusted
for potential sampling bias, the revised estimate of
the odds ratio of the association of human immuno-
deficiency virus (HIV) and clustering increased from
2.7 to 23.6 (Murray and Alland 2002a). No study has
100% case ascertainment, and incomplete sampling
will underestimate the proportion of tuberculosis
cases that are clustered, even if the number of tuber-
62
culosis cases included in the study is large (Glynn et
al.1999a,b).
Biases may also be introduced if a molecular
epidemiologic study does not cover an adequate
time period. 1\vo population-based studies that have
been ongoing for more than 10 years (San Francisco
[Small et al. 1994; Jasmer et al. 1999] and the Neth-
erlands [van Soolingen et al. 1999]), showed that the
percentage of clustered strains was high during the
first 2 years and declined thereafter. Clustering based
on less than 2 years of sampling is unlikely to identify
the source and secondary cases in a chain of trans-
mission, and will likely underestimate the amount of
recent transmission.
5.3.6
Whole Genome and Microarray Technologies
Until recently, the main purpose of developing typing
techniques using molecular markers was to differen-
tiate between strains of M. tuberculosis and to iden-
tify chains of transmission. There is rising interest in
identifying the genotype of specific strains and their
correlations with specific clinical manifestations and
other phenotypes, such as increased infectivity, viru-
lence, or hyper-mutability.
DNA sequencing is still too expensive and too
complex to be applied to make direct comparisons
of genomic sequences of multiple strains of M. tuber-
culosis, and it is likely that all strains vary to some
degree at the nucleic-acid base level. However, it is
now possible to analyze short segments of DNA for
sequence similarities and differences. For example,
genomic fragments can be amplified using PCR, and
an automated DNA-sequencing procedure can be
used to directly sequence the PCR-amplified DNA
fragment. This approach allows a DNA fragment of
300- to 500-bp to be sequenced in 24 h (Tracy and
Mulcahy 1991). In the future, improved automation
of target amplification and direct sequence analysis
may make this a practical approach for laboratories.
Strains of M. tuberculosis can also be compared
based on the whole genome sequence using DNA
microarrays and DNA chip technology. These tech-
niques allow simultaneous detection of the variation
at many different genomic sites by analysis of the
amount and specific location of mycobacterial DNA
(Gingeras et al. 1998). Briefly, the genome of a strain
of M. tuberculosis is compared with that of a known,
sequenced reference strain by hybridization on an
array containing thousands of oligonucleotides on a
limited surface. Differences in gene expression and
K. DeRiemer and C. 1. Daley
genomic deletions can be detected relative to the
reference strain, depending on the array. Because
deletions in the M. tuberculosis genome rarely occur
independently at exactly the same chromosomal
locus, they can be considered unique and irrevers-
ible genetic events. The number and distribution of
deletions provide a genomic pattern that can be used
to construct phylogenetic relationships and to deter-
mine whether the loss of specific genes is related to
the phenotype of a strain, such as its transmissibility
or antigenicity (Salamon et al. 2000; Kato-Maeda et
al. 2001).
5.4
Lessons Learned Using Molecular
Epidemiology
Molecular genotyping has revolutionized our abil-
ity to track strains of M. tuberculosis as they spread
through a community. Studies of the molecular
epidemiology of tuberculosis have elucidated both
suspected and unsuspected transmission, identified
weaknesses in conventional contact investigations,
identified risk factors for recent infection with rapid
progression to disease, demonstrated exogenous
reinfection with different strains, and documented
the existence of laboratory cross-contamination as a
cause for false-positive cultures.
5.4.1
Contact and Outbreak Investigations
Conventional tuberculosis contact investigations use
the"stone-in-the-pond" or concentric circle approach
to collect information and to screen household con-
tacts, coworkers, and increasingly distant contacts
for tuberculosis infection and disease (Veen 1992).
However, the concentric circle method may not be
adequate in many out-of-household settings and
there are considerable limitations with this approach.
Molecular epidemiology has helped to identify some
of these limitations.
In low-incidence areas such as San Francisco
(Small et al. 1994), Zurich (Pyffer et al. 1998), and
Amsterdam (van Deutekom et al. 1997), a relatively
small proportion (5-10%) of the cases that had
identical IS6110-based genotyping patterns were
actually identified as a contact by the source case.
This suggests that unsuspected transmission of
tuberculosis occurs and is not easily traced by con-
The Molecular Epidemiology of Tuberculosis
ventional contact tracing investigations. In a 5-year,
population-based study in the Netherlands, contact
investigations of persons in five of the largest clusters
identified epidemiological links between them based
on time, place, and risk factors. However, tuberculosis
transmission also occurred through only short-term,
casual contact that was not easily detected in routine
contact investigations (van Soolingen et al. 1999). In
a high-incidence area in Barcelona, Spain (163 tuber-
culosis cases/lOO,OOO population), there was 61.5%
concordance between the genotype pattern and con-
ventional contact tracing (Solsona et al. 2001). In that
study, the authors concluded that conventional con-
tact tracing was useful for identifying new tubercu-
losis cases, but it did not provide much information
about the chains of transmission and how to block or
prevent them.
In a population-based molecular epidemiological
study in an urban community in the San Francisco
Bay area, 75 (33%) of 221 cases had the same strain
of M. tuberculosis. Thirty-nine (53%) of the seventy-
three patients developed tuberculosis because they
were not identified as contacts ofsource case-patients;
twenty case-patients (27%) developed tuberculosis
because of delayed diagnosis of their sources; and
thirteen case-patients (18%) developed tuberculosis
because of problems associated with the evaluation
or treatment of contacts; and one case-patient (1 %)
developed tuberculosis because of delay in being
elicited as a contact. This study suggested that even
in a community that has implemented the essential
elements of tuberculosis control, ongoing transmis-
sion of M. tuberculosis will continue to produce cases
unless patients are diagnosed earlier and contacts are
more completely identified (Chin et al. 2000).
Some populations, such as the urban homeless,
present unique challenges for those conducting
conventional contact investigations. However, stud-
ies that incorporate genotyping are able to provide
information about the chains of transmission in
these groups (Barnes et al. 1996; Gutierrez et al. 1998;
Kimerling et al. 1998; Lemaitre et al. 1998; Dobbs et al.
2001). A prospective study of tuberculosis transmis-
sion in Los Angeles, California, USA identified 162
patients who had culture-positive tuberculosis, and
interviewed the patients to identify their contacts
and whereabouts (Barnes et al. 1997). Genotyping
using 156110 RFLP typing and PGRS analysis were
performed on all isolates of M. tuberculosis from
these patients. Patients whose isolates had identical
or closely related genotyping patterns were consid-
ered clustered, and the degree of homelessness and
the instance of a patient using daytime services at
63
three shelters were independently associated with
clustering. Ninety-six (59%) of one hundred and
sixty-two patients were in eight clusters, but only
two of the ninety-six clustered persons named others
in the cluster as contacts. Thus, traditional contact
investigations did not reliably identify patients
infected with the same strain of M. tuberculosis.
This study demonstrated that locations at which the
homeless congregate are important sites of tubercu-
losis transmission for both homeless persons, who
sleep at the shelters, and non-homeless persons, who
use the facilities during the day. The authors recom-
mended that strategies to reduce tuberculosis trans-
mission should be based on locations rather than on
personal contacts.
Additional studies support the idea that specific
locations can be associated with recent or ongoing
tuberculosis transmission. A study among predomi-
nantly HIV-seropositive gay men in Houston, Texas,
found that barhopping, or visiting many bars in
the same neighborhood each night, was associated
with being in an IS6110 RFLP cluster, representing
ongoing tuberculosis transmission (Yaganehdoost
et al.1999).A population-based molecular epidemio-
logic study of tuberculosis transmission in Orizaba,
southeastern Mexico, identified clandestine bars as
the main site of ongoing tuberculosis transmission
and bar patrons as the main sources of tuberculosis
transmission in the community (Garda-Garda et al.
2000). In a case-control study of tuberculosis among
non-institutionalized HIV-infected patients in King
County, Washington, USA, six HIV-infected persons
with the same strain of M. tuberculosis had contact
at one or more of three bars and this was the only
identifiable epidemiologic link among them (Tabet
et al. 1994).
Additional tools, such as geographic information
systems (GIS) data, can add to molecular epidemio-
logic studies. In a 30-month prospective, city-wide
study of all tuberculosis cases in Baltimore, Mary-
land, USA, using traditional contact investigations
and IS6110-based genotyping, 46% (84/182) of iso-
lates were clustered and 32% (58/182) were defined as
being recently transmitted. Only 24% (20/84) of clus-
tered cases had an identifiable epidemiologic link of
recent contact. Using GIS data, there was significant
spatial aggregation of the 20 clustered cases with epi-
demiologic links in geographic areas of the city with
low socioeconomic status and high drug use (Bishai
et al.1998). These findings and those described previ-
ously suggest that location-based control efforts may
be more effective than contact tracing for early iden-
tification of cases, at least in some populations.
64
When a second case of tuberculosis is recognized as
part of a contact investigation, it is generally assumed
that the two cases harbor the same strain of M. tuber-
culosis, with one as the source-case and the other as
the secondary case. Several studies have used geno-
typing to determine whether the contacts that had
tuberculosis were, in fact, secondary cases or whether
they were unrelated to the source case. Of 11,211 con-
tacts evaluated in San Francisco, USA, there were 66
pairs of culture-positive index and contact cases (Behr
et al. 1998). In the 54 pairs for whom genotyping was
available, the index and contact cases were infected
with a different strain of M. tuberculosis 30% of the
time. Unrelated infections were more common among
foreign-born, particularly Asian, contacts. The drug
susceptibility pattern was more likely to be discordant
in those pairs who harbored different strains. Similar
findings have been reported from Denmark and the
Netherlands (Sebek 2000). These studies highlight
the fact that secondary cases are not always related to
the presumed index case and that in areas with high
rates of drug resistance, providers should be careful
when assuming that the drug susceptibility patterns
between the pairs are the same.
Genotyping has been particularly useful in defin-
ing the presence and scope of tuberculosis outbreaks
by identifying otherwise unsuspected and unde-
tected transmission in the community. Traditionally,
outbreak investigations depended on contact tracing
in the field. This strategy can be ineffective in tuber-
culosis outbreaks if patients do not live in stable
settings and either do not know or are unwilling to
reveal the names of contacts. By identifying tubercu-
losis patients whose isolates have identical strains of
M. tuberculosis, genotyping can infer epidemiologic
links or connections between individuals and can
highlight locations or settings where tuberculosis
transmission is occurring.
An outbreak of tuberculosis among the residents
of a housing facility for HIV-positive persons in
San Francisco, California, USA, was one of the first
molecular epidemiologic studies that demonstrated
the usefulness of genotyping techniques and helped
to validate the IS611 a-based typing method (Fig. 5.3)
(Daley et al. 1992). The isolates from persons who were
part of the institutional outbreak had identical IS611 a
RFLP banding patterns, but tuberculosis patients who
were unrelated to the outbreak had different strain
typing patterns. This study was timportant because
it validated the IS6110 RFLP strain typing method
in a well-characterized outbreak setting prior to its
application in other epidemiologic studies and public
health settings. The study also demonstrated the speed
K. DeRiemer and C. L. Daley
with which tuberculosis can spread through an HIV-
infected population and the importance of specific
locations as sites of transmission.
Molecular epidemiology studies have demon-
strated the impact that a single patient can have on the
epidemiology of a community. A single patient was
directly or indirectly responsible for 6% of the tuber-
culosis cases in San Francisco, including those in the
HIV residential care facility noted above (Daley et al.
1992). In another report, IS6110 RFLP analysis showed
that a single homeless tuberculosis patient with highly
infectious pulmonary tuberculosis likely infected 42%
(41197) of contacts who were regular customers and
employees of a neighborhood bar that he patronized
and caused disease in 14 (34%) of these contacts. Iso-
lates of M. tuberculosis were available for 12/14 of the
tuberculosis cases; all had identical IS611 aRFLP band
patterns (Kline et al. 1995).
In many instances, outbreak investigations are able
to detect and document unsuspected transmission. For
example, an early study using genotyping ofM. tubercu-
losis documented the spread of infection and the devel-
opment of disease to specific individuals who mixed
with marginalized drug users, mainly restaurant work-
ers (Genewein et al. 1993). Transmission links between
individuals with only casual contact were detected. The
study elucidated the details and dynamics of a com-
munity sub-epidemic of tuberculosis in a way that
went beyond traditional contact tracing and routine
surveillance, and facilitated public health interventions
designed to limit the tuberculosis transmission.
Molecular epidemiologic techniques have confirmed
suspected and unsuspected tuberculosis transmission
in places such as crack houses (Leonhardt et al. 1994),
sites of illegal floating card games (Bock et al. 1998),
schools (Kim et al. 2001; Bauer et al. 2000), hospitals
(Edlin et al. 1992; Haas et al. 1998), and clinics. Geno-
typing has been used in several tuberculosis outbreaks
in prisons to identify the magnitude of transmission
among jail inmates and prisoners, among inmates and
guards, and from persons in the jail or prison to mem-
bers of the community (Ferreira et al. 1996; Chaves et
al. 1997; March et al. 2000; Hanau-Bercot et al. 2000;
Mohle-Boetani et al. 2002). Tuberculosis transmission
has also been demonstrated among groups such as
church choirs (Mangura et al. 1998), interstate trans-
gender social networks (Anonymous 2000b), and renal
transplant patients (Jereb et al. 1993), as well as from
patient to health care provider (Wilkinson et al. 1997)
and from health care providers to patients (Frieden et al.
1996a,b; Ikeda et al.1995). Some quite unusual sources
of tuberculosis transmission have been confirmed
with molecular epidemiologic methods. For example,
The Molecular Epidemiology of Tuberculosis 65

processing contaminated medical waste resulted in
transmission of M. tuberculosis to at least one medical
waste treatment facility worker (Johnson et al. 2000).
Genotyping was also used to document unsuspected
bronchoscopy-related transmission and the cross-con-
tamination ofpatients (Agerton et al.1997; Michele et al.
1997; Ramsey et al. 2002).
In some instances, molecular typing may reveal
outbreaks caused by more than one strain of M.
tuberculosis and therefore, more than one likely
source of infection. In other situations, molecular
typing can effectively rule out an outbreak and thus
avoid an expensive epidemiologic investigation and
inappropriate interventions. Because of its useful-
ness, genotyping of M. tuberculosis has become a
routine tool for outbreak investigations in many
hospitals and health departments.
5.4.2
Community Epidemiology and Tuberculosis
Control
Tuberculosis results from rapid progression from a
recently acquired infection, reactivation of LIBI, or
occasionally from exogenous reinfection (Small and
Fujiwara 2001). Most molecular epidemiology stud-
ies have assumed that the proportion of clustered
isolates in a population reflects the amount of recent
or ongoing transmission of M. tuberculosis. Some
studies established epidemiological links between
clustered tuberculosis cases, inferring that the tuber-
culosis cases are part of a chain of transmission from
a single common source or from several common
sources (Genewein et al. 1993; Small et al. 1994).
The number and proportion of tuberculosis cases
that are in clusters, representing recent or ongoing
tuberculosis transmission, have varied from study
to study (Table 5.1). The frequency of clustering has
varied from 17-18% in low incidence areas such as
Vancouver, Canada and Norway to 30-40% in urban
areas in the United States. Among gold miners in
South Africa, 50% of tuberculosis patients were
found to be in clusters. However, it is difficult to
make meaningful cross-study comparisons because
the studies differ in a number of important ways,
including the population studied, the proportion of
all tuberculosis cases studied, the duration of the
study, the definition of clustering, and the method of
secondary typing used.

Table 5.1. Frequency of clustering and risk factors for clustering in selected studies

Study (period) Location Study population N Ever Risk factors for clustering
clustered

Small et al. 1994 San Francisco, USA Community-based 473 40% Acquired immune deficiency syndrome
(1991-1992) US-born
Alland et al. 1994 New York City, USA Hospital-based 104 38% HIV-seropositive
(1989-1992) Hispanic ethnicity
Younger age
Drug-resistant tuberculosis
Low income
Bishai et al. 1998 Baltimore, USA Community-based 182 46% Intravenous drug use
(1994-1996)
Bauer 1998 Denmark, including Country-based 1549 49% Analysis not performed
(1992-1996) Greenland
van Soolingen et al. Netherlands Country-based 4266 46% Male gender
1999 (1993-1997) Urban residence
Dutch and Surinamese nationality
Long term residence in Netherlands
Godfrey-Faussett et al. South Africa Gold miners 419 50% Treatment failure
2000 (1995) Time spent working in mines
Dahle et al. 2001 Norway Country-based 698 18% Analysis not performed
(1994-1998)
Hernandez-Garduno Vancouver, Canada Community-based 793 17% Canadian-born aboriginals
et al. 2002 (1995-1999) Canadian-born non-aboriginals
Injection drug users
Diel et al. 2002 Hamburg, Germany Community-based 423 34% Alcohol abuse
(1997-1999) History of contact tracing
Unemployment
Lockman et al. 2001 Botswana Community-based 301 42% Imprisonment
(1997-1998)
66 K. DeRiemer and C. L. Daley

The proportion and rate of clustering has been interaction between years of residence in the United
used as an indicator to assess the performance of States and HIV infection; tuberculosis among foreign-
tuberculosis control programs. In an evaluation of born persons was more likely to result from recent
tuberculosis transmission over a seven-year period in transmission among those who were HIV-infected
San Francisco, the number and proportion of clustered and more likely to result from the reactivation of
tuberculosis cases declined, particularly among the LTBI among those who were not infected with HIV
US-born population. This was attributed to the imple- (Geng et al. 2002). The investigators recommended
mentation of targeted tuberculosis prevention and that tuberculosis prevention and control strategies
control programs such as screening high-risk popu- need to be targeted to the large number of foreign-
lations and the implementation of directly observed born persons in New York City who have LTB!.
therapy to ensure high cure rates (Jasmer et al. 1999). The factors associated with clustering may be
A recent study in New York City showed that as tuber- complex. For example, in a study of South Africa gold
culosis case rates fell from recent high levels, the pro- miners, tuberculosis patients who had failed treat-
portion of cases caused by clustered isolates (i.e., cases ment at entry to the study were more likely to be in
of tuberculosis due to recent transmission) decreased. clusters (adjusted OR = 3.41,95% confidence interval
The proportion of tuberculosis cases caused by recent CI 1.25-9.27), and patients with multidrug-resistant
transmission dropped from 63.2% in 1993 to 31.4% in tuberculosis were more likely to have failed tuberculo-
1999 (Geng et al. 2002). They also reported that tuber- sis treatment but less likely to be clustered than those
culosis was unlikely to result from recent transmission with a drug-susceptible strain (OR = 0.27, 95% CI
in persons born outside the United States. By contrast, 0.09-0.83). HIV seropositivity was common (53.6%)
an 8-year study in Greenland showed that the annual but was not associated with clustering. Apparently,
incidence of tuberculosis increased from 85/100,000 persistently infectious individuals who had previously
in 1990 to 172/100,000 in 1997 and the percentage of failed treatment were responsible for one third of the
culture positive tuberculosis cases in RFLP clusters tuberculosis cases in this population. In addition, the
increased to 85%, reflecting micro-epidemics among World Health Organization's targets of70% case detec-
adults and young children in small, isolated settle- tion and 85% cure (World Health Organization 2002)
ments (Soborg et al. 2001). were not sufficient to interrupt tuberculosis transmis-
Conventional epidemiological methods can be used sion; indicators that are more closely linked to the
in combination with molecular typing techniques to rate of ongoing tuberculosis transmission, are needed
identify the risk factors associated with recent infec- (Godfrey-Faussett et al. 2000).
tion and rapid progression to disease (Table 5.1).
Some of the risk factors that are associated with recent
infection are specific to a particular community, 5.4.3
whereas others are common to tuberculosis patients Exogenous Reinfection with M. tuberculosis
in different geographical areas. For example, young
age, being in an ethnic minority group, homelessness, Molecular typing can determine whether a patient
and substance use have been identified in more than with a recurrent episode of tuberculosis has a relapse
one study as being associated with recent infection with the previous strain of M. tuberculosis or exog-
(Small et al. 1994; Alland et al. 1994; Friedman et al. enous reinfection with a new strain (Table 5.2). In one
1995b; Barnes et al.1997; van Deutekom et al.1997). In of the earliest studies using genotyping to examine
San Francisco, for example, among persons less than reinfection, Godfrey-Faussett and colleagues reported
60 years of age, Hispanic ethnicity (odds ratio OR = that the IS6110 RFLP band patterns from patients in
3.3, P = 0.02), black race (OR = 2.3, P = 0.02), birth in Malawi were different in pairs of isolates from four
the United States (OR=5.8, P <0.001) and a diagnosis patients. The isolates were from different body sites
of acquired immune deficiency syndrome (OR = 1.8,P in two patients, and from different episodes in the
= 0.04) were independently associated with being in a other patients (Godfrey-Fausset and Stoker 1992).
cluster (Small et al.1994). Since then, several groups have shown that exog-
In a recent study in New York City (Geng et al. enous reinfection can occur in both immunocom-
2002), birth outside the United States, age of 60 years promised and immunocompetent persons. More
or older, and diagnosis after 1993 were factors inde- recently, researchers have tried to quantify the rate
pendently associated with having a unique strain, at which exogenous reinfection occurs and assess
while homelessness was associated with clustering its potential impact on tuberculosis transmission
or recent transmission. There was also a significant dynamics. In Cape Town, South Africa, where there
The Molecular Epidemiology of Tuberculosis 67

Table 5.2. The frequency of exogenous reinfection in selected studies. NA not available, HIV+ Seropositive for the human immu-
nodeficiency virus (HIV), HIV- Seronegative for HIV, AIDS Acquired immune deficiency syndrome, TB Tuberculosis
Study (period) Study site Study population Patients with Patients Patients with
two episodes of with reinfection
tuberculosis or genotyp-
two isolates ing
Small et aI. 1993a Kings County Hospital, AIDS patients with positive culture 17 6 0
(1987-1991) New York City for> 1 year or
Increasing drug resistance 31 11 4 (36%)
Godfrey-Faussett et aI. Nairobi, Kenya TB recurrences NA 5 1 (20%)
1994 (1989-1991)
Das et aI. 1995 Madras, India Recurrence or 30 30 11 (37%)
isolated positive culture 32 32 29 (91%)
Gilks et aI. 1997 Nairobi, Kenya HIV + and HIV- women 4 4 2 (50%)
(1989-1992)
Sudre et aI. 1999 Switzerland HIV cohort with two isolates 20 20 2 (10%)
(1988-1996)
Van Rie et aI. 1999 Cape Town, South Recurrent TB 48 16 12 (75%)
(1992-1998) Africa
Chaves et aI. 1999 Madrid, Spain HIV infected Spanish inmates 11 9 2 (22%)
(1993-1994) who remained culture positive for
>4 months
Sonnenberg et aI. 2000 Gauteng Province, HIV + and HIV- gold miners 57 48 2 (4%)
(1995) South Africa
Lourenc;o et aI. 2000 Rio de Janeiro, Brazil HIV + patients with multiple 12 12 3 (25%)
(1990-1994) isolates
Caminero et aI. 2001 Gran Canaria Island, Two positive cultures> 12 months 23 18 8 (44)
(1991-1996) Spain apart
Fitzpatrick et aI. Kampala, Uganda; HIV + and HIV - TB recurrences NA 40 9 (23)
2002
Bandera et aI. 2001 Lombardy, Italy Recurrent TB separated >6 months NA 32 5 (16)
(1995-1999)
de Viedma et aI. 2002 Madrid, Spain HIV + and HIV - cases with two 172 43 14 (33)
(1988-1999) isolates> 100 days apart
Kriiiiner et aI. 2002 Tartu, Estonia Treatment failures 35 11 11 (100)
(1994-1999)

is a high incidence of tuberculosis and ongoing
transmission, 16 of 698 patients had more than one
episode of tuberculosis of whom 75% (12/16) had
pairs of isolates of M. tuberculosis with different
IS6110 -based genotyping patterns (van Rie et al.
1999). Episodes of tuberculosis reinfection in areas
with a low incidence of tuberculosis, such as the
Netherlands (de Boer and van Soolingen 2000), the
Canary Islands (Caminero et al. 2001), and Northern
Italy (Bandera et al. 2001), are uncommon compared
with those in high-incidence regions. In the study in
Northern Italy, a four-fold increased risk for reinfec-
tion was observed among immigrant patients com-
pared with Italian subjects. In contrast, a higher risk
of relapse rather than reinfection was evidenced in
HIY-positive subjects and in patients infected with
multidrug-resistant tuberculosis.
A person can be simultaneously infected with more
than one strain of M. tuberculosis. Multiple infections
were demonstrated in a patient in San Francisco, Cali-
fornia (Yeh et al. 1999), in two patients who worked in
a medical-waste processing plant in Washington State
(Braden et al. 2001) and among prisoners in Spain
(Chaves et al. 1999). These observations indicate that
simultaneous infections with multiple strains of M.
tuberculosis occur in immunocompetent hosts and
may be responsible for conflicting drug-susceptibility
results. Knowing whether mixed infection is occurring
is important because the susceptibility patterns of the
two strains may be different and different treatment
regimens may be needed (Niemann et al. 2000). In
addition, some cases of suspected exogenous reinfec-
tion may be due to initial infections that include more
than one strain.
5.4.4
Geographical Distribution and Dissemination
of M. tuberculosis
Surveillance networks incorporating conventional
and molecular epidemiologic techniques were estab-
68
lished during the 1990s by the Centers for Disease
Control and Prevention (CDC) and the European
community to track long-distance dissemination of
tuberculosis and to help suggest rational tuberculosis
prevention and control strategies to interrupt such
transmission (Glynn et al. 1999a; Castro and Jaffe
2002; Crawford et al. 2002). Several of the strains
identified in outbreaks have been associated with
large clusters that are widely dispersed both geo-
graphically and temporally, possibly because they are
either more transmissible or they are more likely to
cause disease once transmitted than are other strains
(Murray and Nardell 2002b).
In an early comparison of the strains present in
New York City and San Francisco, California, USA,
the most widely prevalent strain of M. tuberculosis
from New York City was isolated from only one of
755 patients in San Francisco-a traveling salesman
(Casper et al. 1996). A comparison of the strains of
M. tuberculosis from San Francisco and the East Bay,
two distinct regions separated by the San Francisco
Bay, only 53 of 724 (7.3%) had genotype patterns that
matched at least one isolate from the other region.
Of 375 isolates from San Francisco with unique band
patterns, only nine (2.4%) matched patterns of East
Bay isolates. These population-based data suggest
that in the San Francisco Bay Area, M. tuberculosis
does not rapidly spread across geographic boundar-
ies and tuberculosis control should focus on trans-
mission within defined areas (Bradford et al. 1998).
The most widely reported example of the geo-
graphical dissemination of a particular strain of M.
tuberculosis is that of the "Beijing" and "W strains"
(Bifani et al. 2002; Glynn et al. 2002) A multidrug-
resistant strain of M. tuberculosis called the "W-
strain" caused many cases of tuberculosis and deaths
attributable to tuberculosis among patients and
healthcare workers in nosocomial outbreaks and in
other institutional settings in New York City during
the 1990s (Anonymous 1993; Bifani et at. 1996).
This same strain was later found in other parts of
the United States (Agerton et al. 1999). By the late-
1990s, the "w strain" was recognized as a member
of the "Beijing family" of strains. A study performed
in Beijing, China in 1995 reported that 85% of the
isolates were strains with greater than 66% similarity
among their IS6110 RFLP patterns (van Soolingen
et al. 1995). This "Beijing family" of strains was also
detected in high proportions among strains in other
parts of Asia (Anh et al. 2000), the former Russian
Federation (Kurepina et al. 2000), Estonia (Kriiiiner
et al. 2001), Latin America (Diaz et al. 1998), and
Europe (Niemann et al. 2000).
K. DeRiemer and C. 1. Daley
Beijing strains, including the "w strain" and its
variants, have an insertion of IS611 0 in the genomic
dnaA-dnaN locus (Kurepina et al. 1998). Based on
several early technical studies and a review of 16
studies of the Beijing or W strains that gave results on
their spoligotyping, the W family and Beijing family
strains have an identical, characteristic spoligotype
based on DNA polymorphism in the direct repeat
region that only contain spacers 35-43 (van Soolin-
gen et al. 1995; Bifani et al. 1999; Sola et al. 1999; Soini
et al. 2000; van Crevel et al. 2001; Glynn et al. 2002).
Because this family of strains is so widespread and
is often drug-resistant, there is concern that these
strains are biologically hyper-mutable, hyper-viru-
lent, and have a predilection for acquiring drug resis-
tance. But as of this writing, comparable data are not
available for other, Widespread strains.
There may be a link between geographic location
and the number of IS6110 copies. In low incidence
areas with a high proportion of tuberculosis cases
attributable to reactivation of LTBI and immigration
(e.g. the Netherlands, San Francisco, Switzerland,
and Norway), the IS6110 RFLP patterns are highly
polymorphic. But in some areas of Asia, a high pro-
portion of the isolates of M. tuberculosis contain no
copies or very few copies ofIS611O. Specific clones or
families of clones may be predominant in a specific
geographical area. For example, two distinct popula-
tions of isolates were detected in Guinea-Bissau, a
former Portuguese colony, one West African and the
other European (Kallenius et al.1999). In the western
Pacific region, the isolates from eight countries could
be divided into three groups based on the mode of the
copy number (Park et al. 2000). In Tunisia, 62% of the
isolates belong to three genotype families that share
greater or equal to 65% similarity in the IS611 0 RFLP
patterns and in Ethiopia, 52% of the isolates belong
to four genotype families (Hermans et al. 1995). As
noted in the studies described above, there can be
significant geographic variation in IS611 O-based
genotype patterns. Therefore, it will be important
to evaluate the geographic variation in new typing
systems to determine their utility in different regions
and populations.
5.4.5
Evaluation of Laboratory Cross-contamination
Although laboratory cross-contamination of M. tuber-
culosis isolates was demonstrated during the 1980s
using bacteriophage typing (Jones 1988), genotyping
has become the most widely used tool for determin-
The Molecular Epidemiology of Tuberculosis 69

ing whether or not laboratory cross-contamination sary, and potentially toxic treatment for the patient, and
has occurred. Careless specimen processing, contami- increased costs and burdens for both patients and local
nated reagents (de C Ramos et al.1999), and contami- health care providers (Burman et al.1997a; Burman and
nation with H37Ra, an avirulent reference strain of M. Reeves 2000; Northrup et al. 2002).
tuberculosis that is used for standardization of labora-
tory procedures, can cause laboratory cross-contami-
nation. A small but important proportion of cases
with laboratory cross-contamination were detected 5.5
in every institution that looked for it (Anonymous The Future of Molecular Epidemiology
2000a; Perfecto et al. 2000; Breese et al. 2001; Chang et
al. 2001; Jasmer et al. 2002). As a result, some settings Molecular genotyping, in combination with conven-
routinely use genotyping to evaluate specimens for tional epidemiologic investigations, will continue to be
possible laboratory cross-contamination. used to detect laboratory contamination, to describe
Early investigations focused on the isolates of disease outbreaks, and to identify previously unsus-
M. tuberculosis that were processed together in the pected chains and locations of tuberculosis transmis-
laboratory and had identical IS6110 RFLP patterns, sion in the community. In addition, molecular epide-
but were from at least one otherwise asymptomatic miologic methods will continue to play an important
patient (Small et al. 1993b; Dunlap et al. 1995). In role identifying appropriate public health interven-
New York City, an isolate was collected from every tions and measuring their impact. The development of
patient with a positive culture for M. tuberculosis real-time amplification based genotyping techniques
during a I-month period, including both incident will likely improve our ability to do effective, timely
and prevalent cases, and IS61 10 RFLP analyses were contact and outbreak investigations.
performed. Of the 441 patients, 3% had clinical, labo- Recent sequencing of the genome of M. tuber-
ratory, and RFLP evidence of falsely positive cultures culosis has demonstrated that the IS6110 element
for M. tuberculosis (Frieden et al. 1996b). The geno- is not distributed at random in the genome of the
typing of all M. tuberculosis isolates from a I-year tubercle bacillus (McHugh and Gillespie 1998). The
period in a 700-bed urban hospital in Chicago, USA, insertion sequence, and thus the number of RFLP
revealed only one possible instance of nosocomial bands, is significantly more likely to be present at
transmission and five (2.7%) false-positive M. tuber- some positions than would be expected by chance
culosis cultures from 183 patients (French et al.1998). alone. This may interfere with the epidemiological
Among isolates collected prospectively over 5 years interpretation of relatedness of strains. Improved,
from a municipal health department laboratory, four more precise molecular typing techniques based on
percent (8/199) of the culture-positive patients were specific genomic phenomenon such as point muta-
identified as probable or definite false-positives. tions, polymorphisms, and deletions will likely to be
Laboratory cross-contamination should be sus- developed in the near future.
pected when there is a single positive culture in a clini- Current molecular epidemiologic approaches will
cally well patient with negative acid fast bacillus (AFB) soon intersect with developments in mycobacte-
smears and no other evidence of tuberculosis. If the iso- rial genomics and other disciplines in many excit-
late was cultured within one week of another specimen ing ways. The availability of the complete genome
and has an identical genotype pattern, then laboratory sequence of M. tuberculosis now makes it possible
cross contamination is likely. Laboratory cross-contam- to use a variety of typing techniques that can help
ination should also be suspected when there is a sudden distinguish between different species and clonal
increase in culture positive isolates without an epidemi- groupings of strains with specific phenotypic char-
ological or clinical explanation. Published studies have acteristics such as transmissibility, pathogenicity,
used IS6110 RFLP typing, VNTR typing (Gascoyne- or resistance to antimicrobial agents. Genome-wide
Binzi et al. 2001), or spoligotyping (Nivin et al. 2000) analyses using techniques such as microarrays for
to detect and evaluate laboratory cross-contamination. gene expression profiling or the detection of genomic
Regardless of the typing method used, investigations deletions can be used to evaluate specific strains and
of possible cross-contamination are based on good to determine the genetic basis of important pheno-
communication between the laboratory and clinicians. typic traits. As molecular typing techniques improve,
When laboratory or inter-patient cross-contamination molecular epidemiology will become an increasing
occurs, it can result in an incorrect diagnosis, unneces- important tool in the prevention and control of
sary clinic visits and consultations, lengthy, unneces- tuberculosis, globally.
70
References
Agasino CB et al (1998) Epidemiology of Mycobacterium
tuberculosis strains in San Francisco that do not contain
IS611O.lnt J Tuberculosis Lung Dis 2:518-520
Agerton T et al (1997) Transmission of a highly drug-resistant
strain (strain WI) of Mycobacterium tuberculosis. Commu-
nity outbreak and nosocomial transmission via a contami-
nated bronchoscope. J Am Med Assoc 278:1073-1077
Agerton TB et al (1999) Spread of strain W, a highly drug-
resistant strain of Mycobacterium tuberculosis, across the
United States. Clin Infect Dis 29:85-92
Alland D et al (1994) Transmission of tuberculosis in New York
City. An analysis by DNA fingerprinting and conventional
epidemiologic methods. N Engl J Med 330:1710-1716
Anh DD et al (2000) Mycobacterium tuberculosis in Beijing gen-
otypes emerging in Vietnam. Emerg Infect Dis 6:302-305
Anonymous (1993) Outbreak of multidrug-resistant tuber-
culosis at a hospital- New York City, 1991. MMWR Morb
Mortal Weekly Rep 42:427-433
Anonymous (2000a) Misdiagnoses of tuberculosis resulting
from laboratory cross-contamination of Mycobacterium
tuberculosis cultures - New Jersey, 1998. MMWR Morb
Mortal Weekly Rep 49:413-416
Anonymous (2000b) HIV-related tuberculosis in a transgen-
der network- Baltimore, Maryland and New York City area,
1998-2000. MMWR Morb Mortal Weekly Rep 49:317-320
Bandera A et al (2001) Molecular epidemiology study of
exogenous reinfection in an area with a low incidence of
tuberculosis. J Clin MicrobioI39:2213-2218
Barnes PF et al (1996) Transmission of tuberculosis among the
urban homeless. J Am Med Assoc 275:305-307
Barnes PF et al (1997) Patterns of tuberculosis transmission in
Central Los Angeles. J Am Med Assoc 278:1159-1163
Bauer J et al (1998) Results from 5 years of Nationwide DNA
Fingerprinting of Mycobacterium tuberculosis Complex
isolates in a country with a low incidence of M. tuberculosis
infection. J Clin Micro 36:305-308
Bauer J et al (1999) Usefulness of spoligotyping to discrimi-
nate IS6110 low-copy-number Mycobacterium tuberculosis
complex strains cultured in Denmark. J Clin Microbiol 37:
2602-2606
Bauer J et al (2000) A prospective evaluation of the clinical value
of nation-wide DNA fingerprinting of tuberculosis isolates
in Denmark. Int J Tuberculosis Lung Dis 4:295-299
Behr MA et al (1998) Predictive value of contact investiga-
tion for identifying recent transmission of Mycobacterium
tuberculosis. Am J Respir Crit Care Med 158:465-469
Bifani PJ et al (1996) Origin and interstate spread of a New
York City multidrug-resistant Mycobacterium tuberculosis
clone family. J Am Med Assoc 275:452-457
Bifani PJ et al (1999) Identification of a W variant outbreak of
Mycobacterium tuberculosis via population-based molecu-
lar epidemiology. J Am Med Assoc 282:2321-2327
Bifani PJ et al (2002) Global dissemination of the Mycobacte-
rium tuberculosis W-Beijing family strains. Trends Micro-
bioi 10:45-52
Bishai WR et al (1998) Molecular and geographic patterns
of tuberculosis transmission after 15 years of directly
observed therapy. J Am Med Assoc 280:1679-1684
Bock NN et al (1998) Outbreak of tuberculosis associated with
a floating card game in the rural south: lessons for tubercu-
losis contact investigations. Clin Infect Dis 27:1221-1226
K. DeRiemer and C. L. Daley
Bonora S et al (1999) Comparative evaluation of ligation-
mediated PCR and spoligotyping as screening methods
for genotyping of Mycobacterium tuberculosis strains. J
Clin MicrobioI37:3118-3123
Braden CR et al (1997) Interpretation of restriction fragment
length polymorphism analysis of Mycobacterium tubercu-
losis isolates from a state with a large rural population. J
Infect Dis 175:1446-1452
Braden CR et al (2001) Simultaneous infection with multiple
strains of Mycobacterium tuberculosis. J Infect Dis 33:e42-e47
Bradford WZ et al (1998) Dissemination of Mycobacterium
tuberculosis across the San Francisco Bay Area. J Infect
Dis 177:1104-1107
Breese PE et al (2001) The effect of changes in laboratory prac-
tices on the rate of false-positive cultures for Mycobacte-
rium tuberculosis. Arch Pathol Lab Med 125:1213-1216
Burman WJ, Reeves RR (2000) Review of false-positive cul-
tures for Mycobacterium tuberculosis and recommenda-
tions for avoiding unnecessary treatment. Clin Infect Dis
31:1390-1395
Burman WJ et al (1997a) The incidence of false-positive cul-
ture for Mycobacterium tuberculosis. Am J Respir Crit Care
Med 155:321-326
Burman WJ et al (1997b) DNA fingerprinting with two probes
decreases clustering of Mycobacterium tuberculosis. Am J
Respir Crit Care Med 155:1140-1146
Caminero JA et al (2001) Exogenous reinfection with tuber-
culosis on a European island with a moderate incidence of
disease. Am J Respir Crit Care Med 163:717-720
Casper C et al (1996) The transcontinental transmission of
tuberculosis: a molecular epidemiological assessment. Am
J Publ Health 86:551-553
Castro KG, Jaffe HW (2002) Rationale and methods for the
National Tuberculosis Genotyping Surveillance Network.
Emerg Infect Dis 8:1188-1191
Cave MD et al (1991) IS611O: conservation of sequence in the
Mycobacterium tuberculosis complex and its utilization in
DNA fingerprinting. Mol Cell Probes 5:73-80
Chang CL et al (2001) False-positive growth of Mycobacterium
tuberculosis attributable to laboratory contamination con-
firmed by restriction fragment polymorphism analysis. Int
J Tuberculosis Lung Dis 5:861-867
Chaves F et al (1996) Usefulness of the secondary probe
pTBN12 in DNA fingerprinting of Mycobacterium tuber-
culosis. J Clin MicrobioI34:1118-1123
Chaves F et al (1997) A longitudinal study of transmission of
tuberculosis in a large prison population. Am J Respir Crit
Care Med 155:719-725
Chaves F et al (1999) Evidence of exogenous reinfection and
mixed infection with more than one strain of Mycobac-
terium tuberculosis among Spanish HIV-infected inmates.
AIDS 13:615-620
Chin DP et al (2000) Spread of Mycobacterium tuberculosis
in a community implementing recommended elements of
tuberculosis control. J Am Med Assoc 283:2968-2974
Cole ST et al (1998) Deciphering the biology of Mycobacterium
tuberculosis from the complete genome sequence. Nature
393:537-544
Collins DM,de Lisle GW (1985) DNA restriction endonuclease
analysis of Mycobacterium bovis and other members of the
tuberculosis complex. J Clin Microbiol 21:562-564
The Molecular Epidemiology of Tuberculosis 71

Collins DM, de Lisle GW (1987) BCG identification by Fitzpatrick LK et al (2002) An investigation of suspected
DNA restriction fragment patterns. J Gen Microbiol 133: exogenous reinfection in tuberculosis patients in Kampala,
1431-1434 Uganda. Int J Tuberculosis Lung Dis 6:550-552
Cousins DV et al (1992) Use of a repetitive element isolated Ferreira MM et al (1996) Tuberculosis and HIV infection
from Mycobacterium tuberculosis in hybridization studies among female inmates in Sao Paulo, Brazil: a prospective
with Mycobacterium bovis: a new tool for epidemiological cohort study. J Acquir Immunodef Syst 13:177-183
studies of bovine tuberculosis. Vet MicrobioI37:1-17 Foxman B, Riley L (2001) Molecular epidemiology: focus on
Crawford J, Bates JH (1984) Phage typing of mycobacteria. In: infection. Am J EpidemioI153:1135-1141
Kubica 6P, Wayne LG (eds) The Mycobacteria: a sourcebook, French AL et al (1998) Use of DNA fingerprinting to assess
part A, section 6, Marcel Dekker, New York, pp 123-132 tuberculosis infection control. Ann Intern Med 129:
Crawford JT et al 1M (2002) National tuberculosis genotyp- 856-861
ing and surveillance network: design and methods. Emerg Frieden TR et al (1996a) A multi-institutional outbreak of
Infect Dis 8:1192-1196 highly drug-resistant tuberculosis: epidemiology and clini-
Dahle UR et al (2001) Molecular epidemiology of Myco- cal outcomes. J Am Med Assoc 276:1229-1235
bacterium tuberculosis in Norway. J Clin Microbiol 39: Frieden TR et al (1996b) The molecular epidemiology of
1802-1807 tuberculosis in New York City: the importance of noso-
Daley CL et al (1992) An outbreak of tuberculosis with acceler- comial transmission and laboratory error. Tubercle Lung
ated progression among persons infected with the human Dis 77:407-413
immunodeficiency virus. An analysis using restriction-frag- Friedman CR et al (1995a) Double-repetitive-element PCR
ment-Iength polymorphisms. N Engl J Med 326:231-235 methods for subtyping Mycobacterium tuberculosis clini-
Das S et al (1995) IS6110 restriction fragment length poly- cal isolates. J Clin Microbiol 33: 1383-1384
morphism typing of clinical isolates of Mycobacterium Friedman CR et al (1995b) Transmission of multidrug-resis-
tuberculosis from patients with pulmonary tuberculosis in tant tuberculosis in a large urban setting. Am J Respir Crit
Madras, South India. Tubercle Lung Dis 76:550-554 Care Med 152:355-359
De Boer AS, van Soolingen D (2000) Recurrent tuberculosis Garcia-Garcia M et al (2000) The role of core groups in trans-
due to exogenous reinfection (letter). N Engl J Med 342: mitting Mycobacterium tuberculosis in a high prevalence
1050-1051 community in Southern Mexico. Int J Tuberculosis Lung
De Boer AS et al (1999) Analysis of rate of change of IS6110 Dis 4:12-17
RFLP patterns of Mycobacterium tuberculosis based on Gascoyne-Binzi DM et al (2001) Rapid identification oflabora-
serial patient isolates. J Infect Dis 80:1238-1244 tory contamination with Mycobacterium tuberculosis using
De C Ramos M et al (1999) Extensive cross-contamination of variable number tandem repeat analysis. J Clin Microbiol
specimens with Mycobacterium tuberculosis in a reference 39:69-74
laboratory. J Clin MicrobioI37:916-919 Genewein A et al (1993) Molecular approach to identifying
De Viedma DG et al (2002) Tuberculosis recurrences: route of transmission of tuberculosis in the community.
reinfection plays a role in a population whose clinical! Lancet 342:841-844
epidemiological characteristics do not favor reinfection. Geng E et al (2002) Changes in the transmission of tubercu-
Arch Intern Med 162:1873-1879 losis in New York City from 1990 to 1999. N Engl J Med
Diaz R et al (1998) Molecular epidemiology of tuberculosis 346: 1453-1458
in Cuba outside of Havana, July 1994-June 1995; utility of Gilks CF et al (1997) Recent transmission of tuberculosis in
spoligotyping versus IS6110 restriction fragment length a cohort of HIV-1-infected female sex workers in Nairobi,
polymorphism. Int J Tuberculosis Lung Dis 2:743-750 Kenya. AIDS 11:911-918
Diel R et al (2002) Epidemiology of tuberculosis in Hamburg, Gingeras TR et al (1998) Simultaneous genotyping and spe-
Germany: long-term population-based analysis applying cies identification using hybridization pattern recognition
classical and molecular epidemiologic techniques. J Clin analysis of generic Mycobacterium DNA arrays. Genome
Microbiol 40:532-539 Res 8:435-448
Dobbs KG et al (2001) Value of Mycobacterium tuberculosis fin- Glynn JR et al (1999a) Interpreting DNA fingerprint clusters
gerprinting as a tool in a rural state surveillance program. of Mycobacterium tuberculosis. European concerted action
Chest 120:1877-1882 on molecular epidemiology and control of tuberculosis. Int
Driscoll JR, McGarry MA, Taber HW (1999) DNA typing of J Tuberculosis Lung Dis 3:1055-1060
a nonviable culture of Mycobacterium tuberculosis in a Glynn JR. Vynnycky E, Pine PE (1999b) Influence of sampling
homeless shelter outbreak. J Clin Microbiol 37:274-275 on estimates of clustering and recent transmission of
Dunlap NE et al (1995) Laboratory contamination with Myco- Mycobacterium tuberculosis derived from DNA finger-
bacterium tuberculosis cultures. Am J Respir Crit Care Med printing techniques. Am J EpidemioI149:366-371
152: 1702-1704 Glynn JR et al (2002) Worldwide occurrence of BeijingIW
Edlin BR et al (1992) An outbreak of multidrug-resistant strains of Mycobacterium tuberculosis: a systematic review.
tuberculosis among hospitalized patients with the Emerg Infect Dis 8:843-849
acquired immunodeficiency syndrome. N Engl J Med 326: Godfrey-Faussett, Stoker NG (1992) Aspects of tuberculo-
1514-1521 sis in Africa. 3. Genetic 'fingerprinting' for clues to the
Eisenach KD, Crawford JT, Bates JH (1988) Repetitive DNA pathogenesis of tuberculosis. Trans R Soc Trop Med Hyg
sequences as probes for Mycobacterium tuberculosis. J 86:472-475
Clin Microbiol 26:2240-2245 Godfrey-Faussett P et al (1994) Recurrence of HIV-related
Engel HW (1978) Mycobacteriophages and phage typing. Ann tuberculosis in an endemic area may be due to relapse or
Microbiology 129:75-90 reinfection. Tubercle Lung Dis 75:199-202
72
Godfrey-Faussett P et al (2000) Tuberculosis control and
molecular epidemiology in a South African gold-mining
community. Lancet 356:1066-1071
Goyal M et al (1999) Spoligotyping in molecular epidemiology
of tuberculosis in Ghana. J Infect 38:171-175
Grange JM et al (1978) The correlation of bacteriophage types of
Mycobacterium tuberculosis with guinea-pig virulence and
in vitro-indicators of virulence. J Gen MicrobioI108:1-7
Groenen PM et al (1993) Nature of DNA polymorphism in the
direct repeat cluster of Mycobacterium tuberculosis; appli-
cation for strain differentiation by a novel typing method.
Mol MicrobiollO:1057-1065
Gruft H et al (1984) Phage-typing and drug-resistance pat-
terns as tools in mycobacterial epidemiology. Am Rev
Respir Dis 130:96-97
Gutierrez MC et al (1998) Molecular fingerprinting of Myco-
bacterium tuberculosis and risk factors for tuberculosis
transmission in Paris, France, and surrounding area. J Clin
Microbiol 36:486-492
Haas WH et al (1993) Mixed-linker polymerase chain reacton:
a new method for rapid fingerprinting of isolates of the
Mycobacterium tuberculosis complex. J Clin Microbiol 31:
1293-1298
Haas DW et al (1998) Nosocomial transmission of a drug-sen-
sitive W-variant Mycobacterium tuberculosis strain among
patients with acquired immunodeficiency syndrome in
Tennessee. Infect Contr Hosp EpidemioI19:635-639
Hanau-Bercot B et al (2000) A one-year prospective study
(1994-1995) for a first evaluation of tuberculosis transmission
in French prisons. Int J Tuberculosis Lung Dis 4:853-859
Heersma HF, Kremer K, van Embden JD (1998) Computer
analysis of IS6110 RFLP patterns of Mycobacterium tuber-
culosis. Methods Mol Bioi 101:395-422
Hermans PW et al (1991) Insertion element IS987 from Myco-
bacterium bovis BCG is located in a hot-spot integration
region for insertion elements in Mycobacterium tuberculo-
sis complex strains. Infect Immun 59:2695-2705
Hermans PW, van Soolingen D, van Embden JD (1992) Char-
acterization of a major polymorphic tandem repeat in
Mycobacterium tuberculosis and its potential use in the
epidemiology of Mycobacterium kansasii and Mycobacte-
rium gordonae. J BacterioI174:4157-4165
Hermans PW et al (1995) Analysis of the population struc-
ture of Mycobacterium tuberculosis in Ethiopia, Tunisia,
and The Netherlands: usefulness of DNA typing for global
tuberculosis epidemiology. J Infect Dis 171:1504-1513
Hernandez-Garduno E et al (2002) Predictors of clustering of
tuberculosis in Greater Vancouver: a molecular epidemio-
logic study. Can Med Assoc J 167:349-352
Ikeda RM et al (1995) Nosocomial tuberculosis: an outbreak
of a strain resistant to seven drugs. Infect Contr Hosp Epi-
demioI16:152-159
Jasmer RM et al (1999) A molecular epidemiologic analysis
of tuberculosis trends in San Francisco, 1991-1997. Ann
Intern Med 130:971-978
Jasmer RM et al (2002) A prospective, multicenter study of
laboratory cross-contamination of Mycobacterium tuber-
culosis cultures. Emerg Infect Dis 8:1260-1263
Jereb JA et al (1993) Nosocomial outbreak of tuberculosis in
a renal transplant unit: application of a new technique
for restriction fragment length polymorphism analysis
of Mycobacterium tuberculosis isolates. J Infect Dis 168:
1219-1224
K. DeRiemer and C. L. Daley
Johnson KR et al (2000) Transmission of Mycobacterium tuber-
culosis from medical waste. J Am Med Assoc 284:1683-1688
Jones WD Jr (1988) Bacteriophage typing of Mycobacterium
tuberculosis cultures from incidents of suspected labora-
tory cross-contamination. Tubercle 69:43-46
Kallenius G et al (1999) Evolution and clonal traits of Myco-
bacterium tuberculosis complex in Guinea-Bissau. J Clin
Microbiol 37:3872-3878
Kamerbeek J et al (1997) Simultaneous detection and strain
differentiation of Mycobacterium tuberculosis for diagnosis
and epidemiology. J Clin MicrobioI35:907-914
Kato-Maeda M et al (2001) Comparing genomes within the spe-
cies Mycobacterium tuberculosis. Genome Res 11:547-554
Kim SJ et al (2001) Transmission of Mycobacterium tuberculo-
sis among high school students in Korea. Int J Tuberculosis
Lung Dis 5:824-830
Kimerling ME et al (1998) Restriction fragment length poly-
morphism screening of Mycobacterium tuberculosis isolates:
population surveillance for targeting disease transmission in
a community. Int J Tuberculosis Lung Dis 2:655-662
Kline SE, Hedemark LL, Davies SF (1995) Outbreak of tuber-
culosis among regular patrons of a neighborhood bar. N
Engl J Med 333:222-227
Kremer K et al (1999) Comparison of methods based on
different molecular epidemiologic markers for typing of
Mycobacterium tuberculosis complex strains: interlabora-
tory study of discriminatory power and reproducibility. J
Clin Microbiol 37:2607-2618
Kriiiiner A et al (2001) Spread of drug-resistant pulmonary
tuberculosis in Estonia. J Clin Microbiol 39:3339-3345
Kriiiiner A et al (2002) Use of molecualr techniques to distin-
guish between treatment failure and exogenous reinfection
with Mycobacterium tuberculosis. Clin Infect Dis 35:146-155
Kurepina NE et al (1998) Characterization of the phyloge-
netic distribution and chromosomal insertion sites of
five IS6110 elements in Mycobacterium tuberculosis: non-
random integrtion on the dnaA-dnaN region. Tubercle
Lung Dis 79:31-42
Kurepina N et al (2000) RFLP and DST data on Mycobacterium
tuberculosis cultures isolated from patients in prison and
civilian populations in Russia. European Respir Soc 2000;
2000 Aug 30-Sept 3; Florence, Italy. Abstract 1537
Lemaitre N et al (1998) Use of DNA fingerprinting for primary
surveillance of nosocomial tuberculosis in a large urban
hospital: detection of outbreaks in homeless people and
migrant workers. Int J Tuberculosis Lung Dis 2:390-396
Leonhardt KK et al (1994) A cluster of tuberculosis among
crack house contacts in San Mateo County, California. Am
J Publ Health 84:1834-1836
Lockman S et al (2001) Molecular and conventional epide-
miology of Mycobacterium tuberculosis in Botswana: a
population-based perspective study of 301 pulmonary
tuberculosis patients. J Clin MicrobioI39:1042-1047
Lourenl1o MCS et al (2000) Genotypic patterns of multiple
isolates of M. tuberculosis from tuberculous HIV patients.
Trop Med Int Health 5:488-494
Mangura BT et al (1998) Mycobacterium tuberculosis miniepi-
demic in a church gospel choir. Chest 113:234-237
March F et al (2000) Predictors of tuberculosis transmission
in prisons: an analysis using conventional and molecular
methods. AIDS 14:525-535
Mazars E et al (2001) High-resolution minisatellite-based
typing as a portable approach to global analysis of Myco-
The Molecular Epidemiology of Tuberculosis
bacterium tuberculosis molecular epidemiology. Proc Nat!
Acad Sci USA 98:1901-1906
McHugh TD, Gillespie SH (1998) Nonrandom association of
IS6110 and Mycobacterium tuberculosis: implications for
molecular epidemiological studies. J Clin Microbiol 36:
1410-1413
Michele TM et al (1997) Transmission of Mycobacterium
tuberculosis by a fiberoptic bronchoscope. Identification
by DNA fingerprinting. J Am Med Assoc 278:1093-1095
Mohle-Boetani JC et al (2002) Tuberculosis outbreak in a
housing unit for human immunodeficiency virus-infected
patients in a correctional facility: transmission risk factors
and effective outbreak control. Clin Infect Dis 34:668-676
Murray M (2002) Sampling bias in the molecular epidemiol-
ogy of tuberculosis. Emerg Infect Dis 8:363-369
Murray M, Alland D (2002a) Methodological problems in the
molecular epidemiology of tuberculosis. Am J Epidemiol
155:565-571
Murray M, Nardell E (2002b) Molecular epidemiology of
tuberculosis: achievements and challenges to current
knowledge. Bull WHO 80:477-482
Niemann S et al (2000) Double infection with a resistant and
a multidrug-resistant strain of Mycobacterium tuberculosis.
Emerg Infect Dis 6:548-551
Nivin B et al (2000) Use of spoligotype analysis to detect labo-
ratory cross-contamination. Infect Contr Hosp Epidemiol
21:525-527
Northrup JM et al (2002) Estimated costs of false laboratory
diagnoses of tuberculosis in three patients. Emerg Infect
Dis 8:1264-1270
Park YK, Bai GH, Kim SJ (2000) Restriction fragment length
polymorphism analysis of Mycobacterium tuberculosis iso-
lated from countries in the western Pacific region. J Clin
MicrobioI38:191-197
Perfecto B, Dorronsoro I, Lopez-Goni I et al (2000) Confir-
mation by molecular typing of cross-contamination in a
mycobacteria laboratory. Enferm Infec Microbiol Clin 18:
12-15
Pfyffer GE et al (1998) Transmission of tuberculosis in the
metropolitan area of Zurich: a 3 year survey based on
DNA fingerprinting. Eur Respir J 11:804-808
Philipp WJ et al (1996) An integrated map of the genome of the
tubercle bacillus, Mycobacterium tuberculosis H37 Rv, and
comparison with Mycobacterium leprae. Proc Natl Acad Sci
USA 93:3132-3137
Poulet S, Cole ST (1995) Characterization of the highly abun-
dant polymorphic GC-rich-repetitive sequence (PGRS)
present in Mycobacterium tuberculosis. Arch Microbiol
163:87-95
Qian Let al (1999) Retrospective analysis of the Beijing family
of Mycobacterium tuberculosis in preserved lung tissues. J
Clin MicrobioI37:471-474
Rado TA et al (1975) World Health Organization studies on
bacteriophage typing of mycobacteria. Subdivision of the
species Mycobacterium tuberculosis. Am Rev Respir Dis
111:459-468
Ramsey AH et al (2002) An outbreak of bronchoscopy-related
tuberculosis infections due to lack of bronchoscope leak
testing. Chest 121:976-981
Reddi PP, Talwar GP, Khandekar PS (1988) Repetitive DNA
sequence of Mycobacterium tuberculosis: analysis of dif-
ferential hybridization pattern with other mycobacteria.
Int J Lepr Mycobact Dis 56:592-598
73
Ross BC et al (1992) Molecular cloning of a highly repeated
DNA element from Mycobacterium tuberculosis and its use
as an epidemiological tool. J Clin Microbiol 30:942-946
Salamon H et al (2000) Detection of deleted genomic DNA
using a semi-automated computational analysis of
GeneChip data. Genome Res 10:2044-2054
Sebek M (2000) DNA fingerprinting and contact investigation.
Int J Tuberculosis Lung Dis 4: [2 Suppl 1]:S45-S48
Skuce RA et al (2002) Discrimination of Mycobacterium tuber-
culosis complex bacteria using novel VNTR-PCR targets.
Microbiology 148:519-528
Small PM, Fujiwara PI (200!) Management of tuberculosis in
the United States. N Engl J Med 345:189-200
Small PM et al (1993a) Exogenous reinfection with multi-
drug-resistant Mycobacterium tuberculosis in patients with
advanced HIV infection. N Engl J Med 328:1137-1144
Small PM et al (1993b) Molecular strain typing of Myco-
bacterium tuberculosis to confirm cross-contamination
in the mycobacteriology laboratory and modification of
procedures which minimize occurrence of false-positive
cultures. J Clin Microbiol 31:1677-1682
Small PM et al (1994) The epidemiology of tuberculosis in San
Francisco. A population-based study using conventional
and molecular methods. N Engl J Med 330:1703-1709
Soborg C et al (2001) Doubling of the tuberculosis incidence in
Greenland over an 8-year period (1990-1997). Int J Tuber-
culosis Lung Dis 5:257-265
Soini H et al (2000) Characterization of Mycobacterium
tuberculosis isolates from patients on Houston, Texas, by
spoligotyping. JClin Microbiol 38:669-676
Soini H et al (200!) Transmission dynamics and molecular
charactization of Mycobacterium tuberculosis isolates with
low copy numbers of IS611O. J Clin MicrobioI39:217-221
Sola C et al (1999) Tuberculosis in the Caribbean: using spacer
oligonucleotide typing to understand strain origin and
transmission. Emerg Infect Dis 5:404-414
Solsona J et al (2001) Molecular and conventional epidemiol-
ogy of tuberculosis in an inner city district. Int J Tubercu-
losis Lung Dis 5:724-731
Sonnenberg P et al (2000) Tuberculosis treatment failure and
drug resistance - same strain or reinfection? Trans R Soc
Trop Med Hyg 94:603-607
Stauffer F et al (2000) Implementation of a molecular typing
system to support epidemiological investigations in the
tuberculosis health care system in Vienna. Wien Klin
Wochenschr 112:791-797
Sudre P et al (1999) Molecular epidemiology of tuberculosis
among HIV-infected persons in Switzerland: a country-
wide 9-year cohort study. Infection 27:323-330
Supply P et al (2002) Minisatellite-based linkage disequilib-
rium analysis of strains from a high tuberculosis incidence
area supports clonal evolution of Mycobacterium tubercu-
losis. Abstract book, 5th international conference on the
pathogenesis of mycobacterial infections, 27-30 June 2002;
Saltsjobaden, Stockholm, Sweden. Abstract 02
Swaminathan B, Matar GM (1993) Molecular typing methods.
In: Persing DH et al (eds) Diagnostic molecular microbiol-
ogy: principles and practices. American Society for Micro-
biology, Washington DC, pp 26-50
Tabet SR et al CM (1994) Restriction fragment length poly-
morphism analysis detecting a community-based tuber-
culosis outbreak among persons infected with human
immunodeficiency virus. J Infect Dis 169:189-192
74
Taylor GM et al (1999) Genotypic analysis of Mycobacterium
tuberculosis from midieval human remains. Microbiology
145:899-904
Thierry D et al (1990) IS611O,an IS-like element of Mycobacte-
rium tuberculosis complex. Nucleic Acids Res 18:188
TracyTE, Mulcahy LS (1991) Asimple method for direct automated
sequencing of PCR fragments. BioTechniques 11:68-75
Van Crevel R et al (2001) Mycobacterium tuberculosis Beijing
genotype strains associated with febrile response to treat-
ment. Emerg Infect Dis 7:1-4
Van Deutekom H et al (1997) A molecular epidemiological
approach to studying the transmission of tuberculosis in
Amsterdam. Clin Infect Dis 25:1071-1077
Van Embden JD et al (1993) Strain identification of Mycobacte-
rium tuberculosis by DNA fingerprinting: recommendations
for a standardized methodology. J Clin MicrobioI31:406-409
Van Rie A et al (1999) Exogenous reinfection as a cause of
recurrent tuberculosis after curative treatment. N Engl J
Med 341:1174-1179
Van Soolingen D et al (1991) Occurrence and stability of inser-
tion sequences in Mycobacterium tuberculosis complex
strains; evaluation of insertion sequence-dependent DNA
polymorphism as a tool in the epidemiology of tuberculo-
sis. J Clin Microbiol 29:2578-2586
Van Soolingen D et al (1993) Comparison of various repetitive
DNA elements as genetic markers for strain differentiation
and epidemiology of Mycobacterium tuberculosis. J Clin
MicrobioI31:1987-1995
Van Soolingen D et al (1995) Predominance of a single geno-
type of Mycobacterium tuberculosis in countries of east
Asia. J Clin Microbiol 33:3234-3238
Van Soolingen D et al (1999) Molecular epidemiology of tuber-
culosis in the Netherlands: a nationwide study from 1993
through 1997. J Infect Dis 180:726-736
K. DeRiemer and C. 1. Daley
Veen J (1992) Microepidemics of tuberculosis: the stone-in-
the-pond principle. Tubercle Lung Dis 73:73-76
Wilkinson D et al (1997) Nosocomial transmission of tubercu-
losis in Africa documented by restriction fragment length
polymorphism. Trans R Soc Trop Med Hyg 91:318
Wilson SM, Goss S, Drobniewski F (1998) Evaluation of strat-
egies for molecular fingerprinting for use in the routine
work of a Mycobacterium tuberculosis reference unit. J Clin
MicrobioI36:3385-3388
Woelffer GB et al (1996) A computer-assisted molecular
epidemiologic approach to confronting reemergence of
tuberculosis. Am J Med Sci 311:17-22
World Health Organization (2002) Global tuberculosis con-
trol: surveillance, planning, financing. WHO report 2002.
Geneva, Switzerland, WHO/CDS/TBI2002.295
Yang ZH et al (2000) Spoligotyping and polymorphic GC-rich
repetitive sequence fingerprinting of Mycobacterium tuber-
culosis strains having few copies ofIS6110. J Clin Microbiol
38:3572-3576
Yaganehdoost A et al (1999) Complex transmission dynam-
ics of clonally related virulent Mycobacterium tuberculosis
associated with barhopping by predominantly human
immunodeficiency virus-positive gay men. J Infect Dis
180:1245-1251
Yeh RW et al (1998) Stability of Mycobacterium tuberculosis
DNA genotypes. J Infect Dis 177:1107-1111
Yeh RW, Hopewell PC, Daley CL (1999) Simultaneous infection
with two strains of Mycobacterium tuberculosis identified
by restriction fragment length polymorphism analysis. Int
J Tuberculosis Lung Dis 3:537-539
Zhang Y et al (1992) DNA polymorphisms in strains of
Mycobacterium tuberculosis analyzed by pulsed-field gel
electrophoresis: a tool for epidemiology. J Clin Microbiol
30:1551-1556
6 Molecular Epidemiology of Mycobacterium bovis
ROBIN A. SKUCE and SYDNEY D. NEILL

CONTENTS that includes farmed and feral animals, wildlife and

also humans (O'Reilly and Daborn 1995). Its epi-
6.1 Introduction 75
demiological pattern in developing and developed
6.2 Bovine Tuberculosis 75
6.3 Epidemiology in Developed Countries 76 countries can be very complex. It is classified as a
6.4 Epidemiology in Developing Countries 77 List B disease by the Office International des Epi-
6.5 Epidemiological Typing 78 zootiques (OlE) and considered of socio-economic
6.6 The Genetics of the M. tuberculosis importance as well as public health significance, as
Complex Bacteria 78
it impacts significantly on the international trade of
6.7 Strain Typing Methodology 79
6.8 Standardisation and Method Performance animals and animal products and hence localliveli-
Evaluation 84 hoods (Cousins 2001; Cousins and Roberts 2001).
6.9 M. tuberculosis Molecular Epidemiology:
Lessons Learned 85
6.10 M. bovis Molecular Epidemiology:
Studies and Trends to Date 85
6.11 Conclusions 88 6.2
References 89 Bovine Tuberculosis

Bovine tuberculosis is a chronic infection of animals,

particularly cattle, and is caused by the tubercle bacil-
lus Mycobacterium bovis. It is usually characterised
6.1 by formation of nodular granulomas or tubercles,
Introduction particularly in the lungs and in the lymph nodes of
the respiratory tract but lesions also may occur in
In 1998 the World Health Organisation (Cosivi et al. the mesenteric lymph nodes, liver, spleen and other
1998) estimated that the incidence of human tuber- organs. Presentation of the disease is variable with
culosis would be 88 million, resulting in 30 million an absence of clinical signs in less advanced infec-
deaths for the period 1990-1999 and the majority tion, often the case in the majority of infected cattle
of cases would be in developing countries. In these in developed countries with regular testing strategies
countries, tuberculosis, caused by Mycobacterium (Neill 1994). It may also present non-specifically, with
bovis, is present in animals. However, surveillance clinical signs such as weakness, anorexia, emaciation
and control programmes for animal tuberculosis in and coughing in advanced cases. In most countries
these countries are often inadequate or non-existent. diagnosis of tuberculosis in cattle is usually by tuber-
Consequently, the epidemiology of M. bovis in public culin skin testing, measuring delayed-type hypersensi-
health issues remains largely unknown (Cosivi et al. tivity. However, in countries where prevalence of infec-
1998). M. bovis has an exceptionally broad host range tion is extremely low or freedom from infection has
been declared, meat inspection alone is employed for
diagnosis and surveillance. More recently, application
of assays for bovine interferon-y has been employed
R.A.SKUCE,BSc,PhD to indicate infection (Neill et al. 1994). Supporting
Veterinary Sciences Division, Stoney Road, Stormont, Belfast, laboratory confirmation using bacteriological culture
BT4 3SD, Northern Ireland
techniques is also often employed.
S. D. NEILL, BSc, PhD
Veterinary Sciences Division, Department of Agriculture and Bovine tuberculosis still can have serious, adverse,
Rural Development, Stoney Road, Stormont, Bellfast BT4 3SD, financial consequences in agricultural regions of the
UK world (Bennett et al.1999; Cousins 2001; Cousins and

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
76
Roberts 2001; Goodchild and Clifton-Hadley 2001)
and is considered a potential human health risk in
some countries (Cosivi et al. 1998). Farmers incur
financial losses from decreases in meat and milk
production, carcass condemnations and restrictions
on animal movement. Bovine tuberculosis presents
a serious barrier to cattle trade within and between
countries and there is considerable expenditure, usu-
ally by governments, on testing and compensation.
Most European countries, through the introduction
of voluntary and subsequently compulsory eradica-
tion programmes, experienced dramatic success
in reducing the incidence of bovine tuberculosis.
However, despite initial progress, eradication has not
yet been achieved in all countries (European Com-
mission Report 1998). Some European countries,
e.g., the United Kingdom, have considerable regional
variability with sporadic outbreaks, particularly in
South-West England and Northern Ireland.
Mycobacterium bovis infection in humans may
result from ingestion of contaminated milk, inhala-
tion via aerosols or through contact with infected
clinical material. Pulmonary disease in humans, due
to M. bovis is clinically, radiologically and pathologi-
cally indistinguishable from pulmonary tuberculosis
caused by M. tuberculosis (O'Reilly and Daborn 1995).
However, transmission of M. bovis from human to
human is considered much less efficient than such
transmission of M. tuberculosis (van Soolingen
2001). Evidence for transmission of M. bovis between
humans is considered rare and largely anecdotal and
equally rare are reports of humans infecting cattle
(O'Reilly and Daborn 1995).
Mycobacterium bovis tuberculosis was once
common in children, affecting the cervical lymph
nodes, the intestinal tract or the meninges, as a result
of feeding infected cows milk. By 1937 up to 25% of
tuberculosis cases in the USA and UK were due to M.
bovis infections (Cousins 2001). The introduction of
pasteurisation for milk and milk products, together
with abattoir meat inspections significantly reduced
this incidence and bovine tuberculosis in children
is currently negligible. It was the recognition of the
public health risk that prompted the introduction
of control and eradication programmes for bovine
tuberculosis in many countries in the early part of
the last century (Cousins 2001). The success of such
schemes has resulted in a low prevalence, or absence,
of M. bovis infection in most national cattle herds,
particularly in countries with advanced agriculture
practices. This was reflected in a dramatic drop in
the incidence of human tuberculosis due to M. bovis.
However, the zoonotic nature of M. bovis still places
R. A. Skuce and S. D. Neill
farmers, abattoir workers and veterinarians at risk
(Grange 2001).
6.3
Epidemiology in Developed Countries
In developed regions of the world, several countries
have been unable to eradicate this disease, despite
rigorously implementing comprehensive and costly
strategies. The primary route of infection for cattle
is respiratory but alimentary infection, cutaneous,
congenital and genital routes have been recorded
(Neill 1994). Social behaviour of cattle in an environ-
ment of intensive farming and a high degree of live
cattle trade, together with the potential of M. bovis
survival on pasture and inanimate objects contrib-
ute to exposure and facilitate subsequent spread of
infection (Morris et al. 1994). Many countries, pre-
viously endemic for bovine tuberculosis, have now
been declared tuberculosis-free, following intensive
test and slaughter policies (Cousins and Roberts
2001). However, where there is interplay between
infected wildlife and domestic animals, control has
proven problematic. It is now evident that better
understanding of this disease and its epidemiology
is essential to overcome the impeded progress in
some countries.
Postulations on the residue of infection have
included deficiencies in skin test specificity and sen-
sitivity, poor cattle testing and the presence of res-
ervoirs of M. bovis in cattle and wildlife. Traditional
farming practices with significant cattle movement,
intensive farming with high stocking densities and
poor boundary fencing between farms allowing
cattle 'nosing' between neighbouring farm herds
have also been cited (Neill 1994; O'Reilly and Daborn
1995). In Ireland, inter-bovine spread of tuberculosis
was identified as the single most important source of
infection (Neill 1994).However,wildlife reservoirs are
increasingly being considered a significant risk and
responsible for protracted eradication programmes
(Cheeseman et al. 1989; Livingstone 1992; Gallagher
and Clifton-Hadley 2001). In the United Kingdom
and Ireland, badgers (Meles meles) and feral deer
(Cervus elaphus) are recognised wildlife reservoirs
of M. bovis (Cheeseman et al. 1989; Hughes and
Rogers 1994; Gallagher and Clifton-Hadley 2001).
Badgers are indigenous and a protected species in
the United Kingdom and Ireland with estimates in
Great Britain of approximately 250,000 badgers, in
Northern Ireland 30,000 (Hughes and Rogers 1994)
Molecular Epidemiology of Mycobacterium bovis
and 250,000 in the Republic of Ireland. The contribu-
tion of wildlife to tuberculosis in cattle in the United
Kingdom and Ireland is not known precisely and this
can be controversial with regard to the badger. One
epidemiological study of risk factors for tuberculo-
sis in dairy herds in Northern Ireland suggests that,
although infected cattle may playa significant role
in the transmission of tuberculosis, the importance
of badgers may have been underestimated (Denny
and Wilesmith 1999). In New Zealand, in marked
contrast to Australia, tuberculosis in cattle is also
exacerbated by a wildlife reservoir and maintenance
host, the possum (Trichosurus vulpecula). Endemic
infection in possums constitutes a major obstacle
to disease control and eradication in some areas
(Morris et al. 1994; de Lisle et al. 2001).
Globally,M. bovis has been estimated to contribute
3.1 % of all human tuberculosis cases (2.1 % of pul-
monary tuberculosis and 9.4% of extra-pulmonary
tuberculosis) (Cosivi et al. 1998). In developed coun-
tries, such cases are now relatively rare and estimated
at less than 1% of all tuberculosis cases, and are prob-
ably due to reactivation of dormant lesions amongst
the elderly (Grange 2001). M. bovis has been isolated
from HIV-infected individuals in some developed
countries, with an additional serious complication
of high primary resistance to isoniazid, streptomycin
and pyrazinamide (Guerrero et al. 1997). Although M.
bovis is thought to be less virulent and also less trans-
missible in humans than animals, human-to-human
transmission of multi-drug resistant (MDR) M. bovis
strains, over a relatively short interval, was demon-
strated in Paris. HIV-induced immunosuppression
may lower host immune responses and lead to overt
disease more rapidly than under normal conditions.
The true incidence of M. bovis infection in humans
appears to be underreported, probably owing to
the similar clinical presentation to M. tuberculosis
infection but also to the fact that many diagnostic
laboratories often would not have differentiated M.
tuberculosis and M. bovis isolates.
6.4
Epidemiology in Developing Countries
On the whole, there is very limited epidemiological
information on bovine tuberculosis in cattle in the
developing world (Cosivi et al.1998). Most cattle are
in regions where no formal control measures are
applied and implementation of a recognisable test-
and-slaughter policy would be considered excep-
77
tional. Information on M. bovis prevalence in other
species is also very limited. In a study of 56 African
countries, 33 provided information about M. bovis
infection in animals. However, few reported M.
bovis infection in humans (Kazwala et al. 2001a,b).
The M. bovis cases reported in some developing
countries may just be the tip of the iceberg, as, for
example in Tanzania, medical authorities do not
recognise M. bovis as a risk to human health. There
has also been a lack of interest amongst medical
practitioners and microbiologists in requesting
speciation after primary isolation of mycobacteria
from patients. In many developing countries there is
little or no pasteurisation of milk or milk products
and many people live a pastoral lifestyle, in close
contact with their cattle, with resultant exposure to
tuberculosis where present. Milk-borne infection
is the main cause of non-pulmonary tuberculosis
in areas where bovine tuberculosis is common
and uncontrolled (Daborn et al. 1996). In Africa,
approximately 85% of cattle and 82% of humans live
in areas where bovine tuberculosis is only partially
controlled (Cosivi et al. 1998). Approximately 90%
of the total milk produced in Africa is consumed
fresh or soured. Therefore, the consumption of
unpasteurised milk, contaminated by M. bovis, is a
likely route of zoonotic tuberculosis transmission
to humans. In the remaining regions of these coun-
tries, control programmes operate only partially. It
is evident that in Africa, Asia and Latin America/
Caribbean, significant human populations work in
close physical contact with animals, and therefore
zoonotic tuberculosis represents a significant, but
in reality, unquantified risk.
The global HIV epidemic adds a further complica-
tion. Tuberculosis is a major opportunistic infection,
if not the most common infection, for HIV-infected
individuals, the majority of whom live in develop-
ing countries. In those countries in which M. bovis
infection is present in animals and where favourable
conditions exist for zoonotic transmission, zoonotic
tuberculosis is a serious public health threat. Zoo-
notic tuberculosis in HIV-positive individuals also
resembles M. tuberculosis infection. WHO estimate
that 70% of people (6.6 million) co-infected with
tuberculosis and HIV live in sub-Saharan Africa
(Cosivi et al.1998).
In African countries, M. bovis infects a higher
proportion of exotic dairy breeds (Bos taurus) than
crossbred beef cattle and indigenous zebu cattle (Bos
indicus) (Cosivi et al. 1998). In intensive conditions,
a death rate from tuberculosis of up to 60% in zebu
cattle has been recorded, with animal crowding con-
78 R. A. Skuce and S. D. Neill

tributing to disease transmission. Tuberculosis con- 6.6

trol and eradication strategies, although understood, The Genetics of the Mycobacterium
are inconsistently applied and usually not sustained tuberculosis Complex Bacteria
in developing countries, largely due to lack of finance
and trained professionals. Therefore there is obvi- The M. tuberculosis complex (Rastogi et al. 2001)
ously an increased likelihood of underestimation of comprises M. tuberculosis, M. africanum, M. bovis,
zoonotic tuberculosis, the full economic implications M. microti and M. bovis BCG (Behr 2001), as well
of which is often overlooked. as newly characterised bacteria M. canetti (van
Soolingen et al. 1997b) and M. caprae (Aranaz et
al. 1999). Isolates with characteristics intermediate
between M. tuberculosis and M. bovis have also been
6.5 reported (Kallenius et al. 1999). The bacteria in the
Epidemiological Typing complex are usually regarded as subspecies and are
characteristically 99.9% similar at the nucleotide
Advances in molecular technologies have enabled level, with identical16S rRNA sequences (Brosch et
specific identification and subsequent recognition al. 2002). However, there are distinct phenotypic dif-
of strains of many pathogens involved in animal and ferences between the 'subspecies', not least their host
human infections (van Belkum et al. 2001; van Belkum range and pathogenicity. M. bovis has the largest host
2002). 'Molecular epidemiology' is the integration of range within the M. tuberculosis complex, infecting
these techniques with conventional epidemiological many mammalian species, including man (O'Reilly
approaches for understanding better disease distri- and Daborn 1995), whereas M. tuberculosis is almost
bution in populations (Small and van Embden 1994). exclusively a human pathogen.
Such studies are often designed to allow inference The M. tuberculosis complex genome sequencing
of patterns of disease transmission from similarities projects represent significant landmarks (Cole et al.
between genetic patterns generated for infectious 1998). The sequencing projects,M. tuberculosis strain
pathogens (Salamon et al. 2000). Using such genetic H37Rv and M. bovis strain AF2122/97 at the Sanger
marker-assisted molecular epidemiology, it is now Centre (http://www.sanger.ac.uk. Cambridge, UK)
possible to differentiate reproducibly between strains and M. tuberculosis strain CSU93 at The Institute
of micro-organisms and to show spatial and temporal for Genome Research (http://www.tigr.org) have
aspects of transmission, for example, between ani- provided a hugely valuable research resource and a
mals. Epidemiological typing systems are generally unique insight into the biology of these major patho-
used to study population dynamics and the spread gens. This should lead to major advances in diagnos-
of bacteria that undergo clonal reproduction both in tic, vaccination and chemotherapeutic options. The
nature and in the clinical setting and often include M. tuberculosis H37Rv genome is relatively large for
studies in bacterial population genetics and patho- an obligate pathogen, reflecting the complex bio-
genesis, providing 'early warning' through local, chemistry of its intracellular lifestyle and providing
regional or national surveillance and in outbreak an insight into its origins. The genome is GC rich and
investigations (Streulens 1996). it contains several different mobile genetic elements,
Typing systems are applied primarily to assist epi- such as insertion sequences and prophages, as well
demiological studies and in generating hypotheses as different classes of repetitive DNA (Gordon et al.
about: 2001).
• the extent of epidemic spread of microbial Until very recently it was thought that M. tuber-
clone(s) in an exposed population; culosis evolved from M. bovis, by adaptation of an
• the number of clones involved in transmission animal pathogen to the human host. Recent studies
and infection; have provided compelling genetic evidence to sug-
it the identity of the source(s) of contamination and gest that the members of the M. tuberculosis complex
vehicles of transmission; shared a common ancestor, which was likely to have
• the identification and monitoring of reservoirs of resembled M. tuberculosis or M. canetti and was
epidemic clone(s) in the population and/or the probably already a human pathogen (Brosch et al.
environment; 2002). Comparative genomics has shown that the
G the evaluation of the efficacy of control measures M. bovis genome is smaller than M. tuberculosis,
aimed at containing or interrupting the spread of which is intriguing due to its wider host range. This
epidemic clone(s). comparative approach has also identified a series of
Molecular Epidemiology of Mycobacterium bovis
insertion and deletion events between members of
the complex (Gordon et al. 1999a), whose distribu-
tion points to a common ancestor and has been used
to derive an entirely new evolutionary scenario. M.
bovis now appears to be the final member of a sepa-
rate lineage that branched from the progenitor of M.
tuberculosis and is represented by M. africanum, M.
microti and M. bovis isolates having undergone suc-
cessive loss of DNA segments. This may have contrib-
uted to clonal expansion and the selection of more
successful pathogens in some hosts. This new data
is inconsistent with the view that the proposed M.
africanum - M. microti - M. bovis lineage could have
spread and adapted to their wide host range within
the suggested 15,000-20,000 years of speciation of
the M. tuberculosis complex.
6.7
Strain Typing Methodology
Phenotypic procedures based on physiological or bio-
chemical characteristics and antibiotic susceptibility
are often capable of discriminating between strains
of micro-organism. However, this has not been the
case for the M. tuberculosis complex (Saunders 1995)
and in 1965 a WHO working group was established
to facilitate development and standardisation of
methods for typing mycobacteria. Phage suscepti-
bility typing rapidly became a valuable epidemio-
logical tool, which yielded data that contributed to
the understanding of tuberculosis transmission and
pathogenesis and provided the basis for current
approaches to human tuberculosis control. However,
the method is relatively crude and much greater dis-
crimination can be achieved using molecular meth-
ods, which now supersede phage typing.
The molecular epidemiology of M. tuberculosis in
humans is much more advanced than that of M. bovis
in cattle, largely because a robust and standardised
strain typing method has been applied widely (van
Soolingen 2001). Most M. bovis isolates are not
amenable to this technique. Therefore, alternative
methods have had to be developed and evaluated.
The availability of genome sequences has led to the
development of a generation of new molecular mark-
ers, which have the potential to offer high-resolution
genotyping and high-throughput analysis.
Collins and colleagues at AgResearch, Wallaceville,
New Zealand pioneered bacterial DNA restriction
enzyme fragment analysis, or restriction enzyme
analysis (REA), as the earliest typing method applied
79
successfully to the organisms of the M. tuberculosis
complex (Collins and de Lisle 1984, 1985). In REA,
DNA extracted from batch cultures, using deter-
gent and protease digestion followed by solvent
extraction, is digested with each of three restriction
enzymes (BstEII, Pvull and Bell), prior to electropho-
retic separation (Fig. 6.1). Ethidium bromide staining
of DNA fragments generates a complex DNA band-
ing pattern, similar to a bar-code fingerprint, with
observable differences in the fragment pattern,
especially in the higher molecular weight fragments.
This produced a highly discriminating typing system
for the M. tuberculosis complex, most notably for M.
bovis. Approximately 2000 M. bovis isolates have been
REA typed over the last 20 years, and more than 165
stable REA types have been identified (Collins 1998).
However, the REA technique is not widely used, due
to difficulties in accurately reproducing the complex
fingerprint patterns and in naming, recording and
databasing them. It is a relatively cumbersome tech-
nique and technology transfer and inter-laboratory
comparisons have proven particularly difficult.
A logical modification of the REA technique
is pulsed field gel electrophoresis (PFGE), where
rare-cutting restriction enzymes are used to digest
genomic DNA immobilised in agarose blocks
(Hughes et al. 2001). The resulting large DNA frag-
ments are then separated by electrophoresis in a
dedicated apparatus (PFGE Fig.6.1). The smaller
number of larger DNA fragments generated sig-
nificantly simplifies the electrophoretic patterns
produced, although the discrimination is consider-
ably less than with REA. PFGE of M. bovis isolates
has produced moderate discrimination and has been
useful in epidemiological studies (Feizabadi et al.
1996). Significant improvements have been made
recently to the culture and preparation of samples
for PFGE and this should allow better evaluation of
method performance (Hughes et al. 2001).
Researchers have also exploited the finding ofrepet-
itive DNA (Tanaka et al. 2000) within the genomes of
M. tuberculosis complex bacteria to develop an array
of genotyping techniques, based either on DNA ampli-
fication using the PCR or Southern blotting, or a com-
bination of the two. There are basically two types of
repetitive DNA in bacterial genomes: dispersed repeats
and tandem repeats. The former can include mobile
genetic elements, including insertion sequences (ISs)
and prophages (Gordon et al. 1999b). M. tuberculosis
H37Rv contains 56 such IS elements, belonging to
well-defined families. Mobile genetic elements con-
tribute significantly to genetic diversity. Whilst most
of these ISs are located in the same genome position
80 R. A. Skuce and S. D. Neill

Gel-based genotyping

M. bovls
culture
-

00 0
'0 " ,

~ .....f' •

M.bovls
genomic
; .~. \

~.",
;1
DNA

fragmented 9..
-::~. ~~"
DNA ;!; j
'"j . . .J' I I
'.-.1:

--=-1
frrtIgtMnt

1 2 3 1 2 3 rr-n.,.,to 1 2 3
",.",bnJlN
Md IIybrldlN to
- -- probe - -
iPFGE ---
--- --- -- REA • -- --
-
RFLP
- -
- ---- - --
• • • Fig.6.1. Schematic representation of
the main restriction enzyme digest-
based genotyping methods applicable
to the M. tuberculosis complex. Actual
gel images for M. bovis field isolates
PGR8-RFLP
are shown for PFGEl, REA2 and RFLP.
(PFGE gel reproduced with permission

.. ..
:_ _ :":111:" DR·
from VM Hughes. REA gel reproduced
with permission from DM Collins. PFLP
. -......... ~.:---
RFlP gel images reproduced with permission
from M Rogers)

in members of the M. tuberculosis complex, there are
some significant and useful differences between spe-
cies, which can be exploited for diagnosis. Of these,
IS6110 is the most abundant and best-characterised,
belonging to the IS3 family. The IS6110 copy number
varies from 0-25 within the M. tuberculosis complex
bacteria. Extracted genomic DNA from small batch
cultures of M. tuberculosis is digested with PvuII, prior
to electrophoretic separation and Southern blotting
(RFLP Fig. 6.1). The method is standardised by run-
ning the DNA digests a predetermined distance and by
the inclusion oflane origin markers, in-lane molecular
weight standards and reference M. tuberculosis strains
(van Embden et al.1993). Computer-assisted gel docu-
mentation, image analysis and databasing contribute
to the excellent reproducibility of the technique
(Heersma et al.1998). The strain typing of M. tubercu-
losis isolates by IS611O-RFLP is now an internationally
accepted part of standard control policies, a tool for
detecting point source outbreaks and a method for
monitoring the efficacy of human tuberculosis control
programmes (van Soolingen 2001).
Unfortunately, for M. bovis isolates, the IS6110
copy number is much lower, often occurring singly,
especially in isolates from cattle (Collins et al. 1993;
Skuce et al. 1996; Cousins et al. 1998b). Interestingly,
however, the IS6110 copy number is often higher in
M. bovis strains isolated from more 'exotic' animal
species (van Soolingen et al. 1994). Therefore addi-
tional probes are often required to improve discrimi-
nation for low IS611 0 copy number strains. One such
probe is the 'polymorphic GC-rich repeat' sequence
Molecular Epidemiology of Mycobacterium bovis
(PGRS), which is found in multiple copies within
the genomes of M. tuberculosis complex mycobac-
teria, although it is not considered complex-specific,
unlike IS611O. PGRS-RFLP, using the 'Doran repeat'
oligonucleotide probe (Doran et al.1993) to hybridise
to AluI-digested genomic DNA, produces relatively
complex patterns. Most of the visible band differ-
ences are in the higher molecular weight fragments.
The M. tuberculosis complex specific Direct Repeat
(DR) locus (see below) can also be used as a probe
in RFLP analysis (Groenen et al. 1993). DR-RFLP,
using a DR oligonucleotide probe to hybridise Alul-
digested genomic DNA produces a relatively simple
banding pattern. Several studies have used the com-
bined results of IS6110-RFLP, PGRS-RFLP and DR-
RFLP to produce an overall RFLP typing scheme
(Skuce et al. 1996; Costello et al. 1999). Genomic
libraries have revealed promising new probes for
RFLP analysis, such as pUCD (O'Brien et al. 2000a,b).
A bioinformatics search shows that pUCD contains
several polymorphic loci, some of which have been
found by other investigators using entirely different
approaches (Supply et al. 2000).
It is widely accepted that the RFLP techniques,
although highly discriminating, are again relatively
cumbersome, inconvenient and not amenable to high
throughput genotyping and inter-laboratory compari-
sons. Hence, effort has been directed at exploiting the
various classes of DNA repeat using DNA amplifica-
tion procedures. For example, the spacer-oligotyping
technique, (spoligotyping) (Kamerbeek et al. 1997)
uses the PCR to amplify a region of DNA, the so-
called Direct Repeat or DR region, the arrangement
and sequence of which is highly specific to members
of the M. tuberculosis complex (Fig. 6.2). The order of
these spacer sequences is remarkably well conserved,
so much so that a phylogeny and common ancestor to
the M. tuberculosis complex bacteria can be inferred,
with a much larger DR locus containing many more
spacer sequences than is presently seen in modern
members of the complex (van der Zanden et al. 2002).
The DR locus belongs to a recently identified class of
repeat sequences found in bacteria, known as 'spacer
interspersed direct repeats' (SPIDRs) or 'clustered
regularly interspaced short palindromic repeats'
(CRISPRs), which are reminiscent of centromere-like
structures with a possible role in replication parti-
tioning. CRISPR-mediated variation may provide
a bacterial population with the diversity required
to be adaptable in changing environments. The DR
locus is comprised of multiple 36 bp direct repeats
interspersed by non-repetitive 34-41 bp DNA spacer
sequences. Spoligotyping involves amplification of the
81
entire CRISPR locus using the DR as a target for PCR,
followed by hybridisation of the amplified DNA to a set
of spacer oligonucleotides derived from M. tuberculo-
sis and M. bovis BCG strain P3. This reverse-cross blot
hybridisation (Kaufhold et al. 1994) records the pres-
ence, or absence, of spacer DNA sequences and pro-
duces a simple digital pattern, which is readily named
and databased. Spoligotyping is particularly attractive
because the patterns generated may be a presumptive
indicator of the M. tuberculosis complex species ampli-
fied. For example, M. bovis isolates generally lack spac-
ers 39-43 inclusive, as well as spacers 6,9 and 16 (van
Soolingen 200l). Spoligotype patterns are also useful
for identifying M. microti (van Soolingen et al. 1998)
and have proved extremely valuable for typing some
emerging epidemic strains, such as M. tuberculosis W-
Beijing (Glynn et al. 2002). Differences in spoligotypes
are due to deletions of DRs in the CRISPR locus,medi-
ated by transposition ofIS611O, homologous recombi-
nation or replication slippage.
Strain-specific differences in spoligotype are the
result of a sequential loss of single spacers or blocks
of contiguous spacers within the CRISPR locus. It
has been possible, therefore, to infer a sequence of
genetic events separating M. bovis of different spo-
ligotypes. It appears that ST140 is a recent progeni-
tor of several different M. bovis spoligotypes, all of
which are represented in current field isolates. The
molecular clock for the CRISPR locus appears to be
very slow. For example, all M. bovis BCG daughter
strains share the same spoligotype, a spoligotype still
seen in French M. bovis field strains (Haddad et al.
2001). Current phylogenetic clustering tools consider
the gain and loss of individual or contiguous spacers
as an equal event. Due to the proposed mechanism
of CRISPR evolution, this may misrepresent the rela-
tionship between strains.
Spoligotyping is relatively easy to perform and
can be used to detect and type M. tuberculosis com-
plex bacteria directly in a range of clinical samples
(Kamerbeek et al. 1997). However, a major problem
with the current configuration of spoligotyping is
its relatively limited discriminatory power. Even for
M. bovis isolates, spoligotyping is only moderately
discriminating, compared to REA and RFLP tech-
niques (Aranaz et al. 1996; Cousins et al. 1998a). More
extensive sequencing of M. tuberculosis complex
isolates has disclosed additional spacer sequences
(van Embden et al. 2000; Caimi et al. 200l), which
have been incorporated into the second generation
spoligotyping technique. The discrimination and
reproducibility of spoligotyping using these new
spacer sequences is currently being evaluated on
 82 R. A. Skuce and S. D. Neill

Spoligotyping

••'.'.'.'.'." '.·.·J"J~\"......~\".!fWA'M •• ••
• 11.11 •• II • • • • • • • • II •• II . ~•• II • • • • • I I '

a ~
a~'»»lIO:au.a"')j.

~!L- _ _----l

........
PCR products :
Drb Orb Orb Orb Orb

o
1-----0
o o
o o
o

II
L_-===~;;2°U c

~~~~~m~~~
AJlPIycta.",IOC.,..,
U&J:lI. au:. bIoa.r
°.°.•.JI"UT to''noI....'
ID • hat " •• brmt
Spacer Itt • • • ' I UC"• • • N ,. • • "'.I111 .

t
...... .... ...
..... ...... Albom
I.ot...

.... ..... ...

I

.... .......
........ ...
~mm~m
R..... aJ"'llO".'l'Pl,1II ..
labcI.d I'CR poducu and lI7bndd.
.... ...... .....
~"'IsHRP.coq 11M
d IrIS .".,.,... lOll upo. X.,,,, (11.
e

Fig. 6.2. Schematic diagram of the PCR-based spoligotyping technique applicable to M. tuberculosis complex isolates. a Structure
of the DR locus in the genome of M. tuberculosis and M. bovis BCG P3. The green rectangles depict the 36 bp Direct Repeat (DR).
b Schematic representation of the polymorphism in DR regions of different M. tuberculosis complex strains. Blocks of DVR are
missing in one strain compared to another. The spacer order remains the same. c Principle of the in vitro amplification of DNA
within the DR region of M. tuberculosis complex bacteria. The use of the two primers, a and b, for in vitro amplification, will
lead to the amplification of any spacer or stretch of neighbouring spacers and DRs. d Overview of the spoligotyping method.
e Spoligotyping result of M. tuberculosis H37Rv and M. bovis field isolates (rows). A membrane with 43 spacer oligonucleotides
was used (columns). Spacer oligonucleotides were derived from the spacers of M. bovis BCG P3 and M. tuberculosis H37Rv.
Courtesy of Isogen Bioscience B. V.

representative panels of M. tuberculosis complex bac-
teria (van der Zanden et al. 2002). Attempts have also
been made to unify and simplify spoligotype pattern
nomenclature in order to facilitate inter-laboratory
studies of the transmission of strains between and
within different countries (Dale et al. 2000) and
relatively large databases of spoligotypes are being
curated (Sola et al. 2000, 2001).
Genomics and bioinformatics have revealed several
different classes of repetitive DNA within the genomes
of M. tuberculosis H37Rv and M. bovis AF2122/97,
which can now be exploited to develop convenient and
discriminating strain typing tools (Brosch et al. 2000).
Many tuberculosis genes were shown to be comprised
of tandem repeats. The Tandem Repeats Finder soft-
ware programme (Benson 1999) has been used to
predict a number of repeats varying from 9-351 bp.
Fifty-three PE_PGRS genes, 27 PPE proteins and 59
unrelated proteins contained such tandem repeats.
This could represent a major source of genetic diver-
sity and antige'lic v?riation in a pathogen with a strik-
ing lack of neutral, single nucleotide polymorphisms
in key house-keeping genes and antigens (Sreevatsan
t al. 1997; Musser et al. 2000). One such class of repeat,
 Molecular Epidemiology of Mycobacterium bovis 83

which is generating considerable interest at present, was described, although clinical isolates were not
is the so-called mycobacterial interspersed repetitive characterised in the study. Analysis of larger numbers
unit (MIRU) (Supply et al. 2000). MIRUs range in size of M. tuberculosis complex bacteria allowed the com-
from 46-101 bp and fall into three classes depending pilation of allele naming tables, which form the basis
on their length, sequence and organisation. MIRUs of the intuitive MIRU-VNTR nomenclature. Further,
occur in 41 loci, mostly intergenic, in M. tuberculosis novel VNTR-type loci have been described recently in
H37Rv. Most MIRUs contain an ORF and could poten- the M. tuberculosis H37Rv genome (Skuce et al. 2002)
tially encode small peptides. MIRUs often occur as and were evaluated on a reference panel of M. bovis
tandem repeats and can be exploited as strain typing isolates compared to existing VNTRs (Frothingham
targets. Specific PCR primers are designed to flank and Meeker-O'Connell 1998) and spoligotyping data.
such loci and PCR products consequently differ in size Different VNTR-PCRs had different discriminatory
relative to the tandem repeat copy number (Fig. 6.3). capacity and combining VNTR-PCRs produced a
Other loci, equivalent to MIRUs, known as variable more discriminatory technique. The majority of novel
number of tandem repeat loci (VNTR), have also been VNTRs were located within ORFs including several
described (Frothingham and Meeker-O'Connell 1998). representatives of the PPE protein family. For exam-
These MIRU-VNTR loci resemble eukaryotic mini- ple, there were 29 allelic variants of the PPE protein
satellite loci, being comprised of tandem repeat units Rv1917c in the test panel of 100 M. bovis isolates.
of varying size and copy number (Supply et al. 1997; Further, novel MIRU-VNTR loci have been
Frothingham and Meeker-O'Connell 1998; Magdalena subsequently identified in M. tuberculosis genome
et al. 1998a,b). Systematic bioinformatic searches of sequences (Roring et al. 2002). Again, their perfor-
available genome sequence databases have disclosed mance has been systematically evaluated on a test
more such loci within the M. tuberculosis complex panel of M. tuberculosis complex isolates, leading
mycobacteria (Smittipat and Palittapongarnpim 2000) to the recommendation that MIRU-VNTR loci be
and a simple scheme for naming the MIRU-VNTR loci selected, for genotyping studies, based on their proven

MIRU-VNTR genotyping
MIRU·VNTR
target Copy No.

...
Copy No.
Streln
8 b c d
....
~

-.....
strain 8 4
SlzePCR
4
prodUcts
....
~
strain b 3 3

~ (electrophore.'8) 2
strain c 2

a ~
....
strain d

VNTR type VNTR Profile

Strain a ---.- 43451

Strain b 34261 Fig. 6.3. Schematic representation of

PCR-based MIRU-VNTR genotyping
Strain c ---------- 22432 of M. tuberculosis complex isolates.
a Representation of tandem repeat copy
number variation between four hypo-
b Strain d --a-------- ---II- 13451
thetical strains at one MIRV-VNTR
locus and sizing of PCR products to
ETR-A QUB11a QUB3232 deduce copy number. b Hypothetical
MIRU-VNTR profiles derived for four
strains at five loci. c Gel images for the
same six M. bovis field isolates MIRU-
VNTR typed at three loci (ETR-A, QUBI
c la and QVB3232)
84
performance with a representative test panel of iso-
lates and not just on the convenience of using prede-
termined sets of markers. For example, the 12 MIRU
panel is at least as discriminatory as IS611O-RFLP for
M. tuberculosis isolates (Mazars et a1. 2001). However,
not all the MIRU loci are informative with the M.
bovis isolates comprising the test panel (Roring et
a1. 2002). There are now upwards of 30 MIRU-VNTR
loci described within the M. tuberculosis complex.
Their discriminatory ability will depend on the test
panel used. The reproducibility of the MIRU-VNTR
PCRs and allele naming is exceptionally high. These
MIRU-VNTR loci detect differences between M.
bovis BCG daughter strains, entirely consistent with
the proposed genealogy (Magdelena et al. 1998a; Behr
2000; Roring et a1. 2002). A major future objective will
be to determine the pace of change for each of these
loci and to match their properties to specific epide-
miological studies. It can be envisaged that a panel
of MIRU-VNTR loci, with varying molecular clock
speeds, will be used in future epidemiological studies
of the M. tuberculosis complex mycobacteria. These
MIRU-VNTR loci have many of the most desirable
features required to progress rapidly the molecular
epidemiology of the M. tuberculosis complex.
Several other PCR-based techniques have been
developed, based on amplification of DNA fragments
between repetitive DNA elements. Some difficulties
have been experienced with the reproducibility of
such methods on inter-laboratory evaluation (Kremer
et a1. 1999).
6.8
Standardisation and Method
Performance Evaluation
In order that strain-typing results are directly compa-
rable between laboratories, it is important that stan-
dard procedures and analyses are performed. Novel
strain typing methods are often applied without criti-
cal evaluation of their performance characteristics.
They often lack standardisation in technical proce-
dures, reference materials and quality assurance, as
well as in the criteria used for the interpretation of
results. Basic terminology is sometimes used ambig-
uously by different investigators, compounding the
confusion created by variations in methodology
(Struelens 1996). Consequently, progress towards
performance comparison, or standardisation, has
only been achieved for a limited number of bacte-
rial and fungal pathogens (Struelens 1996).
R. A. Skuce and S. D. Neill
Different techniques will have different perfor-
mance characteristics, e.g. discrimination, reproduc-
ibility, epidemiological concordance etc. (see Struel-
ens 1996 for definitions). The most discriminating
and the least practical strain-typing technique would
be to produce the entire nucleotide sequence for
every isolate. Several investigators have recognised
the need for a unified approach to the strain typing
of members of the M. tuberculosis complex mycobac-
teria and M. bovis in particular (Kremer et a1. 1999).
This would comprise standardisation of procedures,
reagents, reference strains, image analysis and
nomenclature. Under the auspices of the Interna-
tional Union Against Tuberculosis and Lung Disease
(IUATLD), one such study (Cousins et a1. 1998a) rec-
ommended IS611O-RFLP to define the copy number
in the test population and that IS611O-RFLP was usu-
ally adequate for M. bovis strains possessing more
than three copies of IS6110. For strains with less
than three copies of IS6110, additional typing was
often warranted and PGRS-RFLP was the method of
choice. Spoligotyping was recommended to replace
DR-RFLP for additional typing.
The DNA typing of M. bovis specifically was also
considered by a multi-centre evaluation of method
performance (Skuce et a1. 1998). The candidate
methods of IS6110-RFLP, PGRS-RFLP, DR-RFLP,
PFGE, spoligotyping, ampliprinting and VNTR
typing (Frothingham and Meeker-O'Connell 1998)
were evaluated. Methods were optimised by work-
ing groups and standard protocols were developed,
which detailed procedures, reference strains and
reagents, as well as making recommendations for
appropriate nomenclature schemes and databasing.
The optimised methods were applied to a coded
test population of M. bovis isolates and the vari-
ous performance criteria evaluated were: typability,
reproducibility, discrimination, concordance and
convenience (Struelens 1996). Methods differed in
their performance characteristics. For example, the
combined IS611O, PGRS, DR-RFLP approach was
found to be highly discriminating and, with care,
was adequately reproducible. However, it was by
far the least convenient. Spoligotyping was highly
reproducible and convenient, but only moderately
discriminating in its current configuration with this
reference panel. Spoligotyping was recommended as
a relatively simple and highly reproducible screening
method with moderate discrimination and high con-
cordance. Where further discrimination is required
the highly reproducible, convenient and concor-
dant VNTR technique (Frothingham and Meeker-
O'Connell 1998), or the RFLP and PFGE methods
Molecular Epidemiology of Mycobacterium bovis
were recommended. Further development and evalu-
ation was merited for several methods, most notably
the variations on VNTR typing, especially with the
imminent release of the genome sequence for the M.
bovis isolate AF2122/97.
In another study, the discriminatory power and
reproducibility of the main candidate DNA finger-
printing methods was assessed recently for M. tuber-
culosis complex bacteria, including a small number of
M. bovis isolates (Kremer et al. 1999). Again the various
RFLP-based methods, using IS611 0, IS1081, PGRS, DR
and (GTGh as probes were found to be reproducible,
although not very convenient. The PCR-based meth-
ods of spoligotyping, VNTR typing (Frothingham
and Meeker-O'Connell 1998) and mixed-linker PCR
were found to be highly reproducible. Subsequently,
semi-automated versions ofVNTR-typing (Hughes et
al. 2000) and MIRU-typing (Mazars et al. 2001) were
developed and have been applied to the same test
panel. They were found to be highly reproducible and
discriminatory, with MIRU-typing as discriminating
as the accepted standard IS611O-RFLP method. Addi-
tionally, IS611O-RFLP and MIRU-typing clustered the
test panel isolates in a highly concordant manner,
making MIRU-typing and possible developments
thereof, an exciting, convenient and portable new tool
with which to tackle molecular epidemiology of the M.
tuberculosis complex.
It is anticipated that the best DNA fingerprint-
ing technique will be that which is based on genetic
events occurring in synchrony with the time scale of
the epidemiological situation. Thus, different epide-
miological questions will be best addressed by dif-
ferent techniques. If the genetic events are occurring
considerably faster than the epidemiological events,
epidemiologically related strains will appear to be
unrelated, whereas if the genetic events are occurring
more slowly than the epidemiological events, then
unrelated strains will appear to be related. A major
future challenge of DNA fingerprinting of M. tuber-
culosis complex bacteria will be to determine the
paces of change of the various markers and to match
these to study specific epidemiological questions.
6.9
Mycobacterium tuberculosis Molecular
Epidemiology: Lessons Learned
The ability to identify transmission events accurately
and to intervene in order to prevent further spread is
critical in effective tuberculosis control programmes
85
(Kulaga et al. 2002). Epidemiological studies have
generally relied on demonstrating plausible oppor-
tunities for infection in order to infer transmission.
Strain typing of M. tuberculosis isolates by IS611O-
RFLP analysis is now well developed (van Soolingen
2001) and is being integrated into control policies,
as a tool for detecting point source outbreaks and
for monitoring the efficacy of tuberculosis control
programmes (Small et al. 1994; Alland et al. 1994;
van Soolingen et al. 1997aj Bauer et al. 1998). An
internationally accepted standard procedure (van
Embden et al. 1993) has been devised to facilitate
inter-laboratory comparisons (van Embden et al.
1996; Heersma et al. 1998).
For M. tuberculosis, strain typing is becoming a
standard tool of public health investigators. If there
is sufficient genetic diversity in a population, find-
ing a group with the same, or highly related, pattern
(cluster) infers epidemiological association, where
individuals are either infected from each other or
from a common source. Such cluster analysis must be
interpreted with caution, especially where the survey
has not been sufficiently comprehensive (Glynn et
al. 1999a,b). Strain typing has been used to identify
clonal transmission, e.g. between neighbours and
within prisons, bars, clinics and hospitals. It has also
been used to identify epidemiologically unsuspected
transmission within groups, or unsuspected ongoing
transmission in certain segments of the popula-
tion, where risk factors, environments and common
sources were missed by conventional contact tracing
(van HeIden 1998). IS611O-RFLP has surprisingly and
repeatedly demonstrated that tuberculosis transmis-
sion can occur without close personal contact (Kline
et al. 1995). Prompt recognition of such events has
helped to arrest such micro-epidemics. Strain typing
has also played a role in identifying problematic pro-
cedures and laboratory cross-contamination in refer-
ence laboratories, where isolation of the organism is
often the only basis for diagnosis and intervention.
6.10
Mycobacterium bovis Molecular
Epidemiology: Studies and Trends to Date
The REA technique has been used in several epi-
demiological studies, most notably in New Zealand
(Collins 1998). It has provided a new insight into
disease dynamics, demonstrating clustering of REA
types in defined geographical areas of New Zealand.
Possums, other wildlife and farmed animals in the
86
same area were often infected with the same REA
type (Collins et al. 1986, 1988, 1994). REA typing was
used to include or exclude possible sources of infec-
tion in specific herds and it has clearly demonstrated
whether infection in a farmed animal has come from
the infected local wildlife reservoir or from infected
cattle or deer brought onto those premises (de Lisle
et al. 1995). REA data has been used to influence the
level of herd testing or wildlife control in specific
areas, e.g. REA and IS6110-RFLP were used to study
the M. bovis strain types present in the MacKenzie
basin in the South Island of New Zealand, a region
that had been free of bovine tuberculosis prior to
1980. The identification of two main groups led to the
suggestion that infection was introduced from two
distinct, as yet unidentified, sources. Strain typing
has, therefore, contributed to determining the extent
and spread of infected wildlife in defined areas (de
Lisle et al. 1995; Collins et al. 1999) and is now con-
sidered an integral component in local bovine tuber-
culosis control schemes (Collins 1998).
Restriction enzyme analysis (REA) was used to
type M. bovis isolates from a longitudinal field study
of endemically infected possum, cattle, feral pigs and
feral ferrets sharing the same habitat (Pfeiffer et al.
1994). A Geographical Information System (GIS)
was used to manage the data, which was mapped on
a terrain model of the study area. Statistical analysis
was used to display the association between such
variables as REA type, sex, age, season of detection
and terrain covered. Four REA types were described,
with two major types each dominating at different
times. The same major types were evident in cattle
when dominant in possums. Some important infer-
ences regarding potential transmission routes for the
wild possum population were possible. Shared REA
types between mother and young identified pseudo-
vertical transmission as an effective transmission
route. Two successive, major cycles of transmission,
showing only limited spatial overlap, were described
within the possum population. REA was also used to
infer that infected possum, not environmental sur-
vival of M. bovis, were responsible for reintroducing
tuberculosis to an area following extensive possum
depopulation (Corner et al., unpublished). Strain
typing has also been used in a targeted fashion to
show that domestic cats were usually infected with
the same REA types as predominant in other local
animal species and that they posed an unsuspected
zoonotic risk (de Lisle et al. 1990).
The application of molecular epidemiology to
bovine tuberculosis is most advanced in New Zealand.
However, several other countries have initiated stud-
R. A. Skuce and S. D. Neill
ies on methodology and preliminary epidemiological
investigations (Collins 1998; Durr et al. 2000a,b; Skuce
and Neill 2001). Despite some similarities, the epide-
miology of bovine tuberculosis in other countries may
well be very different from the New Zealand situation.
For example, in Northern Ireland investigators have
favoured the RFLP and spoligotyping techniques to
disclose a surprisingly large number of M. bovis strain
types in a panel of isolates from different animal spe-
cies and from different geographical and temporal ori-
gins (Skuce et al. 1996; Roring et al. 1998). The RFLP
types and spoligotypes identified did not appear to
be host restricted, being isolated from cattle, badgers,
deer and humans. One strain was over-represented in
the panel of isolates, making up approximately 45% of
RFLP types and approximately 70% of spoligotypes.
Clonal expansion of such a strain is indicative of'suc-
cess' and this strain is likely to be host-adapted to
cattle, although not host-restricted. As with possum
and cattle isolates in New Zealand, badger and cattle
isolates in Ireland (Costello et al.1999) from the same
locality or premises often share the same strain type,
suggesting that a common pool of M. bovis strains is
circulating between domesticated, wild and feral ani-
mals. Without more precise epidemiological data and
statistically valid experimental design it is not possible
yet to demonstrate, or quantify, the relative extent, if
any, of transmission between cattle and wildlife, and
vice versa.
Mycobacterium bovis genotyping, using spoligo-
typing, has also shown that, in a significant number
of herds for which M. bovis isolates are available
for individual animals, multiple spoligotypes exist
within an outbreak, usually one predominant spo-
ligotype plus a few isolates of different spoligotypes
(Skuce et al. 1996). This suggests that either the DR
locus is not sufficiently stable within an outbreak or
that such bovine tuberculosis outbreaks are due to
multiple sources and that some types may be more
transmissible than others. The balance of evidence
suggests that such outbreaks are the result of infec-
tion from multiple sources, a phenomenon which
may be under-reported due to the predominance
of some spoligotypes, e.g. ST140. In a spoligotyping
study of more than 600 M. bovis isolates in Northern
Ireland, ST140 accounted for 70% of isolates, with
ST263 accounting for a further 20%. The remaining
five spoligotypes accounted for the remaining 10%
of isolates (Skuce et al., unpublished). The clonal
expansion of M. bovis isolates with the ST140 geno-
type is indicative of success. It may be speculated
that ST140 spoligotype M. bovis possess some selec-
tive advantage.
Molecular Epidemiology of Mycobacterium bovis
IS611O-RFLP, PGRS-RFLP, DR-RFLP and spoli-
gotyping were evaluated on a panel of 452 M. bovis
isolates recovered from cattle, deer, pigs, sheep and
goat in the Republic of Ireland (Costello et al. 1999).
RFLP identified 85 strains with one type, the same
type as identified in Northern Ireland, accounting for
20%. Spoligotyping identified 20 strains, with ST140
accounting for 52% of isolates. This major type was
recovered from all animal species tested and was
widely distributed geographically, suggesting that
transmission between these animal species was a
factor in the epidemiology of bovine tuberculosis in
Ireland. However, some strain types were geographi-
cally more clustered. Subsequently, IS611O- RFLP,
PGRS-RFLP, DR-RFLP and spoligotyping were com-
pared with pUCD-RFLP on a panel of 299 M. bovis
isolates (O'Brien et al. 2000a). Using a derived statis-
tical function for discrimination, pUCD-RFLP was
found to be more discriminating than the combined
RFLP method and spoligotyping. pUCD-RFLP and
DR-RFLP combined produced a more discriminating
and epidemiologically concordant method.
In Great Britain, the majority of isolates from
recent outbreaks in cattle have been spoligotyped and
mapped using the grid reference of the farm location
(Clifton-Hadley et al.I998). Thirty-four spoligotypes
were identified, 25 in cattle amongst 2668 isolates
collected mainly during 1996 and 1997. In agreement
with the Northern Ireland study, the same two spoli-
gotypes (ST140 and ST263) accounted for 70% of all
isolates. One spoligotype was detected in 623 herds
and 2-5 spoligotypes were detected in the remaining
86 herds, again suggesting multiple source outbreaks.
Fifty-three herds had isolates spoligotyped on repeat
occasions and in all but five instances, at least one
spoligotype was the same as the spoligotype initially
detected. This suggests that new sources were respon-
sible for introducing the different spoligotypes to
retested herds. Through epidemiological trace-back a
source for 426 of the outbreaks was proposed. In situ-
ations where badger M. bovis isolates were recovered
following an outbreak in cattle, 90 of 99 (91%) iso-
lates were of the same spoligotype. It was concluded
that spoligotyping was a useful genotyping method
but further discriminatory ability was desirable. A
study of M. bovis recovered from a well-characterised
badger population suggested that spoligotypes were
relatively stable within a location and over a length
of time, within a defined population and within
individual animals. Using Geographical Informa-
tion Systems (GISs) to manage the molecular typing
data, a marked degree of geographical clustering was
demonstrated for several spoligotypes, which can be
87
interpreted as evidence for a stable source of infec-
tion for cattle, such as wildlife.
In marked contrast to the spoligotyping studies
in Northern Ireland, the Republic of Ireland, Great
Britain and several other countries, an extensive spo-
ligotyping study of 1,349 M. bovis isolates collected
mainly from cattle in France over a 20-year period
(1979-2000) demonstrated that the technique had
very acceptable and useful discriminatory capacity
(Haddad et al' 2001). A large number (161) of spoli-
gotypes were found, relatively more than were found
in Great Britain, with two types responsible for 26 and
12% of isolates, respectively. The main spoligotype
pattern was described as BCG-like. This may not be
surprising, since M. bovis BCG was originally derived
from a French M. bovis isolate and was proposed as the
progenitor of further spoligotypes, which were linked
through the unidirectional loss of CRISPR spac-
ers. Spoligotypes were again mapped and there was
marked geographical clustering for several of them.
The remarkable diversity in spoligotypes was inter-
preted as reflecting the very low disease prevalence.
Additional factors, such as traditional trading in a
large number of different cattle breeds were also cited.
It is interesting to note that the 'classical European' M.
bovis BCG-like spoligotype pattern now dominates in
northern Cameroon and that the introduction of Cha-
rolais cattle from France in 1913 is an alleged source.
Although a small sample number, the spoligotype pat-
terns in Cameroon appear relatively homogeneous.
France is still the main exporter of cattle to Cameroon
(Njanpop-Lafourcade et al' 2001).
Molecular typing of M. bovis isolates in Tanzania
(Kazwala et al' 1998) led to the hypothesis that there
were two main categories of strains: autochthonous
strains (atypical cultural properties) and strains
imported from Europe (with classical cultural proper-
ties). Bovine tuberculosis has also become a very sig-
nificant disease problem of wildlife in the major game
reserves of South Africa, severely affecting nature
conservation and eco-tourism. The African buffalo
(Syncerus caffer) is a valuable big-game species which
is susceptible to M. bovis infection, making them a risk
for domestic cattle. Tuberculosis-free buffalo are also
more valuable, a major driving force in illegal trade.
The buffalo populations in Hluluwe-Umfolozi Game
Reserve (HUP) and the Kruger National Park (KNP)
are infected with M. bovis and act as a maintenance
host, responsible for spill-over of infection into local
wildlife. IS611 0- RFLP has been used to trace the origi-
nal source of infection of buffalo populations in the
KNP and in tracing the origin of infected animals that
were moved to other parts of South Africa. Therefore,
88 R. A. Skuce and S. D. Neill

the identification of the probable origin of animals ogy, evolution and pathogenesis of a major pathogen
may be possible by studying the strain types har- such as M. bovis. When integrated with developments
boured by animals, where more conventional animal in animal traceability (Houston 2001) and technolo-
identification methods have either failed or not been gies such as Global Positioning Systems (GPSs) and
applied (Vosloo et al. 2001). Geographical Information Systems (GISs), which
IS6110-RFLP and DR-RFLP were used to type are available in developed countries, they should
M. bovis isolates from cattle in Mexico and Texas provide powerful and challenging new informa-
(Perumaalla et al. 1996, 1999). The disclosure of tion with which to inform control strategies better.
multi-copy IS6110 strains implicated wildlife res- It is probable however that, as with M. tuberculosis,
ervoirs in disease transmission, a theory which has this equivalent of contact tracing may not be very
not been proven yet. In Argentina, the disclosure effective at detecting casual contacts and that strain
of 'cattle' strain types infecting humans identified typing data will provide alternative explanations for
an occupational zoonotic risk (van Soolingen et al. proposed sources and transmission routes.
1994). This contrasts with the current situation in These new tools should also prove valuable for
the Netherlands, where bovine tuberculosis has been identifying sources and routes of M. bovis trans-
successfully eradicated. Strain typing of M. bovis mission, thus providing baseline information on
isolates from humans showed that they were infected prevalence and distribution of M. bovis strain types
with 'multi-copy' IS611 0 strains, which were likely to in animal species. Integration into imaginative and
have been due to exposure to infected 'exotic' animals focussed epidemiological studies should help clarify
in zoos or animal parks. specific issues such as the contributions from inter-
Australia has succeeded in eradicating bovine bovine transmission and from wildlife. Availability
tuberculosis. However, prior to this, M. bovis strain of more precise strain recording should be employed
typing was used to propose potential sources and also to investigate: 'persistence' of infection within
transmission routes for infection on specific farm a herd; the issue of recrudescence of infection and
premises (Cousins et al. 1993, 1998b). Additionally, strain reintroduction, and therefore may ultimately
strain typing has been used to investigate other zoo- help to elucidate further the issue of latency in cattle
notic risks such as tuberculous captive seals, which referred to by Pollock and Neill (2002)
were shown to have infected a seal trainer having the Additionally, it would be of significant value to
same strain (Cousins et al. 1990). REA was used in know if particular M. bovis strains may be associated
Sweden to show that M. bovis was independently, and with particular epidemiological features observed
inadvertently, introduced into Sweden and New Zea- in outbreaks of this disease and if specific strains
land from Great Britain (Bolske et al. 1995). unusually influence the pathogenesis. Correlation
studies could be devised to compare biological prop-
erties of M. bovis isolates from, e.g., non-lesioned
cattle, from weak tuberculin-reacting or tuberculin-
6.11 negative animals exhibiting disease and from large
Conclusions multi-reactor outbreaks of tuberculosis. Identifica-
tion of key biological properties that influence strain
Obviously, the medical and veterinary pnonties behaviour in such ways could have significant impor-
between developing and developed countries can tance in informing control strategies, particularly at
differ significantly and hence their attitudes and the local level.
responses to animal tuberculosis and any potential For developing countries, molecular technologies
resulting zoonosis. Molecular epidemiology is a could be of equal value in the longer term. How-
relatively new discipline and particularly so when ever, in many instances, their use in surveillance is
considering M. bovis. There is now a need to inte- likely to be prohibitively expensive for general and
grate these new developments, which until recently widespread application. Nonetheless, collection and
have been largely 'technology-driven', with classical collation of standard data could inform on mat-
epidemiological studies. ters of animal and human health, which would be
Genetic tools to facilitate accurate diagnosis and a significant advance. The findings from molecular
high-resolution identification of species and strains epidemiological studies in developed countries will
of the M. tuberculosis complex now exist and are eventually yield significant information, which could
amenable to high-throughput studies. These should have an impact if extrapolated to developing regions
provide important new information about the biol- of the globe.
Molecular Epidemiology of Mycobacterium bovis
Acknowledgements. The authors would like to
acknowledge the contributions of colleagues within
the bovine tuberculosis statutory and research
programme at Veterinary Sciences Division (VSD),
Department of Agriculture and Rural Development
(Northern Ireland) and the Department of Veteri-
nary Science at the Queen's University of Belfast to
some of the work discussed in this chapter. This work
was funded by the Department of Agriculture and
Rural Development (DARD, Northern Ireland), the
Department of Environment, Food and Rural Affairs
(DEFRA, Great Britain) and the European Union. The
authors wish to thank Malcolm Taylor (VSD, DARD)
for his assistance with figures and graphics.
References
Alland D, Kalkut GE, Moss AR et al. (1994) Transmission of
tuberculosis in New York City. An analysis by DNA finger-
printing and conventional epidemiologic methods. N Engl
J Med 330:1710-1716
Aranaz A, Liebana E, Mateos A et al. (1996) Spacer oligo-
nucleotide typing of Mycobacterium bovis strains from
cattle and other animals: a tool for studying epidemiology
of tuberculosis. J Clin Microbiol 34:2734-2740
Aranaz A, Liebana E, Gomez-Mampaso E et al. (1999)
Mycobacterium tuberculosis subsp. caprae subsp nov.: a
taxanomic study of a new member of the Mycobacterium
tuberculosis complex isolated from goats in Spain. Int J Syst
Bacteriol 49: 1263-1273
Bauer J, Yang Z, Poulsen S, Andersen AB (1998) Results from
5 years of nationwide DNA fingerprinting of Mycobacte-
rium tuberculosis complex isolates in a country with a low
incidence of M. tuberculosis infection. J Clin Microbiol 36:
305-308
Behr MA (2001) A historical and molecular phylogeny of BCG
strains. Vaccine 17:915-922
Bennett R, Christiansen K and Clifton-Hadley R (1999)
Preliminary estimates of the direct costs associated with
endemic diseases of livestock in Great Britain. Prev Vet
Med 39:155-171
Benson G (1999) Tandem repeat finder: a program to analyse
DNA sequences. Nucleic Acids Res 27:573-580
Bolske G, Englund L, Wahlstrom H, de Lisle GW, Collins DM
and Croston PS (1995) Bovine tuberculosis in Swedish deer
farms: epidemiological investigations and tracing using
restriction fragment analysis. Vet Rec 136:414-417
Brosch R, Gordon SV, Eiglmeier T, Garnier T, Tekaia F, Yeramian
E, Cole ST (2000) Genomics, biology and evolution of the
Mycobacterium tuberculosis complex. In: HatfuII GF, Bloom
WR (eds) Molecular genetics of Mycobacteria. ASM Press,
Washington DC, pp19-36
Brosch R, Gordon SV, Marmiesse M et al. (2002) A new evo-
lutionary scenario for the Mycobacterium tuberculosis
complex. Proc Natl Acad Sci USA 99:3684-3689
Caimi K, Romano MI, Alito A, Zumarraga M, Bigi F, Cataldi
A (2001) Sequence analysis of the direct repeat region in
Mycobacterium bovis. J Clin Microbiol 39:1067-1072
89
Cheeseman CL, Wilesmith JW, Stuart FA (1989) Tuberculosis:
the disease and its epidemiology in the badger, a review.
Epidemiol Infect 103:113-125
Clifton-Hadley RS, Inwald J, Archer J, Hughes S, Palmer N,
Sayers AR, Sweeney K, van Embden JDA (1998) Recent
advances in DNA fingerprinting using spoligotyping:
epidemiological applications in bovine TB. Cattle Pract
6:79-82
Cole ST, Brosch R, Parkhill J et al. (1998) Deciphering the
biology of Mycobacterium tuberculosis from the complete
genome sequence. Nature 393:537-544
Collins DM (1998) Molecular epidemiology: Mycobacterium
bovis. In: Rutledge C, Dale J (eds) Mycobacteria-molecular
biology and virulence. Blackwell Science, Boston
Collins DM, de Lisle GW (1984) DNA restriction endonuclease
analysis of Mycobacterium tuberculosis and Mycobacterium
bovis BCG. J Gen Microbiol 130:1019-1021
Collins DM, de Lisle GW (1985) DNA restriction endonuclease
analysis of Mycobacterium bovis and other members of the
tuberculosis complex. J Clin Microbiol 21:562-564
Collins DM, de Lisle GW, Gabric DM (1986) Geographic distri-
bution of restriction types of Mycobacterium bovis isolates
from brush-tailed possums (Trichosurus vulpecula) in New
Zealand. J Hyg 96:431-438
Collins DM, Gabric DM, de Lisle GW (1988) Typing of Myco-
bacterium bovis isolates from cattle and other animals in
the same locality. NZ Vet J 36:45-46
Collins DM, Erasmuson SK, Stephens DM, Yates GF, de Lisle
GW (1993) DNA fingerprinting of Mycobacterium bovis
strains by restriction fragment analysis and hybridisa-
tion with the insertion elements IS1081 and IS6110. J Clin
MicrobioI31:1143-1147
Collins DM, Radford AJ, de Lisle GW, Billman-Jacobe H (1994)
Diagnosis and epidemiology of bovine tuberculosis using
molecular biological approaches. Vet Microbiol 40:83-94
Collins DM et al. (1999) DNA typing of Mycobacterium bovis
isolates from the Castlepoint area in New Zealand
Cosivi 0, Grange JM,Daborn CJ et al. (1998) Zoonotic tubercu-
losis due to Mycobacterium bovis in developing countries.
Emerg Infect Dis 4(1) CDC URL: http://www.cdc.gov/
ncidodleid/vol4no 11 cosivi.htm
Costello E, O'Grady D, Flynn E et ai. (1999) Study of restriction
fragment length polymorphism analysis and spoligotyping
for epidemiological investigation of Mycobacterium bovis
infection. J Clin MicrobioI37:3217-3222
Cousins DV (2001) Mycobacterium bovis infection and control
in domestic livestock. Rev Sci Tech Off Int Epiz 20:71-85
Cousins DV, Roberts JL (2001) Australia's campaign to
eradicate bovine tuberculosis: the battle for freedom and
beyond. Tuberculosis 81:5-15
Cousins DY, Francis BR, Gow BL et al' (1990) Tuberculosis in
captive seals. Bacteriological studies on an isolate belong-
ing to the Mycobacterium tuberculosis complex. Res Vet Sci
48:196-200
Cousins DV, Williams SN, Ross BC, Ellis TM (1993) Use of a
repetitive element isolated from Mycobacterium tuberculo-
sis in hybridization studies with Mycobacterium bovis as a
new tool for epidemiological studies of bovine tuberculo-
sis. Vet MicrobioI37:1-17
Cousins DV, Skuce RA, Kazwala RR, van Embden JD (1998a)
Towards a standardized approach to DNA fingerprint-
ing of Mycobacterium bovis. Int J Tuberc Lung Dis 2:
471-478
90
Cousins DV, Williams S, Liebana E et al. (1998b) Evaluation of
four DNA typing techniques in epidemiological investiga-
tions of bovine tuberculosis. 1Clin Microbiol 36: 168-174
Daborn q, Grange 1M and Kazwala RR (1996) The bovine
tuberculosis cycle: and African perspective. 1 Appl Bacte-
riol81 [Suppl):S27-S32
Dale IW, Brittain D, Cataldi AA et al. (2001) Spacer oligonucle-
otide typing of bacteria of the Mycobacterium tuberculosis
complex: recommendations for standardised nomencla-
ture. Int 1Tuberc Lung Dis 5:216-219
De Lisle GW, Collins DM, Loveday AS, Young WA, Iulian AF
(1990) A report of tuberculosis in cats in New Zealand,
and the examination of strains of Mycobacterium bovis
by DNA restriction endonuclease analysis. NZ Vet 1 38:
10-13
De Lisle GW, Yates GF, Collins DM, MacKenzie RW, Crews KB,
Walker R (1995) A study of bovine tuberculosis in domestic
animals and wildlife in the MacKenzie Basin and surround-
ing areas using DNA fingerprinting. NZ Vet 143:266-271
De Lisle GW, Mackintosh CG and Bengis RG (2001) Mycobac-
terium bovis in free-living and captive wildlife, including
farmed deer. Rev Sci Tech Off Int Epiz 20:86-111
Denny GO, Wilesmith IW (1999) Bovine tuberculosis in North-
ern Ireland: a case-control study of herd risk factors. Vet
Rec 144:305-310
Doran TI, Hodgson AL, Davies IK, Radford AI (1993) Charac-
terisation of a highly repeated DNA sequence from Myco-
bacterium bovis. FEMS Microbiol Lett 111:147-152
Durr PA, Clifton-Hadley RS, Hewinson RG (2000a) Molecular
epidemiology of bovine tuberculosis II: applications of
genotyping. Rev Sci Tech Off Int Epiz 19:689-701
Durr PA, Hewinson RG, Clifton-Hadley RS (2000b) Molecular
epidemiology of bovine tuberculosis I: Mycobacterium
bovis genotyping. Rev Sci Tech Off Int Epiz 19:675-688
European Commission Report (1998) Trends and sources of
zoonotic agents in animals, feedstuffs, food and man in the
European Union in 1997. Document no VI/8495/98-Rev 2
Feizabadi MM, Robertson ID, Cousins DV, Hampson DI (1996)
Genomic analysis of Mycobacterium bovis and other mem-
bers of the Mycobacterium tuberculosis complex by isoen-
zyme analysis and pulsed-field gel electrophoresis. 1 Clin
MicrobioI34:1136-1142
Frothingham R, Meeker-O'Connell WA (1998) Genetic diver-
sity in the Mycobacterium tuberculosis complex based on
variable numbers of tandem DNA repeats. Microbiology
144:1189-1196
Gallagher 1, Clifton-Hadley RS (2001) Tuberculosis in badgers:
a review of the disease and its significance for other ani-
mals. Res Vet Sci 69:203-217
Glynn IR, Bauer 1, de Boer S, Borgdorff MW, Fine PEM, God-
frey-Faussett P, Vynnycky E (1999a ) Interpreting DNA
fingerprint clusaters of Mycobacterium tuberculosis. Int 1
Tuberc Lung Dis 3:1055-1060
Glynn IR, Vynnycky E, Fine PEM (1999b ) Influence of sam-
pling on estimates of clustering and recent transmission of
Mycobacterium tuberculosis derived from DNA fingerprint
techniques. Am 1 EpidemioI149:366-371
Glynn IR, Whiteley 1, Bifani PI et al. (2002) CDC-worldwide
occurrence of Beijing/W strains of Mycobacterium tuber-
culosis: a systematic review. CDC URL:http://www.cdc.gov/
ncidod/EID/voI8n08/02-0002-G.htm
Goodchild AV, Clifton-Hadley RS (2001) Cattle-to-cattle trans-
mission of Mycobacterium bovis. Tuberculosis 81:23-41
R. A. Skuce and S. D. Neill
Gordon SV, Brosch R, Billault A, Garnier T, Eiglmeier K, Cole
ST (1999a) Identification of variable regions in the genome
of tubercle bacilli using artificial chromosome arrays. Mol
Microbiol 32:643-655
Gordon SV, Heym B, Parkhill J, Barrell B, Cole ST (1999b) New
insertion sequences and a novel repeated sequence in the
genome of Mycobacterium tuberculosis H37 Rv. Microbiol-
ogy 145:881-892
Gordon SV, Eiglmeier K, Garnier T, Brosch R, Parkhill 1, Barrell
B, Cole ST, Hewinson RG (2001) Genomics of Mycobacte-
rium bovis. Tuberculosis 81:157-163
Grange 1M (2001) Mycobacterium bovis infection in human
beings. Tuberculosis 81:71-77
Groenen PMA, Bunschoten AE, van Soolingen D and van
Embden IDA (1993) Nature of DNA polymorphisms in the
direct repeat cluster of Mycobacterium tuberculosis: appli-
cation for strain differentiation by a novel typing method.
Mol Microbioll0:1057-1085
Guerrero A, Cobo 1, Fortun 1et al. (1997) Nosocomial transmis-
sion of Mycobacterium bovis resistant to 11 drugs in people
with advanced HIV-l infection. Lancet 350:1738-1742
Haddad N, Ostyn A, Karoni C, Masselot M, Thorel MF, Hughes
SL, Inwald 1, Hewinson RG, Durand B (2001) Spoligotype
diversity of Mycobacterium bovis strains isolated in France
from 1979-2000.1 Clin MicrobioI39:3623-3632
Heersma HF, Kremer K, van Embden ID (1998) Computer
analysis of IS611 0 RFLP patterns of Mycobacterium tuber-
culosis. Methods Mol Bioi 101:395-422
Houston R (2001) A computerised database system for bovine
traceability
Hughes MS and Rogers M (1994) Vaccination of the badger
(Meles meles) against Mycobacterium bovis. Vet Microbiol
51:363-379
Hughes SL, Frothingham R, Inwald 1 et al. (2000) The develop-
ment of semi-automated fluorescent techniques for improved
molecular typing of Mycobacterium bovis. Abstract of the 5th
international meeting on bacterial epidemiological markers
(IMBEM), Noordwijkerhout, Netherlands, Sept 2000
Hughes VM, Stevenson K, Sharp 1M (2001) Improved prepa-
ration of high molecular weight DNA for pulsed-field gel
electrophoresis from mycobacteria. 1 Microbiol Methods
44:209-215
Kallenius G, Koivula T, Ghebremichael S et al. (1999) Evolution
and clonal traits of Mycobacterium tuberculosis complex in
Guinea-Bissau. 1Clin MicrobioI37:3872-3878
Kamerbeek 1, Schouls LM, Kolk A et al. (1997) Simultaneous
detection and strain differentiation of Mycobacterium
tuberculosis for diagnosis and epidemiology. J Clin Micro-
bioi 35:907-914
Kaufhold A, Podbielski A, Baumgarten G, Blokpoel M, Top 1,
Schouls L (1994) Rapid typing of group A Streptococci by the
use of DNA amplification and non-radioactive allele specific
oligonucleotide probes. FEMS Microbiol Lett 119:19-26
Kazwala RR, Daborn q, Kusiluka LIM, Iiwa SFH, Sharp 1M,
Kambarage DM (1998) Isolation of Mycobacterium species
from raw milk of pastoral cattle of the Southern HigWands
of Tanzania. Trop Anim Health Prod 30:233-239
Kazwala RR, Daborn q, Sharp 1M et al. (2001a) Isolation of
Mycobacterium bovis from human cases of cervical adeni-
tis in Tanzania: a cause for concern? lnt J Tuberc Lung Dis
5:87-91
Kazwala RR, Kambarage DM, Daborn q et al. (2001b) Risk
factors associated with the occurrence of bovine tubercu-
Molecular Epidemiology of Mycobacterium bovis
losis in cattle in the Southern Highlands of Tanzania. Vet
Res Commun 25:609-614
Kline SE, Hedemark LL, Davies SF (1995) Outbreak of tuber-
culosis among regular patrons of a neighbourhood bar. N
Engl J Med 333:222-227
Kremer K, van Soolingen D, Frothingham R et al. (1999) Com-
parison of methods based on different molecular epidemi-
ological markers for typing of Mycobacterium tuberculosis
complex strains: interlaboratory study of discriminatory
power and reproducibility. J Clin MicrobioI37:2607-2618
Kulaga S, Behr M, Musana K et al. (2002) Molecular epidemiology
of tuberculosis in Montreal. Can Med Assoc J 167:353-354
Livingstone PG (1992) Tuberculosis in New Zealand-current
status and control policies. Surveillance 19:14-18
Magdalena J, Vachee A, Supply P, Locht C (1998a) Identification
of a new DNA region specific for members of Mycobacte-
rium tuberculosis complex. J Clin MicrobioI36:937-943
Magdalena J, Supply P, Locht C (1998b) Specific differentiation
between Mycobacterium bovis BCG and virulent strains of
the Mycobacterium tuberculosis complex. J Clin Microbiol
36:2471-2476
Mazars E, Lesjean S, Banuls A-L, Gilbert M, Vincent V, Gicquel
B, Tibayrenc M, Locht C, Supply P (2001) High-resolution
minisatellit-based typing as a portable approach to global
analysis of Mycobacterium tuberculosis molecular epidemi-
ology. Proc Natl Acad Sci USA 98:1901-1906
Morris RS, Pfeiffer DU, Jackson R (1994) The epidemiology of
Mycobacterium bovis infections. Vet MicrobioI40:153-177
Musser JM, Amin A, Ramaswamy S (2000) Negligible genetic
diversity of Mycobacterium tuberculosis host immune
system protein targets: evidence of limited selective pres-
sure. Genetics 155:7-16
Neill SD (1994) Pathogenesis of Mycobacterium bovis infection
in cattle. Vet MicrobioI40:41-52
Neill SD, Cassidy J,Hanna J, Mackie DP, Pollock JM, Clements
A, Walton E, Bryson DG (1994) Detection of Mycobacte-
rium bovis infection in skin test-negative cattle with an
assay for bovine interferon-gamma. Vet Rec 50:229-232
Njanpop-Lafourcade BM, Inwald J, Ostyn A, Durand B, Hughes
S, Thorel M-F, Hewinson G, Haddad N (2001) Molecular
typing of Mycobacterium bovis isolates from Cameroon. J
Clin Microbiol 39:222-227
O'Brien R, Flynn 0, Costello E, O'Grady D, Rogers M (2000a)
Identification of a novel DNA probe for strain typing Myco-
bacterium bovis by restriction fragment length polymor-
phism analysis. J Clin MicrobioI38:1723-1730
O'Brien R, Danilowiez BS, Bailey L et al. (2000b) Character-
ization of the Mycobacterium bovis restriction frament
length polymorphism DNA probe pUCD and performance
comparison with standard methods. J Clin Microbiol 38:
3362-3369
O'Reilly LM, Daborn CJ (1995) The epidemiology of Myco-
bacterium bovis infections in animals and man: a review.
Tuberc Lung Dis 76 [SuppI1]:1-46
Perumaalla VS,Adams LG,Payeur J et al. (1996) Molecular epi-
demiology of Mycobacterium bovis in Texas and mexico. J
Clin Microbiol 34:2066-2071
Perumaalla VS, Adams LG, Payeur J et al. (1999) Molecular
fingerprinting confirms extensive cow-to-cow intra-herd
transmission of a single Mycobacterium bovis strain. Vet
Microbiol 70:269-276
Pollock JM, Neill SD (2002) Review: Mycobacterium bovis
infection and tuberculosis in cattle. Vet J 163:115-127
91
Pfeiffer DU, Jackson R, de Lisle GW, Collins DM, Morris RS
(1994) The use of genetic markers to clarify tuberculosis
epidemiology. Kenya Vet 18:293-295
Rastogi N, Legrand E, Sola C (2001) The mycobacteria: an
introduction to nomenclature and pathogenesis. Rev Sci
Tech OffInt Epi 20:21-54
Roring S, Brittain D, Bunschoten AE et al. (1998) Spacer
oligonucleotide typing of Mycobacterium bovis isolates
compared to typing by restriction fragment length poly-
morphism analysis using PGRS, DR and IS611 0 probes. Vet
MicrobioI61:111-120
Roring S, Scott A, Brittain D, Walker I, Hewinson G, Neill S,
Skuce R (2002) Development of variable-number tandem
repeat typing of Mycobacterium bovis: comparison of
results with those obtained by using existing exact tandem
repeats and spoligotyping. J Clin MicrobioI40:2126-2133
Salamon H, Behr MA, Rhee JT, Small PM (2000) Genetic dis-
tances for the study of infectious disease epidemiology. Am
J EpidemioI151:324-334
Saunders NA (1995) State of the art typing of Mycobacterium
tuberculosis. J Hosp Inf 29:169-176
Skuce RA, Neill SD (2001) Molecular epidemiology of Myco-
bacterium bovis: exploiting molecular data. Tuberculosis
81:169-175
Skuce RA, Brittain D, Hughes MS, Neill SD (1996) Differentia-
tion of Mycobacterium bovis isolates from animals by DNA
typing. J Clin MicrobioI34:2469-2474
Skuce RA, Brittain D, van Embden JD, Smith T, Sharp M,
Hewinson G, Rogers M, Garcia Marin J-F, Neill SD (1998)
Development of novel standardised methodology and
nomenclature for the identification of Mycobacterium
bovis strains. EU Contract SMT4 CT96 2097 Final Report
(EU Commission)
Skuce RA, McCorry TP, McCarroll JF, Roring SM, Scott AN, Brit-
tain D, Hughes SL, Hewinson RG and Neill SD (2002) Dis-
crimination of Mycobacterium tuberculosis complex bacteria
using novel VNTR-PCR targets. MicrobioI148:519-528
Small PM, van Embden JD (1994) Molecular epidemiology of
tuberculosis. In: Bloom BR (ed) Tuberculosis: pathogenesis,
protection and control. American Society for Microbiology
Small PM, Hopewell PC, Singh SP et al. (1994) The epidemiol-
ogy of tuberculosis in San Francisco. A population-based
study using conventional and molecular methods. N Engl
J Med 330:1703-1709
Smittipat N, Palittapongarnpim P (2000) Identification of pos-
sible loci of variable number of tandem repeats in Myco-
bacterium tuberculosis. Tubercle Lung Dis 80:69-74
Sola C, Filliol I, Legrand E, Rastogi N (2000) Recent develop-
ments of spoligotyping as applied to the study of epidemi-
ology, biodiversity and molecular phylogeny of the Myco-
bacterium tuberculosis complex. Path Bioi 48:921-932
Sola C, Filliol I, Gutierrez M C et al. (2001) Spoligotype data-
base of Mycobacterium tuberculosis: biogeographic distri-
bution of shared types and epidemiologic and phylogenetic
perspectives. Emerg Inf Dis http://www.cdc.gov?ncidod/
eid.vol7n03/solaGl.htm
Sreevatsan S, Pan X, Musser JM (1997) Restricted structural
gene polymorphism in the Mycobacterium tuberculosis
complex indicates evolutionarily recent global dissemina-
tion. Proc Natl Acad Sci USA 94:9869-9874
Struelens MJ (1996) Consensus guidelines for the appropri-
ate use and evaluation of microbial epidemiologic typing
systems. Clin Microbiol Infect 2:2-11
92
Supply P, Magdalena J, Himpens S, Locht C (1997) Identification
of novel intergenic repetitive units in a mycobacterial two-
component system operon. Mol MicrobioI26:991-1003
Supply P, Mazars E, Lesjean S, Vincent V, Gicquel Band Locht C
(2000) Variable human minisatellite-like regions in the Myco-
bacterium tuberculosis genome. Mol MicrobioI36:762-771
Tanaka MM, Small PM, Salamon H, Feldman MW (2000) The
dynamics of repeated elements: applications to the epidemi-
ology of tuberculosis. Proc Natl Acad Sci USA 97:3532-3537
Van Belkum A (2002) Molecular typing of micro-organisms:
at the centre of diagnostics, genomics and pathogenesis of
infectious diseases? J Med MicrobioI51:7-10
Van Belkum A, Struelens M, de Visser A, Verbrugh H,
Tibayrenc M (2001) Role of genomic typing in taxonomy,
evolutionary genetics and microbial epidemiology. Clin
Microbiol Rev 14:547
Van der Zanden AG, Kremer K, Schouls LM, Caimi K, Cataldi
A, Hulteman A, Nageikerke NJ, van Soolingen D (2002)
Improvement of differentiation and interpretability of
spoligotyping for Mycobacterium tuberculosis complex
isolates by introduction of new spacer oligonucleotides. J
Clin Microbiol 40:4628-4639
Van Embden JD, Cave MD, Crawford JT et al. (1993) Strain
identification of Mycobacterium tuberculosis by DNA fin-
gerprinting: recommendations for a standardized method-
ology. J Clin MicrobioI31:406-409
Van Embden JD, van Soolingen D, Heersma HF et al. (1996)
Establishment of a European Network for the surveillance
of Mycobacterium tuberculosis, MRSA and penicillin-
R. A. Skuce and S. D. Neill
resistant pneumococci (letter; comment). J Antimicrob
Chemother 38:905-907
Van Embden JDA, van Gorkom T, Kremer K et al. (2000)
Genetic variation and evolutionary origin of the direct
repeat locus of Mycobacterium tuberculosis complex bac-
teria. J BacterioI182:2393-2401
Van Heiden P (1998) Molecular epidemiology: human tuber-
culosis. In: Rutledge C, Dale J (eds) Mycobacteria-molecu-
lar biology and virulence. Blackwell Science, Boston
Van Soolingen D (2001) Review: molecular epidemiology of
tuberculosis and other mycobacterial infections: main
methodologies and achievements. J Intern Med 249:1-26
Van Soolingen D, de Haas PEW, Haagsma J et al. (1994) Use
of various genetic markers in differentiation of Myco-
bacterium bovis strains from animals and humans and
for studying epidemiology of bovine tuberculosis. J Clin
Microbiol 32:2425-2433
Van Soolingen D, Borgdorff MW, de Haas PEW et al. (1997a)
Molecular epidemiology of tuberculosis in The Nether-
lands: a nationwide study from 1993 through 1997. J Infect
Dis 180:726-736
Van Soolingen D, Hoogenboezem T, de Haas PE et al. (1997b) A
novel pathogenic taxon of the Mycobacterium tuberculosis
complex, Canetti: characterisation of an exceptional isolate
from Africa. Int J Syst BacterioI47:1236-1245
Van Soolingen D, van der Zanden AG, de Haas PW et al. (1998)
Daignosis of Mycobacterium microti infections among
humans by using novel genetic markers. J Clin Microbiol
36: 1840-1845
7 Tuberculosis in Special Groups
and Occupational Hazards
JAIME ESTEBAN
CONTENTS
7.1 Special Groups (1). Diabetics 93
7.2 Special Groups (2). Transplantation 94
7.3 Special Groups (3). Malignant Diseases 96
7.4 Special Groups (4). Drug Abusers 97
7.5 Special Groups (5). Correctional Facilities 98
7.6 Special Groups (6). Shelters 100
7.7 Special Groups (7). Immigrants 100
7.8 Occupational Hazards (1). Health Care Workers 101
7.9 Occupational Hazards (2). Laboratory Workers 104
7.10 Occupational Hazards (3). Silicosis 106
7.11 Occupational Hazards (4). Other Workers 106
7.12 Conclusion 107
References 108
At the beginning of the last quarter of the twentieth
century, tuberculosis seemed to be a disease in decay
in the developed world. The image was that of a slow
but progressive decrease in the prevalence of the dis-
ease due to the availability of a very effective therapy
and the increasing development of hygienic measures.
However, some years later, an increase in the number
of cases of tuberculosis was detected in the USA. This
fact, together with the appearance of cases of multi-
drug-resistant tuberculosis, which were very difficult
to treat, increased the interest of the scientific commu-
nity in this half-forgotten disease. In 1993, the WHO
declared tuberculosis a global emergency because of
the increasing number of cases throughout the world,
and more measures were taken in the fight against the
disease. The fact that tuberculosis is mainly a respira-
tory disease that can be transmitted airborne or by
droplets makes any person in contact with a patient
who suffers from active respiratory tuberculosis sus-
ceptible to infection. However, because of the nature
of transmission and the immune response of the
J. ESTEBAN, MD
Department of Medical Microbiology, Fundaci6n Jimenez
Diaz, Av. Reyes Cat6licos 2, 28040 Madrid, Spain
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
host against the pathogen, some population groups,
such as closed communities (correctional facilities,
shelters, jails, hospitals, miners and others) seem to
present unique characteristics. This is also the case in
patients with underlying conditions that make the risk
of active tuberculosis higher, either by immunosup-
pression or by local factors (silicosis, transplantation,
diabetes, malignancies, drug abuse). The following
pages describe the special characteristics of tubercu-
lous diseases in these groups, with a special emphasis
on protective measures.
7.1
Special Groups (1). Diabetics
Diabetes is a well-known risk factor for tuberculosis.
Diabetics have higher rates of tuberculosis than non-
diabetics (2-5 times higher) (Opsald et al.1961; Small
and Fujiwara 2001; Turner 1951). Although the reason
for this is not well known, there are several hypoth-
eses which try to explain this phenomenon. They
include alterations in pulmonary immune cell func-
tion (including altered polymorphonuclear cell func-
tion, reduced activity of monocyte-macrophages, in
particular impaired phagocytosis of intracellular
organisms and depressed lymphocyte function) and
the effect of diabetic microangiopathy in the capillar-
ies of the alveolar septa and alveolar arterioles. The
latter may affect local response to infection (Koziel
and Koziel 1995).
Tuberculous diabetics have the same symptoms
and signs as non-diabetic patients, and their evolu-
tion seems to be similar. However, several reports
describe atypical radiographic findings in diabetics,
such as a higher prevalence of cavitations and tuber-
culous lesions in lower lobes. These data have been
reported in several studies, which have in common
small sample sizes, the results being statistically non
conclusive (Morris et al. 1992). Two other reports
with a higher number of patients gave different
results. One of them supports the idea of a higher
94
incidence of lower lobe disease in diabetics of all
age groups (Perez-Guzman et al. 2000). The other
report showed no statistical difference between
all the groups, although a more detailed statistical
analysis suggested an association between lower lobe
disease and female gender. Another finding was that
lower lobe disease in patients older than 40 years was
more frequent in diabetics, whereas the frequency of
cavitations seemed to be higher in patients with insu-
lin-dependent diabetes (Bacakoglu et al. 2001). This
report also showed a lower frequency of smear-posi-
tive patients in diabetics compared to non-diabetics.
Given the higher incidence of tuberculosis in
diabetics, tuberculin testing should be carried out
in these patients. However, because of the risk of
anergy, appropriate controls must be done in poorly
controlled diabetics for a proper evaluation of the
results. Prophylactic isoniazid is indicated in patients
with a positive PPD skin test because of the high
risk of reactivation of the disease. Treatment regi-
mens are otherwise similar to those of non-diabet-
ics. As peripheral neuropathy is a secondary effect
of isoniazid, pyridoxine should be given with the
antibiotic and all these patients should be carefully
monitored for clinical symptoms of both neuropathy
and hepatic disease (Koziel and Koziel 1995).
Therapy of tuberculosis is the same as for non-
diabetic patients. However, there are data suggest-
ing higher relapse rates among diabetics, so longer
courses of therapy may be indicated in order to mini-
mise relapse incidence (Kameda et al. 1990). More-
over, a recent report establishes a higher incidence
of multidrug-resistant tuberculosis among diabetics
(Bashar et al. 2001). However, no other reports have
confirmed this finding, so changes in therapeutic
protocols are not currently indicated unless specific
epidemiological data suggests a high rate of multi-
drug-resistant tuberculosis in the area. In this case,
specific measures for proper management of the
cases must be implemented and, probably, antimi-
crobial susceptibility testing would be indicated for
all the isolates of M. tuberculosis from these patients.
7.2
Special Groups (2). Transplantation
Transplantation is a therapeutic procedure that is
increasingly being employed worldwide for numer-
ous diseases. However, it implies several procedures
to minimise the risk of rejection and most of them
produce an alteration in the immunity of the patient.
J. Esteban
The impairment of immune function is a circum-
stance that must be taken into account in all these
patients because it increases the risk of infections,
caused by both common and rare organisms. The
degree of immunosuppression and other compli-
cations is quite different between bone marrow
transplants and solid organ transplants and so is the
risk of infection, which is higher for bone marrow
transplantation. This is so because immunosuppres-
sive therapy is given before and after the transplant;
the risk is higher in allogeneic transplants than in
autotransplants. In the pretransplantation period
the presence of neutropenia and altered anatomical
barriers are the main risk factors for development of
infection. In the pre-engraftment period, aplasia due
to the conditioning regimen employed, represents the
major immune defect. In the third phase, neutrope-
nia abates and the anatomic barriers are restored,
although immune alterations persist. The presence
of graft-versus-host disease and its therapy in this
period represents further complications that increase
the risk of infection. Later, in a fourth period, chronic
graft-versus-host disease represents almost the only
risk factor for the development of infection (Sable
and Donowitz 1994).
Tuberculosis (a disease that is closely related to
immune function) was expected to be more frequent
among transplanted patients due to the broad impair-
ment of host immunity that they suffer. However, ini-
tial reports showed lower rates than those expected,
although they were higher than those for the general
population (Kurzrock et al.1984). An initial explana-
tion for this finding was that these studies were done
in countries with low global rates for tuberculosis
(Roy and Weisdorf 1997). However, recent reports
from countries with higher prevalence rates suggest
that the initial findings should be reviewed and that
tuberculosis rates in this group of patients are in fact
similar to the rates of the general population (Aljurf
et al. 1999; Budak-Alpdogan et al. 2000; George et al.
2001; Ip et al. 1998).
Despite these data on incidence, tuberculosis in
patients that underwent bone marrow transplanta-
tion had special characteristics compared to those
of general population, perhaps due to the immune
impairment that these patients suffer. There is a high
number of cases of extrapulmonary or disseminated
disease in many reports (Aljurf et al. 1999; George et
al. 2001), although in others respiratory disease is
the most frequent form of the disease (Ip et al. 1998;
Kurzrock et al. 1984). However, even respiratory
cases can have atypical presentations (Kurzrock et
al. 1984). These facts, together with the relatively low
Tuberculosis in Special Groups and Occupational Hazards
incidence of the disease can often mislead the phy-
sician, and in many cases therapy is delayed. Some
cases are diagnosed after the patient's death. Because
of the severity of the disease in these cases, an aggres-
sive approach should be used to diagnose the disease
as soon as possible.
Common therapeutic regimens can be used for
these patients. These regimens, however, must be
started as soon as possible, as diagnostic tests are not
always available on time.
A recent official report establishes the measures for
prevention of tuberculosis in this group (Centers for
Disease Control 2000). Measures include screening of
candidates by careful medical history and tuberculin
skin testing by the Mantoux method. For immunocom-
promised patients, a positive test is defined as 5 mm or
more of induration. Persons with a positive test or a
history of positive test must be evaluated for detection
of active tuberculosis. If no disease is detected, a full 9-
month course of isoniazid chemoprophylaxis must be
performed. This regimen is also indicated for patients
who have been in close contact with an active tubercu-
losis case, regardless of skin test status. Routine anergy
testing is not indicated in these patients. An alterna-
tive for isoniazid therapy is a 2-month course of daily
pyrazinamide/rifampin, although there are only lim-
ited data about the safety and efficacy of this. It must
also be taken into account that rifampin interacts with
many drugs, such as cyclosporine or corticosteroids,
so routine use of this regimen is not recommended.
BCG vaccination is also contraindicated in these
patients, because the BCG strain might cause severe
(even fatal) disease in immunocompromised patients
(Abramowsky et al.1993).
Solid organ transplantation includes a broad
range of procedures with variable degrees of immu-
nosuppression that depends on the organ to be trans-
planted. Prevalence of tuberculosis among trans-
planted patients in developed countries is as high as
1% (Patel and Paya 1997), which is higher than that
for the general population in these countries (Singh
and Paterson 1998).
Most cases of tuberculosis have been reported in
renal transplant recipients, although cases have also
been reported in liver, heart and lung transplanted
patients (Singh and Paterson 1998). Unlike other
infections, tuberculosis in these patients is not related
to the time since the transplantation procedure (Patel
and Paya 1997). However, the disease appeared earlier
in heart, liver and lung transplant recipients than in
renal ones.
Although transmission of tuberculosis through
the graft has been reported (Graham et al. 2001;
95
Mourad et al. 1985), this mechanism of infection
is rare, and reactivation and primary infection are
the most common mechanisms of acquisition of the
disease. There have also been reports of outbreaks of
tuberculosis in transplantation programmes (Sund-
bery et al.199l).
The clinical aspects of the disease include a high
number of cases of extrapulmonary or dissemi-
nated tuberculosis (49%) (Singh and Paterson 1998).
Transplantation of organs that imply a high level of
immunosuppression, such as heart transplant, show
an even higher proportion of extrapulmonary cases
(Munoz et al. 1995). Among extrapulmonary cases,
presentations include gastrointestinal, hepatic,
skin, muscle, osteoarticular, genitourinary, ganglial,
central nervous system and other more uncommon
sites of disease such as larynx, tonsils or eye (Patel
and Paya 1997; Singh and Paterson 1998). Clinical
data from these patients were similar to those from
the general population. Mortality of tuberculosis in
transplanted patients is 30%. Disseminated disease,
prior rejection and treatment with OKT3 or anti-T
cell antibodies are significant predictors of mortality
in these patients.
Clinicians should suspect tuberculosis in patients
who present with fever and radiographic infiltrates
or symptoms of extrapulmonary disease. Smears for
acid-fast stains and mycobacterial cultures should
be performed on appropriate specimens. Pulmonary
cases require an aggressive approach if the results of
sputum testing are inconclusive because these patients
have a high mortality risk. Procedures to be performed
include bronchoscopy, including bronchoalveolar
lavage or transbronchial biopsy, or even open lung
biopsy (Patel and Paya 1997). Other techniques (bone
marrow biopsy, liver or other organ biopsies, blood
cultures for mycobacteria) are also useful when extra-
pulmonary or disseminated tuberculosis is suspected,
although the results of some of these (for example,
blood cultures) can take a long time (Esteban et al.
200lb). Samples should also be sent for pathologi-
cal evaluation, looking for suggestive lesions such as
granuloma formation.
The treatment of transplanted patients with
tuberculosis must be performed with at least two
bactericidal drugs and during a longer period of
time (l2 months) than that for common regimens
(Patel and Paya 1997). Hepatotoxicity is a side effect
of great importance in the context of liver trans-
plantation, so these patients must be monitored
carefully, including the performance of biopsies to
detect hepatitis or graft rejection. Published data has
shown that hepatotoxicity requiring discontinuation
96 J. Esteban

of isoniazid therapy occurs in 41 % of liver transplant

Patients with leukaemia have a higher risk of devel-
patients. The percentages of hepatotoxicity for other oping disseminated or extrapulmonary disease com-
patients were lower (Singh and Paterson 1998). It is pared to those with other haematological tumours
also of interest to note that antimycobacterial ther- (Libshitz et al. 1997; Morii and Narita 1998). Another
apy has important drug interactions. Rifampin can report mentions a high incidence of tuberculosis in
induce hepatic enzymes and accelerate metabolism hairy cell leukaemia (Advani and Banavali 1989),
of drugs such as cyclosporine or corticoids, so close with incidences as high as 50% for patients suffering
monitoring of drug levels must be performed (Patel from this disease. The incidence of these manifesta-
and Paya 1997). tions is also higher in patients with lymphomas than
Tuberculin skin testing is indicated both in trans-in the general population, especially in Hodgkin
plant recipients and, if possible, in donors. A historylymphoma. However, in these diseases extrapulmo-
of tuberculosis exposure or past disease must also nary tuberculosis is usually a focal disease and not a
be recorded. Although no study has demonstrated disseminated one (Melero et al. 1992).
the efficacy of isoniazid prophylaxis in tuberculin In bone marrow transplanted patients (a group
skin test positive transplanted patients, this therapy of patients that often suffer from malignancies as
has been reported to minimise the risk of develop- well) the disease has unique characteristics that
ing tuberculosis in renal patients (John et al. 1994). make diagnosis very difficult, such as involvement of
However, the indication of chemoprophylaxis must extrapulmonary organs or atypical signs and symp-
be balanced against the risk of hepatotoxicity. If toms. Atypical radiographic signs can also appear in
possible, isoniazid should be given several months pneumonia (Libshitz et al. 1997). Some of these pre-
prior to transplantation. Other patients that should sentations can go unrecognised until the death of the
be considered as candidates for chemoprophylaxis patient (Libshitz et al. 1997; Morii and Narita 1998;
include those with skin test conversion, a history of Muller et al. 1980; Rosenthal et al. 1975). An aggres-
inadequately treated tuberculosis, close contacts with sive approach must be taken for an early diagnosis of
active tuberculosis cases, patients with chest radio- these patients, including the taking of biopsies or the
graph abnormalities or the immunocompromised performance of bronchoscopic procedures. Proper
host. Prophylaxis should also be considered for therapy must then be performed using common
patients who receive grafts from donors with a his- schemes. Screening using the Mantoux technique
tory of tuberculous infection or disease (Lichtenstein must be done in patients who are going to receive
and MacGregor 1983; Qunibi et al. 1991). aggressive chemotherapy, as the latter can induce
severe immunosuppression (Morii and Narita 1998).
In patients with a positive skin test, chemotherapy
with isoniazid is indicated.
7.3 In solid organ cancers, tuberculosis is mainly
Special Groups (3). Malignant Diseases pulmonary, and extrapulmonary or disseminated
disease is less common (Libshitz et al. 1997). The
Tuberculosis in cancer patients does not seem to have highest incidence is seen in head and neck cancer,
very different characteristics to that in the general lung cancer and gastrointestinal cancer (Kumar et
population, despite an increased incidence of the dis- al. 1999; Libshitz et al.I997). Pulmonary tuberculosis
ease (Libshitz et al. 1997). However, some neoplastic and lung cancer can appear simultaneously, and in
diseases have special characteristics that can affect some series the relation between these diagnoses is
the course of the disease. more frequent than other associations (Libshitz et
Haematological malignancies have a higher al. 1997). Nevertheless, in other series the incidence
incidence of tuberculosis than solid organ cancers of simultaneous diseases and sequential diseases is
(Libshitz et al. 1997; Melero et al. 1992). Among similar (Tamura et al. 1999). The location of tubercu-
these malignancies, acute leukaemia and Hodgkin losis and lung cancer can be the same (in one series
lymphoma have an even higher incidence than other M. tuberculosis was cultured from collapsed lobes in
diseases (Advani and Banavali 1989; Libshitz et al. patients with lung cancer (Libshitz et al. 1997» or
1997; Melero et al.I992). The patients with these hae- different (Tamura et al. 1999). A normal radiographic
matological disorders have impaired immunity. This pattern can appear if HIV infection is associated with
is due to the disease itself and to cancer treatments, cancer and tuberculosis (Libshitz et al.I997). Because
which can include radiotherapy, chemotherapy, of the risk of simultaneous infection and neoplastic
corticosteroids and even bone marrow transplant. disease, samples for diagnosis should be sent to both
Tuberculosis in Special Groups and Occupational Hazards
microbiology and pathology laboratories. If cancer
and tuberculosis appear in a sequential pattern, the
disease seems to be more severe. This could be due
to a greater debilitation of the patients during cancer
therapy (Libshitz et al. 1997).
7.4
Special Groups (4). Drug Abusers
Drug abusers have distinct characteristics that affect
the course and management of concurrent tubercu-
losis. Many of them are outcasts and have antisocial
behaviours, such as criminality or erratic behaviour
(including auto-medication or even trading with the
pills to obtain money or drugs), which are a serious
hazard to the correct treatment of any disease. An
important number of them also have other important
risk factors, such as HIV infection, malnutrition or
bad hygienic practices. The close association between
HIV infection and intravenous drug abuse makes it
difficult to differentiate between the role of the two
factors in the development of the disease.
Relatively little is known about the immune
system in drug abusers. In the case of HIV-infected
patients, the effects on the immune system of the
HIV are so important that they minimise other pos-
sible defects. However, in HIV-non-infected patients,
immune defects can also be found. In this group of
intravenous drug abusers serum IgM and IgG levels
are frequently elevated (Brown et al. 1974), and this
elevation is usually accompanied by autoantibod-
ies. The repeated injection of antigens is a possible
explanation for the hypergammaglobulinaemia. T-
cell function is also altered, as several in vitro studies
with T-Iymphocytes have shown (Levine and Brown
2001). Delayed hypersensitivity skin test reactions
are sometimes diminished or absent. Both increases
and decreases in the T-Iymphocyte cell count have
been reported. Morphine depresses several mono-
cyte functions affecting antiviral defence and it also
inhibits the monocyte response to activating stimuli
(Stoll-Keller et al. 1997). Some studies describe the
effects of morphine and, to a lesser degree, metha-
done on the neutrophil-monocyte system. The clini-
cal implications of these in vitro findings are, how-
ever, unknown (Levine and Brown 2001).
The relation between tuberculosis and intrave-
nous drug abuse has been well known for many years
(Perlman et al. 1995). Recent reports show rates of
44% positive Mantoux test for HIV-negative intrave-
nous drug users (Daley et al.1998) and a 6.8% rate for
97
tuberculous disease among intravenous drug abusers
(Friedman et al. 1996). Other studies have described
drug abuse as a statistically significant risk factor for
tuberculosis in the general population (Iftigo-Marti-
nez et al. 2000; Moro et al. 2002). Drug abuse is linked
to unemployment and low socio-economic status. In
these population groups tuberculosis is a prevalent
disease. Many intravenous drug users also have a
previous history of imprisonment or homelessness.
These circumstances increase the risk of tuberculo-
sis, as will be explained further in the chapter.
The AIDS epidemic has also affected the relation
between drugs and tuberculosis. AIDS patients have a
high risk for development of tuberculosis, both when
they are positive for the tuberculin skin test and when
they are anergic (Selwyn et al. 1992). In these patients,
"atypical" presentations of the disease are also found,
including a higher percentage of extrapulmonary
or disseminated disease (Havlir and Barnes 1999;
Hill et al. 1991; Small and Fujiwara 2001). In recent
years, outbreaks of multidrug-resistant tuberculosis
with extremely high mortality have been reported
(Perlman et al. 1995). In these outbreaks, primary
progressive disease has been described and many of
the patients in these outbreaks were also intravenous
drug abusers.
Diagnosis of tuberculosis in intravenous drug
abusers is affected by the presence of concomitant
HIV infection and also by the association of other
infections at the moment of diagnosis. This can mask
clinical data, so clinicians should rule out tuberculo-
sis in any drug abuser who presents with fever (Ham-
burg and Frieden 1994). Screening of tuberculous
infection is also difficult in these patients as HIV-
infected patients are generally anergic. In this setting
a single skin test does not predict the development
of tuberculosis (Friedman et al. 1996; Perlman et al.
1995; Selwyn et al. 1992). Moreover, recent data sug-
gest that anergy tests are not reproducible, so it is not
recommended to perform these tests in HIV patients
anymore (Havlir and Barnes 1999).
The use of chemoprophylaxis in infected patients
has been proved to be effective in preventing the
development of active disease (Gourevitch et al.1998;
Selwyn et al.1992; Snyder et al.I999). However, recent
studies have shown that chemoprophylaxis does not
reduce the incidence of disease in HIV anergic
patients, so in these cases it is not currently indicated
(Daley et al.1998; Havlir and Barnes 1999).
Treatment of the active disease has also been
shown to be effective (Small and Fujiwara 2001),
although in HIV patients some modifications should
be made. In these patients, therapy with protease
98
inhibitors and reverse-transcriptase inhibitors for
the treatment of HIV infection complicates the
therapy of tuberculosis because they interact with
rifampin. The administration of this antibiotic can
result in sub-therapeutic levels of antiretroviral
drugs. In these cases, rifabutin, a rifampin derivative,
can be used with efficacy instead of rifampin in the
common regimens (Havlir and Barnes 1999; Small
and Fujiwara 2001).
However, the main problem in performing a proper
anti-tuberculous therapy remains in the low compli-
ance of these patients. A great number of intravenous
drug abusers abandon therapy, and the sequels of
this include therapeutic failure, prolonged infectiv-
ity, increased rates of transmission and the develop-
ment of drug-resistant strains (Perlman et al. 1995).
Several mechanisms have been developed in order to
increase compliance among these patients. Directly
observed therapy (DOT) is a strategy that has proved
useful for the treatment of low compliance patients
(Gasner et al. 1999; Perlman et al. 1995). However,
some patients refuse any therapy. Legal measures
have been developed for these cases, including the
arrest of these patients during the treatment period
(Gasner et al. 1999). These measures, however, must
be restricted to only a few "impossible" patients who
refuse any therapeutic measure (including DOTs), so
arrest should remain only as the last resort to cure
patients in tuberculous programmes.
7.5
Special Groups (5). Correctional Facilities
Prisons and other correctional facilities, such as jails,
penal colonies and prisoner of war camps, represent a
very special community for different reasons. Prison-
ers are often young males from socio-economically
disadvantaged groups, poor educational backgrounds
and ethnic minorities, including illegal immigrants.
Prison conditions frequently include overcrowding,
high incidence of HIV-infected individuals and low
standards of hygiene. In addition, several behaviours,
such as drug abuse and sex between men, often pro-
mote illness. Many of these circumstances, for exam-
ple overcrowding, increase the risk of transmission
of airborne infectious diseases, and HIV infection is
considered the strongest known risk factor for the
development of tuberculosis (Centers for Disease
Control 1996). Moreover, correctional facilities are
not closed environments; so many people circulate
between the prison and the open society (Fig. 7.1).
J. Esteban
These include liberated prisoners, but also workers
and visitors. A history of previous imprisonment
is one of the statistically significant risk factors for
tuberculosis in some reports (Inigo-Martinez et al.
2000; Moro et al. 2002).
Prisons are among the communities with the
highest incidence of tuberculosis. Many reports have
published annual case notification rates as high as
7200 cases/l00,000 (Coninx et al. 2000; Koffi et al.
1997; Wares and Clowes 1997) and even in developed
countries, rates are several times higher than those of
the general populations (Jones and Schaffner 2001;
Fernandez-Martin et al. 2000; Centers for Disease
Control 1996) (Table 7.1). An obvious consequence
of this is the high risk of becoming infected for all
the people who dwell in these institutions, especially
when the period of imprisonment is long. Molecular
epidemiology analysis confirmed internal transmis-
sion of tuberculosis inside correctional facilities,
with more then 25% of the patients appearing in clus-
ters (Fernandez-Martin et al. 2000; Laniado-Laborin
2001). This is true even if the time that the prisoner
spends in the correctional facility is shorter than in
prisons (Jones et aI.1999).
Other problems include delay in the diagnosis
of sputum smear-positive cases, failures in medical
services, transfer of infectious patients inside and
between prisons, and patients with low compliance for
therapy. The diagnosis of these patients can be difficult
because of the concomitant existence of other factors,
such as HIV infection, that can increase the possibil-
ity of extrapulmonary, disseminated or even atypical
forms of the disease. Other factors that could affect
the diagnosis are of another nature: in some prisons,
patients with disease sell smear-positive sputa to other
prisoners so that they can avoid hard labour (Drob-
niewski 1995) or have access to better living conditions
Non-infected Prisoners
i __________
Infected"Prisoners ----------. Prisoners with active tuberculosis
Non-infeited prison workers
t
Infected prison workers Outer Society
~
Prison workers with active disease
/
_ : Risk of contagion
-+ : Possible evolution of the patients
Fig.7.1. Relation between different kinds of patients in the
correctional setting
Tuberculosis in Special Groups and Occupational Hazards
Table 7.1. Incidence of tuberculosis in correctional facilities
Location of correctional facility Annual rate
(year) (per 100,000)
Tomsk, Russia (1996) 7,000
Azerbaijan (1994) 4,667
Bouake, Cote d'ivoire (1990-1995) 7,200
Madrid, Spain (1997-1998) 693
Memphis, USA (1995-1997) 274
New York, USA (1993) 139.3
in the prison hospital. Power structures in prisons also
have direct implications in the management of tuber-
culosis, affecting detection and treatment of patients
(Reyes and Coninx 1997). The therapy of diagnosed
cases is also problematic. Some patients use the pills
for trading with other prisoners or even correctional
workers. Other prisoners manage to obtain the pills
and take the drugs without proper medical and con-
trol. This inadequate therapy can lead to the develop-
ment of resistant strains. This risk must be added to
the bad compliance of many patients that implies the
abandonment of therapy after a few weeks, especially
if the prisoner is liberated and no specific programme
exists to co-ordinate the correctional medical struc-
ture efforts with the open society structures. Some-
times, when anti-tuberculous drugs are scarce, a black
market for these drugs can appear, especially if tuber-
culosis is an important problem in the facility (Coninx
et al. 2000). This fact can be of extreme importance in
the prisons of developing countries, where drug are
not always available.
Mobility of prisoners is another factor that can
extend the disease between prisons, as prisoners can
be transferred when they are still infectious (Reyes
and Coninx 1997). This contributes to the transmis-
sion of multidrug-resistant strains, as in New York in
1991 (Drobniewski 1995).
Proper measures must be implemented to mini-
mise the problem of tuberculosis inside correctional
facilities. These measures are based in a few clear
points: rapid detection of patients with tuberculous
disease, isolation and therapy for these cases, detec-
tion of infected patients and the institution of che-
moprophylaxis when indicated.
The pivotal measure for tuberculosis control is the
active detection of infectious cases. This detection
must be done using screening tests, such as symptom
screening, tuberculin skin test or chest radiograph
screening (Centers for Disease Control 1996). The
selection of the initial screening test must be done
according to the individual characteristics of each
centre.
99
Symptom screening is useful in areas where tuber-
culosis prevalence is high. The presence of prolonged
productive cough, fever, haemoptysis and other sys-
temic symptoms (malaise, weight loss, etc.) suggest
the possibility of active tuberculosis. In these cases,
proper diagnostic measures must be performed,
including radiological and microbiological tests, and
also a careful medical history taken about previous
episodes or treatments.
The performance of chest radiographs for all
symptomatic patients is always indicated. There
are also studies showing cost-effective use of chest
radiographs for all the prisoners in areas with high
prevalence (Jones and Schaffner 2001). The results of
chest radiography must be interpreted together with
a careful medical history, and other diagnostic tests
are indicated if there is suspicion of active tubercu-
losis.
The Mantoux test is the preferred tuberculin skin
test. However, its interpretation must be cautious in
this setting, because there are areas where BCG vac-
cination is currently used. The high prevalence of
HIV-infected patients can also affect the results, as
severely immunocompromised patients have nega-
tive test results. All patients with positive skin test
should be medically evaluated and other tests (like
radiographs or sputum analysis) must be performed
if indicated. Negative skin test individuals without
anergy should be tested again periodically.
Screening tests should be done both for correc-
tional inmates and for correctional workers, because
both groups have a high risk of infection. The earlier
the detection of these patients, the less is the risk of
infection for other people, as diagnosis must lead to
proper isolation measures and initiation of therapy.
Isolation measures are recommended for patients
with respiratory tuberculosis. They will need trans-
port to hospital if the prison lacks proper isolation
measures.
Therapy must be performed according to common
protocols. It must be directed observed therapy
(DOT), as there is a low level of compliance in many
patients. DOT assures therapy of the patients and
minimises the risk of developing resistances.
Preventive therapy with isoniazid (6-12 months)
is recommended also for infected individuals with-
out active disease. Anergic HIV-infected patients
should also receive prophylaxis because of the high
risk of development of active tuberculosis in these
patients (Selwyn et al. 1992).
Co-ordination between correctional administra-
tions and health programmes outside correctional
facilities is essential, as many patients abandon therapy
100
when they are liberated. Active searching and follow-
up of these patients by tuberculosis programmes in the
different communities should ensure the proper ther-
apy of these patients and minimise the risk of relapse,
therapeutic failures or development or resistances
(Maher 1998; Centers for Disease Control 2000).
7.6
Special Groups (6). Shelters
Homelessness is one of the risk factors for the devel-
opment of tuberculosis, with estimated prevalences
from 1.6% to 6.8% (Centers for Disease Control
1992a). Among these individuals, other risk factors
for tuberculosis are found, such as malnutrition, poor
hygiene, immigration from countries with high prev-
alence of the disease, substance abuse (both alcohol
and/or intravenous drugs) and, in recent years, HIV
infection (Centers for Disease Control 1992a; Moss et
al. 2000). It is difficult to determine the true role of
each of these factors, as they all playa role. However,
an important fact is the recent finding of a high pro-
portion of clustering when molecular epidemiology
studies are applied to this population (Gutierrez et
al. 1998; Moss et al. 2000). These data suggest a high
proportion of transmission of tuberculosis in this
population. Shelters and other provisional dwellings
are considered as foci where transmission of the
disease occurs and shelter-associated tuberculosis
outbreaks have been described (Nardell et al. 1986).
Because of the difficulties of establishing an ade-
quate prevention programme in shelters, measures
should be directed at rapid diagnosis and treatment of
infectious cases (Stead 1989). Clinical characteristics
of tuberculosis in this group are similar to the gen-
eral population and variations in them are due to the
presence of other concomitant factors, such as HIV
infection. Thus, people with symptoms and signs of
pulmonary tuberculosis should be studied with chest
radiographs and, if available, proper microbiological
studies. When the diagnosis of tuberculosis is made,
therapy should be started using common therapeu-
tic regimens (Centers for Disease Control 1992a).
The most important variable in the outcome of these
patients is the low compliance that many of them
have. Directly observed therapy is indicated for these
difficult cases, and even hospitalisation throughout
all the treatment is indicated in individuals that have
mental diseases that incapacitate them to follow the
therapy. Several incentives have been recommended
to increase the adherence to therapy, such as travel
J. Esteban
tickets, food, clothes or even money. These incentives
have proven to be effective in increasing the compli-
ance of these patients (Centers for Disease Control
1992a; Moss et al. 2000).
Contact study of the infective cases is more dif-
ficult. The mobility of this population makes them
difficult to study. Other troublesome factors are the
use of aliases by the patient or the potentially infected
persons. Tuberculin skin test positive patients who
were treated with isoniazid are not entirely free of
risk, as shown by one of the first studies to describe
an outbreak among these populations (Nardell et al.
1986). The frequent description of clustering sup-
ports this finding, even in populations with high
prevalence of latent infection (Moss et al. 2000).
However, if tuberculin skin testing is performed, the
concomitant presence of factors such as HIV infec-
tion among these patients must be taken into account
(Centers for Disease Control 1992a).
Structural measures in the shelters have also been
considered as useful to prevent transmission of the
disease. The use of ultraviolet lamps has been recom-
mended to disinfect the air. These lamps, however,
must be installed appropriately to minimise the risk
of eye and skin lesions. Enhanced ventilation systems
could also help to reduce the airborne transmission
of the disease (Centers for Disease Control 1992a).
7.7
Special Groups (7). Immigrants
In the early years of the twenty-first century, immi-
gration from many countries to more developed
ones has become a problem of great importance. The
migratory movements of populations between coun-
tries have become more frequent, and this represents
an easy route for transmission of infectious diseases
from one place to another. Tuberculosis prevalence
is higher in countries from Africa and Asia, accord-
ing to the WHO data (World Health Organization
2001). Movements affecting people from these high
prevalence areas are of great interest because they
can affect the control programmes of other countries.
As a matter of fact, they have been considered the
cause of resurgence of the disease in Western coun-
tries (Rieder et al. 1994). In our hospital, the relative
percentage of immigrants with tuberculosis has gone
from 5.05% in 1995 to 21.8% in 2000. This data prob-
ably explains the maintenance of tuberculosis rates
in our area, despite the decreasing incidence in the
local population (Dr. M.C. Alvarez-Castillo, personal
Tuberculosis in Special Groups and Occupational Hazards 101

communication, and Epidemiological Bulletins from grants and transmission between them (Lillebaek et
the Comunidad de Madrid) (Figs. 7.2 and 7.3). al. 2001).A possible explanation is the low social con-
Illegal immigration adds more problems to the tact that exists between the two communities. Other
control of the potentially infected population. A reports suggest that tuberculosis in immigrants is the
study from Spain, a country frequently crossed by result of reactivation and not of primary disease due
immigrants en route to other European countries, to recent infection (Chin et al. 1998; Moro et al. 2002;
detected that 78% of the immigrants with active Samper et al. 1998).
tuberculosis were illegal ones (Huerga et al. 2000). Drug resistance among these patients is also a risk
This population involves more factors that compli- to be taken into account. Resistance rates in many
cate the management of patients, such as overcrowd- countries from which immigrants go to Western states
ing, language difficulties, poverty, fear to go to health are high (The WHO/IUATLD Global Project on Anti-
care facilities and frequent imprisonment. All these tuberculosis Drug Resistance Surveillance 1994-1997,
factors probably affect the therapy compliance in 1997). When these patients develop tuberculous dis-
these patients, and compliance is even lower for che- ease this data is of great importance in the planning
moprophylaxis of infected patients without active of proper therapeutic regimens. Multidrug-resistance
disease (Duran et al. 1996). However, strict follow-up is also important in these patients (Fig. 7.4). Antimi-
and other complementary measures, such as the use crobial susceptibility testing is indicated in patients
of the patient's language for examinations, made it who come from world areas with high resistance rates
possible to increase the compliance of tuberculous against anti-tuberculous drugs, as these rates are car-
patients to 78% (Huerga et al. 2000). ried on by this group of patients (Huerga et al. 2000).
The potential transmission of tuberculosis from
immigrants to local population is a theoretical
risk. However, a Danish study showed only minor
transmission between these populations. This study 7.8
performed a molecular epidemiology analysis Occupational Hazards (1).
using both RFLP and spoligotyping, and found only Health Care Workers
nine minor clusters in both Danish and immigrant
patients. Many other clusters were found, suggesting Among the different professions, health care workers
the existence of predominant strains between immi- are at this time almost unanimously considered a spe-

50
45
40
35
30
25
20
15 District 7
10 Fig. 7.2. Incidence of tuberculosis (No. cases/lOO,OOO
5 inhabitants). Comunidad de Madrid and District 7.
o 1996-1999
1996 1997 1998 1999

Fundaci6n Jimenez Diaz Fig. 7.3. Percentage of tuberculosis in

immigrants. Fundaci6n Jimenez Diaz,
5L--~-­ Comunidad de Madrid and District 7-
o Madrid (District 7 includes Fundaci6n
1996 1997 1998 1999 Jimenez Diaz) 1996-1999
102
Fig.7.4. Multidrug-resistant tuberculosis in an immigrant
patient
cial kind of worker with a higher risk of tuberculous
infection than other workers or the general popula-
tion. However, this common idea has not always been
considered an unquestionable fact. It was not until
the decade of 1950s that the increased risk for health
care workers was recognised, in the years before that
there were even several authors that considered the
health care worker a profession with a lower risk of
being infected than other people (Sepkowitz 1994).
Nowadays, at least in developed countries, health
care facilities are institutions where contact between
workers and tuberculous patients is a frequent
matter, chiefly in those hospitals where the number
of tuberculous patients is high. In our experience, the
majority of diagnoses of tuberculosis were made in
the hospitals, where patients go when they are acutely
(or even moderate chronically) sick. These are also
the institutions where tuberculous patients spent at
least the first days of their therapy. In some cases,
the patients could spend several days undiagnosed,
until a proper diagnosis was made after performing
several tests of different kinds in different areas of
the hospital. Even in a few cases death occurred and
the diagnosis was delayed until autopsy or cultures
gave the final results (Esteban et al. 2001bj Flora et al.
1990; Kramer et al.1990; Rosenthal et al.I975). Delay
in diagnosis is a problem common in those hospitals
or institutions where tuberculosis prevalence is low
J. Esteban
and it affects only a minority among all the patients.
However, this delay could have very important con-
sequences. Health care workers could be infected
because no proper preventive measures were taken
before the diagnosis, as has been demonstrated in
epidemic outbreaks (Frampton 1992; Griffith et al.
1995; Pearson et al. 1992) and, recently, in an envi-
ronment where there was no outbreak (Greenaway
et al. 2002).
However, it is important to discriminate between
the risk of infection and the risk of developing the
active disease. Historical data concerning the last
topic showed an increased incidence in nurses that
could be as high as 500 times higher than in the
general population (Sepkowitz 1994). Recent reports,
however, have shown incidences analogous to that of
the overall population (Capewell et al. 1988) or even
lower (Raitio and Tala 2000). The low prevalence of
tuberculosis and other risk factors for developing dis-
ease (such as HIV infection), together with BCG vac-
cination, has been considered in one of these reports
as the explanation for this phenomenon (Raitio and
Tala 2000). Speed in the detection of infectious cases
and, again, BCG vaccination, are the reasons shown
in the other article (Capewell et al.1988). Moreover, it
must be taken into account that health care workers
are usually healthy young people, without clear risk
factors for developing tuberculosis, except for a few
individuals (the healthy worker effect; Menzies et al.
1995). Probably the reasons for these data are difficult
to find and possibly they are multifactorial. However,
the risk of developing tuberculosis has been clearly
recognised in several reports of outbreaks, including
some due to multidrug-resistant strains (Griffith et
al. 1995; Menzies et al.1995), so it must still be consid-
ered, mainly in some special circumstances.
The risk of infection has also been considered
another hazard of work in health care workers since
the same time that the risk of tuberculous disease
was acknowledged. Historical data documented con-
versions as high as 100% among health care workers
(Sepkowitz 1994). More recent data indicate a high
risk of infection among these workers (Louther et al.
1997; Menzies et al. 1995; Schwartzman et al. 1996),
although clearly lower than the previous reports.
Global annual infection risk has been reported
between 0.2-10% (Menzies et al.1995). This risk could
be even higher in outbreaks (Griffith et al.1995; Pear-
son et al. 1992; Wenger et al. 1995). However, not all
health care workers, or even all health care facilities,
had identical risks. More detailed studies indicated
that the risk of infection is related to the number
of tuberculous patients admitted by the hospital,
Tuberculosis in Special Groups and Occupational Hazards
the contact between these patients and the different
workers and other factors such as inadequate ventila-
tion or delay in the diagnosis (Boudreau et al. 1997;
Greenaway et al. 2002; Louther et al. 1997; Menzies
et al. 1995,2000; Schwartzman et al. 1996). Hospitals
with fewer than ten tuberculous patients/year of less
than one patient with active disease/l00 workers per
year had fewer Mantoux conversions than those with
a higher number of admissions due to tuberculosis
(Menzies et al.1995; Schwartzman et al.I996). Nurses
are among those categories of workers with high con-
version rates in many studies (Boudreau et al. 1997;
Griffith et al. 1995; Ktsanes et al. 1986; Louther et
al. 1997; Schwartzman et al. 1996), even when they
are still students (Esteban et al. 2001a; Lainez et al.
1999; Sepkowitz 1994; Weiss 1973). However, other
conditions such as housekeeping, ward clerks and
workers in a bronchoscopy department have been
documented as having high risk of infection (Bou-
dreau et al. 1997; Louther et al. 1997; Schwartzman
et al. 1996).
Preventive measures has also been described for
health care workers since the decade of 1930 (Sep-
kowitz 1994). Nowadays, several guidelines have
been published where the measures for protecting
the workers against infection are clearly specified
(Centers for Disease Control 1994; Joint Tuberculo-
sis Committee of the British Thoracic Society 2000;
Moreno-Guillen et al. 1997), although these are not
always adequately implemented in daily routine
(Sutton et al. 2000). These measures have obviously
been developed for respiratory disease, which is con-
sidered as contagious because tuberculosis is trans-
mitted through respiratory secretions in most cases,
although there are reports of infection due to extra-
pulmonary tuberculosis (Frampton 1992; Hutton et
al.I990). Two important groups of measures could be
defined: proper isolation measures and a programme
of tuberculin screening for the workers.
Isolation measures (Table 7.2) include both struc-
tural and personal protective measures. Structural
interventions include the isolation of infectious
patients in individual rooms with proper ventilation.
Several reports of outbreaks defined poor ventilation
as the cause, at least in part, of the contagion (Men-
zies et al. 1995). Another report also described poor
ventilation as an important risk factor for infection in
a non-outbreak setting (Menzies et al. 2000). Rooms
with negative pressure are therefore recommended
for isolation (Centers for Disease Control 1994).
These rooms must have the door closed and at least
six changes of air per hour (Centers for Disease Con-
troI1994). Other structural measures, such as the use
103
Table 7.2. Recommended isolation measures for patients with
tuberculosis
Subject Specific measure
Patient lodgings Private room with negative pressure.
Door must be closed
Respiratory protection Biosafety respirators (N95)
Patient transport Limited to strictly necessary.
The patient wears a surgical mask
during transportation
Standard precautions Hand washing
Gloves when necessary
Face protection when necessary
Clean, non-sterile gown
Proper management of sharp devices
Minimise aerosols or other exposures
to samples
of ultraviolet lamps, have also been recommended as
useful and cheaper than the negative pressure isola-
tion rooms (Hannan et al. 2000; Menzies et al.I995).
Protective measures for individuals are directed
at blocking the access of infectious particles to the
lower respiratory tract. For this purpose, wearing a
mask is the measure recommended. However, not all
the masks offer similar protection. Classical surgical
masks are designed to avoid contamination from the
person who wears it. However, they are not designed
to protect that person, and there are reports where
failure of these masks to protect health care workers
has been documented (Sokolove et al.I994). Because
of this fact, surgical masks are only recommended to
be worn by the infectious patients when they are out-
side the isolation room (Centers for Disease Control
1994; Moreno-Guillen et al. 1997). Dust and HEPA
respirators are designed to minimise the risk of a
particle of a minimal size (l fl) reaching the alveolus.
Because there are differences in the percentage of
filtration efficacy, Centers for Disease Control and
other protocols recommend the wearing of respira-
tors with at least 95% of efficiency and the capacity
to filter particles of 1 fl. These recommendations
have been adopted by other guidelines (Centers for
Disease Control 1994; Moreno-Guillen et al. 1997).
However, there are reports where the efficacy of the
different kinds of respiratory protection is similar
(Catanzaro 1995). Another difficulty is how to evalu-
ate one single measure if other additional isolation
measures have been taken, and it is impossible to
estimate the degree of protection that is due to each
single measure. Because of all these data, it seems
wise to use the recommended respirators if possible,
but when they are not available, wearing a surgical
mask probably offers a certain degree of protection
that is better than nothing.
104
The other great pillar in the measures for protec-
tion of health care workers is the programme for
detection of recent infections and treating them
through tuberculin skin testing and proper isoniazid
chemotherapy (Bolyard et al. 1998; Centers for Dis-
ease Control 1994; Joint Tuberculosis Committee of
the British Thoracic Society 2000; Moreno-Guillen
et al. 1997). The programme is based on the perfor-
mance of a two-step initial test for those workers who
are negative in the first test to minimise the risk of a
booster reaction that could be considered as a con-
version, and the periodical re-testing of those work-
ers that were negative in the previous test (Bolyard
et al. 1998). However, in our experience and also the
experience of others (Clague et al. 1991), it could be
very difficult to develop a tuberculin programme in
hospital. The tuberculin skin test is different from
other tests performed in a protocol of health control
because it must be performed on one day and read
on another. This fact seems to be of great impor-
tance, and there are many workers that do not go to
the reading of the test because they estimate that the
result "is negative". Moreover, if it is difficult to con-
vince the workers to perform one test, the two-step
test for negative ones is even more difficult to per-
form, so, in some guidelines, this two-step test is not
recommended for all the workers and must be indi-
vidually evaluated (Moreno-Guillen et al. 1997). The
periodicity to perform the test is another problem,
because if it is performed too frequently (according
to the worker's idea), the worker will not go to have
the test the next time.
BCG vaccination is recommended in some guide-
lines (Joint Tuberculosis Committee of the British
Thoracic Society 2000), although many others do not
recommend it for all workers (Centers for Disease
Control 1994). However, the use of BCG vaccination
could be indicated for some health care workers in
some special conditions, such as outbreaks of multi-
drug-resistant tuberculosis (Bolyard et al. 1998). As
in the case of the use of respiratory protections, in
these cases a certain degree of protection is better
than nothing.
7.9
Occupational Hazards (2).
Laboratory Workers
Laboratory workers are a special category among
health workers because their risk has special charac-
teristics. Laboratory workers are not usually exposed
J. Esteban
to contagion by the usual route (contact with a
patient with active respiratory tuberculosis), but
they are exposed to infection through manipulation
of clinical samples or cultures. Because tuberculosis
is a disease mainly airborne transmitted, procedures
that imply aerosol formation have infection risk. In a
laboratory, aerosol-producing activities include sub-
culturing and streaking cultures, cooling and flaming
loops, pipetting for mixing microbial suspensions,
needle and syringe manipulations, centrifugation of
samples and cultures, mixing and homogenising pro-
cedures (use of shakers, homogenisers, blenders, son-
icators and other mixing instruments such as vortex)
pouring fluids, open culture containers (including
plates or tubers), spillage of infectious material,
lyophilisation and filtration under vacuum and even
egg inoculation and harvesting (Sewell 1995). Other
risk procedures include manipulation of samples or
tissues from infected patients for processing, as done
in microbiology and pathology departments. Even
direct inoculation of cultures has also been reported
as the cause of tuberculosis in a laboratory and must
be taken into account when working with pure cul-
tures of the organism (Peerbooms et al. 1995). The
recent introduction of automated liquid cultures for
rapid detection of mycobacteria introduces another
potential risk, because inoculation of samples and
extraction of liquid from positive cultures are done
using syringes and this manipulation has the risk of
creating aerosols (Richmond et al. 1996).
Several reports have stated that laboratory work-
ers have a risk of infection two to nine times higher
than the general population (Grist and Emslie 1994;
Harrington and Shannon 1976; Sewell 1995) There is
a report in which incidence of infection was as high
as 100 times the frequency observed in the general
population (MUller 1988). The progressive lowering of
the incidence of tuberculosis has been claimed as the
cause of the decrease in the number of these infections.
This is seen in pathology laboratories rather than in
microbiology ones (Grist and Emslie 1994). In a recent
report, laboratory personnel did not have the highest
rates of infection among health care workers, and their
rate was even lower than the global rate (4.4 conver-
sions per 100 person/year versus 5.2 conversions per
100 person/year) (Louther et al. 1997). Probably, all
these data suggest the efficacy of the safety measures
implemented in the laboratories.
Mycobacterium tuberculosis is an organism included
in the biosafety level III in most guidelines for labora-
tory safety when manipulation of cultures is performed
(Anonymous 1997; Richmond et al. 1996; Richmond
and McKinney 1999), although procedures such as
Tuberculosis in Special Groups and Occupational Hazards
preparation of smears for acid-fast stains only require
biosafety level II measures (Richmond and McKinney
1999). The organism is considered as an agent with
potential for aerosol transmission and the disease
that it produces may have serious or even lethal con-
sequences, but the risk of infection is only individual,
and the population as a whole has only a minimal risk
(Richmond and McKinney 1999; World Health Orga-
nization 1994). Measures to minimise this risk include
a broad spectrum that goes from rooms with negative
pressure that must be away from the common labora-
tory to the use of biosafety cabinets, protective masks,
gloves and other protective clothes (Table 7.3).
The biosafety level III laboratory must have special
design features. A double door access to the laboratory
is recommended. The mycobacteria laboratory must
also be separated from the laboratory and the ventila-
tion must be balanced to provide directional airflow
into the working room (negative pressure), although
in some legislation this measure is only recommended
and not obligatory (Anonymous 1997). In others this
measure is always considered as necessary (Richmond
and McKinney 1999; World Health Organization 1994).
In the era of multidrug-resistant tuberculosis, and
because of the risk of dissemination of these organ-
isms, negative pressure is a desirable objective to be
achieved for all the laboratories that perform manipu-
lation of mycobacteria cultures. However, this measure
is far from being available in many laboratories. In a
recent survey performed by the Sociedad Madrileiia
de Microbiologia Clinica (Society for Clinical Micro-
biology of Madrid), in the year 2001 only 4 out of 15
mycobacteriology laboratories from the Madrid Com-
Table 7.3. Recommended laboratory facilities for working with Mycobacterium tuberculosis complex organisms in laboratories
(Richmond and McKinney 1999; Anonymous 1997; World Health Organization 1994)
Measures USA
Physical separation from other working areas Yes
or corridors
Filtration of air through HEPA filters Yes, for exhausted
air from cabinets
Restricted access Yes
Negative airflow into laboratory Yes
Surfaces easily disinfected Yes
The working area should be sealed for disinfection Yes
Surfaces resistant to chemicals Yes
Security storage of biological agents Not stated
Window or other system for control of the workers Not stated
Laboratory with full exclusive equipment Not stated
Use of biosafety cabinets for manipulation Yes
of specimens and cultures
Autoclave or incinerator available Yes
Self-closing, double-door access Yes
105
munity had negative pressure (data reported in the
2002 congress of the SMMC).
Biosafety cabinets (BSCs) are a measure included as
necessary in all the legislation and recommendations
(Anonymous 1997; Richmond et al. 1996; Richmond
and McKinney 1999; World Health Organization
1994). The design of these containment devices dif-
fers in significant ways, although both BSC classes
II and I protect the laboratory worker and the envi-
ronment (Richmond et al. 1996). Class II devices are
also designed for protecting samples and cultures
from contamination. In the case of mycobacterial
laboratories, both BSC classes I and II are adequate for
biosafety level II measures, but when biosafety level III
measures are required (manipulation of cultures for
identification or susceptibility testing, for example) a
BSC class II or even class III is recommended as neces-
sary (Richmond and McKinney 1999).
Protective clothing for working is another impor-
tant protective measure for workers. Wearing gloves,
solid-front gowns, scrub suits or coveralls is recom-
mended, and this equipment should not be used
outside the biosafety level III laboratory. Reusable
equipment must be disinfected before been laun-
dered. Frequent hand washing is also recommended
(Richmond and McKinney 1999) and in some guide-
lines foot-operated lavatories are also suggested
as necessary (World Health Organization 1994).
Respiratory protection is recommended only when
BSCs are not available (Richmond and McKinney
1999). However, because of the risk of accidents due
to compromise of the containment capacity of BSCs
or other potential accidents (Richmond et al. 1996),
European Union W.H.O.
Desirable Yes
Yes, for exhausted air Recommended
Yes Yes
Desirable Yes
Yes, for workbench, Yes
floor, walls and ceilings
Desirable Yes
Yes Yes
Yes Not stated
Yes Not stated
Yes Not stated
Yes Yes
Yes Yes
Not stated Yes
106
respiratory protection for working at biosafety level
III is recommended, even if a BSC is currently used.
Together with the biosafety level II or III measures,
periodic tuberculin skin testing is recommended for
laboratory workers (Richmond and McKinney 1999).
This procedure should be repeated at least annually
for all the workers who are Mantoux negative, and
probably the frequency of re-testing should be higher
for those who work in high risk areas. If an accident
with a high level of exposure potential occurs, all
negative tuberculin skin test workers should be re-
tested, and then re-tested again at 3 month intervals
until no new conversions are detected (Richmond et
al. 1996). Chemotherapy for infected workers should
be performed according to the general criteria.
7.10
Occupational Hazards (3). Silicosis
The relation between tuberculosis and silicosis has
been known for many centuries. In the past century
many reports established high rates of tuberculosis
for silicotic patients throughout the world (Rosen-
man et al. 1997; Small and Fujiwara 2001; Snider
1978). Local alterations of the lung, the favourable
conditions for contagion in the mines, the difficul-
ties for a rapid diagnosis and the high prevalence of
tuberculosis in the population may explain the high
rates of tuberculous disease among these patients.
Important local alterations include the impairment
of macrophage function, a cell that plays a signifi-
cant role in the pathogenesis of the disease. Small
particles of silica dust are ingested by macrophages.
Then, the ingested particles disrupt the phagosomal
membranes and their enzymes are released, causing
the death of the affected cells. The lysed cells stimu-
late the migration of more macrophages, the prolif-
eration of fibroblasts and the production of collagen.
Sublethal doses of silica dust also affect the ability of
macrophages to inhibit the growth of Mycobacterium
tuberculosis (Snider 1978).
The diagnosis of tuberculosis in silicotic patients
may be a difficult matter. Symptoms are usually
non-specific and could be due to other conditions
or diseases. Acid-fast stains from the sputa of these
patients are negative in many cases, probably because
the tuberculous bacilli hardly reach the bronchial
tree due to silicotic fibrosis.
Chest roentgenogram is considered of great value
in the diagnosis (Barras 1970). The appearance of
new infiltrates, cavities, pleural effusion, coalescence
J. Esteban
of nodules and the development of bronchial steno-
sis are radiographic criteria that strongly suggest
the diagnosis of tuberculosis (Snider 1978), even in
the presence of negative microbiological diagnosis
techniques results. Accordingly, clinical and radio-
logical data should be used to establish a diagnosis
and begin adequate therapeutic regimens in these
patients.
Therapy of tuberculosis among these patients
includes longer therapeutic regimens with the same
drugs as for non-silicotic patients. Longer regimens
are necessary due to the lower penetration of the
drugs in the silicotic nodules. A study in Hong-Kong
showed that the standard 6-month short course
therapy is inadequate for these patients, as 22% of
patients had bacteriological relapses in the 6-month
therapy group compared to 7% of bacteriological
relapses in the 8-month therapy group (Hong Kong
Chest Service/Tuberculosis Research Centre and
Madras/British Medical Research Council 1991). As
a general recommendation, in the silicotic patient
the sterilising phase must be prolonged and patients
should be treated for at least 9 months. If pyrazin-
amide is not included in the regimen, therapy with
isoniazid and rifampin must be prolonged for a mini-
mum of 12 months (Snider 1978). Retreatment regi-
mens must also be prolonged as much as 12 months
and the follow-up of these patients should also be
prolonged for 5 years.
Preventive chemoprophylaxis with isoniazid has
also been considered of great value in diminishing
the rates of active disease. A study that compared
the treatment of 825 silicotic patients with 2,408
untreated silicotic patients showed that preventive
therapy with isoniazid decreases the risk of develop-
ing active disease from 2.66% to 0.32% (annual rates)
(Snider 1978). Shorter courses of chemoprophylaxis
with several drugs have also achieved lower rates of
tuberculosis than in non-treated patients, despite
its efficacy being unacceptably lower (Cowie 1996).
Other studies have shown the efficacy of isoniazid
chemotherapy. However, the difficulties are the same
as for the therapy of tuberculous disease and longer
regimens are indicated for these patients.
7.11
Occupational Hazards (4). Other Workers
Other groups of workers also have high rates of
tuberculous diseases. Farm workers and other work-
ers involved in animal management are considered a
Tuberculosis in Special Groups and Occupational Hazards
group with high risk for tuberculosis (McKenna et
al. 1996). Specific measures for the prevention of the
disease in these groups have been established. Among
these, it has been reported that disease rates are even
higher in migrant farm workers (Centers for Disease
Control 1992b; Ciesielski et al. 1991; Garcia et al. 1996;
Hibbs et al. 1989; McCurdy et al. 1997; Schulte et al.
2001). This population has unique characteristics,
such as overcrowding, close contact between indi-
viduals and high rates of immigrants from countries
with a high prevalence of tuberculosis (Centers for
Disease Control 1992b; World Health Organization
2001). Prevention and control of tuberculosis among
these workers includes the following measures: rapid
diagnosis and therapy of tuberculous patients, contact
investigation (Mantoux test) and chemoprophylaxis of
infected persons, and screening and proper therapy in
immunosuppressed workers, including HIV-infected
ones. Diagnosis causes no difficulties, as these patients
suffer from a similar disease to that suffered among the
general population. Therapy should not be performed
with short-course standard schemes and, if needed,
should include directly observed therapy (Centers for
Disease Control 1992b).
In some cases, the disease can be caused by Myco-
bacterium bovis rather than Mycobacterium tubercu-
losis in workers in contact with animals (Dalovisio et
al. 1992; Fanning and Edwards 1991; Pillai et al. 2000;
Thompson et al.1993). In these cases, therapy schemes
should consider that M. bovis strains are intrinsically
pyrazinamide-resistant, so this drug cannot be used.
Proper veterinary control measures, such as detection
and management of tuberculous animals, could also
be useful in the control of the infection.
Workers with pneumoconiosis other than silicosis
also have high rates of tuberculous disease (Mos-
quera et al. 1994; Taguchi et al. 2000). However, in
these patients the characteristics differ from those
of silicotic patients, probably because of the implica-
tions of silica particles in the pathogenesis of the dis-
ease, as has been previously explained (Snider 1978).
Differences in the incidence of tuberculosis can also
be found between the different kinds of pneumoco-
niosis, depending on the material inhaled (Starzynski
et al. 1996). The rates of tuberculosis are higher in
workers of the metallurgical industry than in those
from refractory material manufacturing plants (such
as ceramics). Nine-month therapeutic regimens
(2 months of daily streptomycin, isoniazid, rifampin
and pyrazinamide followed by 7 months of daily iso-
niazid and rifampin) have proven effective in these
patients with only a 5% relapse rate and an identical
amount of therapeutic failures (Lin et al. 1987).
107
Another profession that has been reported to have
unusually high rates of tuberculosis is that of funeral
director (McKenna et al.1996). Tuberculous infection
has also been reported to be very frequent in other
funeral home employees (Gershon et al.1998).A pos-
sible explanation is the contact with infected corpses,
in which the tuberculous bacilli can be cultured even
if they have been embalmed (Weed and Boggenstoss
1951). This theory has been recently demonstrated
in a case of tuberculosis acquired by a funeral direc-
tor, which was confirmed by both conventional and
molecular epidemiology (Lauzardo et al. 2001). Peri-
odic tuberculin skin testing and the use of respira-
tory protection has been advised for these workers
because of the high incidence rates of disease and
infection (Gershon et al. 1998).
Another study has reported a higher risk of tuber-
culosis in unemployed people (McKenna et al. 1996).
This is probably linked to neglect and other risk fac-
tors that are found among unemployed people and
that are different to the mere occupational ones.
7.12
Conclusion
Tuberculosis is perhaps one of the oldest known
human diseases. However, in the beginning of the
twenty-first century, and despite the existence of a
very effective therapy, this disease is still affecting
humans as one of the most frequent and lethal ones.
Airborne transmission favours infection in those
circumstances where a concentration of tuberculous
patients occurs especially if there is overcrowding,
poor hygiene and bad ventilation. Moreover, recently
developed aggressive therapies that are used for other
diseases favour the development of active disease
through alterations in the normal defences of the
organism. The knowledge of these facts must lead
to the development of effective measures to control
the expansion of the disease. However, we must not
forget that the most effective measure for the con-
trol of the infection is early diagnosis and therapy
of patients with active disease. Continuous efforts
must be made in this direction in order to minimise
the risk of healthy people acquiring this disease that
has been called "captain among these men of Death"
(Daniel 1997).
Acknowledgements. I want to acknowledge Dr. Fran-
cisco G. Santos-O'Connor for his help with the Eng-
lish language of this document.
108
References
Abramowsky C, Gonzalez B, Sorensen RU (1993) Disseminated
Bacillus Calmette-Guerin infections in patients with pri-
mary immunodeficiencies. Am J Clin PathoI100:52-56
Advani SH, Banavali SD (1989) Pattern of infection in hema-
tologic malignancies: an Indian experience. Rev Infect Dis
11 [SuppI7):SI621-S1628
Aljurf M et al. (1999) Mycobacterium tuberculosis infection
in allogeneic bone marrow transplantation patients. Bone
Marrow Transplant 24:551-554
Anonymous (1997) REAL DECRETO 664/1997, de 12 de mayo.
Protecci6n de los trabajadores contra los riesgos relacio-
nados con la exposici6n a agentes biol6gicos durante el
trabajo. Boletin Oficial del Estado, Spain
Bacakoglu F et al (2001) Pulmonary tuberculosis in patients
with diabetes mellitus. Respiration 68:595-600
Barras G (1970) Silico-tuberculose en Suisse. Schweiz Med
Wochenschr 100:1802
Bashar M, Alcabes P, Rom WN, Condos R (2001) Increased
incidence of multidrug-resistant tuberculosis in diabetic
patients on the Bellevue Chest Service, 1987-1997. Chest
120:1514-1519
Bolyard EA et al. (1998) Guideline for infection control in
health care personnel, 1998. American Journal of Infection
Control 26:289-354
Boudreau AY et al. (1997) Occupational risk fo Mycobacterium
tuberculosis infection in hospital workers. Am J Indust Med
32:528-534
Brown SM et al. (1974) Immunologic dysfunction in heroin
addicts. Arch Intern Med 134:1001-1006
Budak-Alpdogan T et al. (2000) The frequency of tuberculosis
in adult allogeneic stem cell transplant recipients in Turkey.
Bioi Blood Marrow Transplant 6:370-374
Capewell S, Leaker AR, Leitch AG (1988) Pulmonary tubercu-
losis in health service staff-is it still a problem? Tubercle
69:113-118
Catanzaro A (1995) Preventing nosocomial transmission of
tuberculosis. Lancet 345:204-205
Centers for Disease Control (1992a) Prevention and control of
tuberculosis among homeless persons. Recommendations
of the advisory council for the elimination of tubercu-
losis. MMWR 41 http://www.cdc.gov/mmwr/preview/
mmwrhtmll00019922.htm
Centers for Disease Control (1992b). Prevention and control of
tuberculosis in migrant farm workers. Recommendations
of the Advisory Council for the Elimination of Tuberculo-
sis. MMWR 41:1-15
Centers for Disease Control (1994) Guidelines for preventing
the transmission of Mycobacterium tuberculosis in health
care facilities. MMWR 43:1-132
Centers for Disease Control (1996) Prevention and control
of tuberculosis in correctional facilities: recommenda-
tions of the Advisory Council for Elimination of Tuber-
culosis. MMWR 45 http://www.cdc.gov/mmwr/preview/
mmwrhtrnllOOO 19922.htm
Centers for Disease Control (2000) Guidelines for preventing
opportunistic infections among hematopoietic stem cell
transplant recipients: recommendations of Centers For
Disease Control, the Infectious Disease Society of America,
and the American Society pf Blood and Marrow Transplan-
tation. MMWR 49:43-44
Chin DP et al. (1998) Differences in contributing factors to
J. Esteban
tuberculosis incidence in US-born and foreign-born per-
sons. Am J Respir Crit Care Med 158:1797-1803
Ciesielski SD, Seed JR, Esposito DH, Hunter N (1991) The epi-
demiology of tuberculosis among North Carolina migrant
farm workers. JAMA 265:1715-1719
Clague JE, Fields P, Graham DR, Davies PD~ (1991) Screening
for tuberculosis: current practices and attitudes of hospital
workers. Tubercle 72:265-267
Coninx R, Maher D, Reyes H, Grzemska M (2000) Tuberculosis
in prisons in countries with high prevalence. Br Med J 320:
440-442
Cowie RL (1996) Short course chemoprophylaxis with rifam-
picin, isoniazid and pyrazinamide for tuberculosis evalu-
ated in gold miners with chronic silicosis: a double-blind
placebo controlled trial. Tubercle Lung Dis 77:239-243
Daley CL et al. (1998) Incidence of tuberculosis in injection
drug users in San Francisco. Impact of anergy. Am J Respir
Crit Care Med 157:19-22
Dalovisio J, Stetter M, Mikota-Wells S (1992) Rhinoceros'rhi-
norrea: cause of an outbreak of infection due to airborne
Mycobacterium bovis in zookeepers. Clin Infect Dis 15:
598-600
Daniel TM (1997) Captain of death: the story of tuberculosis.
University of Rochester Press, Rochester
Drobniewski F (1995) Tuberculosis in prisons-forgotten
plague. Lancet 346:948-949
Duran E et al. (1996) Tuberculosis in recently arrived immi-
grants to Barcelona. Med Clin (Bare) 106:525-528
Esteban J et al. (2001a) Tuberculous infection in student nurses
in Madrid, Spain. Int J Tuberc Lung Dis 6:583-584
Esteban J et al. (2001b) Mycobacterium tuberculosis bactere-
mia in a universitary hospital. Int J Tuberc Lung Dis 6:
763-768
Fanning A, Edwards S (1991) Mycobacterium bovis infection
in human beings in contact with elk (Cervus elaphus) in
Alberta, Canada. Lancet 338:1253-1255
Fernandez-Martin JI et al. (2000) Transmission of tuberculo-
sis in the prisons of Madrid (Spain). Med Clin (Bare) 115:
246-250
Flora GS, Modilevsky T,Antoniskis D, Barnes PF (1990) Undi-
agnosed tuberculosis in patients with Human Immunode-
ficiency Virus infection. Chest 98:1056-1059
Frampton MW (1992) An outbreak of tuberculosis amon hos-
pital personnel caring for a patient with a skin ulcer. Ann
Intern Med 117:312-313
Friedman LN, Williams MT, Singh TP, Frieden TR (1996)
Tuberculosis, AIDS, and death among substance abusers
on welfare in New York City. N Engl J Med 334:828-833
Garcia JG, Matheny-Dresser KS, Zerr AD (1996) Respiratory
health of Hispanic migrant farm workers in Indiana. Am J
Indust Med 29:23-32
Gasner MR et al. (1999) The use of legal action in New York
City to ensure treatment of tuberculosis. N Engl J Med 340:
359-366
George B et al. (2001) Tuberculosis among allogeneic bone
marrow transplant recipients in India. Bone Marrow
Transplant 27:973-975
Gershon RR et al. (1998) Tuberculosis risk in funeral home
employees. J Occup Environ Med 40:497-503
Gourevitch MN, Alcabes P, Wasserman WC, Arno PS (1998)
Cost-effectiveness of directly observed chemoprophylaxis
of tuberculosis among drug users at high risk for tubercu-
losis. Int J Tuberc Lung Dis 2:531-540
Tuberculosis in Special Groups and Occupational Hazards
Graham JC et al. (2001) Tuberculosis transmitted through
transplantation. J Infect 43:251-254
Greenaway C et al. (2002) Delay in diagnosis among hospi-
talizad patients with active tuberculosis-Predictors and
outcomes. Am J Respir Crit Care Med 165:927-933
Griffith DE et al. (1995) Tuberculosis outbreak among health-
care workers in a community hospital. Am J Respir Crit
Care Med 152:808-811
Grist NR, Emslie JAN (1994) Association of clinical patholo-
gists' surveys of infection in British clinical laboratories,
1970-1989. J Clin PathoI47:391-394
Gutierrez MC, et al. (1998) Molecular fingerprinting of Myco-
bacterium tuberculosis and risk factors for tuberculosis
transmission in Paris, France, and surrounding area. J Clin
Microbiol 36:486-492
Hamburg MA, Frieden TR (1994) Tuberculosis transmission
in the 1990s. N Engl J Med 330:1750-1751
Hannan MM et al. (2000) Hospital infection control in an era
of HIV infection and multi-drug resistant tuberculosis. J
Hosp Infect 44:5-11
Harrington JM, Shannon HS (1976) Incidence of tuberculosis,
hepatitis, brucellosis, and shigellosis in British medical
laboratory workers. Br Med J 1:759-762
Havlir DV, Barnes PF (1999) Tuberculosis in patients with
Human Immunodeficiency Virus infection. N Engl J Med
340:367-373
Hibbs J, Yeager S, Cochran J (1989) Tuberculosis among
migrant farm workers. JAMA 262:1775
Hill AR et al. (1991) Disseminated tuberculosis in the Acquired
Immunodeficiency Syndrome era. Am Rev Respir Dis 144:
1164-1170
Hong Kong Chest Service/Tuberculosis Research Centre, Madras/
British Medical Research Council (1991) A controlled clinical
comparison of 6 and 8 months of antituberculosis chemo-
therapy in the treatment of patients with silicotuberculosis
in Hong Kong. Am Rev Respir Dis 143:262-267
Huerga H, L6pez-Velez R, Navas E, G6mez-Mampaso E (2000)
Clinicoepidemiological features of immigrants with tuber-
culosis living in Madrid, Spain. Eur J Clin Microbiol Infect
Dis 19:236-240
Hutton MD et al. (1990) Nosocomial transmission of tubercu-
losis associated with a draining abscess. J Infect Dis 161:
286-295
Hiigo-Martinez J et al. (2000) Recent transmission of tuber-
culosis in Madrid (Spain): usefulness of molecular tech-
niques. Med Clin (Bare) 115:241-245
Ip MS et al. (1998) Risk factors for pulmonary tuberculosis in
bone marrow transplant recipients. Am J Respir Crit Care
Med 158:1173-1177
John GT et al. (1994) Cyclosporine immunosuppression and
mycobacterial infections. Transplantation 58:247-249
Joint Tuberculosis Committee of the British Thoracic Society
(2000) Control and prevention of tuberculosis in the United
Kingdom: code of practice 2000. Thorax 55:887-901
Jones TF, Schaffner W (2001) Miniature chest radiograph
screening for tuberculosis in jails. A cost-effectiveness
analysis. Am J Respir Crit Care Med 164:77-81
Jones TF et al. (1999) Transmission of tuberculosis in a jail.
Ann Intern Med 131:557-563
Kameda K, Kawabata S, Musuda N (1990) Follow-up study
of short-course chemotherapy of pulmonary tuberculosis
complicated with diabetes mellitus. Kekkaku 65:791-803
Koffi N et al. (1997) Smear positive pulmonary tuberculosis in
109
a prison setting: experience in the penal camp of Bouake,
Ivory Coast. Int J Tuberc Lung Dis 1:250-253
Koziel H, Koziel MJ (1995) Pulmonary complications of dia-
betes mellitus. Pneumonia. Infect Dis Clin North Am 9:
65-96
Kramer F et al. (1990) Delayed diagnosis of tuberculosis in
patients with Human Immunodeficiency Virus infection.
Am J Med 89:451-456
Ktsanes VK, Williams WL, Boudreaux VV (1986) The cumula-
tive risk of tuberculin skin test conversion for five ears of
hospital employment. Am J Publ Health 76:65-67
Kumar RR, Shafiulla M, Sridhar H (1999) Association of tuber-
culosis with malignancy at KIMIO - an oncology centre.
Indian J Pathol MicrobioI42:339-343
Kurzrock R et al. (1984) Mycobacterial pulmonary infections
after allogeneic bone marrow transplantation. Am J Med
77:35-40
Lainez RM et al. (1999) Infecci6n tuberculosa en estudiantes
de enfermeria. Prevalencia y virajes durante 3 afios de
seguimiento. Med Clin (Bare) 113:685-689
Laniado-Laborin R (2001) Tuberculosis in correctional facili-
ties: a nightmare without end in sight. Chest 119:681-683
Lauzardo M, Lee P, Duncan H, Hale Y (2001) Transmission of
Mycobacterium tuberculosis to a funeral director during
routine embalming. Chest 119:640-642
Levine DP, Brown PD (2001) Infections in injection drug users.
In: Mandell GL, Douglas B (eds) Principles and practice of
infectious diseases, 4th edn. Williams and Wilkins, Phila-
delphia, pp 3112-3126
Libshitz HI, Pannu HK, Elting LS, Cooksley CD (1997) Tuber-
culosis in cancer patients: an update. J Thorac Imaging 12:
41-46
Lichtenstein IH, MacGregor RR (1983) Mycobacterial infec-
tions in renal transplant recipients: report of five cases and
review of the literature. Rev Infect Dis 5:216-226
Lillebaek T et al. (2001) Risk of Mycobacterium tuberculosis
transmission in a low-incidence country due to immigration
from high incidence areas. J Clin MicrobioI39:855-861
Lin TP et al. (1987) Short-course chemotherapy of pulmonary
tuberculosis in pneumoconiotic patients. Am Rev Respir
Dis 136:808-810
Louther J et al. (1997) Risk of tuberculin conversion according
to occupation among health care workers at a New York
City hospital. Am J Respir Crit Care Med 156:201-205
Maher D, Grzemska M, Coninx R, Reyes H (1998) Guidelines
for the control of tuberculosis in prisons. WHO/TB/98.250.
World Health Organisation, Geneva
McCurdy SA, Arretz DS, Bates RO (1997) Tuberculin reactiv-
ity among California Hispanic migrant farm workers. Am
J Indust Med 32:600-605
McKenna MT, Hutton M, Cauthen G, Onorato 1M (1996)
The association between occupation and tuberculosis. A
population based survey. Am J Respir Crit Care Med 154:
587-593
Melero M, Gennaro 0, Dominguez C, Sanchez-Avalos JC
(1992) Tuberculosis in patients with lymphomas. Medicina
(Buenos Aires) 52:291-295
Menzies D, Fanning A, Yuan L, Fitzgerald M (1995) Tuberculo-
sis among health care workers. N Engl J Med 332:92-98
Menzies D et al. (2000) Hospital ventilation and risk for
tuberculous infection in Canadian health care workers.
Ann Intern Med 133:779-789
Moreno-Guillen S et al. (1997) Control de la transmisi6n noso-
110
comial de la tuberculosis. Programa regional de prevenci6n
y control de la tuberculosis en la Comunidad de Madrid.
Direcci6n General de Prevenci6n y Promoci6n de la Salud.
Consejeria de Sanidad y Servicios Sociales, Comunidad de
Madrid, Madrid, pp 65-79
Morii T, Narita N (1998) Mycobacterial infection in patients
with hematologic disorders. Nippon Rinsho 56:3209-3211
Moro ML et al. (2002) Two-year population-based molecular
epidemiological study of tuberculosis transmission in the
Metropolitan Area of Milan, Italy. Eur J Clin Microbiol
Infect Dis 21:114-122
Morris JT, Seaworth BJ, McAllister CK (1992) Pulmonary
tuberculosis in diabetics. Chest 102:539-541
Mosquera JA, Rodrigo L, Gonzalvez F (1994) The evolution of
pulmonary tuberculosis in coal miners in Asturias, north-
ern Spain. An attempt to reduce the rate over a 15-year
period, 1971-1985. Eur J Epidemioll0:291-297
Moss AR et al. (2000) Tuberculosis in the homeless. A prospec-
tive study. Am J Respir Crit Care Med 162:460-464
Mourad G et al. (1985) Transmission of Mycobacterium tuber-
culosis with renal allografts. Nephron 41:82-85
Muller HE (1988) Laboratory-acquired mycobacterial infec-
tion. Lancet ii:331
Muller K, Fopp M, Senn HJ (1980) Unrecognized atypical
tuberculosepsis in generalized hematologic neoplasms.
Schweiz Med Wochenschr 110:1815-1817
Munoz P et al. (1995) Tuberculosis in heart transplant recipi-
ents. Clin Infect Dis 21:398-402
Nardell E, McInnis B, Thomas B, Weidhaas S (1986) Exogenous
reinfection with tuberculosis in a shelter for the homeless.
N Engl J Med 315:1570-1575
Opsald R, Riddervald HO, Aas TW (1961) Pulmonary tuber-
culosis in mitral stenosis and diabetes. Acta Tuberc Scand
4:290
Patel R, Paya CV (1997) Infections in solid-organ transplant
recipients. Clin Microbiol Rev 10:86-124
Pearson ML et al. (1992) Nosocomial transmission of mul-
tidrug-resistant Mycobacterium tuberculosis. A risk to
patients and health care workers. Ann Intern Med 117:
191-196
Peerbooms PGH et al. (1995) Laboratory-acquired tuberculo-
sis. Lancet 345:1311-1312
Perez-Guzman C, Torres-Cruz A, Villarreal-Velarde H, Vargas
MH (2000) Progressive age-related changes in pulmonary
tuberculosis images and the effect of diabetes. Am J Respir
Crit Care Med 162:1738-1740
Perlman DC et al. (1995) Tuberculosis in drug users. Clin
Infect Dis 21:1253-1264
Pillai SD et al. (2000) Failure to identify non-bovine reser-
voirs of Mycobacterium bovis in a region with a history of
infected dairy-cattle herds. Prev Vet Med 43:53-62
Qunibi WY et al. (1991) Mycobacterial infection after renal
transplantation-report of 14 cases and review of the litera-
ture. Quart J Med 77:1039-1060
Raitio M, Tala E (2000) Tuberculosis among health care work-
ers during three recent decades. Eur Respir J 15:304-307
Reyes H, Coninx R (1997) Pitfalls of tuberculosis programmes
in prisons. Br Med J 315:1447-1450
Richmond JY, McKinney RW (1999) Biosafety in Microbiologi-
cal and Biomedical Laboratories. US Government Printing
Office, Washington DC
Richmond JY, Knudsen RC, Good RC (1996) Biosafety in the clini-
cal mycobacteriology laboratory. Clin Lab Med 16:527-550
J. Esteban
Rieder HL et al. (1994) Tuberculosis control in Europe and
international migration. Eur Respir J 7:1545-1553
Rosenman KD, Reilly MJ, Kalinowski DJ, Watt FC (1997) Sili-
cosis in the 1990 s. Chest 111 :779-786
Rosenthal T, Pitlik S, Michaeli D (1975) Fatal undiagnosed
tuberculosis in hospitalized patients. J Infect Dis 131
[Suppl]:S51-S56
Roy V, Weisdorf D (1997) Mycobacterial infections follow-
ing bone marrow transplantation: a 20 year retrospective
review. Bone Marrow Transplant 19:467-470
Sable CA, Donowitz GR (1994) Infections in bone marrow
transplant recipients. Clin Infect Dis 18:273-284
Samper S et al. (1998) The molecular epidemiology of tuber-
culosis in Zaragoza, Spain: a retrospective epidemiological
study in 1993. Int J Tuberc Lung Dis 2:281-287
Schulte JM, Valway SE, McCray E, Onorato 1M (2001) Tuber-
culosis cases reported among migrant farm workers in the
United States, 1993-1997. J Health Care Poor Underserved
12:311-322
Schwartzman K, Loo V, Pasztor 1, Menzies D (1996) Tuberculo-
sis infection among Health Care Workers in Montreal. Am
J Respir Crit Care Med 154:1006-1012
Selwyn PA et al. (1992) High risk of active tuberculosis in
HIV-infected drug users with cutaneous anergy. JAMA
268:504-509
Sepkowitz KA (1994) Tuberculosis and the health care worker:
a historical perspective. Ann Intern Med 120:71-79
Sewell DL (1995) Laboratory-associated infections and bio-
safety. Clin Microbiol Rev 8:389-405
Singh N, Paterson DL (1998) Mycobacterium tuberculosis
infection in solid-organ transplant recipients: impact
and implications for management. Clin Infect Dis 27:
1266-1277
Small PM, Fujiwara PI (2001) Management of tuberculosis in
the United States. N Engl J Med 345:189-200
Snider DE (1978) The relationship between tuberculosis and
silicosis. Am Rev Respir Dis 118:455-460
Snyder DC et al. (1999) Tuberculosis prevention in methadone
maintenance dinics. Effectiveness and cost-effectiveness.
Am J Respir Crit Care Med 160:178-185
Sokolove PE, Mackey D, Wiles J, Lewis RJ (1994) Exposure of
emergency department personnel to tuberculosis: PPD
testing during an epidemic in the community. Ann Emerg
Med 24(3):418-421
Starzynski Z, Marek K, Kujawska A, Szymczak W (1996)
Mortality among different occupational groups of workers
with pneumoconiosis: results from a register-based cohort
study. Am J Indust Med 30:718-725
Stead WW (1989) Special problems in tuberculosis. Clin Chest
Med 10:397-405
Stoll-Keller F et al. (1997) Effects of morphine on purified
human blood monocytes: modifications of properties
involved in antiviral defenses. Int J Immunopharmacol
19:95-100
Sundbery R et al. (1991) A tuberculosis outbreak in a renal
transplant program. Transplant Proc 23:3091-3092
Sutton PM, Nicas M, Harrison RJ (2000) Tuberculosis iso-
lation: comparison of written procedures and actual
practices in three California hospitals. Infect Contr Hosp
EpidemioI21:28-32
Taguchi 0 et al. (2000) Mixed dust fibrosis and tuberculosis
in comparison with silicosis and macular pneumoconiosis.
Am J Indust Med 37:260-264
Tuberculosis in Special Groups and Occupational Hazards
Tamura A et al. (1999) Active pulmonary tuberculosis in
patients with lung cancer. Kekkaku 74:797-802
The WHO/IUATLD Global Project on Anti-tuberculosis Drug
Resistance Surveillance 1994-1997 (1997) Anti-tuberculo-
sis drug resistance in the world. World Health Organiza-
tion, Geneva
Thompson PJ et al. (1993) Seals, seal trainers, and mycobacte-
rial infecion. Am Rev Respir Dis 147:164-167
Turner WM (1951) Pulmonary tuberculosis and diabetes mel-
litus. Quart J Med 26:31
Wares DF, Clowes CI (1997) Tuberculosis in Russia. Lancet
350:957
111
Weed L, Boggenstoss A (1951) The isolation of pathogens from
tissues of embalmed human bodies. Am J Clin Pathol 21:
1114-1120
Weiss W (1973) Tuberculosis in student nurses at Philadelphia
General Hospital. Am Rev Respir Dis 107:136-139
Wenger PN et al. (1995) Control of nosocomial transmission of
multidrug-resistant Mycobacterium tuberculosis among health-
care workers and HIV-infected patients. Lancet 345:235-240
World Health Organization (1994) Manual de bioseguridad en
ellaboratorio. World Health Organization, Geneve
World Health Organization (2001) Global tuberculosis control.
WHO report 2001. WHO/CDS/TBI2001.287. World Health
Organization, Geneva
 Microbiology, Immunology,
Pathogenesis and Pathology
8 Microbiology of Tuberculosis
A. Ow OSOBA

CONTENTS 8.5.7.1 Luciferase-based Reporter Phage 128

8.5.7.2 Epsilometer (E-) test 128
8.1 Aetiology and Morphology 115 8.5.7.3 Hybridisation Protection Assay (HPA) 128
8.2 Characteristics of Mycobacteria 116 8.6 Serological Diagnosis of Tuberculosis 129
8.3 Laboratory Diagnosis of Mycobacteria 117 8.6.1 Immunochromatographic Tests 130
8.3.1 Specimen Collection 117 8.6.1.1 ICT Tuberculosis Test 130
8.3.2 Microscopic Examination 118 8.6.1.2 Rapid Test TB 130
8.3.2.1 New AFB Detection Methods 119 8.6.2 Enzyme-linked Immunosorbent Assays 130
8.3.2.2 Specimen Preparation and Culture 119 8.6.2.1 Tuberculosis IgA EIA 130
8.3.2.3 Liquefaction and Decontamination 119 8.6.2.2 The Pathozyme TB Complex Test 130
8.3.3 Culture Media for the Isolation 8.6.2.3 PATHOZYME-MYCO IgG, IgA,
of Mycobacteria 120 and IgM Tests 130
8.3.3.1 Solid Media 120 References 131
8.3.3.2 Liquid Automated Culture Systems 122
8.3.3.3 Inoculation and Incubation Procedure 123
8.3.4 Identification of Mycobacteria Species 123
8.4 Molecular Techniques for Identification
of Mycobacteria 123 8.1
8.4.1 Nucleic Acid Probes for Actively Growing Aetiology and Morphology
Mycobacteria 123
8.4.2 The Accuprobe System 124
Tuberculosis is caused by Mycobacterium tubercu-
8.4.3 The BDProbe Tec ET 124
8.4.4 The Amplified Mycobacterium Tuberculosis losis. The word tuberculosis is derived from the
Direct (MTD) Test 124 word tubercle, meaning a small lump or nodule
8.4.5 Amplicor M. Tuberculosis (PCR) Test 125 (Wolinsky 1988). The disease was first described by
8.4.6 Ligase Chain Reaction (LCR) 125 Jean-Antoine Villemin, a French military doctor in
8.4.7 Other Molecular Techniques 125
1868, when he successfully transmitted the infec-
8.5 Antimicrobial Susceptibility Testing 126
8.5.1 Conventional Methods of Susceptibility tion to rabbits by inoculating them with material
Testing 126 from man and cattle. He later showed that scrofula
8.5.2 Radiometric Method of Drug Susceptibility (tuberculous cervical lymphadenitis) and pulmo-
Testing 126 nary tuberculosis were different manifestations of
8.5.3 Non-Radiometric Methods of Drug
the same disease (Grange 1984). The organism was
Susceptibility Testing 127
8.5.4 The Mycobacterial Growth Indicator Tube first isolated in pure culture by Robert Koch in 1882.
(MGIT) System 127 The characterisation of the isolate led to the famous
8.5.5 The MB/BacT Microbial Detection System 127 Koch's postulates, providing concrete evidence of M.
8.5.6 Genetic Approach to the Detection of Drug tuberculosis as the causative agent of tuberculosis. In
Resistance by M. tuberculosis 127
1896, Lehmann and Neumann described the genus of
8.5.6.1 Line Probe Assay (LIPA) 128
8.5.7 Newer Techniques of Drug Susceptibility Mycobacterium, which included M. tuberculosis and
Testing 128 M. leprae, the leprosy bacillus.
Mycobacterium tuberculosis has been classified in
the genus of Mycobacterium, which belongs to the
family of Mycobacteriaceae of the order of Actino-
mycetales. In this order are the families of Actino-
mycetaceae and Streptomycetaceae. In the order of
A. O. OSOBA, MD, FRCPath (UK), FFPath (Ireland), FWACP Actinomycetaceae are the genera of Actinomyces
Consultant & Head of Microbiology, King Khalid National and Nocardia, while the order of Streptomycetaceae
Guard Hospital, P.O. Box 9515, Jeddah 2143, Saudi Arabia includes the genus of Streptomyces.
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
116 A.O.Osoha

8.2
Characteristics of Mycobacteria
Superflcialliplds
(Myoosldes. cord lactor
Mycobacteria are primarily slow-growing intracellu- and 5ulpholipids)

lar curved bacilli. Their main characteristic feature

is their distinctive staining properties. When stained
with carbolfuchsin or other fluorochrome dyes, they
resist decolorisation with acid and alcohol. Hence
they are sometimes referred to as "acid-fast bacilli"
(AFB). Robert Koch not only cultured M. tuberculosis
but was also able to stain the organism with alkaline
solution of methylene blue for 24 hours. The tech-
nique was later improved by Ehrlich using a hot
solution of aryl ethylene dye fuchsin. This procedure Peptidoglycan
was later modified by Ziehl and Neelsen (ZN), whose (Murein layer)

names are still today used to describe the staining "'~--f-- Upoarabinomannan

technique used for staining Mycobacteria (Grange

1984).
Another characteristic feature of the Mycobacteria
is their unique and complex cell wall. The cell walls
are rich in lipid complexes such as peptidoglycolip-
ids (mycosides), cord factor and sulpholipids, which Fig. 8.1. A diagrammatic representation of the mycobacterial
give the organism its shape, rigidity and colonial cell wall
characteristics (Fig. 8.1). These mycosides are struc-
turally and functionally similar to the 0 antigens
of Gram-negative bacteria and determine the sero-
agglutination and bacteriophage susceptibility of the
organism. The cord factor (trehalose dimycolate) and
sulpholipids have toxic properties but their contribu-
tion to virulence has not been elucidated (Davies
et al. 1996). Recently, some workers have provided
evidence to suggest that lipoarabinomannan plays
a significant role in virulence (Moulding 1999). The
unique, lipid-rich outer membrane confers on the
organism an effective barrier to prevent entry of
antimicrobials into the cell and helps to resist phago-
cytosis as well as the immune system of the host.
(Fig. 8.2 + 8.3)
The Mycobacteria are aerobic, non-sporing and
non-motile organisms. They appear as thin slightly
curved rods and, when stained by the ZN stain, may
appear beaded. They measure 0.2-0.6 by 1.0-10.0 11m
in size. They are only lightly stained or weakly Gram-
positive by Gram's stain, which is used to stain most
bacteria, due to the high lipid content of their cell Fig. 8.2. False electron micrograph of the surface of Mycobac-
wall. They are characteristically slow growing on terium tuberculosis (After Medicine Digest, February, 1996,
laboratory media and have a generation time of Cover page)
12-36 hours.
Mammalian tuberculosis is caused by M. tubercu- lower mammals. They are collectively referred to as
losis (human tubercle bacillus), M. bovis (the bovine M. tuberculosis complex. M. tuberculosis grows very
tubercle bacillus), M. africanum, found in equatorial slowly at 35°C and may take 3-6 weeks to produce
Africa and intermediate between the above two spe- visible colonies, which are buff-coloured, rough and
cies, and M. microti, a rare pathogen of voles and friable.
Microbiology of Tuberculosis 117

Human infections may also be produced by 8.3

Mycobacteria other than the M. tuberculosis com-
plex. These have been referred to as atypical, anony-
mous or MOTT (Mycobacterium other than typical
tuberculosis). However the preferred terminology
nowadays is non-tuberculous Mycobacteria (NTM)
(Grange 1984). The most widely used classification
of Mycobacteria is that of Runyon, which is based on
clinical and microbiological features of the organ-
isms, such as rate of growth, ability to grow and
produce pigment in the presence or absence of light.
These features provide an important method of dif-
ferentiating between the various species in a clinical
laboratory (Table 8.1) (Hopewell and Bloom 1994).
The characteristic features of the NTM are shown in
Table 8.1.
Laboratory Diagnosis of Mycobacteria
The confirmation of the clinical diagnosis of myco-
bacterial infection depends on the identification of
the mycobacteria in smears, isolation on laboratory
media, identification of the species and susceptibility
testing of the isolate.
8.3.1
Specimen Collection
In order to make an accurate diagnosis of mycobac-
terial infection, adequate and proper collection of
specimens submitted for examination must be made.

Table 8.1. Runyon's Classification of Non-tuberculous Mycobacteria

Runyon's group Species Rate of Clinical features

and pigment production growth (days) and site of lesion

Group I
Photochromogens M. kansasii 10-21 Rare contaminant; lung disease
M. marinum 7-14 Rare contaminant; skin and soft tissue
Yellow/orange M.simiae 7-14 Rare human pathogen; lung
pigment after M. asiaticum 7-14 Rarely pathogenic
exposure to light
Group II
Scotochromogens M. scrofulaceum 10-28 Rare human pathogen; lymphadenitis
M.szulgae 12-28 Rare human pathogen; lung disease
Orange/red M.gordonae 10-28 Rarely pathogenic; environmental
pigment in the contaminant
dark M. flavescens 7-10 Rarely pathogenic; environmental
contaminant
Group III
Non- M. avium- 10-21 Lung disease in immunocompromised
photochromogens intracellulare patients
(MAC)
(Smooth and M. ulcerans 28-60 Buruli ulcer (Africa)
cream coloured M. xenopi 14-28 Optimal growth at 42°C; lung disease
colonies) Environmental contaminant

}
M. malmoense Rare human pathogen; lung disease
18-84
M. haemophilum Optimal growth at 20-32°C; skin and
soft tissue disease in
immunocompromised host
M. terrae-complex 10-21 Rarely pathogenic; environmental
contaminant
M. gastri 10-21 Rarely pathogenic; environmental
contaminant
M. flavescens Rarely pathogenic
M. triviale 10-21 Environmental contaminant
Group IV
M. fortuitum 3-7 Environmental contaminants; ulcers
Rapid growers M. chelonae 3-7 Environmental contaminant; ulcers
(Growth in few M. smegmatis 3-7 Rarely pathogenic
days) M.phlei 3-7 Rarely pathogenic
118 A. O. Osoba

In addition, the specimens must be submitted in the

correct containers, held in the appropriate conditions
and sent to the laboratory for processing as soon as
possible, in leak-proof universal containers. If there
is going to be a delay in culture, the sample must be
refrigerated. Swabs are not recommended for isola-
tion of mycobacteria.
a) Pulmonary specimens - 5-10 ml of early morning
sample of sputum collected on three consecutive
days in a sterile container will be adequate. Deep
coughing must be encouraged so that exudative
material is obtained from the lungs. If the patient
is unable to expectorate a satisfactory sample,
deep coughing may be induced by inhalation of
an aerosol of warm, hypertonic (5-10%) saline.
The specimen should be properly labelled as
,induced sputum' so that it will not be rejected by
the laboratory as saliva. Saliva and nasal drain-
age are unsatisfactory samples. Similarly 5-10 ml
bronchoalveolar lavage submitted in a sterile con-
tainer will be adequate for processing.
b) Extrapulmonary specimens - For urine samples
at least 40 ml of early morning samples collected
on three consecutive days in a sterile container is
recommended. Two millilitres of cerebrospinal
fluid and 5-10 ml of gastric aspirates are usually Fig. 8.3. ZieW Neelsen stain of sputum sample showing slender
acid-fast bacilli stained red against a blue background
required. Gastric lavage is particularly useful in
children who do not produce adequate sputum
specimens and who usually swallow coughed
secretions. Hospitalisation of the child may be acid alcohol, rinsed again with tap water and then
necessary so that the lavage can be collected counter-stained with methylene blue or malachite
very early in the morning. Synovial and pleural green for 2 minutes. The Mycobacteria are stained
fluid require at least 5 ml of the sample in a ster- red against a blue background. This so-called
ile container. Tissues, fine needle aspirates and "acid-fast" staining procedure is based on the fact
bone marrow samples can also be processed for that Mycobacteria are capable of retaining the red
mycobacteria, if submitted in sterile containers or colour of the carbolfuchsin despite treatment with
inoculated at the bedside in liquid culture media acid alcohol (Fig. 8.3). The smears are then examined
(Shinnick and Good 1999). under a light microscope with a 100x objective oil
immersion. At the higher magnification, the smear
should be read in three parallel sweeps of the long
8.3.2 axis, with a minimum of 300 fields being examined
Microscopic Examination or for a duration of 15 minutes. Usually, 5,000-10,000
organisms per millilitre of sample are required
After decontamination and concentration of the before they can be detected under the microscope.
samples, two staining procedures are commonly used This makes the ZN stain less sensitive than cultures,
for the identification of mycobacteria in specimens. although a properly performed smear has a positive
These are the ZieW-Neelsen stain and the aurarnine- predictive value of over 90% and specificity of over
rhodamine stain. 90% in pulmonary tuberculosis (Shinnick and Good
1999; Marshall and Shaw 1997). Usually, 40 -70% of
Ziehl-Neelsen Stain. The smears are heat-fixed patients with M. tuberculosis isolated in culture will
and then flooded with carbolfuchsin and heated give a positive ZN stain reaction, while the rest will
intermittently for 1 minute. The smears are then be ZN-negative. The advantage of the ZN stain is
rinsed with tap water and then decolourised with that it is cheap, the result can be made available in
Microbiology of Tuberculosis
a few hours and, more importantly, it can be used
to determine the infectivity of a patient suspected
of pulmonary open tuberculosis, based on clinical
findings or chest radiography. (Fig. 8.6).
Auramine-rhodamine Stain. This is a fluorochrome
stain employing auramine-rhodamine fluorescent
dyes. The smear is flooded with auramine stain
for 10 minutes and then rinsed with tap water.
It is then decolourised with 5% acid alcohol for
2 minutes, rinsed again with tap water and then
counter-stained for 5 minutes with potassium per-
manganate. The slide is allowed to air-dry and is
then examined under fluorescent microscopy with
25x and 40x objectives. At lower magnification, the
smear should be read in three parallel sweeps of the
long axis, with a minimum of 30 fields of view. Acid-
fast bacilli appear as bright luminous yellow bacilli
approximately 1-10 flm long, most typically in the
form of slender rods but may appear curved or bent.
Individual bacteria may display beaded and uneven
staining. Some mycobacteria other than M. tubercu-
losis may appear pleomorphic, ranging from long
rods to coccoid forms, with more uniform distribu-
tion of staining properties. A positive fluorochrome
stain requires confirmation by over-staining the
same slide with ZN stain. The fluorochrome stain is
more sensitive and rapid than the ZN stain since the
examiner is looking for yellowish-green fluorescing
bacteria against a dark background and scanning of
the smears is usually at a lower magnification.
8.3.2.1
New Acid-Fast Bacilli (AFBJ Detection Methods
Improved AFB detection methods in clinical speci-
mens have been described, in order to increase the
efficiency of standard microscopic procedures. One of
these tests is the immunomagnetic separation strategy
for capturing and concentrating mycobacteria from
processed specimens. Magnetic polystyrene beads are
coated with antibody to mycobacteria and then mixed
with the sputum sample. The beads now covered with
organisms are then recovered using a magnet. The
organisms are washed off and stained and examined
by fluorescent microscopy. The detection of M. tuber-
culosis by the immunocapture technique has been
improved by the use of magnetic beads coated with
rabbit IgG polyclonal antibody to lipoarabinomannan,
a constituent of the mycobacterial cell wall (Fig. 8.1).
The preliminary evaluation is very impressive but more
clinical laboratory evaluation is still required (Shinnick
and Good 1999).
119
B.3.2.2
Specimen Preparation and Culture
The definitive diagnosis of mycobacterial infec-
tion depends on the detection and recovery of
AFB in clinical specimens. Since mycobacteria are
often present in situations where there is resident
bacterial flora (e.g. sputum) or are present in
small numbers (e.g. CSF), specimens submitted
for examination must be specially prepared for
culture. Three main steps are usually required for
processing specimens:
a) Bacterial flora present in the specimens must be
removed or reduced in numbers before inocula-
tion in special media, since these are faster grow-
ing than mycobacteria;
b) Mycobacteria trapped in mucin must be released
by a liquefaction process without killing the
mycobacteria, for easy identification and culture;
c) Specimens with low bacterial load must be con-
centrated to enable detection by microscopy and
culture;
d) Suitable media (solid and/or liquid media) and
appropriate incubation environment must be chosen
to ensure optimum recovery of mycobacteria.
8.3.2.3
Liquefaction and Decontamination
The selection of a liquefaction and decontamination
process that maintains the viability of the mycobacte-
ria, eliminates the present bacterial flora and liquefies
the mucin in the sample is desirable (Roberts et al.
1991). The three most widely used digestion meth-
ods are NaOH, the Zephiran-trisodium phosphate
method and the NALC-NaOH method (Roberts et al.
1991). The NALC-NaOH method utilises a mucolytic
agent, NALC for digestion and NaOH for decontami-
nation and it is the most commonly used method in
clinical laboratories. The advantage of this method is
that a large number of specimens can be processed
in a short time; concentrated smears can be ready for
staining immediately and it has been shown to give
a high rate of recovery of mycobacteria (Roberts et
al. 1991) (Figs. 8.4, 8.5).
Specimens collected aseptically from sterile body
sites such as CSF, ascitic fluid and synovial fluids
may be inoculated directly into appropriate media.
Specimens other than sputum do not usually require
digestion and decontamination. Urine specimens can
be inoculated directly or treated with 10% CaCl and
then centrifuged for 30 minutes at 2,500xg and the
120 A. O. Osoba

Fig. 8.4. Mycobacterium tuberculosis showing serpentine Fig. 8.6. Ziehl Neelsen stain of sputum showing a mass of acid-
cords (Ziehl Neelsen stain. (After Meylan, 1993) fast bacilli in a patient with open tuberculosis

SMEAR deposit inoculated into appropriate media. For stool

Staining:
Microscopy samples, the specimen can be processed by the NaOH
Ziehl-Neelsen sputum digestion method and then inoculated into
Auramine-Rhodanine appropriate media.
After suitable treatment of the specimens, the sus-
pension should be centrifuged for at least 15 minutes
CULTURE/ TB confirmation of at 2,200-2,500xg or higher and the deposit inoculated
Inoculation +vesmears
of Media by PCR,LCR into appropriate media. All centrifugation should be
carried out in sealed safety cups.

SOLID LIQUID
U 12B
Middlebrook Septi-Check 8.3.3
BacT/Alert Culture Media for the Isolation of Mycobacteria
BAaEC MGIT 960

8.3.3.1
Solid Media

Three types of solid media have been described for

the isolation of Mycobacteria and each has its advan-
tages (Table 8.2).

a) Egg-based media - The most widely known media

is the Lowenstein-Jensen (LJ) medium, which is
an egg-based non-selective medium. LJ medium
has a long shelf life, up to 1 year when refrigerated
* AST =Antimicrobial
and is very cheap to prepare; it is used widely in
susceptibility developing countries. Its disadvantages are that it
Testing may be difficult to distinguish artefacts from true
Fig. 8.5. Processing of clinical specimens for mycobacteria colonies on this medium and the fact that heat
identification and susceptibility testing is required for solidification, which along with
Microbiology of Tuberculosis 121

the presence of albumin inactivates some anti-

tuberculous drugs. On the whole it is good for
the recovery of M. tuberculosis but not so much
for the non-tuberculous mycobacteria (Table 8.2).
It is generally regarded as the ,gold standard' as it
results in a very high recovery rate of mycobacte-
ria (Hopewell and Bloom 1994; Garcia et al. 1998;
Palacios et al. 1999) (Fig. 8.7).
The other types of egg-based media are the
American Thoracic Society medium and a modi-
fied LJ medium by Petragnani, which contains a
different concentration of malachite green used
in the medium.
Conventional culture on solid media, although
inexpensive but specific, requires 1-5 x 102 bacilli
per millilitre for reliable positive results and may
take 6-8 weeks, since the growth of M. tuberculo-
Fig. 8.7. Mycobacterium tuberculosis growing on Lowenstein-
sis species is rarely visible to the naked eye until Jensen's medium. Note buff-coloured colonies produced after
after 18 days of incubation. (Fig. 8.7) (Garcia et al. 4 weeks incubation
1998; Palacios et al. 1999).
b) Agar-based media - the agar-based selective
media are the Middlebrook 7H1O and 7Hll media, c) Biphasic media - Another media used for the cul-
which contain defined salts, vitamins, cofactors, ture of Mycobacteria in some laboratories is the
oleic acid, glycerol, glucose as enrichment, cata- BBL-Septic-Chek AFB system, which is a biphasic
lase and biotin, to stimulate the regrowth of dam- system, containing both agar and broth. It uses
aged bacilli, and albumin to promote growth by the Middlebrook 7H9 broth in its lower chamber
combining with the toxic products in the media and agar slope of Middlebrook 7H 11 in the upper
(Table 8.2). part. Mycobacteria grow readily in this medium

Table 8.2. Different Media used for the Isolation of Mycobacteria

Medium Composition Malachite green Inhibitory agents for

present/absent non-mycobacterial contaminants
(gllOO ml)

Egg-based Fresh whole eggs, defined salts, glycerol, 0.025 None

(non-selective) potato flour
Lowenstein-Jensen
Egg-Based 7H9 broth base, bovine serum albumin, None Polymyxin B, amphotericin B, nalidixic acid,
(liquid -selective) casein hydrolysate, catalase, 14C-labelled trimethoprim, azlocillin
Middlebrook 7H12 palmitic acid
Agar-based medium Defined salts, vitamins, cofactors, oleic 0.00025 Cycloheximide, lincomycin, nalidixic acid
(selective) acid, albumin, catalase, glycerol, glucose
Middlebrook 7H10
Agar-based medium Defined salts, vitamins, cofactors, oleic 0.0025 Carbenicillin, amphotericin B, polymyxin B,
(selective) acid, albumin, catalase, glycerol, glucose, trimethoprim lactate
Middlebrook 7H11 casein hydrolysate
Liquid culture Same as Middlebrook 7H12 above None Same as in 7H12 above
Bactec (Middle-
brook 7H12)
Liquid culture Middlebrook 7H9 broth supplemented None Amphotericin B, azlocillin, nalidixic acid,
MB/BacT (Middle- with growth factors polymyxin B, trimethoprim
brook 7H9 broth)

* Modified from Roberts et al' (1991)

122 A. O. Osoba

and at 2 or 3 days intervals during incubation the 8 weeks (Garcia et al. 1998; Palacios et al. 1999).
broth is made to run over the agar. Colonies are 3) BACTEC MGIT 960 - The MGIT (Mycobacterium
formed on the agar if they contain mycobacteria. growth indicator tube) system (Becton Dickinson,
The incorporation of NAP in the agar in a second Sparks, Md.) uses tubes in which a fluorescent
tube will suppress the growth of M. tuberculosis compound is embedded in silicone on the bottom
and allow quick identification of MBT (Jenkins of tubes. This fluorescent compound is sensitive
1998). to the presence of oxygen dissolved in the broth.
Initially, the large amount of dissolved oxygen
Agar-based media are expensive and have a rela- quenches emissions from the compound and
tively short shelf life, hence are not widely used in little fluorescence can be detected. Later, actively
developing countries. However, they have the advan- respiring Mycobacteria consume the oxygen and
tage of easy identification of microcolonies and allow the fluorescence to be detected. The mean
accurate drug concentrations can be achieved for time to detection (TTD) is about 15.4 days (range
susceptibility testing. 4-47 days (Flanagan et al. 1999). It has the advan-
tage of accommodating 960 tubes, which is suit-
8.3.3.2 able for large volume laboratories (Fig. 8.9).
Liquid Automated Culture Systems
Both the MB/BacT and MGIT systems have the
1) The Bactec 460 TB instrument was the first auto- same limitations of failure to detect mixed cultures,
mated instrument developed by Becton Dickinson, the inability to identify colonial morphology and
Sparks, Md. for the recovery of Mycobacteria from high cost of media and equipment. These preclude
clinical specimens, using a radiometric detection their Widespread use in developing countries. These
system. The liquid medium is an enriched Middle- liquid culture systems increase the recovery of myco-
brook 7H9 base supplemented with bovine serum, bacteria while at the same time decreasing the time of
albumin (fraction V) catalase, casein hydrolysate recovery, since the mycobacterial growth is detected
and 14 C-labelled substrate. The medium is used more rapidly in liquid media. Many clinicallaborato-
for the isolation, differentiation of M. tuberculosis ries now use a combination of solid media and liquid
complex from non-tuberculous mycobacteria and selective media in primary isolation of mycobacteria
drug susceptibility testing. Although it is capable since most species grow in 7-14 days in liquid media
of recovering Mycobacteria in 7-14 days, it has as opposed to 6-8 weeks with solid media.
several disadvantages, such as the inability to
observe colonial morphology and mixed cultures
and high costs. It is labour-intensive and uses
radioactive reagents and needles, which could
lead to cross-contamination. In view of these
limitations, it is now less commonly used.
2) The MB/BacT system (Organon Teknika, Turn-
hout, Belgium) is the first fully automated non-
radiometric system for the culture of Mycobacte-
ria, approved by the FDA in 1996 (Fig. 8.8). It relies
on a continuous colorimetric CO 2 detection to
indicate Mycobacterial growth in a closed system.
The measured values are transmitted continu-
ouslyevery 10 minutes to a computer, which indi-
cates vials with Mycobacterial growth based on a
sophisticated algorithm. The system will recover
Mycobacteria in 10-14 days. In the MB/BacT
system the process bottles contain Middlebrook
7H9 broth, Tween 80, amaranth and an antibiotic
supplement containing amphotericin B, azlocillin
nalidixic acid, polymyxin Band trimethoprim.
The MB/BacT cabinet holds 240 process bottles, Fig.8.8. MB/BacT instrument for automated mycobacterial cul-
which are continuously monitored for up to ture, identification and susceptibility testing (Organon Teknika)
Microbiology of Tuberculosis
Fig. 8.9. BACTEC MGIT 960 instrument for automated culture
and susceptibility testing (Becton Dickinson)
8.3.3.3
Inoculation and Incubation Procedure
Solid Media
Egg-based Medium. After the decontamination and
digestion of the samples, the sediment is inoculated
into Lowenstein-Jensen slope and incubated in the
dark at 35-37°C for 8 weeks, in an atmosphere of
5-10% CO 2 for the first 7 days, with the screw caps
left loose. The cultures are visually inspected each
week for evidence of growth (Fig. 8.7).
Liquid Automated Culture Systems. The supplements
are reconstituted and inoculated into culture pro-
cess bottles. The processed sample from the speci-
men is inoculated into the culture bottles. They are
loaded into the incubator cabinet, which monitors
the growth for 8 weeks. Bottles flagged positive are
checked for purity and presence of acid-fast bacilli
and then subjected to identification tests and suscep-
tibility testing (Figs. 8.8, 8.9).
8.3.4
Identification of Mycobacteria Species
Conventional methods for the identification of Myco-
bacteria are (1) growth rate at different temperatures,
(2) colonial morphology, (3) pigmentation in dark or
light, and (4) biochemical tests. The biochemical tests
commonly used to differentiate Mycobacteria are
123
niacin test, nitrate reduction, susceptibility to T2H,
catalase test, urease test, arylsulphatase test, iron uptake
test, growth on MacConkey agar without crystal violet,
sodium chloride tolerance test, deamination of pyra-
zinamide, tel1urite reduction and Tween 80 hydrolysis.
Mycobacterium tuberculosis can be identified by
its slow growth on solid media,non-photochromoge-
nicity, strongly positive niacin test, negative catalase
test and positive nitrate reduction test.
The p-nitro- -acetylamino-beta-hydroxypropio-
phenone (NAP) test is an intermediate compound in
the synthesis of chloramphenicol and inhibits spe-
cies of the Mycobacterium complex (i.e. M. tubercu-
losis, M. bovis, M. africanum, M. microti and bacilli
Calmette-Guerin). This compound does not inhibit
other mycobacteria. This test can be used to identify
M. tuberculosis in positive cultures in the liquid auto-
mated culture systems in 4-6 days (Siddiqi 1999).
However, although the NAP test has been recom-
mended by the Centers for Disease Control and
Prevention and widely used for the identification of
M. tuberculosis, it is not to be used alone as a confir-
matory test because of some reports of false-positive
results (Rau and Libman 1999) (Fig. 8.5).
8.4
Molecular Techniques for Identification
of Mycobacteria
Since conventional methods for the identification
of Mycobacteria take 2-3 weeks, there has been
increased effort to develop rapid identification tech-
niques, especially in the wake of the resurgence of
tuberculosis as a result of the HIV/AIDS pandemic.
Molecular techniques have been developed in the last
decade with the aim of achieving faster commence-
ment of appropriate treatment, shortening time to
detection of anti-mycobacterial drug resistance, and
instituting appropriate control measures.
8.4.1
Nucleic Acid Probes for Actively Growing
Mycobacteria
Actively growing mycobacteria may be identified by
using commercial DNA probes, which detect spe-
cific DNA or ribonucleic acid (RNA) fragments by
nucleic acid hybridisation. These tests require some
amplification of the bacterial DNA. These methods
include target, probe and signal amplification tech-
124
niques, where ribosomal RNA or different parts of
genomic DNA are selected as targets. The specificity
of the tests is nearly 100%, but their sensitivity is still
relatively lower compared with culture, especially in
smear-negative specimens. The tests are cumber-
some and expensive, which preclude their use in
many developing countries. Probes are now available
for the identification of M. tuberculosis complex, M.
avium complex, M kansasii and M. gordonae isolated
either on solid or liquid media.
8.4.2
The Accuprobe System
In the Accuprobe system (Gen-Probe, San Diego,
Calif.) the bacterial culture is sonicated to break
up the bacterial cells, the lysate is incubated with
an acridinium-Iabelled oligonucleotide probe. This
hybridises to mycobacterial ribosomal RNA. The
unbound oligonucleotide is removed and the bound
probe is measured using a luminescence detection
system. The Accuprobe can identify M. tuberculosis
is under 2 hours. The test has been reported to be
specific and when combined with the Bactec system
can provide speciation of an isolate from a clinical
sample in 14 days. However, false positives have
been reported with M. avium-intercellulare complex
strains (Marshall and Shaw 1997).
8.4.3
The BDProbe Tec ET
The BDProbe Tec (Becton Dickinson, Sparks, Md.)
is a nucleic amplification technique, which is based
on homogeneous strand displacement amplification
(SDA) and fluorescent energy transfer detection in an
instrument system. SDA is an isothermal enzymatic
process that amplifies nucleic acid exponentially.
The process is based on the nicking of a modified
recognition sequence by the restriction endonuclease
BsoBl and the extension and repair of that site by
the DNA polymerase Bst, which synthesises a new
strand of DNA while displacing the existing strand.
The displaced strand can then serve as a template
for further amplification. The process takes place
at a temperature of 52.5°C. An internal amplifica-
tion control system is run with each sample and is
designed to verify the SDA reaction. The assay con-
sists of direct M. tuberculosis complex (DTB) assay
performed on clinical respiratory specimens and the
culture identification (ID) assays for M. tuberculosis
A.O.Osoba
complex, M. avium complex and M. kansasii, per-
formed on isolates grown in liquid or solid media.
The test consists of three main steps: sample process-
ing, priming and warming, and simultaneous ampli-
fication and detection. The whole procedure takes
about 2 hours. While the sensitivity and specificity
of the test is about 100% for smear-positive samples,
however for smear-negative specimens the sensitivity
and specificity are about 85% and 97%, respectively
(Bergmann et al. 2000).
Current probe technology, however, cannot distin-
guish live from dead mycobacteria, which indicates
that while molecular techniques will achieve great
usefulness in the primary diagnosis of mycobacte-
rial infections, cultural methods will be required for
monitoring the response to therapy. Most clinical
laboratories now combine both molecular techniques
and cultural methods in the laboratory diagnosis and
susceptibility testing of Mycobacteria and this prac-
tice is likely to continue in the foreseeable future.
8.4.4
The Amplified Mycobacterium Tuberculosis
Direct (MTD) Test
The MTD test uses a transcription-mediated ampli-
fication (TMA) principle to detect M. tuberculosis
complex directly from respiratory specimens and
provides a billion-fold amplification of the rRNA
targets (MTD, Gen-Probe, San Diego, Calif.). The
amplification process uses a single temperature
throughout the test. The target is 16s RNA. The myco-
bacterial cells are sonicated and the nucleic acids
are released. The specimens are then heated to 42°C
and multiple copies of the mycobacterial RNA are
generated. In the test system, ampIicons are created
through DNA intermediates by a reverse transcrip-
tase. M. tuberculosis complex-specific sequences are
then detected by the Gen-probe chemiluminescent-
labelled DNA probes and the results are measured in
a luminometer. The test is recommended for use only
with smear-positive concentrated specimens such as
bronchoalveolar lavages, bronchial aspirates and tra-
cheal aspirates. The specificity of the test is 97-99%,
while the sensitivity is 82-97%. The advantage of the
test is that it uses a single temperature and a single
tube format. The detection of rRNA accounts for its
high sensitivity. The high reliability of the test makes
it an ideal tool for diagnosing tuberculosis with low
bacterial count. Since the test has a negative predic-
tive value close to almost 100%, it can quickly help
in ruling out tuberculosis from the differential diag-
Microbiology of Tuberculosis
nosis, alternatively, if the smear is ZN positive, it sug-
gests the presence of non-tuberculous mycobacteria
if MTD negative (Pfyffer 1994).
8.4.5
Amplicor M. Tuberculosis (PCR) Test
The Amplicor M. tuberculosis test by Roche Diag-
nostic Systems (Branchburg, N.J.) is based on a
polymerase chain reaction (PCR), nucleic acid
hybridisation for the detection of M. tuberculosis
complex in digested, decontaminated sputum and
bronchial alveolar lavage samples. The test is based
on three major processes: PCR, target amplifica-
tion, hybridisation of the amplified product to a
specific nucleic acid probe and the detection of the
amplified product by colour formation. In the test,
gene-specific primers located in a highly conserved
region of the 14S ribosomal RNA (rRNA) gene of
M. tuberculosis are used to amplify a 584 base-pair
sequence. The DNA-containing sample and reagent
mixture are heated to separate the double-stranded
helix and expose the primer target sequences. As
the mixture cools, the biotinylated primers anneal
to their targets. The thermostable DNA polymerase,
in the presence of excess of deoxynucleoside tri-
phosphates (dNTPs), extends the annealed primers
along the target templates to produce amplicons.
This process is repeated for a number of cycles,
each cycle effectively doubling the amount of
target DNA. After 37 cycles, over a billion copies
can be produced from a single copy of DNA. The
ampIicons are then chemically denatured to form
single strands that are added to a microwell plate
containing a bound, amplicon-specific oligonucle-
otide probe. The biotin-labelled amplicons will then
bind (hybridise) to the amplicon specific probe and
thus be captured on to the plate. The detection of the
amplicons is carried out by an enzyme immunoas-
say technique by the addition of Av-HRP conjugate
and the optical density of the mixture is measured
in an automated microwell plate reader. In smear-
positive samples the sensitivity was found to be 95%
and the specificity to be 98% in one report, while
in smear-negative samples the sensitivity was 75%
and the specificity was 99% (Rau and Libman 1999;
Gilpin et al. 1999; Carpenter et al. 1995). However
there are reports of some current commercial kits
having a sensitivity range of up to 70% for smear-
negative M. tuberculosis samples (Rau and Libman
1999). The assay is rapid and easy to perform, taking
about 8 hours and can be fitted into the normal
125
working hours of a clinical diagnostic laboratory
(Beavis et al. 1995).
8.4.6
Ligase Chain Reaction (LCR)
Ligase chain reaction is a probe amplification tech-
nique incorporating a DNA polymerase and it is based
on sequential rounds of template-dependent ligation
of two adjacent oligonucleotide probes. When a pair
of probes has hybridised to the target sequence on a
single strand of DNA, there is a gap of a few nucleo-
tides between the probes. The addition of a DNA
polymerase into the reaction acts to fill in this gap by
incorporating nucleotides. Once the gap is filled ligase
can covalently join the pair of probes to form an ampli-
fication product that is complimentary to the original
target sequence and can itself serve as a target in sub-
sequent cycles of amplification. The target nucleic acid
sequence for LCR assay for M. tuberculosis (Abbott
LCX probe system, Abbott Diagnostics, Chicago, Ill.)
is found within the single copy chromosomal gene of
M. tuberculosis which encodes for protein antigen b.
This gene sequence is specific to the M. tuberculosis
complex and has been detected in all M. tuberculosis
complex strains. In the assay sediments from a clinical
sample or culture material are sonicated and amplified
in a thermocycler. The amplicons were detected in an
LCX analyser (LCX probe system). In smear-positive
samples the sensitivity was found to be 100% and the
specificity to be 98% in one report, while in smear-
negative samples the sensitivity was 89% and the
specificity was 98% (Gilpin et al. 1999).
8.4.7
Other Molecular Techniques
A number ofmolecular techniques have been described
for the rapid identification of mycobacteria growing on
solid media. These are (1) thin-layer chromatography,
(2) high-performance liquid chromatography, (3) gas-
liquid chromatography, (4) analysis with DNA probes.
Although these procedures can produce rapid identi-
fication of mycobacteria growing on solid media, they
are too insensitive for the identification of isolates in
liquid media, due to the relatively low cell mass often
produced, or they are only capable of identifying a
limited number of species (Taylor et al. 1997).
In summary, these newer techniques will have to
be used to supplement the conventional methods
and need to be interpreted in conjunction with the
126
patient's clinical features, such as medical history,
physical examination, chest radiograph, tuberculin
test and therapeutic response to anti-tuberculous
agents (Pfyffer 1994).
8.5
Antimicrobial Susceptibility Testing
Antimicrobial susceptibility testing (AST) of isolated
Mycobacteria is very important not only for appro-
priate therapy but also for control of the disease and
the identification of resistant strains. Conventional
methods are still the method of choice but these take
3-8 weeks, while the Bactec 460 TB and the liquid
culture systems take 8-12 days (Siddiqi 1999). The
newer methods of susceptibility testing have distinct
advantages and improvements over the older meth-
ods (Inderlied 1994).
8.5.1
Conventional Methods of Susceptibility Testing
There are three methods of conventional suscepti-
bility testing of M. tuberculosis and these are based
on its growth on solid medium containing a fixed
amount of the drug. (Hawkins IE et al.1991; Vareldzis
et al. 1994).
a) Absolute concentration method: medium contain-
ing known concentrations of antimicrobial agents
and control medium without drugs are inoculated
with standardised suspension of M. tuberculosis.
A preliminary reading is made at 2 weeks and a
final reading at 4 weeks. The strain is regarded as
resistant if there is growth greater than 20 colony-
forming units (CPU) at a specific drug concentra-
tion. The method is highly sensitive to variations
in inoculum size.
b) Resistance ratio method: this method is similar
to the absolute concentration method in that
both the drug-free media and the drug-con-
taining media are inoculated with both the test
strain and a standard strain of M. tuberculosis (
H37Rv). Resistance is expressed as the ratio of
the MIC of the test strain divided by the MIC of
the control organism used in the test. If the test
strain has a ratio of 8 or more it is considered
as resistant, while a resistance ratio of 2 or less
is defined as sensitive. A resistance ratio of 2 or
less is equivalent to the presence of less than 20
A. O. Osoba
colonies growing on 0.2 mgtl of isoniazid in the
absolute concentration method. In this method,
it is important that the inoculum be properly
standardised. Variations in the inoculum size are
difficult to prevent (Vareldzis et al. 1994).
c) Proportion method: this is the most widely used
method and it involves the inoculation of both
drug-free media and drug-containing media in
duplicate with various dilutions of the test organ-
ism and the plates incubated. The media used are
the Middlebrook 7HlO or 7H11. The dilutions are
made to produce between 50 and 100 colonies on
at least one of the control media. The CPU on the
drug-containing media is then compared with
that of the drug-free media. Resistance is defined
as Mycobacteria with greater than 1% of the pop-
ulation exhibiting growth in the presence of the
lowest concentration of the drug tested. Results
can be available in 14-21 days (Hawkins et al.
1991). The proportion method is the most widely
used method in the USA, as it provides fairly
accurate and reproducible results for primary
and secondary anti-mycobacterial drugs. How-
ever the drawback of the method is the difficulty
of standardisation of the inoculum and quality
control. In addition, it has not been validated for
testing new agents such as ciprofloxacin, ofloxacin
and rifabutin (Hawkins et al. 1991; Vareldzis et al.
1994).
8.5.2
Radiometric Method of Drug Susceptibility
Testing
Bactec 460 TB System. In 1980, Gardner Middlebrook
introduced the Middlebrook 7H12 liquid medium,
an automatable radiometric detection system for the
growth of Mycobacteria in a liquid selective medium
using the Bactec 460 TB system (Becton Dickinson
Diagnostic Instruments Systems, Towson, Md.). The
system revolutionised the isolation and susceptibility
testing of Mycobacteria. The system is far superior
to the conventional solid media in that it recov-
ers more positive cultures in a significantly shorter
period (2 weeks instead of 6-8 weeks) (Marshall and
Shaw 1997; Siddiqi 1999). The medium contains 14 C_
labelled palmitic acid and the growth of Mycobacte-
ria is detected by metabolic release of radio-labelled
carbon dioxide from the palmitic acid. Radioactivity
of the labelled CO 2 is determined quantitatively and
is a measure of the rate and amount of growth in the
vial. These numbers are designated the "growth index»
Microbiology of Tuberculosis 127

(GI). By daily monitoring of the culture vials, detection 8.5.5

times have been shortened in this system to 7-14 days. The MB/BacT Microbial Detection System
By the addition of anti-mycobacterial drugs into the
vials, inhibition of growth results in a reduction in This is also a non-radiometric, automated colori-
the amount of 14C produced and a reduction of the metric system consisting of a culture bottle with a
GI, compared with the drug-free vial (control). The growth sensor for detecting mycobacterial growth.
Bactec system allows the drug susceptibility testing to The instrument incubates and scans the bottles for
be performed in 12-14 days after isolation and identi- positivity and a computerised data management
fication, with excellent correlation with conventionally system records and reports results. The MB/BacT
determined susceptibility testing (Marshall and Shaw susceptibility test is performed using the MB/BacT
1997). It can be used to test all primary drugs, i.e. process bottles, MB/BacT antibiotic supplement
isoniazid, rifampicin, streptomycin, ethambutol and (MAS) reconstitution fluid and antimicrobial drugs.
pyrazinamide as well as secondary drugs including The drugs are added to the new MB/BacT process
quinolones and rifabutin. The main drawback of the bottles (one drug per bottle and one concentration
method is the reporting of false susceptibility or resis- per drug). These bottles with drugs are inoculated
tance due to testing of mixed populations of M. tuber- with broth from a positive MB/BacT bottle. Next to
culosis and non-tuberculous mycobacteria. However, these bottles with drugs a process bottle without
the critical concentrations have not been established drugs is inoculated with broth from the positive
for the newer agents (Inderlied 1994). MB/BacT bottle. This bottle is considered the growth
control. Growth of the M. tuberculosis isolate in the
bottles with drugs is compared to the growth control
8.5.3 to decide if the isolate is resistant or susceptible. The
Non-Radiometric Methods of Drug Susceptibility M. tuberculosis isolate is considered resistant if the
Testing bottle containing the drug is flagged positive before
or at the same time as the growth control. An isolate
The worldwide resurgence of tuberculosis has under- is considered susceptible if the bottle containing the
scored the necessity to develop rapid methods for drug remains negative during the test period or when
isolation and susceptibility testing of mycobacteria. flagged positive after the growth control. The con-
Because of the concerns for radioactive materials, centration at which the drugs are tested are strepto-
some governments placed restrictions on the Bactec mycin 1 Ilg/ml, isoniazid 1 Ilg/ml, rifampicin 1 Ilg/ml,
460 system culture bottles. This led to the develop- ethambutol 2 Ilg/ml and pyrazinamide 50 Ilg/ml. The
ment ofnon-radiometric culture systems such as the test is concluded in 12 days. The MB/BacT cabinet
Mycobacterial Growth Indicator Tube (MGIT) Bactec holds 240 process bottles, which are continuously
MGIT 960 (Becton Dickinson) and the MB/BacT monitored for up to 8 weeks (Fig. 8.8).
(Organon Teknika, Durham, N.C.).

8.5.4
The Mycobacterial Growth Indicator Tube (MGIT)
System
It is an automated non-radiometric culture system
using Middlebrook 7H9 broth and an oxygen sensi-
tive fluorescence sensor to indicate microbial growth
in a vial. It is similar to the Bactec 460 system. By
the addition of anti-mycobacterial drugs, the actively
respiring mycobacteria consume the oxygen and
allow fluorescence to be detected by the instrument,
indicating resistance. Control bottle without any
drug is included with each run. The cabinet has a
capacity to accommodate 960 bottles (Fig. 8.9). The
mean time to determination of susceptibility or resis-
tance is 10-12 days (Flanagan et al. 1999).
8.5.6
Genetic Approach to the Detection of Drug
Resistance by Mycobacterium tuberculosis
Recent major advances have been made in uncovering
the molecular basis of the resistance of M. tuberculo-
sis to various antimicrobials, viz. isoniazid (Zhang et
al. 1992; Banerjee et al. 1994), streptomycin (Finken
et al.1993), quinolones (Takiff et al.1994), rifampicin
(Telenti 1993a,b, 1997). For example, the mechanism
of rifampicin resistance by M. tuberculosis has been
associated with point mutations and small insertions
or deletions in a limited region of the gene encoding
for the B-subunit of the RNA polymerase (rpoB). In
the case of isoniazid, mutations in the kat G, inh AI
mab A, ahp C and oxy R genes confer resistance on
M. tuberculosis. Based on this approach, commercial
128
tests have become available for use in clinical diag-
nostic laboratories.
8.5.6.1
Line Probe Assay (UPA)
This test is based on the detection of mutations by
hybridisation in the rpo B gene, which confers resis-
tance on M. tuberculosis. Telenti (1993a,b; Cole and
Telenti 1995) developed a screening method using PCR
and single strand conformation polymorphism analy-
sis and Williams (1994) used the PCR and heteroduplex
formation to detect rifampicin resistance in M. tubercu-
losis cultures. In the LIPA test, oligonucleotide probes
are immobilised as parallel lines at known locations
on a nitrocellulose strip (Beenhower et al. 1995). The
amplified material is denatured and incubated with a
LIPA strip in a shaking water bath for 1 hour at 50°C in
1 ml of hybridisation mixture. The hybrids formed are
revealed by an immunoenzymatic procedure, resulting
in a coloured precipitate that enables visual reading of
the results. If a mutation is present in one of the probe
target regions, the mismatch created will prevent the
corresponding probe from hybridising. The test is fast
and easy to perform and allows the simultaneous detec-
tion of M. tuberculosis and the discrimination between
rifampicin-sensitive strains and rifampicin-resistant
strains of M. tuberculosis in clinical samples. The test
can be completed in 2 days and when performed on
cultures it takes about 8 hours. In one report, (Been-
hower et al. 1995) the susceptibility of M. tuberculosis
strains can be correctly determined in up to 97% of the
strains tested. However, the test will fail to determine
the susceptibility of strains utilising a different mecha-
nism of resistance from the rpo B gene (Beenhower et
al. 1995; Drobniewski and Pozniak 1996).
8.5.7
Newer Techniques of Drug Susceptibility Testing
8.5.7.1
Luciferase-based Reporter Phage
This is a molecular technique which has recently been
described to provide a rapid drug susceptibility test-
ing of M. tuberculosis. In this test, the gene encoding
a luciferase enzyme from fireflies was inserted into
a bacteriophage. When the bacteriophage infects M.
tuberculosis, the luciferase enzyme is synthesised and
emits light. Following the exposure of the organism
to anti-mycobacterial agent, the organism dies and no
light is emitted, indicating susceptibility to the drug.
A. O. Osoba
Between 500 and 5,000 mycobacteria infected by this
bacteriophage will produce a clear positive signal and
the test can be completed in 3 hours (Marshall and
Shaw 1997; Roth et al. 1997; Jacobs et al. 1993). This
is a promising test, which will find useful application
in clinical laboratories in the future.
8.5.7.2
Epsilometer (E-) test
The E-test has been used very recently to determine
the Minimum Inhibitory Concentration (MIC) of
many organisms. More recently the technique has
been modified to determine the susceptibility of
M. tuberculosis and M. avium-intercellulare to both
first-line and second-line anti-mycobacterial agents
(Flynn et al. 1997; Koontz et al. 1994; Wanger and
Mills 1994). The E-test ( AB Biodisk, Solna, Sweden)
is based on a stable-gradient strip impregnated
with an antimicrobial drug placed on Middlebrook
7Hll media which has been inoculated with the test
organism, and then incubated for 5 days. The results
are interpreted by applying National Committee for
Clinical Laboratory Standards (NCCLS) criteria. Four
strips can be placed on a 150-mm diameter plate. The
method produces MICs within 1 week of primary iso-
lation, thus allowing early modification of therapy.
8.5.7.3
Hybridisation Protection Assay (HPA)
This is a simple and fast M. tuberculosis drug suscepti-
bility test, which is now commercially available. After
isolation of the organism from clinical specimen, it
is grown on Middlebrook 7H9 broth and incubated
at 37°C for 5-7 days. The inocula are standardised to
McFarland no. 0.5 standard and then mixed with an
anti-mycobacterial drug solution to give a final con-
centration of 0.1 and 1.0 Ilg/ml for isoniazid and 1.0
and 10.0 Ilg/ml for rifampicin. The resulting bacterial
suspension with or without the test drug, is cultured
on an agitating plate at 37°C for up to 5 days. At 0-,
1-,3- and 5-day intervals each sample is removed and
subjected to the HPA test with acridinium ester (AE)-
labelled DNA probe{ AccuProbe; Gen-Probe, Inc. San
Diego, Calif.). The assay results are read in a lumi-
nometer and expressed as relative light units (RLU).
The statistical significance of the RLU between tubes
with and without drugs is calculated (Miyamoto et
al. 1996). In this test, drug-resistant strains of M.
tuberculosis can be accurately detected after 1 day of
incubation with the drug. The HPA test is safe, simple
and fast. It can be incorporated into the routine of
Microbiology of Tuberculosis 129

a clinical diagnostic laboratory. However, it is more serological diagnosis of tuberculosis have been
expensive than the radiometric or conventional agar handicapped by the fact that the pattern of antibody
methods. It may have useful application in some situ- response after M. tuberculosis infection varies from
ations such as testing immunocompromised patients one lesion to another and from patient to patient. In
infected with multidrug-resistant M. tuberculosis. many patients the antibody response is weak and is
In general, a handicap of both genetic and phe- frequently directed predominantly against non-spe-
notypic tests for susceptibility of M. tuberculosis is cific mycobacterial antigens (Woodhead 1992).
that they cannot precisely distinguish between mixed Antigen preparations have contained moieties
cultures that contain 1% to 10% of resistant cells and present in many microbial species that lead to sero-
cultures that contain only susceptible organisms. This logical cross-reactivity. Assays with crude antigens
distinction is of clinical importance since the patient have had poor specificity because of the presence
with a mixed culture would be regarded as having of cross-reacting epitopes between M. tuberculosis
a drug-resistant M. tuberculosis and would be given and other mycobacteria. By means of chemical and
appropriate treatment (Shinnick and Good 1999). immunological methods, several purified antigens,
In conclusion, conventional methods based on such as A60 antigen (Amicosante et al. 1993; Alifano
ZN microscopy and culture on LJ medium with et al. 1994; Charpin et al. 1990) and 38 kDa antigen
biochemical tests to identify the isolates are the (Simonney et al. 1996; Young et al. 1986; Bothamley et
most widely used for the diagnosis of tuberculosis al.1992; Harboe and Wiker 1992) have been prepared
and are still considered the "gold standard" . These which have improved the enzyme-linked immuno-
methods are cost-effective but are rather slow for sorbent assay (ELISA) test chatacteristics (Wilkins
the effective control of tuberculosis. The new tech- and Ivanyi 1990; Jackett et al. 1988). Recently, mouse
niques of nucleic acid amplification offer the pros- monoclonal antibodies (Mabs) to specific epitopes
pect of more rapid diagnosis, susceptibility testing on antigens of M. tuberculosis have been used in a
and improved and accurate epidemiological infor- solid-phase antibody competition test to measure
mation (Marshall and Shaw 1997). The resurgence antibody titres to these determinants in human sera
of tuberculosis has necessitated the development (Jackett et al. 1988; Cocito 1991; Bothamley et al.
of more rapid methods of diagnosis. The objec- 1992). The application of Mabs into the serological
tive of clinical laboratories should be to report the tests for the diagnosis of tuberculosis is a significant
isolation of M. tuberculosis within 10-14 days of the advance because they can be readily prepared in a
receipt of the sample and the result of drug suscep- standard form and are usually free from problems
tibility tests within the next 14 days (Shinnick and of cross-reactivity. They also allow the estimation of
Good 1999; Rau and Libman 1999). both IgG and IgM in a single assay, thereby obviating
the need to purify antigens from M. tuberculosis.
The interest to develop serological tests for the
diagnosis of tuberculosis stems from the fact that
8.6 blood is readily available in any form of the disease
Serological Diagnosis of Tuberculosis for serological testing. Therefore the measurement
of titres of antibody to M. tuberculosis specific
Since 40-60% of patients with pulmonary disease antigens may be of value in the diagnosis of both
and up to 75% of patients with extrapulmonary smear-positive and smear-negative pulmonary
diseases are smear-negative, a number of alternative tuberculosis. Antibody titres may also help to
diagnostic methods using molecular, chromato- identify those patients who should be treated in a
graphic and immunological methods have been hospital, those who require observed chemother-
developed. While these tests are very specific and apy as well as those who might require additional
only moderately sensitive, they rely on the provision drugs to the standard regimen for the therapy of
of adequate material from the site of the lesion. This their infection.
is often difficult, especially in children, and in most In order to meet these objectives, a number of
cases of extrapulmonary tuberculosis, such as cere- serological tests developed for the diagnosis of
bral tuberculoma or tuberculous osteomyelitis and tuberculosis have now become commercially avail-
abdominal tuberculosis (Wilkins and Ivanyi 1990; able. These tests can be divided into two groups, viz.:
Pottumarthy et al. 2000). Previous investigations into (a) Immunochromatographic tests and (b) Enzyme-
tuberculous serology have been faced with problems linked immunosorbent Assay. Some of these will be
of both sensitivity and specificity. These studies on briefly described.
130 A.O.Osoba

8.6.1 8.6.2.3
Immunochromatographic Tests PATHOZYME-MYCO IgG, IgA, and IgM Tests

8.6.1.1 The three assays measure the levels in serum of IgG,

ICT Tuberculosis Test IgA and IgM antibodies, respectively, to two antigens
lipoarabinomannan (LAM) and recombinant 38 kDa
This is a immunochromatographic test in which five antigen. These kits detect infection due to Mycobac-
highly purified antigens (including one of 38 kDa) terium species.
secreted by M. tuberculosis during active infection In a study evaluating the performances of these
are immobilised in four lines on the test strip (ICT tests in various categories of patients (Pottumarthy
Diagnostics, Balgowish, New South Wales, Australia). et al. 2000), it was found that the ICT Tuberculosis
The test detects the presence of immunoglobulin G test followed by PATHOZYME-MYCO IgG had the
(IgG) antibodies to these antigens. best performance characteristics with sensitivi-
ties of 41 % and 55%, respectively, with sera from
8.6.1.2 patients with acute tuberculosis and specificities
Rapid Test TB of 96% and 89%, respectively, with sera from the
Mantoux test controls, and 88% and 90%, respec-
It is a one-step coloured immunochromatographic tively, with sera from anonymous controls. In a
assay (Quorum Diagnostics, Vancouver, British study in the United Arab Emirates (Nsanze et al.
Columbia, Canada). It detects antibodies to the recom- 1997) on a population of patients with tuberculosis
binant 38 kDa antigen from M. tuberculosis expressed and leprosy, using a combination of PATHOZYME
and purified in E coli (Pottumarthy et al. 2000). TB-Complex test and PATHOZYME-MYCO IgG
test, it was found that the tests showed a significant
difference in antibody levels between the patients
8.6.2 with active pulmonary disease, extrapulmonary
Enzyme-linked Immunosorbent Assays tuberculosis and leprosy in comparison with the
control group. The sensitivity of the two tests
8.6.2.1 combined for proven pulmonary tuberculosis was
Tuberculosis IgA EIA about 95% using a higher EIA cut-off of 1.6, while
the specificities were 100%. The sensitivity for
This test detects the IgA antibody to a mycobacterial extrapulmonary tuberculosis at the higher cut-
Kp90 immuno-cross-reactive antigenic compound off was 51 %. The authors conclude that while the
(ImCRAC). tests may be of some value in decision making in
extrapulmonary tuberculosis, they may be useful in
8.6.2.2 serological screening of populations at high risk in
The Pathozyme TB Complex Test endemic areas (Nsanze et al. 1997).
Although a negative result by any of these tests
This test detects serum IgG antibody to a recombi- would be useful in helping to exclude disease in a
nant 38 kDa antigen from M. tuberculosis expressed population with a low prevalence of tuberculosis, a
and purified from E coli, permitting the isolation of positive result may aid in clinical decision making
significant quantities of protein (OMEGA DIAG- when applied to symptomatic patients being evalu-
NOSTICS LTD, Alloa, Scotland). This antigen has ated for active tuberculosis (Pottumarthy et al. 2000).
been reported as the single most important antigen The tests now commercially available confirm a more
for the serological diagnosis of tuberculosis. It is a limited antibody response in smear-negative pulmo-
unique disease-associated protein, which appears to nary disease, but suggest that a battery of monoclo-
be completely specific to the M. tuberculosis complex, nal antibodies, with differing specificities, produces
i.e. M. tuberculosis, M. bovis and M africanum. The better results than single monoclonal antibody
kit is specific for the diagnosis of disease due to M. (Woodhead 1992). However they have the advantage
tuberculosis complex. In this test the test sera are that they are rapid and relatively inexpensive. Over-
diluted 1:50 and standards were provided to gener- all, more research is still required to produce better
ate a semi logarithmic reference curve. The results tests with increased sensitivities in extrapulmonary
are expressed as sero-units. tuberculosis which can be readily incorporated into
clinical diagnostic laboratories.
Microbiology of Tuberculosis
References
Alifano M et al. (1994) ELISA method for evaluation of anti-A
60 IgG in patients with pulmonary and extrapulmonary
tuberculosis. Microbiologica 17:37-44
Amicosante M et al. (1993) Antibody repertoire against the
A60 antigen complex during the course of pulmonary
tuberculosis. Eur Respir J 6:816-822
Banerjee A et al. (1994) InhA, a gene encoding a target for
Isoniazid and Ethionamide in Mycobacterium tuberculosis.
Science 263:227-230
Beavis KG et al. (1995) Evaluation of amplicor PCR for direct
detection of Mycobacterium tuberculosis from sputum
specimens. J Clin MicrobioI33:2582-2586
Beenhower HD et al. (1995) Rapid detection of rifampicin
resistance in sputum and biopsy specimens from tuber-
culosis patients by PCR and line probe assay. Tuberc Lung
Dis 76:425-430
Bergmann JS, Keating WE, Woods GL (2000) Clinical evalua-
tion of the BDProbeTec ET system for rapid detection of
Mycobacterium tuberculosis. J Clin Microbiol 38:863-865
Bothamiey GH et al. (1992a) Clinical value of the measure-
ment of M. tuberculosis-specific antibody in pulmonary
tuberculosis. Thorax 47:9-14
Bothamley GH et al. (1992b) Clinical value of the measure-
ment of Mycobacterium tuberculosis specific antibody in
pulmonary tuberculosis. Thorax 47:270-275
Carpenter E et al. (1995) Diagnosis of tuberculosis by amplicor
Mycobacterium tuberculosis test: a multicenter study. J Clin
MicrobioI33:3106-3110
Charpin D et al. (1990) Value of ELISA using A60 antigen in
the diagnosis of active pulmonary tuberculosis. Am Rev
Respir Dis 142:380-384
Cocito CG (1991) Properties of the mycobacterial antigen
complex A60 and its applications to the diagnosis and
prognosis of tuberculosis. Chest 100:1687-1693
Cole ST, Telenti A (1995) Drug resistance in Mycobacterium
tuberculosis. Eur Respir J B SuppI20:701s-713s
Davies PDO, Girling DJ, Grange JM (1996) Tuberculosis. In:
Weatherall DJ, Ledingham JGG, Warrell DA (eds) Oxford
textbook of medicine. Oxford University Press, Oxford,
pp 638-655
Drobniewski FA, Pozniak AL (1996) Molecular diagnosis,
detection of drug resistance and epidemiology of tubercu-
losis. Br J Hosp Med 56:204-208
Finken M et al. (1993) Molecular basis of streptomycin resis-
tance in Mycobacterium tuberculosis: Alteration of the
ribosomal protein S 12 gene and point mutations within a
functional16S ribosomal RNA pseudoknot. Mol Microbiol
9: 1239-1246
Flanagan PG, Williams R, Paull A (1999) Comparison of two
automated systems for the isolation of mycobacteria from
clinical specimens. Eur J Clin Microbiol Infect Dis 18:
912-914
Flynn et al. (1997) Application of the E test to the antimicrobial
susceptibility testing of Mycobacterium marinum clinical
isolates. J Clin 35:2083-2086
Garcia FG et al. (1998) Evaluation of the MB/BACT automated
mycobacteria culture system versus culture on Lowenstein
medium. Clin Microbiol Infect 4:339-343
Gilpin CM et al. (1999) Comparative study of Amplicor poly-
merase chain reaction and Ligase chain reaction for direct
131
detection of M. tuberculosis in clinical specimens. Saudi
Med J 20:79-84
Grange JM (1984) Tuberculosis. In: Wilson G, Miles A, Parker
MT (eds) Topley and Wison's principles of bacteriology,
virology and immunity, 7th edn. Arnold, London, pp 32-35
Harboe M, Wiker HG (1992) The 38-kDa Protein of Mycobacte-
rium tuberculosis.: a review. J Infect Dis 166:874-884
Hawkins JE, Wallace RJ, Brown BA (1991) Antibacterial
susceptibility tests: Mycobacteria. In: Balows A, Hausler
WJ, Herrmann KL, Shadomy HJ (eds) Manual of clinical
microbiology, 5th edn, chap 114. American Society for
Microbiology, Washington DC, pp 1138-1152
Hopewell PC, Bloom BR (1994) Tuberculosis and other myco-
bacterial diseases, 2nd edn. In: Murray JF, Nadel JA (eds)
Textbook of respiratory medicine. Saunders, Philadelphia,
pp 1094-1117
Inderlied CB (1994) Antimycobacterial susceptibility testing:
present practices and future trends. Eur J Clin Microbiol
Infect Dis 13:980-993
Jackett PS et al. (1988) Specificity of antibodies to immuno-
dominant mycobacterial antigens in pulmonary tubercu-
losis. J Clin MicrobioI26(11):2313-2318
Jacobs WR et al. (1993) Rapid assessment of drug susceptibili-
ties of Mycobacterium tuberculosis by means of luciferase
reporter phages. Science 260:819-822
Jenkins PA (1998) The microbiology of tuberculosis. In: Davies
PO (ed) Clinical tuberculosis. Chapman and Hall, London,
pp 69-79
Koontz FP et al. (1994) Etest for routine clinical antimicrobial
susceptibility testing of rapid growing mycobacteria iso-
lates. Diagn Microbiol Infect Dis 19:183-186
Marshall BG, Shaw RJ (1997) Diagnostic techniques: old and
new. Eur Respir Mon 4:30-50
Meylan PR ASM News vol 59(12), 1993 (Cover page)
Miyamoto J et al. (1996) New drug susceptibility test for Myco-
bacterium tuberculosis using the hybridization protection
assay. J. Clin Microb. 34:1323-1326
Moulding T (1999) Tuberculosis and non nontuberculous
mycobacterial infections, 4th edn. In: Schlossberg D (ed)
Pathogenesis, pathophysiology, and immunology: clinical
orientations. Saunders, Philiadelphia, pp 48-64
Nsanze H et al. (1997) Serodiagnosis of tuberculosis and lep-
rosy by enzyme immunoassay. Clin Microb and Infection.
3:236-239
Palacios JJ et al. (1999) Fully automated liquid culture system
compared with lowenstein-Jensen solid medium for rapid
recovery of mycobacteria from clinical samples. Eur J Clin
Microbiol Infect Dis 18:265-273
Pfyffer GE (1994) Amplification techniques: hope or illusion
in the direct detection of tuberculosis. Med Microbiol Lett
3:335-347
Pottumarthy S, Wells VC, Morris AJ (2000) A comparison of
seven tests for serological diagnosis of tuberculosis. J Clin
MicrobioI38:2227-2231
Rau NV, Libman (1999) Laboratory implementation of the
polymerase chain reaction for confirmation of pulmonary
tuberculosis. Eur J Clin Microbiol Infect Dis 18:35-41
Roberts GD, Koneman EW, Kim YK (1991) Mycobacterium.
In: Balows A, Hausler WJ, Herrmann KL, Shadomy HJ
(eds) Manual of clinical microbiology, 5th edn, chap 34.
American Society for Microbiology, Washington DC, pp
304-318
Roth A, Schaberg T, Mauch H (1997) Molecular diagnosis of
132
tuberculosis: current clinical validity and future perspec-
tives. Eur Respir J 10:1877-1891
Shinnick TM, Good RC (1999) Diagnostic mycobacteriology
laboratory practices. Clin Infect Dis 21 :291-299
Siddiqi SH (1999) Recent advances in the diagnosis of tubercu-
losis: an overview. J King Abdulaziz Univ Med Sci 7:31-35
Simonney N et al. (1996) Comparison of A60 and three gly-
colipid antigens in an ELISA test for tuberculosis. Clin
Microbiol Infect 2:214-222
Takiff HE et al. (1994) Cloning and nucleotide sequence of
the Mycobacterium tuberculosis gyrA and gyrB genes and
characterization of quinolone resistance mutations. Anti-
microbial Agents Chemother 38:773-780
Taylor TB et al. (1997) Routine use of PCR-restriction frag-
ment length polymorphism analysis for identification of
mycobacteria growing in liquid media. J Clin Microbiol
35(1):79-85
Telenti A (1993a) Detection of rifampicin-resistance
mutations in Mycobacterium tuberculosis. Lancet 341:
647-650
Telenti A (1993b) Direct automated detection of Rifam-
picin-resistant Mycobacterium tuberculosis by poly-
merase chain reaction and single-strand conformation
polymorphism analysis. Antimicrob Agents Chemother
37:2054-2058
Telenti A (1997) Genetics of drug resistance in tuberculosis.
Clin Chest Med 18:55-64
A. O. Osoba
Tuberculosis. In: Wyngaarden JB, Smith LH (eds) Cecil text-
book of medicine. Saunders I Harcourt Brace Jovanovich,
Philadelphia, pp 1682-1696
Vareldzis BP et al. (1994) Drug-resistant tuberculosis: labora-
tory issues. World Health Organization recommendations.
Tubercle Lung Dis 75:1-7
Wanger A, Mills K (1994) Etest for susceptibility testing of
Mycobacterium tuberculosis and Mycobacterium avium-
intracellulare. Diagn Microbiol Infect Dis 19:179-181
Wilkins EGL, Ivanyi J (1990) Potential value of serology for
diagnosis of extrapulmonary tuberculosis. Lancet 336:
641-644
Williams DL (1994) Characterization of rifampicin resistance
in pathogenic mycobacteria. Antimicrob Agents Che-
mother 38:2380-2386
Wolinsky E (1988) Tuberculosis. In: Wyngaarden JB, Smith
LH (Eds.) Cecil Textbook of Medicine. Pp 1682-1696. WB
Sounders Co., Harcourt Brace Javonovich, Inc., Philadel-
phia
Woodhead M (1992) New approaches to the rapid diagnosis
of tuberculosis. Thorax 47:264
Young D et al. (1986) Immunological activity of a 38-Kilo-
dalton protein purified from Mycobacterium tuberculosis.
Infect Immun 54(1):177-183
Zhang Y et al. (1992) The catalase-peroxidase gene and iso-
niazid reisstance of Mycobacterium tuberculosis. Nature
358:591-593
9 The Immunology and Pathogenesis of Tuberculosis
GRAHAM A. W. ROOK

CONTENTS 9.2.9.5 Macrophage Apoptosis and the Death

of Bacteria 141
9.1 Introduction 133 9.2.10 Evasion of the Antimicrobial Functions
9.1.1 Latent Tuberculosis and Its Relationship of Macrophages 141
to Persisters After Chemotherapy 134 9.2.11 Bacterial Genetics and Mechanisms of Avoidance
9.1.2 Protection Versus Immunopathology 134 of Destruction by the Type 1 Response 142
9.1.3 The Objectives of the Immunologist 135 9.3 Immunopathology 142
9.2 Immunity to Tuberculosis 135 9.3.1 The Koch Phenomenon; the Response
9.2.1 The Crucial Role of the Type 1 Response Characteristic of Disease 142
for Immunity to Tuberculosis in Mice 135 9.3.2 Tuberculin Response and Protection 142
9.2.2 The Possible Role of Other Types 9.3.3 The Relationship Between the Koch Phenomenon
of T Lymphocyte 135 and the Shwartzman Reaction 143
9.2.2.1 The Detrimental Role of Type 2 Responses 135 9.3.4 A Hypothesis to Explain the Koch
9.2.2.2 Variability of Tuberculosis in Different Mouse Phenomenon 143
Strains 135 9.3.5 Detrimental Roles of Tumour Necrosis Factor-a
9.2.3 The Protective Role of Tumour Necrosis Factor in Human Tuberculosis 144
(TNF-a) 136 9.3.6 Fibrosis 144
9.2.4 Immunity to Tuberculosis in Man 136 9.4 Endocrinology 144
9.2.4.1 Genetics of Susceptibility to Tuberculosis 136 9.4.1 Vitamin D3 Metabolism in Tuberculosis
9.2.4.2 Natural Human Gene Knockouts; Lesions 144
Proof of the Role of IFN-y and IL-12 136 9.4.2 Adrenal Steroids in Tuberculosis 145
9.2.5 Other T Cell Types That May Be Involved 9.4.2.1 The Effects of Stress 145
in Immunity 136 9.4.2.2 The Hypothalamo-Pituitary-Adrenal Axis
9.2.5.1 CD8+ IFN-y-secreting T Cells 136 in Human Tuberculosis 145
9.2.5.2 CD-I-restricted T Cells 137 9.4.2.3 Dysregulation of the Cortisol-Cortisone
9.2.5.3 Gammaldelta (y/'6) T Cells 137 Shuttle 145
9.2.5.4 Regulatory T Cells 138 9.5 Vaccination 145
9.2.6 The Role of TNF-a in Human Tuberculosis 138 9.5.1 Interference by Exposure to Environmental
9.2.7 Type 2 Responses in Human Tuberculosis 138 Mycobacteria 145
9.2.8 The Causes of the Controversy over Expression 9.5.2 Concurrent Parasite Infections 146
ofIL-4 in Human Tuberculosis 139 9.5.3 Vaccine Dose 146
9.2.8.1 The Mechanisms of the Shift Towards 9.5.4 A Possible Novel Approach to Vaccine Design 146
a Type 2 Cytokine Profile 139 9.5.6 Immunotherapy 146
9.2.9 Macrophage Function and M. tuberculosis 139 References 147
9.2.9.1 Uptake of Mycobacteria 139
9.2.9.2 Toll-like Receptors 140
9.2.9.3 Mycobactericidal Mechanisms Within
Macrophages 140
9.2.9.4 Reactive Oxygen and Nitrogen Intermediates 140
9.1
Introduction

Tuberculosis is a global emergency, currently aggra-

G. A. W. ROOK, BA, MB Bchir, MD
Professor, Department Medical Microbiology, Royal Free and
University College Medical School, Windeyer Institute of
Medical Sciences, 46 Cleveland Street, London WIT 4JF, UK
vated by HIV, the break-down of health care systems
and increases in drug resistance (Espinal et al. 2001).
There are now more cases of tuberculosis than ever
before in the history of mankind. The current failure
to control tuberculosis is attributable to several factors
that must be taken into consideration in any account of
the immunology of the disease. These factors are:

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
134
1. The inadequacy of the current vaccine, Bacillus
Calmette-Guerin. The protection given varies from
zero to about 80% (Fine 1995).
2. Inadequate understanding of the biology of the
organism itself. Mycobacterium tuberculosis can
remain viable within the tissues without causing
symptoms for the life of the individual (Wayne
1994) and, at least in countries with low or moder-
ate tuberculosis endemicity, many cases of tuber-
culosis result from reactivation of latent infection
(van Rie et al. 1999).
3. The best available chemotherapy must be contin-
ued for at least 6 months. This treatment regimen
is not a realistic proposition in most developing
countries, or even in the inner cities of rich ones,
because the patients feel well after a few weeks
and stop taking the drugs. WHO now admits that
directly observed therapy short-course (DOTS), in
which the patient is supervised while taking every
dose of therapy, helps but does not solve the prob-
lem (Butler 2000).
4. Our inadequate understanding of the mechanism
of protective immunity, which hampers efforts
to design better vaccines, and the extraordinary
problem of knowing how to select and test the
best vaccine candidates.
5. Inadequate understanding of how to implement
immunotherapy. Immunological manipulations that
correct the patients' failing immune responses may
be the only way to shorten treatment regimens.
9.1.1
Latent Tuberculosis and Its Relationship
to Persisters After Chemotherapy
There are two interrelated reasons for the requirement
for 6-month regimens. The first is obvious and often
discussed. The chemotherapy kills the vast majority
of the bacteria within a few days, but there are sub-
populations of "persisters" (Grange 1992). It is not
clear whether these organisms are in true stationary
phase (Siegele and Kolter 1992) or merely replicating
extremely slowly. Nor is it clear whether these persist-
ers, defined as organisms that survive chemotherapy,
are the same as "latent" or "dormant" bacilli that
remain in the tissues of the approximately 90% of
individuals who show no clinical disease after infec-
tion. These bacteria are evidently controlled by the
immune response, but not killed by it. Most research
has addressed dormant rather than persister bacilli,
though they are likely to be similar populations. It
was assumed that latent bacilli are in old lesions or
G.A,W.Rook
sites of fibrosis or calcification, where oxygen avail-
ability may be low. However, in a forgotten paper
published in 1927, Opie and Aronson reported that
80% of tuberculous lesions were already sterile (as
determined by culture in guinea-pigs) 5 years after
the primary infection, whereas live bacilli could be
found in macroscopically normal lung tissue (Opie
and Aronson 1927).
A major recent advance is the detection, by in situ
polymerase chain reaction of DNA, of M. tuberculosis
in lung tissue from victims of accidental death with
no history or post mortem signs of tuberculosis
(Hernandez-Pando et al. 2000a). When such samples
were taken from countries with endemic tuberculo-
sis (Ethiopia and Mexico), 30-40% of samples were
positive. In contrast, no DNA of M. tuberculosis was
found in samples from Norway, where transmission
within the community ceased several decades ago.
In the positive Ethiopian and Mexican samples, the
DNA was seen in histopathologically normal tissue
and was not confined to macrophages. Cell types
containing M. tuberculosis DNA included type II
pneumocytes, endothelial cells and fibroblasts
(Hernandez-Pando et al. 2000a). These observations
raise the possibility that by entering cells that are not
"professional antigen-presenting cells",M. tuberculo-
sis is able to escape detection and perhaps subvert the
immune response.
9.1.2
Protection Versus Immunopathology
The second reason for the need for prolonged
treatment is usually overlooked. Most tuberculosis
patients have a necrotising pattern of response to
M. tuberculosis, analogous to the phenomenon first
noted by Koch in guinea-pigs (Koch 1891). There is
overwhelming evidence that the Koch phenomenon
is not a correlate of optimal protective immunity to
tuberculosis. Indeed, pre-immunisation of animals
so that they have a Koch phenomenon before they
are challenged with virulent M. tuberculosis results
in a clear and reproducible increase in susceptibility
to the disease, compared to unimmunised controls
(Wilson et al. 1940). This and other aspects of the
Koch phenomenon are discussed in detail later. Its
relevance at this point is that this inappropriate pat-
tern of response may not correct itself rapidly during
treatment. Therefore, if chemotherapy is stopped at
3 months, relapse rates are high even when the che-
motherapy was an optimal rifampicin-containing
one that achieved sputum negativity well before
The Immunology and Pathogenesis of Tuberculosis
3 months, and in spite of the fact that there are very
few live organisms in the patients' tissues at this time
(Balasubramanian et al. 1990).
9.1.3
The Objectives of the Immunologist
The immunologist, therefore, needs to understand
the differences between protective immunity and
the Koch phenomenon, and the factors that deter-
mine which response pattern is present. The ultimate
objectives are better prophylactic vaccines, and also
therapeutic vaccines that can replace the pathologi-
cal response with a protective one very early during
treatment, using immunotherapeutic strategies that
also cause recognition and clearance of persisters
and dormant bacilli.
9.2
Immunity to Tuberculosis
9.2.1
The Crucial Role of the Type 1 Response
for Immunity to Tuberculosis in Mice
The ability to manipulate the immune system of mice
with neutralising antibodies or gene knockout (KO)
has provided strong evidence that, in this species,
immunity to tuberculosis correlates with a Type 1
response. In vivo CD4+ lymphocytes act in concert
with CD8+ cells and with numerous other cell types
including macrophages, B cells and some stromal
cells. Collectively these give rise to two patterns
of cytokine release known as Type I (dominated
by IL-12 from antigen-presenting cells, and by IL-
2, and IFN-y from Thl lymphocytes) and Type 2
(dominated by IL-4, 5 and 13 from Th2lymphocytes)
(Clerici and Shearer 1994; Salgame et al. 1991).
Disruption of the major histocompatibility
(MHC) class II genes or of the gene for the ~ chain
of the a/~ T cell receptor (Ladel et al. 1995), result-
ing in a deficiency of CD4+ a/~ T cells, render mice
susceptible even to the avirulent Bacillus Calmette
Guerin (BCG). Similarly, disruption of the gene for
IFN-y makes mice very susceptible to M. tuberculosis
(death within 3 weeks) and such mice may even die
after many weeks from challenge with BCG (Flynn
et al. 1993). A major inducer of the Type 1 pathway
is IL-12, and IL-12 KO mice are more susceptible to
tuberculosis (Wakeham et al. 1998).
135
9.2.2
The Possible Role of Other Types
ofT Lymphocyte
In addition to conventional CD4+ a/~, Class II MHC-
restricted T cells, several other T cell types are also
involved in the response to mycobacteria, including
CD8+ cells, gamma/delta (y/o) T cells, CDl-restricted
T cells and regulatory T cells. These are discussed later
in relation to the immunology of the disease in man.
9.2.2.1
The Detrimental Role of Type 2 Responses
These data emphasise the crucial role of the Type 1
response. In agreement with this, other data indicate
that the Type 2 response is not only unable to protect
mice but can seriously undermine the efficacy of the
Type 1 response. If a weak Type 2 response to the
shared mycobacterial antigens is deliberately induced
before challenge, mice are found to be strikingly more
susceptible to tuberculosis than are unimmunised con-
trol animals (Hernandez-Pando et al.1997b; Lindblad
et al. 1997). Similarly, in the Balb/c mouse model of
pulmonary TB infection, the appearance of IL-4 in the
lung lesions (as seen by immunohistochemistry and
RT-PCR) coincides temporally and spatially with the
appearance of areas of pneumonia and necrosis, lead-
ing to rapid clinical deterioration and death (Hernan-
dez-Pando et al. 1996). These observations are not
contradicted by a claim that, in IL-4 gene-disrupted
mice, there is no evidence of increased resistance to
the infection (North 1998). The animals used were
resistant anyway and had controlled growth of the
bacteria by 21 days, which is the optimum that can be
achieved (North 1998). In contrast, Balb/c mice control
the infection briefly at 21 days, and then relapse with
progressive disease. This late phase is accompanied by
a switch to Type 2 cytokine production (Hernandez-
Pando et al. 1996, 1997b), but progression is greatly
attenuated if the IL-4 gene is knocked out (Hernandez-
Pando et al., in preparation).
9.2.2.2
Variability of Tuberculosis in Different Mouse Strains
As outlined above, the nature of the disease caused by
M. tuberculosis in mice depends on the mouse strain
used. For instance,A/J mice develop progressive inter-
stitial pneumonitis, while C57BL/6 mice can develop
massive granulomas mediated by a Type 1 response
and so "drown" in Thl lymphocytes (Watson et al.
2000). Balb/c mice start with a Type 1 response and
136
control the infection at 21 days, but then switch to a
Type 2 response and die with an increasing bacterial
load and immunopathology (Hernandez-Pando et
al. 1996, 1997b). Therefore, we cannot be sure which
model is most similar to human tuberculosis without
first studying the human disease.
9.2.3
The Protective Role ofTumour Necrosis Factor
(TNF-a)
This pro-inflammatory cytokine can have either pro-
tective or detrimental effects in murine disease, and
the same is likely to be true in man, as discussed below
in relation to immunopathology. Its effects appear to
depend upon the other cytokines present. In the mouse,
TNF-a is protective early in infection. TNF-a levels are
elevated early (day 3) in mice infected via the intra-
tracheal route and reach a second peak in the third
week, coinciding with mature granuloma formation
(Hernandez-Pando et al. 1997a). TNF-a receptor KO
mice succumb faster to M. tuberculosis infection than
control mice (Flynn et al.1995), and a disruption of the
granulomatous response and increase in mycobacte-
rial load is noted in M. tuberculosis-infected mice when
TNF-a bioactivity is blocked (Adams et al. 1995).
9.2.4
Immunity to Tuberculosis in Man
9.2.4.1
Genetics of Susceptibility to Tuberculosis
Conventional genetic studies of tuberculosis patients
showed that polymorphisms in genes encoding natural-
resistance-associated macrophage protein (NRAMP1),
the IL-l gene cluster, the vitamin D receptor and man-
nose-binding lectin were associated with susceptibility
(Bellamy 2000). The relevance of vitamin D receptor
polymorphisms is increased when there is also vitamin
D deficiency as in the Gujarati population in London
(Wilkinson et al. 2000). However, the effects on disease
susceptibility are small, and so far studies of this type
have cast little light on the mechanisms of immunity.
9.2.4.2
Natural Human Gene Knockouts; Proof of the Role
of IFN-y and IL-12
On the other hand definitive evidence that, as in mice,
the Type 1 response is crucial for immunity to tuber-
G.A. W. Rook
culosis in man has come from the study of children
with genetic defects of the Type 1 cytokine system.
Vaccination with Bacille Calmette-Guerin (BCG), an
avirulent derivative of the organism responsible for
bovine tuberculosis, occasionally causes dissemi-
nated infection. The gene for the IFN-yRI gene in
such a child had a single nucleotide deletion that
resulted in the creation of a premature stop codon
near the N-terminus (Jouanguy et al. 1996). Another
study (Newport et al. 1996) involved four children
from the same small town in Malta who presented
with disseminated mycobacterial infections. The
mycobacterial species isolated were M. fortuitum, M.
avium (two strains) and M. chelonei. One child also
had prolonged salmonellosis. These children had a
single nucleotide substitution (A for C) rather than
a deletion (Newport et al. 1996). It allowed normal
levels of expression of the mRNA, but again intro-
duced a premature stop codon.
Interleukin-12 (IL-12) receptor deficiency has also
been found in otherwise healthy individuals with
mycobacterial infections. Unlike the children with IFN-
yR deficiency these patients are able to form mature
granulomata, but their NK cells and T cells secrete little
IFN-y. Thus, IL-12-dependent IFN-gamma secretion in
humans seems essential in the control of mycobacterial
infections, despite the formation of mature granulomas
(Altare et al.1998; de Jong et al. 1998).
9.2.5
Other T Cell Types That May Be Involved
in Immunity
In addition to conventional CD4+ aJ~, Class II MHC-
restricted T cells, several other T cell types are also
involved in the response to mycobacteria. These are
particularly fascinating because many of them recog-
nise non-peptide antigens. Fig. 9.1 summarises the lym-
phocyte types involved, and some of their functions.
9.2.5.1
CD8+ IFN-y-secreting T Cells
There may be a high frequency of IFN-y-secret-
ing CD8+ cytotoxic T cells in the peripheral blood
of healthy individuals who have been exposed to
tuberculosis (Pathan et al. 2000). This response was
maintained in an asymptomatic contact over 2 years,
so such cytotoxic T cells can clearly persist in indi-
viduals without active tuberculosis. These CD8 cells
recognised a peptide from ESAT-6 (an immunogenic
peptide released very early into cultures of M. tubercu-
The Immunology and Pathogenesis of Tuberculosis 137

Various cell types thai recognise

lipids, glycolipids, Isoprenoids

CD·1 1-1.

IL-12 COB

:-J Tc1 11----+----+ Cytotoxicity - - - - - - ,

Class I
MHC
04
Class II
= IFN'r Activation - - - - - - ,

Antigen·
presenting
ltl- ' - - - - - - - - - - Suppression
cells

~IL-10
T re TGF·~
IL-10

Class II
MHC
Class I
-
fr'L~
Tc
COB
IL·402
' - - - - Fibrosis

Fig.9.1. Multiple lymphocyte types are involved in immunity to mycobacteria. IFN -r, the major activator of macrophages,
may be released by CD4+ Thl cells, but also by CD8+ cytotoxic T cells and a variety of CD-I-restricted T cells that recognise
unconventional antigens such as lipids and glycolipids (see main text). Macrophage activation and macrophage killing by
cytotoxic cells can both lead to bacterial death. If the antigen-presenting cells are secreting IL-IO, rather than IL-12, the T
cells generated will tend to be regulatory T cells (Treg ) or IL-4 (IL-402) secreting CD8+ cells and CD4+ Th2 cells. The presence
of all of these in TB patients is well documented and, together with TNF·a (not shown), they will impair immunity and cause
suppression, macrophage deactivation, fibrosis and immunopathology. The reality is still more complex, because TB patients
also have T cells that make both IFN and lL-lO -r

losis), and other peptides recognised by this cell type
have been defined using cells from patients recovering
from tuberculosis (Cho et al. 2000). These observations
are important since both in mice (Silva and Lowrie
2000) and in man (Cho et al. 2000) CD8+ cytotoxic T
cells that use the granule-mediated pathway of killing
(as opposed to the Fas-FasL pathway) lead to killing
of the contained mycobacteria, too. The role of the
granule contents, particularly granulysin, has been
reviewed (Krensky 2000).
9.2.5.2
CD-l-restricted T Cells
In humans and other large long-lived animals there
are several types of T cell (including a~ T lympho-
cytes negative for both CD4 and CD8 molecules (so-
called double-negative T cells), rb T cells and certain
CD4+, CD8+, CD8a/a+ and NK lymphocytes) that
recognise mycobacteriallipoarabinomannan (LAM)
or mycolic acids presented by the MHC class I-like
molecules of the CDI family (Beckman et al. 1994;
Moody et al. 1997; Sieling et al. 1995). Similar T cells
have been detected that recognise an evolutionarily
conserved family of isoprenoid glycolipids that
include essential components of protein glycosyl-
ation and cell wall synthesis pathways (Moody et
al. 2000).
The role of the CD-l restricted T cells remains
unclear because rodents have a homologue of CD-
Id, but have lost the genes for CD-la, b or c which
are the forms that present mycobacterial lipids and
glycolipids. Thus, their protective role will need to
be tested in guinea-pigs or in large long-lived ani-
mals that have these genes, though it is reported
that some types of human CD-I-restricted T cell
not only produce significant amounts of IFN -y but
also appear able to lyse infected cells and directly
kill intracellular mycobacteria (Sieling et al. 1995;
Stenger et al. 1998).
It is fascinating that the different forms of CD-l
"survey" different intracellular compartments, and
so will encounter products of M. tuberculosis in dif-
ferent ways (Schaible et al. 2000; Sugita et al. 2000).
9.2.5.3
Gamma/delta (r/o) TCells
Like CD8+ T cells, rioT cells can also kill infected cells
by the granule pathway, and can reduce the viability
of the contained bacteria (Dieli et al. 2000b). These
cells are activated during pulmonary tuberculosis and
138 G.A.W.Rook

increase in frequency (Behr Perst et al.1999), but they IL-12 provide definitive evidence of the importance
decline after treatment (Dieli et al. 2000). of Type 1 cytokines, and suggest a close parallel
with the mouse models. Recently the negative role
9.2.5.4 of Type 2 cytokines in human TB, again paralleling
Regulatory T Cells the murine models, has been established (Marchant
et al. 2001; Seah et al. 2000; van Crevel et al. 2000).
A dominant theme in other domains of immunol- Expression of IL-4, whether measured by flow
ogy is the regulatory T cell or "T reg" though they cytometry or by sensitive quantitative RT-PCR
have received little attention in the context of on non-stimulated peripheral blood T cells (Seah
tuberculosis. Treg characteristically release IL-lO and Rook 1999), is increased and correlates with
and TGF-p, and suppress inflammation in models severity of disease and with cavitation (Seah et al.
of allergy, inflammatory bowel disease and auto- 2000; van Crevel et al. 2000). The IL-4 mRNA copy
immunity (Maloy and Powrie 2001). It has been number also correlates with total IgE (Seah et al.
claimed that T cells cloned from the bronchoalveo- 2000) and with levels of soluble CD30 (Seah and
lar lavage of human TB patients make both IFN-y Rook 2000). Thus, although it is true that actual
and IL-l 0, and so may be a form of Treg (Gerosa et al. cytokine levels and mRNA copy numbers are
1999). Other authors have also noted the increased higher for Th-l than for Th-2 cytokines in tuber-
secretion of TGFp and IL-I0, though some of the culosis, the major change in cytokine expression
IL-lO may come from macrophages rather than Treg compared to healthy donors is not, as previously
(Ellner 1997). Inappropriate premature activation of stated, the small decrease in expression of Thl
Treg is clearly a possible mechanism of failure of the cytokines, but rather a massive (80-100x) increase
immune response. in expression of Th2 cytokines (Seah et al. 2000).
Similarly, M. tuberculosis-responding T cells cloned
from the peripheral blood of untreated tuberculo-
9.2.6 sis patients, whether from Africa or Italy, secrete
The Role ofTNF-a. in Human Tuberculosis abundant IL-4 in response to M. tuberculosis. After
treatment, the cytokine profile returns to Type 1
The other cytokine that is clearly implicated in man, (Marchant et al. 2001). Interestingly, many of the
which again shows a parallel with data from the IL-4-secreting cells in TB patients are, in fact,
mouse models discussed above, is tumour necrosis CD8+ (Gerosa et al. 1999; van Crevel et al. 2000).
factor alpha (TNF-a.). Absolute proof of its role in Actually, it is likely that they are in reality secreting
man is available only in relation to latent tubercu- IL-402, the splice variant of IL-4 that drives fibrosis
losis. An antibody that neutralises TNF-a. has been (Atamas et al. 1999; Seah and Rook 1999). Finally,
achieving some success as a treatment for rheu- it has emerged that M. tuberculosis contains com-
matoid arthritis, but a clear consequence in some ponents that drive striking expression of IL-4 in
individuals is the reactivation of latent tuberculosis normal peripheral blood mononuclear cell cultures
(Keane et al. 2001). It is therefore interesting that the in vitro (Seah and Rook 2001).
DNA from M. tuberculosis detected by in situ PCR is These data have resolved a long-running contro-
frequently found in cells situated in tissues devoid versy which deserves explanation. That there is an
of lymphocyte infiltration (Hernandez-Pando et al. IL-4 component in the response of human tubercu-
2000a), so that lymphocyte-derived cytokines may losis patients to M. tuberculosis ought to have been
be at very low levels. On the other hand, numerous accepted 10 years ago, because there is no other
components of M. tuberculosis may directly trigger known explanation for the presence of specific IgE
TNF-a. release from macrophages and other non- antibody (Yong et al. 1989). Interestingly, the other
lymphoid cell types, and this local release of TNF-a. largely Type 2 cytokine-dependent antibody, IgG4,
may maintain latency. is also reported to be increased in patients (Wilsher
et al. 1999). Similarly, immunohistochemistry reveals
IL-4-expressing cells in the lymphoid tissue of tuber-
9.2.7 culosis patients (but not in tissue from patients with
Type 2 Responses in Human Tuberculosis sarcoidosis) (Bergeron et al. 1997).

These observations and, above all, the susceptibil-

ity of children with defective receptors for IFN-y or
The Immunology and Pathogenesis of Tuberculosis 139

9.2.8 Increasing antigen load is likely to be one factor,

The Causes of the Controversy over Expression since the Thl/Th2 balance is strikingly linked to dose
of IL-4 in Human Tuberculosis when immunising with particulate antigens such as
mycobacteria (Hernandez-Pando and Rook 1994), or
Why, then, has the matter been controversial? Firstly, leishmania (Bretscher et al. 1992).
IL-4 is biologically active at much lower concentra- There are now conclusive data showing that M.
tions than IFN-y and has a correspondingly lower tuberculosis contains components that drive IL-4
mRNA copy number, so methods that reliably pick expression and secretion in vitro when cultured with
up IFN-yor its messenger RNA fail to detect biologi- human peripheral blood mononuclear cells (Seah
cally significant levels of IL-4. Moreover the mRNA and Rook 2001). The nature and importance of these
for IL-4 has a very short half-life, so it has often components is currently under investigation.
disappeared from clinical specimens collected in In some populations excessive or inappropriate
the field before they reach the laboratory. Workers contact with environmental mycobacteria may have
who failed to detect IL-4 in either normal donors primed a Type 2 response to the crucial common
or TB patients assumed, with no logical justifica- antigens. This mechanism is clearly demonstrable
tion, that it was not increased in the latter. Secondly, in mice, in which it can massively increase suscepti-
attempts to increase cytokine expression by stimu- bility (Hernandez-Pando et al. 1997b) and has been
lation of the cells in vitro do not yield an accurate suggested, but not conclusively proven, in man (Fine
reflection of the cytokine balance present in vivo, 1995; Stanford et al.I981).
and rapid early production of IFN-y can suppress During active infection, selective apoptosis of Thl-
Th2 cytokine release, as repeatedly demonstrated like T cells may be a factor (Das et al. 1999;Varadhachary
by workers in the field of allergy. Finally, previous and Salgame 1998), as may prostaglandin release
studies failed to take into account the presence of (Hilkens et al.1995; van-der-Pouw-Kraan et al.1995).
the IL-4 splice variant (IL-482). This variant of lL-4 Finally there are now strong reasons for suggesting
may be an inhibitor of IL-4 activity and is always that endocrine interactions with the immune system
co-expressed with IL-4 at about the same level are important. Vitamin D3, cortisol metabolism and
(Arinobu et al. 1999). In lung cells it may be dehydroepiandrosterone levels are discussed in the
expressed at higher levels than IL-4 itself. However, endocrinology section below.
almost every study of IL-4 mRNA levels used prim-
ers that would amplify mRNA for both lL-4 and the
splice variant (Seah and Rook 1999). 9.2.9
Macrophage Function and M. tuberculosis
9.2.8.1
The Mechanisms of the Shift Towards a Type 2 9.2.9.1
Cytokine Profile Uptake of Mycobacteria

What are the likely causes of this shift in cytokine Mycobacteria are taken up by multiple pathways
profile? Several possibilities are shown in Fig. 9.2. including complement receptors and mannose recep-

Fig. 9.2. Mechanisms that may lead to excessive acti-

vation of Type 2 cytokines and immunoregulation in
human tuberculosis. These can impair mycobacteri-
cidal functions of macrophages and, in concert with
TNF-a, contribute to immunopathology. Additional
factors not shown are prostaglandins, and also prior
induction of Th2 response to cross-reactive compo-
nents of environmental mycobacteria
140
tors (Kang and Schlesinger 1998; Schlesinger 1998).
However, this is clearly not the whole story because in
vitro M. tuberculosis can enter a variety of cell types
that do not express these receptors (Filley and Rook
1991; Shepard 1958). The exact mode of uptake must
affect the subsequent fate within the cell (Astarie-
Dequeker et al. 1999).
9.2.9.2
Toll-like Receptors
Much of the initial activation and cytokine response
of macrophages to mycobacteria may be mediated by
interaction with the Toll-like receptors (TLR). These
are members of the interleukin 1 (IL-1) receptor family,
related to Toll, a molecule involved in innate microbial
resistance mechanisms in Drosophila. Both virulent
and attenuated strains of M. tuberculosis can activate
in a TLR-dependent manner. TLR2, but not TLR4,
could confer responsiveness to LAM isolated from
rapidly growing mycobacteria. In contrast, LAM iso-
lated from M. tuberculosis or Bacillus Calmette-Guerin
failed to induce TLR-dependent activation. However,
other soluble and cell wall-associated components
of M. tuberculosis are involved. A soluble heat-stable
and protease-resistant factor was found to mediate
TLR2-dependent activation, whereas a heat-sensitive
cell-associated mycobacterial factor mediated TLR4-
dependent activation (Means et al. 1999). Lipoproteins
can activate via TLR, and several from M. tuberculosis
will drive IL-12 production in this way (Brightbill et al.
1999), perhaps explaining the latter result.
9.2.9.3
Mycobactericidal Mechanisms Within Macrophages
Inhibition or killing ofM. tuberculosis is easily induced
in murine macrophages by exposure to IFN-y but such
1) Activation MACROPHAGE
IFN-y Acidification of phagosomes
lNF-a Maturation to hydrolytic vesicle
Reactive oxygen and
nitrogen derivatives ------l
2) Cytotoxic T cells
~
'YJ Perlorin
Granulysin - - - - - +
3) Induction of Fas
apoptosis -----A--------"
ATP ~ Death of some bacteria and
H20 2 uptake of apoptotic bodies
Fas Ligand by fresh macrophages
Multiple mechanisms attempt to kill M. tuberculosis within macrophages
G.A. W. Rook
effects are extremely difficult to demonstrate convinc-
ingly in human cells (Rook et al. 1985, 1986). Success
has been reported using human alveolar lavage macro-
phages exposed to TNF-a in vitro (Hirsch et al.1994).
It is possible that the major killing mechanism is not
a direct effect of activated macrophages, but rather an
event that occurs during certain types of apoptosis
(see below) or during killing of the macrophage by
cytotoxic T cells that "inject" granulysin and perforin
(see section on CD8+ effector cells above) (Fig. 9.3).
9.2.9.4
Reactive Oxygen and Nitrogen Intermediates
If macrophages do themselves kill M. tuberculo-
sis, these are likely candidate killing mechanisms
(Fig. 9.3) (Chan et a1.1992, 1995). The mechanism of
action of the nitric oxide (NO) is uncertain, because
it has important signalling and second messenger
functions that may be as important as direct toxicity
for the organisms (Roach et a1. 1994). However, KO
mice unable to make NO or other reactive nitrogen
intermediates (RNI) showed no increase in prolif-
eration of M. tuberculosis in the lungs until very
late in the infection, but there was increased growth
in the spleen. In contrast, KO mice unable to make
reactive oxygen intermediates (ROI) had increased
growth of bacilli in the lungs. Interestingly, activa-
tion of macrophages by IFN-y in vitro to control
proliferation of M. tuberculosis was dependent upon
RNI rather than ROI, and so appeared to parallel
immunity in the spleen rather than in the lungs
(Adams et a1. 1997). The role of NO in man remains
unclear. It probably is made by appropriately acti-
vated human macrophages (Nicholson et a1. 1996),
but never in the very large quantities that murine
macrophages can release. There is not universal
agreement about the antimycobacterial relevance
Fig. 9.3. It is still not clear which are the most impor-
tant mechanisms leading to killing of M. tuberculo-
sis, particularly in human cells where killing is very
difficult to achieve in vitro. Candidate pathways are
shown and all are referenced and discussed in the
text. Note that only some inducers of apoptosis
result in reduced viability of the contained bacte-
ria. Further macrophage activation must occur via
the Toll-like receptors but their role in killing is not
yet known
The Immunology and Pathogenesis of Tuberculosis 141

of this mechanism (Aston et al. 1998). There is 9.2.10

some evidence that l,25(OHhD3 may be involved Evasion of the Antimicrobial Functions
in the activation of iNOS in a human monocyte cell of Macrophages
line (Rockett et al. 1998), so this could explain the
antimycobacterial effect of this material in vitro Apoptosis (programmed cell death) is a major regulator
(Rook et al. 1986). ofthe immune system,byproviding differential removal
of different lymphocyte and macrophage subsets. Some
9.2.9.5 human peripheral blood T cells undergo apoptosis in
Macrophage Apoptosis and the Death of Bacteria the presence of antigen from M. tuberculosis (Das et
al. 1999; Hirsch et al. 1999). In contrast, several groups
Large numbers of apoptotic T cells and macrophages report that M. tuberculosis actively inhibits apoptosis
are seen in caseous foci. Immunohistochemistry and of macrophages (Fig. 9.4). Mannose-capped lipoarabi-
in situ hybridisation showed that the macrophages nomannan (Man-LAM) stimulates phosphorylation of
surrounding caseous foci are negative for the anti- Bad, a pro-apoptotic protein, and so inhibits apopto-
apoptotic protein bel2, but positive for the pro-apop- sis (Maiti et al. 2000). Another group has shown that
totic protein bax, while the associated T cells express Man-LAM inhibits apoptosis by a mechanism involv-
IFN-y and FasL (Fayyazi et al. 2000). Macrophage ing calcium-dependent signalling (Rojas et al. 2000).
apoptosis may be beneficial to the host, because Moreover cells containing M. tuberculosis are markedly
death of the infected macrophage can be associated more sensitive to killing by TNF-a (Filley and Rook
with death of the contained mycobacteria (Fig. 9.3). 1991), but infected cells secrete soluble TNF receptors
Moreover, it has been suggested that reduction in (sTNF-R2) that can block their destruction by TNF-a
bacillary numbers is only achieved by apoptosis of (Fratazzi et al. 1999). Thus it might be suggested that
infected monocytes, not by the necrotic mode of M. tuberculosis causes death of a subset of T cells, while
death (Molloy et al. 1994; Oddo et al. 1998). Apopto- preserving some macrophages as hiding places with
sis induced by ATP promotes the killing of virulent reduced microbicidal and antigen-presenting func-
M. tuberculosis within human macrophages (Kusner tion. A point against this hypothesis is that infected
and Adams 2000; Lammas et al. 1997) as does apop- macrophages may also have down-regulated mRNA
tosis induced by Fas ligand (Oddo et al. 1998), and expression of bcl-2, an inhibitor of apoptosis (Klingler
hydrogen peroxide-induced apoptosis also causes et al. 1997). It may be that the organism is preferentially
mycobactericidal effects (Lauchumroonvorapong et inducing forms of apoptosis that leave the organisms
al.1996). Thus, perhaps M. tuberculosis has strategies unharmed.
that oppose macrophage apoptosis in order to limit In addition to opposing apoptosis, mycobacteria
the bacterial killing that can accompany this event. have numerous other strategies for avoiding being

M. tuberculosis avoids being killed by macrophages

1) Mechanisms during uptake
Uptake via Mannose receptor
Block Ca++ signalling by complement receptor

Mannose Cell not activated Macrophage

receptor
2) Resistance to killing
I Blocked Tough cell w S) Increased release
soluble TNFf11
Fig. 9.4. This figure should be compared with Fig. 3.
function SuperoXJde dismu1aSe Each of the killing pathways in Fig. 3 is opposed by
Free radlClll scavengers IL·l0
TGF·p some mechanisms in Fig. 4. Short thick lines indicate

H~"C?~ 8
IL-6
blocked pathways. Most of the figure is self-explana-
tory. Uptake via mannose receptors may avoid trig-
3) Alterations to vesicles gering killing mechanisms. The vacuoles containing
Failed maluratJon 6) Decreased release
Allered fUSlOtl panern or expression mycobacteria lie within the sorting/recycling endo-
Blocked proton pump Faa somal machinery of the macrophage and fail to
SHP-l TNF-C
4) Control cell signalling ..... acidify and fuse with acidic lysosomes. But they do
Actrvate phosphotyroslne phosphalase fuse with LAMP-I-containing endosomes and they

9
Block signalling via IFN-y receplOfS release vesicles containing LAM. Thus their fusion
Illock an1Jgetl presa<llatJon
Receptor pattern is altered by the pathogen. (LAM lipoarabi-
Peptide + Class II MHC nomannan, LAMP-I lysosome-associated protein 1)
SHP-I is a phosphotyrosine phosphatase
142
killed by phagocytes (Fig. 9.4) (Chan et al. 1991), M.
tuberculosis may be taken up via mannose receptors
that fail to trigger killing events (Astarie-Dequeker
et al. 1999). It also inhibits complement-receptor-
mediated Ca++ signalling, which may contribute to
the failure of killing mechanisms (Malik et al. 2000).
Mycobacteria can inhibit acidification of the phago-
some (Sturgill-Koszycki et al. 1994) and modify
intracellular trafficking of vacuoles so they behave
like part of the endosomal recycling compartment
rather than as toxic phagolysosomes (Xu et al. 1994).
These vacuoles release quantities of lipoarabino-
mannan (LAM) which insert into glycosylphosphati-
dylinositol (GPI)-rich domains in the cell membrane
(Ilangumaran et al. 1995). LAM is itself a GPI of
unusual glycan structure, with the ability to modify
numerous macrophage functions including the abil-
ity to respond to IFN -y and the ability to present
antigen (reviewed in Ilangumaran et al. 1995). The
last point may be relevant to the apparent inability of
long-term mycobacterium-infected macrophages to
present antigen to CD4+ T cells (Pancholi et al.1993).
Thus the organisms may hide, undetected, in modi-
fied macrophages.
The pattern of cytokine release from infected mac-
rophages changes so that macrophage activation is
diminished and T cell recruitment impaired (Fig. 9.4).
Recruitment of Thl lymphocytes requires IL-12 pro-
duction, which is inhibited by increased production of
TGF-~ and IL-IO (Ellner 1997; Fulton et al.1998; Toossi
et al.1997) and IL-6 release may also be a factor. TGF-~
and IL-I0 also impair macrophage microbicidal func-
tion and the IL-I0 contributes to increased release of
TNF receptor-2, which blocks the activating role of
TNF-a (Balcewicz-Sablinska et aI.1998).
9.2.11
Bacterial Genetics and Mechanisms of Avoidance
of Destruction by the Type 1 Response
There is intense interest in the bacterial factors
that allow M. tuberculosis to resist destruction by
the immune system in mice after the development
of the full Type 1 response at about 21 days. The
recent publication of the entire sequence of the M.
tuberculosis genome has allowed a variety of genetic
techniques to be applied. Mutations that inactivate
mycolic acid cyclopropane synthase (Glickman et
al. 2000) or isocitrate lyase, which is required for
the metabolism of fatty acids (McKinney et al.
2000), both result in decreased ability to persist.
Others have expressed candidate virulence genes,
G.A. W. Rook
or plasmid libraries of M. tuberculosis DNA in the
avirulent organism M. smegmatis and then looked
for increased ability to survive inside macrophages
(Miller and Shinnick 2000; Wei et al. 2000). A review
of the comparative genomics of mycobacteria gives
an overview of the techniques used and the results
obtained (Brosch et al. 2000).
9.3
Immunopathology
9.3.1
The Koch Phenomenon; the Response
Characteristic of Disease
As outlined in the introduction, Koch noted that
4-6 weeks after establishment of infection in guinea-
pigs, intradermal challenge with whole organisms or
culture filtrate resulted in necrosis locally and in the
original tuberculous lesion (Koch 1891). Similar
phenomena occur in humans. The tuberculin test is
frequently necrotic in subjects who are, or have been,
tuberculous, whereas necrosis does not occur when
positive skin-tests to tuberculin are elicited in normal
BCG recipients or in tuberculoid leprosy patients.
Koch sought to exploit this phenomenon for the
treatment of tuberculosis and found that injection of
larger quantities of culture filtrate (Old Tuberculin)
subcutaneously into tuberculosis patients would
evoke necrosis in established tuberculous lesions
at distant sites (Anderson 1891). This resulted in
necrosis, sloughing and "cure" of the lesions of
skin tuberculosis (Lupus vulgaris, usually caused
by bovine strains), but when similar necrosis was
evoked in deep lesions in the spine or lungs, the
results were disastrous and merely provided further
necrotic tissue in which the bacteria could proliferate
(Fig. 9.5). This treatment was therefore abandoned,
but it provided crucial clues as to the nature of the
immunopathology of tuberculosis.
9.3.2
Tuberculin Response and Protection
The error in Koch's thinking was highlighted in the
1940s. When guinea-pigs were pre-immunised so that
they had powerful Koch phenomena in response to
small doses of tuberculin, they became more suscep-
tible to tuberculosis than were unimmunised control
animals (Wilson et al.1940). Obviously this was seen
The Immunology and Pathogenesis of Tuberculosis 143

Fever, rigors. necrosis and
sloug/ling 01 the skln lesion,
and dangerous necrosis of
other lesions In lUngs or spine.
Fig. 9.5. Koch's failed immunotherapeutic treatment for
tuberculosis, similar to the phenomenon he had previ-
ously described in guinea-pigs. It achieved necrosis and
sloughing of the chronic lesions of cutaneous tubercu-

24-48hrs
.~
~ losis (Lupus vulgaris), but the simultaneous necrosis
and liquefaction of cryptic lesions in deep tissues was
disastrous and the treatment was abandoned

only if the challenge infection was into the lungs or by 9.3.4

deep intramuscular injection, so that necrosis could
not result in shedding of the infected tissue. Opti-
mal protection was seen in animals with moderate
positive responses. This experiment is illustrated in
Fig. 9.6 (Wilson et al.I940),and makes the point, also
apparent from human epidemiology (Fine 1993), that
the most protective response correlates with a small
positive induration in the Mantoux test (delayed type
hypersensitivity or DTH) rather than with the mas-
sive necrotic Koch phenomenon.
9.3.3
The Relationship Between the Koch Phenomenon
and the Shwartzman Reaction
Koch's discovery that soluble bacterial material could
trigger necrosis in a distant tuberculous site (in addi-
tion to the site of injection) has some parallels with
the Shwartzman reaction (Shwartzman 1937), and
subsequent experiments strengthen the parallel still
further. For instance injections of endotoxin-rich
material into a distant site (instead of the tuberculin
used by Koch) will also trigger necrosis in tuberculous
lesions (Bordet 1931; Debonera et al.1932; Shands and
Senterfitt 1972). These observations are compatible
with the view that tuberculous lesions are susceptible
to superimposed cytokine-mediated damage.
A Hypothesis to Explain the Koch Phenomenon
There is increasing evidence for increased susceptibility
to cytokine-mediated tissue necrosis of tissues under-
going inflammation mediated simultaneously by Type
1 and Type 2 cytokines. This is easily demonstrated in
Balblc mice with pulmonary tuberculosis (Hernandez-
Pando et al. 1995). During the first 3 weeks, DTH sites
were not sensitive to local injection of as much as 1 ~g
of recombinant TNF-a. This is the period of Type 1
response (Orme et al.1993). After 50 days, the animals
enter a phase of slowly progressive disease accompa-
nied by increasing Th2 cytokine production, seen as
IL-4 positive cells in the lesions. In these animals DTH
sites become TNF-a-sensitive (Hernandez-Pando et al.
1997b; Hernandez-Pando and Rook 1994). Recently we
have found that this state of TNF-a-sensitivity of the
DTH sites does not occur in tuberculous IL-4 KO mice,
but is restored if the IL-4 KO mice receive an intrave-
nous injection of a persistent form of IL-4 (Hernan-
dez-Pando et al., in preparation). This propensity for
TNF-a toxicity in the presence of IL-4 is supported
by several examples in non-mycobacterial models.
Lawrence and co-workers, studying Trichinella spira-
lis infection in mice, have found that the enteropathy
caused by TNF-a is dependent upon IL-4 (Lawrence et
al. 1998). Other murine studies have also revealed that,
in certain mouse strains, DTH caused by Th2 cells cor-

Result of inlramuscular
infecllon with 18
, . . . . . - - - - - - - . . . , MORE susceptible
~ ... 'Koch" than unimmunised
~ ooMro~
I
Immunise with killed r- Fig.9.6. Large numbers of outbred guinea-pigs were
M. tuberculosis by Rank immunised with killed M. tuberculosis by a variety of

W
various protocols \ according to ~ Positive
tuberculin ~
R' t t
es,s an protocols an d then tested with a range of dilutions of
• ~ response tuberculin, before they were challenged by the intramus-
cular route with a very small does of M. tuberculosis. The
most protected animals were those with small positive
.. Negative Susceptible
tuberculin responses, whereas animals with exquisite
Unimmunised
controls ~+----_Negative Susceptible
tuberculin sensitivity (Koch phenomenon) were more
susceptible to the infection that were unimmunised
controls (Wilson et al. 1940)
144
relates with TNF-a production by the cells (Muller et al.
1994). Finally, if an important component of the Koch
phenomenon is cytokine-mediated damage in a site of
mixed Type l/Type 2 inflammation, then the toxicity of
Koch's "treatment" for tuberculosis is easily explained
(Figs. 9.5, 9.2) (Anderson 1891).
9.3.5
Detrimental Roles ofTNF-a in Human Tuberculosis
The previous section suggests that a component of the
immunopathological state in human tuberculosis may
result from the simultaneous presence of Type 1 and
Type 2 cytokines and TNF-a. An increase in plasma
TNF-a levels has been associated with clinical deterio-
ration in patients with severe tuberculosis (Bekker et al.
1998). Both thalidomide and pentoxifylline, as inhibi-
tors of TNF-a release, have been tried as treatments
for the cachexia of chronic disease (Haslett 1998). Tha-
lidomide, which reduces IL-6, IL-lO as well as TNF-a
levels, and reduced lung pathology in a murine model
(Moreira et al. 1993), was clearly clinically beneficial in
the human disease (Tramontana et al. 1995).
9.3.6
Fibrosis
The fibrosis is another aspect of the immunopathol-
ogy of human tuberculosis that requires explana-
tion. The dominant cytokine in tuberculosis, IFN-y,
opposes fibrosis, so why does fibrosis occur? In fact,
in all human diseases characterised by marked pulmo-
nary fibrosis there is expression of Type 2 cytokines
(systemic sclerosis, idiopathic pulmonary fibrosis,
radiation-induced pulmonary fibrosis, chronic lung
allograft rejection [reviewed in Lee et al. 2001]) and
the same is true in the mouse in the bleomycin model
(Lee et al. 2001) and in schistosomiasis, where fibro-
sis and tissue remodelling do not occur if induction
of Type 2 cytokines by the ova is blocked by pre-
immunisation with ova+IL-12. These observations
have led to the "Type 2 cytokine hypothesis offibrosis"
(Lee et al. 2001). Type 2 cytokines such as IL-4, IL-482
and IL-13 activate fibroblasts and promote collagen
formation. IL-13 also drives formation of TGF-~
and, by increasing release of the enzymes that activate
TGF-~ (such as MMP-9) while down-regulating latent
TGF-~-binding protein (LTBP), it also activates the
TGF-~ (Lee et al. 2001). Significantly, the increase in
expression of Type 2 cytokines in human TB extends
to IL-13 which, like IL-4 and IL-482, is increased 80-100
G.A. W. Rook
fold compared to controls matched for age, race and
gender (Seah et al. 2000).
9.4
Endocrinology
There are several endocrine and metabolic changes in
tuberculosis that may contribute to the failure of the
Type 1 immune response to control the infection, and
to the increasing level of Type 2 cytokine expression.
9.4.1
Vitamin D3 Metabolism in Tuberculosis Lesions
The macrophages of tuberculosis patients, following
activation by IFN-y, express an active la-hydroxylase,
and rapidly convert 25(OH)-vitamin-D3 to calcitriol
(Rook 1988; Rook et al. 1986). This is a potent phe-
nomenon, leading occasionally to leakage of calcitriol
into the periphery and to hypercalcaemia, though it
has in the past been difficult to understand its role in
the disease (Rook 1988). It now seems possible that this
is a feed-back mechanism that tends to down-regulate
Th1 and enhance Th2 responses, because calcitriol
inhibits production of IFN-y and IL-2, and increases
production ofIL-4 and IL-5 (Daynes et al. 1991; Rigby
et al. 1987). This may well be related to the ability of
calcitriol to inhibit release of IL-12 (Lemire 1995).
In the 1940s attempts were made to treat tubercu-
losis with vitamin D. When patients with skin tuber-
culosis (Lupus vulgaris, often due to M. bovis) were
treated with this vitamin, the chronic non-healing
granulomatous lesions often underwent necrosis fol-
lowed by resolution (Macrae 1947). However necrosis
and liquefaction also occurred in deep lesions in the
spine and lungs (Brincourt 1967), so the result was
as disastrous as the use of Koch's immunotherapy
described above (Anderson 1891). The mechanism of
this effect remains unknown, but an increase in Type
2 cytokine expression in granulomata rich in Type
1 cytokines and TNF-a would be expected to cause
necrotising immunopathology as discussed earlier.
The Immunology and Pathogenesis of Tuberculosis
9.4.2
Adrenal Steroids in Tuberculosis
9.4.2.1
The Effects of Stress
Glucocorticoids cause a switch to Type 2 cytokine
production (Brinkmann and Kristofic 1995; Ramirez
et al. 1996), probably because of effects on dendritic
cells, which secrete less IL-12 and more IL-1O in their
presence (Vieira et al. 1998; Visser et al.1998), but glu-
cocorticoids also down-regulate the antimycobacterial
effects of macrophages (Rook et al. 1987). It is there-
fore not surprising that reactivation or progression of
infection with tuberculosis is sensitive to glucocorti-
coid therapy and to activation of the hypothalamo-
pituitary-adrenal axis. Exposing humans to the stress
of war or poverty (Spence et al. 1993), or cattle to the
stress of transportation are enough to cause reactiva-
tion of disease. The disease-promoting effect of stress
(Brown et al.1993; Tobach and Bloch 1956) and exces-
sive activation of the hypothalamo-pituitary-adrenal
axis (Hernandez-Pando et al.1998a,b) have been dem-
onstrated under more controlled conditions in mice.
9.4.2.2
The Hypothalamo-Pituitary-Adrenal Axis
in Human Tuberculosis
In tuberculosis patients who are rested and acclima-
tised to hospital, the cortisol diurnal rhythm is normal
and so are the responses of the adrenals to CRH and to
very low doses (i.e. physiological) of ACTH (Rook et al.
2000). The total 24-hour cortisol output may be normal
or modestly raised. However, there is a change in the
pattern of metabolism of cortisol indicating a large
alteration in the equilibrium point of the cortisol-cor-
tisone shuttle, as discussed below (Rook et al' 1996).
9.4.2.3
Dysregulation of the Cortisol-Cortisone Shuttle
A major mechanism for the regulation of local tissue
cortisol levels is the inter-conversion of active cortisol
(II-hydroxy) and inactive cortisone (II-keto). Thus,
effective cortisol concentrations in different organs
can be very different from the values found in the
serum. Gas chromatography and mass spectrometry
revealed a striking excess of metabolites of cortisol
relative to metabolites of cortisone in 24-hour urine
collections from tuberculosis patients (Rook et al.
1996). Subsequent analysis of alveolar lavage samples
from tuberculosis patients and controls has revealed
145
that the site of abnormal conversion of inactive cor-
tisone to active cortisol in the patients is the infected
lung itself (Baker et al. 2000; Rook et al. 2000). This may
be explained by the observation that TNF-a and IL-
1~ both increased the expression levels and reductase
activity of ll~-HSD-l in a cell line in vitro (Escher et
al. 1997). The result is a local increase in cortisol levels
that is not apparent from measurements of serum
cortisol. This cortisol excess will cause a shift towards
Type 2 cytokine expression, deactivation of the anti-
mycobacterial effects of macrophages, increased IL-
10 and increased TGF-~, so it may contribute to the
changes seen in the human disease.
9.5
Vaccination
The protective efficacy of BCG vaccination varies from
80% protection to no protection at all in different popu-
lations (reviewed in Fine 1993). The need to understand
this variability before undertaking long and expensive
trials of novel vaccines in man is obvious. Several
hypotheses are currently under investigation.
9.5.1
Interference by Exposure to Environmental
Mycobacteria
This hypothesis assumes that the common antigens,
shared by all mycobacterial species, are relevant to
protection. We have known for many years that the
Bacillus Calmette-Guerin (BCG) is as effective a vaccine
against leprosy as it is against tuberculosis, although M.
leprae is an entirely different species (Fine et aI. 1986).
BeG must therefore be able to work through common
epitopes. Similarly, there is evidence that contact with
an environmental organism leading to mycobacterial
skin-test positivity is protecting the population of
Malawi from both tuberculosis and leprosy (Fine et
aI. 1994). In mice powerful protective or "anti-protec-
tive" effects can be induced with an environmental
mycobacterial saprophyte, simply by altering the
immunisation protocol so as to induce Type 1or Type 2
response (Hernandez-Pando et al. 1997b). These effects
are obviously due to the common antigens. It is also
significant that tuberculosis patients who still main-
tain necrotising skin-test positivity to antigens of M.
tuberculosis itself, have diminished or absent skin-test
responses to environmental saprophytes (Kardjito et al.
1986), whereas these do evoke responses in protected
146
populations. The heat shock proteins (hsp), which are
very highly conserved across species, are important
examples of this concept. Indeed, the 65 kDa heat
shock protein of M. leprae can protect mice against
M. tuberculosis (Silva and Lowrie 1994), as can DNA
vaccines based on its sequence. It is, therefore, logical
to suggest that an apparent reduction in the efficacy
of BCG could occur either because the environmental
saprophytes were themselves protecting, or because
they were priming deleterious patterns of response
(e.g. Type 2) to the common antigens and so blocking
the efficacy of the BCG. These two effects were sug-
gested to be occurring in different countries (Stanford
et al. 1981) and both effects are easily demonstrated
experimentally (Hernandez-Pando et al. 1997b).
9.5.2
Concurrent Parasite Infections
Some authors suggest that the presence of concomitant
parasite infections may cause a systemic bias towards
Th2 responses that undermines the ability of BCG to
induce a Thl response to mycobacterial antigens.
9.5.3
Vaccine Dose
Another suggestion is that BCG would be more reli-
able if used at a very low dose, because the dose at
which the vaccine starts to evoke a deleterious Type
2 component may be much lower in some individuals
than in others. This idea is derived from vaccination
studies with Leishmania in different mouse strains,
where a dose that evokes a Type 1 response in some
strains is too high, and so Type 2-inducing, in others
(Bretscher et al. 1992). Because BCG is a live vaccine
it should still be effective at very low doses. In each
individual it should proliferate to the level at which
a mycobactericidal Thl response is induced (Power
et al. 1998). In deer 5XI04 or 5xl07 is protective but
5x108 is less effective, so there might be some truth
in this idea (Griffin et al. 1999).
9.5.4
A Possible Novel Approach to Vaccine Design
Enormous effort has gone into testing purified or
recombinant antigens and seeking those that evoke the
most potent Type 1 response, with maximal output of
IFN-yfrom human or murine cells. These antigens are
G.A.W.Rook
then used as recombinant vaccinia or salmonella or as
the purified protein plus adjuvant, or as DNA vaccines.
The problem is that almost all such vaccines show
some efficacy in mouse models, but rarely work as
well as BCG, a vaccine already known to fail in man.
This approach might be flawed. Virtually all
patients develop a strong Type 1 response, so that
a vaccine that merely accelerates the establishment
of this Type 1 response provides marginal benefit. It
may be that the solution is to identify the antigens
and epitopes of M. tuberculosis shown to drive the
"subversive"Type 2 component (Seah and Rook 2001),
by screening for IL-4 induction, and then incorporate
these in a vaccine with a potent Type 1 adjuvant, such
as IL-12, or with an adjuvant capable of activating
regulatory T cells that suppress Type 2 responses
(Zuany-Amorim et al. 2002). The aim would be to sup-
press the Type 2 response to these antigens, or to force
the establishment of a Type 1 response to them, even
if they prove to be common mycobacterial antigens to
which there is a pre-existing Type 2 response evoked
by environmental species. Then if the individual were
subsequently exposed to M. tuberculosis, the organ-
ism would be unable to activate the subversive IL-4-
secreting component. This novel approach is under
investigation and should, moreover, provide a vaccine
suitable for use as an immunotherapeutic.
9.5.6
Immunotherapy
This is perhaps the most important issue facing
workers in the field of tuberculosis. As outlined in the
introduction, DOTS helps but fails to solve the prob-
lem, and multi-drug-resistant disease is an increasing
threat. Immunotherapy is the only solution, and the
need for this approach has been recognised since the
time of Robert Koch (Stanford et al. 1993).
One potential approach to immunotherapy is the
direct use of cytokines in patients, either systemically
or given by aerosol. IL-2, IFN-y, IL-12 and GM-CSF
have all been investigated (reviewed in Holland
2000). Their potential roles in therapy, other than as a
potential adjunct to drug treatment in multi-resistant
cases, have yet to be elucidated.
Another approach that gives striking results in
mice is the use of dehydroepiandrosterone or of
the very similar androstenediol. These compounds
oppose a subset of the effects of glucocorticoids, and
can reverse the switch towards Type 2 cytokine pro-
file in Balblc mice, but this has not been tested in man
(Hernandez-Pando et al.I998ab).
The Immunology and Pathogenesis of Tuberculosis 147

Immunotherapy is unlikely to be achieved by References

simple antigen preparations, because there needs
to be an immunoregulatory component with Abou-Zeid C et al. (1997) Induction of a Type 1 immune
down-regulation of pre-existing Type 2 cytokine responses to a recombinant antigen from Mycobacterium
production. This may explain why antigen prepa- tuberculosis expressed in Mycobacterium vaccae. Infect
Immun 65:1856-1862
rations that can protect against tuberculosis when Adams LB et al. (1995) Exacerbation of acute and chronic murine
used as vaccines before challenge can fail to act tuberculosis by administration of a tumour necrosis factor
as therapeutics in mouse models (Turner et al. receptor-expressing adenovirus. J Infect Dis 171:400-405
2000). Interestingly, the only preparations shown Adams LB et al. (1997) Comparison of the roles of reactive
oxygen and nitrogen intermediates in the host response
to be effective immunotherapeutics in tuberculous
to Mycobacterium tuberculosis using transgenic mice.
mice are killed Mycobacterium vaccae (Hernandez- Tubercle Lung Dis 78:237-246
Pando et al. 2000b; Rook and Hernandez-Pando Altare F et al. (1998) Impairment of mycobacterial immunity
1996; Zhang et al. 2000) and more recently DNA in human interleukin-12 receptor deficiency. Science 280:
vaccines encoding common mycobacterial antigens 1432-1435
(Lowrie et al. 1999). Attempts at immunotherapy Anderson MC (1891) On Koch's treatment. Lancet i:651-652
Arinobu Y et al. (1999) Antagonistic effects of an alternative
using the environmental saprophyte, M. vaccae, splice variant of human IL-4, IL-4 delta 2, on IL-4 activi-
followed studies of the influence of such organisms ties in human monocytes and B cells. Cell Immunol 191:
on disease susceptibility and on the efficacy of BCG 161-167
vaccination (Stanford et al. 1981), a concept that Arkwright PD, David TJ (2001) Intradermal administration
of a killed Mycobacterium vaccae suspension (SRL 172)
is now widely accepted (Fine 1995). Two studies
is associated with improvement in atopic dermatitis in
of single injections of heat-killed M. vaccae have children with moderate-to-severe disease. J Allergy Clin
been carried out to Good Clinical Practice (GCP) in Immunol107:531-534
human tuberculosis patients also receiving DOTS. Astarie-Dequeker C et al. (1999) The mannose receptor medi-
The results are conflicting (Group 1999; Johnson ates uptake of pathogenic and nonpathogenic mycobacte-
ria and bypasses bactericidal responses in human macro-
et al. 2000), though the striking success of recent
phages. Infect Immun 67:469-477
studies in retreatment cases (National Cooperation Aston C et al. (1998) Early inhibition of mycobacterial growth
Group on Clinical Study of Mycobacterium vaccae by human alveolar macrophages is not due to nitric oxide.
and Luo 2001) and multi-drug-resistant (MDR) TB Am J Respir Crit Care Med 157:1943-1950
(Luo et al. 2000) in China suggest that further trials Atamas SP et al. (1999) Production of Type 2 cytokines by
CD8+ lung cells is associated with greater decline in
of multiple doses should be undertaken. In view of
pulmonary function in patients with systemic sclerosis.
the inevitable difficulty of demonstrating an effect Arthritis Rheum 42: 1168-1179
when a single dose is superimposed upon DOTS Baker RW et al. (2000) Increased cortisol-cortisone ratio in
under clinical trial conditions in patients with acute pulmonary tuberculosis. Am J Respir Crit Care Med
drug-sensitive infections, there is a clear need for 162:1641-1647
Balasubramanian R et al. (1990) Five year results of a 3 month
a trial of multiple doses in MDR disease, as used in
and two 5 month regimens for the treatment of sputum-
the pilot non-GCP studies in MDR TB (Farid et al. positive pulmonary tuberculosis in South India. Tubercle
1994). M. vaccae has appropriate properties because, 71:253-258
although it is used killed, it not only evokes Type Balcewicz-Sablinska MK et al. (1998) Pathogenic Mycobacte-
1 responses (Abou-Zeid et al. 1997) and cytotoxic rium tuberculosis evades apoptosis of host macrophages by
release of TNF-R2, resulting in inactivation of TNF-alpha. J
CD8+ T cells that kill target cells infected with M.
ImmunoI161:2636-2641
tuberculosis (Skinner et al. 1997), but also acts as a Beckman EM et al. (1994) Recognition of a lipid antigen by
potent non-specific down-regulator of pre-existing CD I-restricted alpha beta+ T cells. Nature 372:691-694
Th2 responses both in mice (Wang and Rook 1998) Behr Perst SI et al. (1999) Phenotypically activated gammad-
and in clinical trials in human patients with asthma elta T lymphocytes in the peripheral blood of patients with
tuberculosis. J Infect Dis 180: 141-149
or eczema (Arkwright and David 2001; Camporota
Bekker L et al. (1998) Selective increase in plasma tumour
et al. 2000). Interestingly, it is now clear that the necrosis factor-a and concomitant clinical deterioration
mode of action of M. vaccae is mainly the induction after initiating therapy in patients with severe tuberculosis.
of regulatory T cells that suppress Type 2 responses J Infect Dis 178:580-584
(Zuany-Amorim et al. 2002). It can also induce Thl Bellamy R (2000) Identifying genetic susceptibility factors
for tuberculosis in africans: a combined approach using a
recognition of common mycobacterial antigens in
candidate gene study and a genome-wide screen. Clin Sci
HIV+ individuals, which may prove to be an impor- (Colch) 98:245-250
tant requirement in developing countries (Waddell Bergeron A et al. (1997) Cytokine patterns in tuberculous and
et al. 2000). sarcoid granulomas. J ImmunoI159:3034-3043
148
Bordet P (1931) Contribution a l'etude de l'allergie. CR Soc
Bioi 107:622-623
Bretscher PA et al. (1992) Establishment of stable, cell-medi-
ated immunity that makes "susceptible" mice resistant to
Leishmania major. Science 257:539-542
Brightbill HD et al. (1999) Host defense mechanisms triggered
by microbial lipoproteins through toll-like receptors. Sci-
ence 285:732-736
Brincourt J (1967) Le calciferol a-toil une action liquefiante sur
Ie caseum? Poumon Coeur 23:841-851
Brinkmann V, Kristofic C (1995) Regulation by corticosteroids
of Th1 and Th2 cytokine production in human CD4+ effec-
tor T cells generated from CD45RO' and CD45RO+ subsets.
J ImmunoI155:3322-3328
Brosch R et al. (2000) Comparative genomics of the mycobac-
teria. Int J Med MicrobioI290:143-152
Brown DH et al. (1993) Regulation of mycobacterial growth
by the hypothalamus-pituitary-adrenal axis: differential
responses of Mycobacterium bovis BCG-resistant and -
susceptible mice. Infect Immun 61:4793-4800
Butler D (2000) Consortium aims to kick-start TB research.
Nature 403:692
Camporota L et al. (2000) Effects of intradermal Mycobac-
terium vaccae on allergen-induced airway responses and
IL-5 generation by PBMC in mild to moderate asthma. Am
J Respir Crit Care Med 161:477
Chan J et al. (1991) Lipoarabinomannan, a possible virulence
factor involved in persistence of Mycobacterium tuberculo-
sis within macrophages. Infect Immun 59:1755-1761
Chan J et al. (1992) Killing of virulent Mycobacterium tuber-
culosis by reactive nitrogen intermediates produced by
activated murine macrophages. J Exp Med 175:1111-1122
Chan J et al. (1995) Effects of nitric oxide synthase inhibitors
on murine infection with Mycobacterium tuberculosis.
Infect Immun 63:736-740
Cho S et al. (2000) Antimicrobial activity of MHC class 1-
restricted CD8+ T cells in human tuberculosis. Proc Natl
Acad Sci USA 97:12210-12215
Clerici M, Shearer GM (1994) The Th1-Th2 hypothesis of HIV
infection: new insights. Immuno1 Today 15:575-581
Das G et al. (1999) Apoptosis of Th I-like cells in experimental
tuberculosis (TB). Clin Exp ImmunoI115:324-328
Daynes RA, Meikle AW, Araneo BA (1991) Locally active
steroid hormones may facilitate compartmentalization of
immunity by regulating the types of lymphokines pro-
duced by helper T cells. Res Immunol 142:40-45
De Jong R et al. (1998) Severe mycobacterial and Salmonella
infections in interleukin-12 receptor-deficient patients. Sci-
ence 280: 1435-1438
Debonera G, Tzortakis N, Falchetti E (1932) Inflammation et
phenomene de Shwartzman. CR Soc Bioi 109:24-26
Dieli F et al. (2000a) Ligand-specific alphabeta and gammad-
elta T cell responses in childhood tuberculosis. J Infect Dis
181:294-301
Dieli F et al. (2000b) Vgamma9/Vdelta2 T lymphocytes reduce
the viability of intracellular Mycobacterium tuberculosis.
Eur J ImmunoI30:1512-1519
Ellner JJ (1997) Regulation of the human immune response
during tuberculosis. J Lab Clin Med 130:469-475
Escher G et al. (1997) Tumour necrosis factor a and interleukin
Ib enhance the cortisone/cortisol shuttle. J Exp Med 186:
189-198
Espinal MA et al. (2001) Global trends in resistance to antitu-
G.A. W. Rook
berculosis drugs. World Health Organization-International
Union against Tuberculosis and Lung Disease Working
Group on Anti-Tuberculosis Drug Resistance Surveillance.
N Engl J Med 344:1294-1303
Farid R et al. (1994) Mycobacterium vaccae immunotherapy
in the treatment of multi-drug-resistant tuberculosis: a
preliminary report. Iran J Med Sci 19:37-39
Fayyazi A et al. (2000) Apoptosis of macrophages and T cells
in tuberculosis associated caseous necrosis. J Pathol 191:
417-425
Filley EA, Rook GAW (1991) Effect of mycobacteria on sen-
sitivity to the cytotoxic effects of tumor necrosis factor.
Infect Immun 59:2567-2572
Fine PE (1993) Immunities in and to tuberculosis: implications
for pathogenesis and vaccination. In: McAdam KPWJ, Porter
JDH (eds) Tuberculosis: back to the future. Proceedings of
the London School of Hygiene and Tropical Medicine 3rd
Annual Public Health Forum. Wiley, Chichester, pp 54-78
Fine PE (1995) Variation in protection by BCG: implications of
and for heterologous immunity. Lancet 346:1339-1345
Fine PE et al. (1986) The protective efficacy of BCG against
leprosy in Northern Malawi. Lancet ii:499-502
Fine PE et al. (1994) Delayed type hypersensitivity, myco-
bacterial vaccines and protective immunity. Lancet 344:
1245-1249
Flynn JL et al. (1993) An essential role for interferon gamma
in resistance to Mycobacterium tuberculosis infection. J Exp
Med 178:2249-2254
Flynn JL et al. (1995) Tumor necrosis factor-alpha is required
in the protective immune response against Mycobacterium
tuberculosis in mice. Immunity 2:561-572
Fratazzi C et al. (1999) Macrophage apoptosis in mycobacterial
infections. J Leukoc Bioi 66:763-764
Fulton SA et al. (1998) Regulation of interleukin-12 by inter-
leukin-lO, transforming growth factor-beta, tumor necrosis
factor-alpha, and interferon-gamma in human monocytes
infected with Mycobacterium tuberculosis H37Ra. J Infect
Dis 178:1105-1114
Gerosa F et al. (1999) CD4(+) T cell clones producing both
interferon-gamma and interieukin-lO predominate in
bronchoalveolar lavages of active pulmonary tuberculosis
patients. Clin Immunol 92:224-234
Glickman MS, Cox JS, Jacobs WR Jr (2000) A novel mycolic
acid cyclopropane synthetase is required for coding, per-
sistence, and virulence of Mycobacterium tuberculosis.
Mol Cell 5:717-727
Grange JM (1992) The mystery of the mycobacterial persister.
Tubercle Lung Dis 73:249-251
Griffin JF et al. (1999) Vaccine protocols to optimise the pro-
. tective efficacy of BCG. Tubercle Lung Dis 79:135-143
Group DIT (1999) ImmunotherapywithMycobacterium vaccae
in patients with newly diagnosed pulmonary tuberculosis:
a randomised controlled trial. Lancet 354:116-119
Haslett PA (1998) Anticytokine approaches to the treatment of
anorexia and cachexia. Semin OncoI25:53-57
Hernandez-Pando R, Rook GAW (1994) The role of TNF-a in
T cell-mediated inflammation depends on the Thl/Th2
cytokine balance. Immunology 82:591-595
Hernandez-Pando R et al. (1995) Adrenal changes in murine
pulmonary tuberculosis; a clue to pathogenesis? FEMS
Immunol Med MicrobioI12:63-72
Hernandez-Pando R et al. (1996) Correlation between the
kinetics of Th1lTh2 cells and pathology in a murine model
The Immunology and Pathogenesis of Tuberculosis 149

of experimental pulmonary tuberculosis. Immunology 89: Krensky AM (2000) Granulysin: a novel antimicrobial peptide
26-33 of cytolytic T lymphocytes and natural killer cells. Bio-
Hernandez-Pando R et al. (1997a) Analysis of the local kinetics chern PharmacoI59:317-320
and localisation of interleukin 1a, tumour necrosis factor a Kusner DJ, Adams J (2000) ATP-induced killing of virulent
and transforming growth factor b during the course ofexper- Mycobacterium tuberculosis within human macrophages
imental pulmonary tuberculosis. Immunology 90:507-516 requires phospholipase D. J ImmunoI164:379-388
Hernandez-Pando R et al. (1997b) Pathogenesis of tuberculosis Ladel CH, Daugelat S, Kaufmann SH (1995) Immune response
in mice exposed to low and high doses of an environmental to Mycobacterium bovis bacille Calmette Guerin infection
mycobacterial saprophyte. Infect Immun 65:3317-3327 in major histocompatibility complex class 1- and II- defi-
Hernandez-Pando R et al. (1998a) Emergent therapeutic cient knock-out mice: contribution of CD4 and CD8 T cells
properties of a combination of glucocorticoid and anti- to acquired resistance. Eur J ImmunoI25:377-384
glucocorticoid steroids in tuberculous Balb/c mice. Quart Lammas DA et al. (1997) ATP-induced killing of mycobac-
J Med 91:755-766 teria by human macrophages is mediated by purinergic
Hernandez-Pando R et al. (1998b) The effects of androstene- P2Z(P2X7) receptors. Immunity 7:433-444
diol and dehydroepiandrosterone on the course and cyto- Lauchumroonvorapong P et al. (1996) H202 induces mono-
kine profile of tuberculosis in Balb/c mice. Immunology cyte apoptosis and reduces viability of Mycobacterium
95:234-241 avium-M. intracellulare within cultured human mono-
Hernandez-Pando R et al. (2000a) Persistence of DNA from cytes. Infect Immun 64:452-459
M. tuberculosis in superficially normal lung tissue during Lawrence CE et a1. (1998) IL-4-regulated enteropathy in
latent infection. Lancet 356:2133-2138 an intestinal nematode infection. Eur J Immunol 28:
Hernandez-Pando R et al. (2000b) Interactions between hor- 2672-2684
mone-mediated and vaccine-mediated immunotherapy for Lee CG et al. (2001) Interleukin-13 induces tissue fibrosis by
pulmonary tuberculosis in Balblc mice. Immunology 100: selectively stimulating and activating transforming growth
391-398 factor beta(1). J Exp Med 194:809-821
Hilkens CM et al. (1995) Differential modulation of T helper Lemire JM (1995) Immunomodulatory actions of 1,25-dihy-
type 1 (ThI) and T helper type 2 (Th2) cytokine secretion droxyvitamin D3. J Steroid Biochem Mol Bioi 53:599-602
by prostaglandin E2 critically depends on interleukin-2. Lindblad EB et al. (1997) Adjuvant modulation of immune
Eur J Immunol 25:59-63 responses to tuberculosis subunit vaccines. Infect Immun
Hirsch CS et al. (1994) Complement receptor mediated uptake 65:623-629
and Tumour Necrosis Factor a-mediated growth inhibition Lowrie DB et a1. (1999) Therapy of tuberculosis in mice by
of Mycobacterium tuberculosis by human alveolar macro- DNA vaccination. Nature 400:269-271
phages. J ImmunoI152:743-753 Luo Y, Lu S, Guo S (2000) Immunotherapeutic effect of Myco-
Hirsch CS et al. (1999) Apoptosis and T cell hyporesponsive- bacterium vaccae on multi-drug resistant pulmonary
ness in pulmonary tuberculosis. J Infect Dis 179:945-953 tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 23:85-88
Holland SM (2000) Cytokine therapy of mycobacterial infec- Macrae DE (1947) Calciferol treatment of Lupus vulgaris. Br
tions. Adv Intern Med 45:431-452 Med J 59:333-338
Ilangumaran S et a1. (1995) Integration of mycobacterial Maiti D, Bhattacharyya A, Basu J (2000) Lipoarabinomannan
lipoarabinomannans into glycosylphosphatidylinositol- from Mycobacterium tuberculosis promotes macrophage
rich domains of Iymphomonocytic cell plasma mem- survival by phosphorylating Bad through a phosphati-
branes. J ImmunoI155:1334-1342 dylinositol 3-kinaselAkt pathway. J Bioi Chern
Johnson JL et al. (2000) Randomised controlled trial of Myco- Malik ZA, Denning GM, Kusner DJ (2000) Inhibition of Ca(2+)
bacterium vaccae immunotherapy in non human immuno- signaling by Mycobacterium tuberculosis is associated
deficiency virus infected ugandan adults with newly diag- with reduced phagosome-lysosome fusion and increased
nosed pulmonary tuberculosis. J Infect Dis 181:1304-1312 survival within human macrophages. J Exp Med 191:
Jouanguy E et al. (I996) Interferon-g receptor deficiency in an 287-302
infant with fatal Bacille Calmette-Guerin infection. N Engl Maloy KJ, Powrie F (2001) Regulatory T cells in the control of
J Med 335:1956-1961 immune pathology. Nat ImmunoI2:816-822
Kang BK, Schlesinger LS (1998) Characterization of mannose Marchant A et al. (2001) Polarization of PPD-specific T-cell
receptor-dependent phagocytosis mediated by Mycobacte- response of patients with tuberculosis from ThO to Thl
rium tuberculosis lipoarabinomannan. Infect Immun 66: profile after successful antimycobacterial therapy or in
2769-2777 vitro conditioning with interferon-alpha or interleukin-12.
Kardjito T et al. (1986) A comparison of the responsiveness Am J Respir Cell Mol BioI 24:187-194
to four new tuberculins among Indonesian patients with McKinney JD et al. (2000) Persistence of Mycobacterium tuber-
pulmonary tuberculosis and healthy subjects. Eur J Respir culosis in macrophages and mice requires the glyoxylate
Dis 69:142-145 shunt enzyme isocitrate lyase. Nature 406:735-738
Keane J et a1. (2001) Tuberculosis associated with infliximab, Means TK et al. (1999) Human toll-like receptors mediate cel-
a tumor necrosis factor alpha-neutralizing agent. N Engl J lular activation by Mycobacterium tuberculosis. J Immunol
Med 345:1098-1104 163:3920-3927
Klingler K et al. (1997) Effects of mycobacteria on regulation Miller BH, Shinnick TM (2000) Evaluation of Mycobacterium
of apoptosis in mononuclear phagocytes. Infect Immun tuberculosis genes involved in resistance to killing by
65:5272-5278 human macrophages. Infect Immun 68:387-390
Koch R (1891) Fortsetzung tiber ein Heilmittel gegen Tuber- Molloy AP, Laochumroonvorapong P, Kaplan G (1994) Apop-
culose. Dtsch Med Wochenschr 17:101-102 tosis but not necrosis of infected monocytes is coupled
150
with killing of intracellular Bacille Calmette-Guerin. I Exp
Med 180:1499
Moody DB et al. (1997) Structural requirements for glycolipid
antigen recognition by CD1b-restricted T cells. Science
278:283-286
Moody DB et al. (2000) COle-mediated T-cell recognition
of isoprenoid glycolipids in Mycobacterium tuberculosis
infection. Nature 404:884-888
Moreira AL et al. (1993) Thalidomide exerts its inhibitory
action on tumor necrosis factor alpha by enhancing mRNA
degradation. I Exp Med 177:1675-1680
Muller KM et al. (1994) H-2D haplotype-linked expression and
involvement of TNF-a in Th2 cell-mediated tissue inflam-
mation. I ImmunoI153:316-324
National Cooperation Group on Clinical Study of Mycobac-
terium Vaccae V, Luo Y (2001) The immunotherapeutic
effect of Mycobacterium vaccae vaccine on initially treated
pulmonary tuberculosis. Zhonghua Jie He He Hu Xi Za
Zhi 24:43-47
Newport M et al. (1996) A mutation in the interferon -g recep-
tor gene and susceptibility to mycobacterial infection. N
Engl I Med 335:1941-1949
Nicholson S et al. (1996) Inducible nitric oxide synthase
in pulmonary alveolar macrophages from patients with
tuberculosis. I Exp Med 183:2293-2302
North RJ (1998) Mice incapable of making IL-4 or IL-10 dis-
play normal resistance to infection with Mycobacterium
tuberculosis. Clin Exp ImmunoI113:55-58
Oddo M et al. (1998) Fas ligand-induced apoptosis of infected
human macrophages reduces the viability of intracellular
Mycobacterium tuberculosis. I IrnmunoI160:5448-5454
Opie EL,Aronson ID (1927) Tubercle bacilli in latent tubercu-
lous lesions and in lung tissue without tuberculous lesions.
Arch Pathol Lab Med 4:1-21
Orme 1M et al. (1993) Cytokine secretion by CD4 T lympho-
cytes acquired in response to Mycobacterium tuberculosis
infection. I ImmunoI151:518-525
Pancholi P et al. (1993) Sequestration from immune CD4+
T cells of mycobacteria growing in human macrophages.
Science 260:984-986
Pathan AA et al. (2000) High frequencies of circulating IFN-
gamma-secreting CD8 cytotoxic T cells specific for a novel
MHC class I-restricted Mycobacterium tuberculosis epitope
in M. tuberculosis-infected subjects without disease. Eur J
ImmunoI30:2713-2721
Power CA, Wei G, Bretscher PA (1998) Mycobacterial dose
defines the ThI/Th2 nature of the immune response
independently of whether immunization is administered
by the intravenous, subcutaneous, or intradermal route.
Infect Immun 66:5743-5750
Ramirez F et al. (1996) Glucocorticoids promote a Th2 cyto-
kine response by CD4+ T cells in vitro. I Immunol 156:
2406-2412
Rigby WF et al. (1987) Comparison of the effects of 1,25-dihy-
droxyvitamin D3 on T lymphocyte subpopulations. Eur I
ImmunoI17:563-566
Roach TIA, Chatterjee D, Blackwell 1M (1994) Induction of
early-response genes KC and IE by mycobacterial lipo-
arabinomannans: regulation of KC expression in murine
macrophages by Lsh/fty/Bcg (candidate Nramp). Infect
Immun 62:1176-1184
Rockett KA et al. (1998) 1,25-dihydroxyvitamin D3 induces
nitric oxide synthase and suppresses growth of Mycobac-
G.A. W. Rook
terium tuberculosis in a human macrophge-like cell line.
Infect Immun 66:5314-5321
Rojas M et al. (2000) Mannosylated lipoarabinomannan
antagonizes Mycobacterium tuberculosis-induced macro-
phage apoptosis by altering Ca+2-dependent cell signaling.
I Infect Dis 182:240-25 I
Rook G (1988) The role of vitamin D in tuberculosis (edito-
rial). Am Rev Respir Dis 138:768-770
Rook G, Hernandez-Pando R (1996) The pathogenesis of
tuberculosis. Annu Rev Microbiol 50:259-284
Rook G et al. (1985) I-A restricted activation by T cell lines of
anti-tuberculosis activity in murine macrophages. Clin Exp
ImmunoI59:414-420
Rook G et al. (1986) Vitamin D3, gamma interferon, and
control of proliferation of Mycobacterium tuberculosis by
human monocytes. Immunology 57:159-163
Rook G et al. (1987) A direct effect of glucocorticoid hor-
mones on the ability of human and murine macrophages
to control the growth of M. tuberculosis. Eur I Respir Dis
71:286-291
Rook G et al. (1996) Urinary steroid metabolites in tuberculo-
sis; a new clue to pathogenesis. Quart I Med 89:333-341
Rook Get al. (2000) Local regulation of glucocorticoid activity
in sites of inflammation; insights from the study of tuber-
culosis. Ann NY Acad Sci 917:913-922
Salgame PR et al. (1991) Differing Iymphokine profiles of
functional subsets of human CD4 and CD8 T cell clones.
Science 254:279-282
Schaible UE et al. (2000) Intersection of group I CDI molecules
and mycobacteria in different intracellular compartments
of dendritic cells. I ImmunoI164:4843-4852
Schlesinger LS (1998) Mycobacterium tuberculosis and the com-
plement system. Trends MicrobioI6:47-49; discussion 49-50
Seah GT, Rook GA (1999) A sensitive, non-radioactive quan-
titative method for measuring IL-4 and IL-4delta2 mRNA
in unstimulated cells from multiple clinical samples, using
nested RT-PCR. J Immunol Methods 228:139-149
Seah GT, Rook GAW (2000) High levels of mRNA encoding
IL-4 in unstimulated peripheral blood mononuclear cells
from tuberculosis patients revealed by quantitative nested
RT-PCR; correlations with serum IgE levels. Scand J Infect
Dis 33:106-109
Seah GT, Rook GAW (2001) IL-4 influences apoptosis of myco-
bacterium-reactive lymphocytes in the presence of TNF-a.
J ImmunoI167:1230-1237
Seah GT, Scott GM, Rook GA (2000) Type 2 cytokine gene acti-
vation and its relationship to extent of disease in patients
with tuberculosis. J Infect Dis 181:385-389
Shands IW, Senterfitt VC (1972) Endotoxin-induced hepatic
damage in BCG-infected mice. Am J Pathol 67:23-40
Shepard CC (1958) A comparison of the growth of selected
mycobacteria in HeLa, monkey kidney, and human amnion
cells in tissue culture. I Exp Med 107:237-246
Shwartzman G (1937) Phenomenon of local tissue reactivity
and its immunological, pathological, and clinical signifi-
cance. Hoeber, New York
Siegele DA, Kolter R (1992) Life after log. I Bacteriol 174:
345-348
Sieling PA et al. (1995) CD I-restricted T cell recognition of
microbial Iipoglycan antigens. Science 269:227-230
Silva CL, Lowrie DB (1994) Asingle mycobacterial protein (hsp65)
expressed by a transgenic antigen-presenting cell vaccinates
mice against tuberculosis. Immunology 82:244-248
The Immunology and Pathogenesis of Tuberculosis
Silva CL, Lowrie DB (2000) Identification and characteriza-
tion of murine cytotoxic T cells that kill Mycobacterium
tuberculosis. Infect Immun 68:3269-3274
Skinner MA et al. (1997) Immunization with heat-killed
Mycobacterium vaccae stimulates CD8+ cytotoxic T cells
specific for macrophages infected with Mycobacterium
tuberculosis. Infect Immun 65:4525-4530
Spence DP et al. (1993) Tuberculosis and poverty. Br Med J
307:759-761
Stanford JL, Shield MJ, Rook GAW (198l) How environmental
mycobacteria may predetermine the protective efficacy of
BCG. Tubercle 62:55-62
Stanford JL et al. (1993) Old plague, new plague and a treat-
ment for both? AIDS 7:1275-1277
Stenger S et al. (1998) An antimicrobial activity of cytolytic T
cells mediated by granulysin. Science 282:121-125
Sturgill-Koszycki S et al. (1994) Lack of acidification in
Mycobacterium phagosomes produced by exclusion of the
vesicular proton-ATPase. Science 263:678-681
Sugita M et al. (2000) CDlc molecules broadly survey the
endocytic system. Proc Natl Acad Sci USA 97:8445-8450
Tobach E, Bloch H (1956) Effect of stress by crowding prior
to and following tuberculous infection. Am J Physiol 187:
399-402
Toossi Z et al. (1997) Modulation of IL-12 by transform-
ing growth factor-beta (TGF-beta) in Mycobacterium
tuberculosis-infected mononuclear phagocytes and in
patients with active tuberculosis. J Clin Lab Immunol
49:59-75
Tramontana JM et al. (1995) Thalidomide treatment reduces
tumor necrosis factor alpha production and enhances
weight gain in patients with pulmonary tuberculosis. Mol
Med 1:384-397
Turner J et al. (2000) Effective preexposure tuberculosis vac-
cines fail to protect when they are given in an immuno-
therapeutic mode. Infect Immun 68:1706-1709
Van Crevel R et al. (2000) Increased production of interleukin
4 by CD4+ and CD8+ T cells from patients with tuberculo-
sis is related to the presence of pulmonary cavities. J Infect
Dis 181:1194-1197
Van Rie A et al. (1999) Exogenous reinfection as a cause of
recurrent tuberculosis after curative treatment. N Engl J
Med 341:1174-1179
Van-der-Pouw-Kraan TC et al. (1995) Prostaglandin-E2 is a
potent inhibitor of human interleukin 12 production. J
Exp Med 181:775-779
Varadhachary AS, Salgame P (1998) CD95 mediated T cell
apoptosis and its relevance to immune deviation. Onco-
gene 17:3271-3276
Vieira PL et al. (1998) Glucocorticoids inhibit bioactive IL-
151
12p70 production by in vitro- generated human dendritic
cells without affecting their T cell stimulatory potential. J
ImmunoI161:5245-5251
Visser J et al. (1998) Differential regulation of interleukin-1O
(IL-I0) and IL-12 by glucocorticoids in vitro. Blood 91:
4255-4264
Waddell RD et al. (2000) Safety and immunogenicity of a five-
dose series of inactivated Mycobacterium vaccae vaccine
for the prevention of HIV-associated tuberculosis. Clin
Infect Dis 30 [SuppI3]:S309-S315
Wakeham J et al. (1998) Lack of both types 1 and 2 cytokines,
tissue inflammatory responses, and immune protection
during pulmonary infection by Mycobacterium bovis bac-
ille Calmette-Guerin in IL-12-deficient mice. J Immunol
160:6101-6111
Wang CC, Rook GAW (1998) Inhibition of an established
allergic response to ovalbumin in Balblc mice by killed
Mycobacterium vaccae. Immunology 93:307-313
Watson VE et al. (2000) Apoptosis in Mycobacterium tubercu-
losis infection in mice exhibiting varied immunopathology.
Journal of Pathology 190,211-20
Wayne LG (1994) Dormancy of Mycobacterium tuberculosis
and latency of disease. Eur J Clin Microbiol Infect Dis 13:
908-914
Wei J et al. (2000) Identification of a Mycobacterium tubercu-
losis gene that enhances mycobacterial survival in macro-
phages. J Bacteriol 182:377-384
Wilkinson RJ et al. (2000) Influence of vitamin D deficiency
and vitamin D receptor polymorphisms on tuberculosis
among Gujarati Asians in west London: a case-control
study. Lancet 355:618-621
Wilsher ML et al. (1999) Human in vitro immune responses
to Mycobacterium tuberculosis. Tubercle Lung Dis 79:
371-377
Wilson GS, Schwabacher H, Maier I (1940) The effect of the
desensitisation of tuberculous guinea-pigs. J Pathol Bac-
teriol 50:89-109
Xu S et al. (1994) Intracellular trafficking in Mycobacterium
tuberculosis and Mycobacterium avium-infected macro-
phages. J Immunol 153:2568-2578
Yong AJ et al. (1989) Total and antimycobacterial IgE levels
in serum from patients with tuberculosis and leprosy.
Tubercle 70:273-279
Zhang L, Ma W, Da J (2000) Mycobacterium vaccae influences
the kinetics of Thl/Th2 cells and expression of iNOS in a
marine model of experimental tuberculosis. Zhonghua Jie
He He Hu Xi Za Zhi 23:43-46
Zuany-Amorim C et al' (2002) Suppression of airway eosino-
philia by killed Mycobacterium vaccae-induced allergen-
specific regulatory T-cells. Nat Medicine 8:625-629
10 Pathology of Tuberculosis
MOHAMMED AKHTAR and HADEEL AL MANA

CONTENTS 10.1
Pathobiology of Tuberculosis
10.1 Pathobiology of Tuberculosis 153
10.2 Microscopic Pathology 154
10.3 Identification of Mycobacteria Following inhalation, the droplet nuclei usually
in Tissue Section 155 implant in the bronchioles and alveoli. The tubercu-
10.4 Macroscopic Pathology 156 lous bacilli first elicit a polymorphonuclear leukocyte
10.4.1 Lung 156 reaction. Later the bacilli are ingested by pulmonary
10.4.2 Lymph Nodes 157
macrophages. If the organisms are virulent or if the
10.4.3 Intestine 157
10.4.4 Urinary Tract 157 inhaled dose is large, the bacilli may multiply within
10.4.5 Male Genital Tract 158 the phagocytic vacuoles of the macrophages. The
10.4.6 Female Genital Tract 158 resulting lesion may grow due to influx of additional
10.4.7 Central Nervous System 159 monocytes/ macrophages from the blood stream. At
10.5 Miliary Tuberculosis 159
this stage, this immune response is mainly due to a
10.6 Disseminated Bacille Calmette-Guerin
(BCG) 160 tissue-damaging delayed hypersensitivity response,
10.7 Role of Fine-Needle Aspiration Biopsy which kills the bacilli-laden macrophages thus pro-
in the Diagnosis of Tuberculosis 160 ducing a solid caseating center containing extracellu-
References 161 lar bacilli. The next stage of the lesion depends upon
the quality and intensity of cell mediated immunity. If
only poor cell mediated immunity response develops,
Tuberculosis is caused by Mycobacterium tuberculo- the bacilli escape from the periphery of the caseat-
sis, a slender, acid-fast aerobic organism which can ing center and are ingested by macrophages. Again
be transmitted by inhalation, by ingestion or, rarely, the delayed hypersensitivity response kills these
by direct implantation. A fetus can be infected in macrophages causing enlargement of the caseous
utero transplacentally. Patients with active pulmo- center and progression of the disease. If, on the other
nary tuberculosis are the major source of infection hand, an effective cell-mediated immunity response
and inhalation is overwhelmingly the most frequent is mounted, the lesion is surrounded by highly acti-
route. Droplet nuclei containing a few organisms vated macrophages, i.e. epithelioid cells, which ingest
become airborne during coughing, speaking or even and destroy the bacilli often arresting the lesion at
singing and may remain suspended in the room air the subclinical stage.
for hours. In many cases, however, the bacilli continue to
multiply extracellularly. Additional delayed hyper-
sensitivity reaction may cause local tissue damage,
thus resulting in progressive enlargement of the
lesion with extension into surrounding tissues. With
the passage of time, the areas of caseation undergo
liquefaction. The cause of liquefaction is not known.
M. AKHTAR, MD, FCAP, FRCPA, FRCPath Possibly hydrolytic enzymes released from the dead
Chairman, Department of Pathology and Laboratory Medi- inflammatory cells and products of cell wall of the
cine, King Faisal Specialist Hospital and Research Centre, P.O. bacilli may playa role. Liquefaction plays an impor-
Box 3354, MBCIO, Riyadh 11211, Saudi Arabia tant role in the progression of the disease in two
H. AL MANA, MD
Fellow, Department of Pathology ad Laboratory Medicine,
important ways. First, the liquidized caseous mate-
King Faisal Specialist Hospital and Research Centre, P.O. Box rial is frequently, but not always, an excellent growth
3354, Riyadh 11211, Saudi Arabia medium for tubercle bacilli. Second, the liquidized

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
154 M. Akhtar and H. Al Mana

material causes further tissue damage and may easily cells are believed to be formed by the fusion of several
erode into adjacent structures, such as blood vessels epithelioid cells. The granulomas are surrounded by a
and bronchi, thus spreading the disease to other zone of lymphocytic cells, which are an integral part
locations. Excessive localized tissue damage may also of the cell-mediated immunity against mycobacterial
result in cavitation (Allison et al. 1962; Shima et al. infection. Several subtypes of lymphocytes have been
1972; Dannenberg 1993,1999; Dannenberg and Rook identified including natural killer cells, cytotoxic T
1994; Fayyazi et al. 2000). cells and type I and type II T helper cells among
others (Dannenberg 1999).
Most of the granulomas in mycobacterial infec-
tions are characterized by the presence of central
10.2 areas of caseation necrosis and are termed as
Microscopic Pathology caseating granulomas. (Figs. 10.2-1004) In these
granulomas, the epithelioid cells form a palisading
The morphologic hallmark of tissue response to arrangement around the caseating center. Other
mycobacterial infection is inflammation which is granulomas may lack central caseation and are called
composed of nodular microscopic lesions called non-caseating granulomas. (Fig. 10.1) Histologically,
granulomas. Granulomas are usually small «2 mm) caseation necrosis consists of amorphous, coarsely
and are composed of aggregates of epithelioid cells granular, acellular debris. Liquefaction of the caseat-
surrounded by a rim of lymphocytes (Fig. 10.1). The
epithelioid cells are highly activated macrophages
which are characterized by abundant slightly eosin-
ophilic granular cytoplasm and a relatively small,
usually eccentric, nucleus. The epithelioid cells may
superficially mimic epithelial cells. At ultrastructural
level, these cells lack junctional complexes and their
cytoplasm is rich in lysosomes and abundant rough
endoplasmic reticulum. These cells are less actively
phagocytic than the typical macrophages, but have
potent bactericidal function due to the abundant
lysosomal content. A striking morphologic feature
of granulomas is the presence of multinucleated
giant cells. Some of the multinucleated cells, which
have nuclei evenly dispersed through the cytoplasm,
Fig. 10.2. Granulomas in a lymph node composed of epithe-
are called foreign body giant cells. Others have lioid cells and a few giant cells. The granulomas are non-
nuclei disposed in a horse-shoe arrangement and caseating, except for one on the left which reveals extensive
are termed Langhans giant cells. Both types of giant caseation. Hematoxylin and eosin stain magnification IOOx

Fig. IO.I. Photomicrograph featuring a lymph node with gran- Fig. 10.3. A large caseating granuloma with palisading epithe-
ulomas composed of epithelioid cells. No caseation is present. lioid cells and multinucleated giant cells. Hematoxylin and
Hematoxylin and eosin stain magnification 200x eosin stain magnification 200x
Pathology of Tuberculosis
Fig. lOA. A caseating granuloma with a Langhans giant cell.
Hematoxylin and eosin stain magnification 300x
Fig. 10.5. A non-caseating granuloma undergoing fibrosis
Fig. 10.6. Acid-fast bacilli stained by Ziehl-Neelsen stain. Diff-
Quik stain 400x
155
ing centers may result in cavitation. As the granuloma
matures, it becomes surrounded by a layer of fibrosis
which may gradually extend towards the center of the
granuloma. (Fig.IO.S) Healed granulomas are virtu-
ally completely replaced by fibrous tissue although
occasional epithelioid cells or giant cells may be
preserved. These fibrosed granulomas usually lack
any organisms. Cavitary lesions heal by fibrosis and
calcification, both of which may be quite extensive in
longstanding lesions.
10.3
Identification of Mycobacteria
in Tissue Section
Histopathologic diagnosis of tuberculosis depends
on the demonstration of granulomatous inflamma-
tion with caseation. However, similar morphologic
changes may also be encountered in association
with other infections, especially fungal organisms.
Therefore, demonstration of the offending organism
may be crucial for arriving at a definitive diagnosis of
tuberculosis. If the tissue is received in a fresh state,
part of the material may be submitted for micro-
biologic cultures. In addition, histologic sections
of the tissue may be stained using special staining
procedures for the demonstration of tubercle bacilli.
Ziehl-Neelsen stain is the most commonly used stain
for identification of tubercle bacilli in tissue sections.
Owing to the waxy nature of the cell wall component,
mycobacteria stain with the red dye carbolfuchsin
and retain the stain after washing with acid alcohol
(Stevens and Frances 1996). This indicates that the
mycobacteria are acid-fast as compared to other
bacilli, which are usually not acid-fast. When acid-
fast bacilli (AFB) are detected in the granulomatous
area, it is highly specific for mycobacterial infection.
In Ziehl-Neelsen stain, tubercle bacilli are identified
as a straight or curved rod 0.2-0.sx2-S 11m, that stain
uniformly bright red but often have a beaded appear-
ance. (Fig. 10.6) The organisms are usually located at
the junction of viable tissue with the necrotic zone.
Ziehl-Neelsen stain, however, is relatively insensitive.
It has been estimated that between 105_10 6 bacteria
per milliliter of tissue is required before they can be
visualized in a section stain by Ziehl-Neelsen stain
(Jagirdar and Zagzag 1996). Auramine-rhodamine
stain is a more sensitive stain for mycobacteria, but
it may not be so specific. Furthermore, it requires a
fluorescent microscope for detection of the organ-
isms (Stevens and Frances 1996).
156 M. Akhtar and H. Al Mana

Monoclonal antibodies directed to different pro-
teins of M. tuberculosis have been employed by some
workers for identification of these organisms. However,
several technical problems remain to be addressed
(Barbolini et al. 1989). Recently polymerase chain
reaction (PCR) for identification of mycobacterium
tuberculosis complex DNA has been employed suc-
cessfully on formalin-fixed paraffin-embedded
material as well as lesion material obtained by fine
needle aspiration biopsy. These studies have provided
evidence for higher sensitivity and specificity of PCR
amplification for identification of M. tuberculosis in
biopsy material (Vago et al. 2000). Ligase chain reaction
(LCR) is another diagnostic alternative for detection of
tubercle bacilli. This technique uses a combination of
polymerase and ligase for amplification of target DNA
sequences in the chromosome of M. tuberculosis. This
technique may be applied in body fluids as well as
material from tissue biopsies. It is commercially avail-
able as a semi-automatic assay which is rapid, sensitive
and is claimed to be quite specific. However, additional
studies are required to determine the usefulness of this
technique for identification of tuberculous bacilli in
material derived from paraffin embedded tissue sec-
tions (Ruiz-Serrano et al. 1998).
node component. The parenchymal lesion can occur
anywhere in the lung but is usually seen in the middle
and lower lobes. On gross examination, it is a well
circumscribed, often subpleural, nodule. (Fig. 1O.7)
Initially, these lesions are centrally necrotic but later
they became fibrotic and calcified. The lymph node
component of Ghon complex is found in the draining
hilar or intrapulmonary nodes (Allen 1995).
Post primary tuberculosis represents recurrence of
pulmonary tuberculosis after previous sensitization
to M. tuberculosis during primary infection. These
lesions usually involve apical and posterior seg-
ments of the lungs and are grossly characterized by
fibrotic and caseous areas, frequently with cavitation.
(Figs. 10.8, 1O.9) These cavities may be quite extensive
and fibrosis and calcification of their walls may prevent
collapse. In severe cases, cavities may replace the entire
lobe and may extend to the other parts of the lung. In

10.4
Macroscopic Pathology

The macroscopic appearance of an organ or tissue

involved by tuberculosis has several important clues
which raise suspicion of tuberculosis. The presence of Fig. 10.7. A localized subpleural nodule in the lower lobe of
caseation necrosis, cavitation and enlarged regional right lung, probably representing primary tuberculosis in a
lymph nodes with similar areas of caseation necro- 20-year-old man
sis should strongly suggest tuberculosis. However, the
macroscopic appearance may be somewhat modified
according to the location and the organ involved.

10.4.1
Lung

The gross appearance of the lung involved by tubercu-

losis depends upon whether it is primary tuberculosis,
which is encountered in patients who have no previous
exposure to M. tuberculosis infection, or whether it is
post primary tuberculosis in a patient who has had a
previous encounter with tuberculosis infection.
The clinical and pathologic manifestation of pri- Fig. 10.8. Wedge resection of the upper lobe of the right lung
mary tuberculosis is the Ghon complex, which consists featuring multiple cavitary lesions extensively replacing the
of a parenchymal component and a prominent lymph lung parenchyma
Pathology of Tuberculosis 157

volume of the diseased tissue in regional lymph

nodes than at the original site of infection. Infected
lymph nodes commonly become greatly enlarged. On
sectioning, the nodes are found to be replaced by light
yellow areas of caseation necrosis. (Fig. 10.10) Some
of the nodes may be virtually completely replaced
by caseating necrosis, which may also extend to
adjacent extranodal tissues. Several lymph nodes
may be matted together to form a large mass super-
ficially mimicking a neoplasm. Overlying skin may be
stretched or ulcerated, resulting in sinus tracts.

Fig. 10.9. Cut surface of left lung showing congestion of the 10.4.3
lung parenchyma. Several yellowish areas in the upper part
Intestine
of the lung represent caseating lesions of tuberculous bron-
chopneumonia
The ileocecal region is the most common site for gas-
trointestinal tuberculosis. The earliest changes occur
most frequently in the terminal ileum or cecum in the
form of small nodules which project into the intestinal
lumen and may show evidence of ulceration. As the
inflammatory process progresses, characteristic mor-
phologic changes of ulcerative, ulceratohypertrophic
or hypertrophic disease occur. In the ulcerative type,
the disease is mostly limited to mucosa and submu-
cosa with transverse ulcers which may involve the
intestine circumferentially. Mesenteric lymph nodes
are frequently involved by the disease and may show
massive enlargement. The hypertrophic form (seen in
10% of patients) is characterized by extensive inflam-
mation and fibrosis of all layers of the bowel wall.
Fibrous adhesions involving bowel, mesentery and
Fig. 10.10. Multiple matted lymph nodes extensively replaced
lymph nodes may result in a mass closely resembling
by caseating tuberculous lesions a neoplasm. Stricture of the bowel wall may lead to
obstruction. The ulceratohypertrophic type of disease
may have features of both ulcerative and hypertrophic
the higWy susceptible, highly sensitized individual, the disease (Yangking and Chandrasama 1999). Colonic
tuberculous infection may spread rapidly throughout tuberculosis may involve any part of the colon. In the
large areas of the lung and produce a diffuse bron- acute phase, localized ulceration of the mucosa occur.
chopneumonia or lobe-exudative consolidation. The ulcers tend to be transverse. In the chronic phase,
Longstanding cavities may be secondarily infected by extensive caseation and fibrosis may be seen. Ano-
bacteria or fungi (Scully et al. 1990; Anonymous 1970). rectal tuberculosis may be an ulcerative type similar
Some of the fungal infections maybe in the form of to that seen in the colon or it may be characterized
fungal balls. Rarely, tuberculous lesions may co-exist by complex fistulae. The latter type of involvement
with carcinoma (Ting et al. 1976). may be difficult to distinguish from Crohn's disease
(Kobayashi and Chandrasama 1999).

10.4.2
Lymph Nodes 10.4.4
Urinary Tract
A nearly uniform event following infection by M.
tuberculosis is spread from the primary focus to The renal parenchyma initially becomes infected by
regional lymph nodes. This often results in a greater hematogenous spread, which may involve both the
158 M. Akhtar and H. AI Mana

kidneys. The kidney may be enlarged or decreased

in size with irregular scarring on the surface (Farrow
1997). The lesions usually start in the medullary
region, later leading to papillary necrosis and exten-
sion of the disease into the renal pelvis. A cut sur-
face reveals dilation of the renal pelvis and calyxes,
which are filled by light yellow caseous material.
(Fig.lO.11) Distinction between renal tuberculosis
and xanthogranulomatous pyelonephritis may be
difficult. In advanced cases, the renal parenchyma
is replaced by caseous material leaving a thin rim
of atrophic renal tissue at the periphery. (Fig. 10.12)
Localized lesions may heal by calcification. In the
urinary bladder, early macroscopic changes are typi- Fig. 10.11. Cut surface of nephrectomy specimen with marked
dilatation of calyxes and the renal pelvis. Some of the calyxes
cally disposed around the ureteral orifices. There is
are lined by light yellow caseous material. Renal parenchyma
marked edema and hyperemia, later changing to reveals variable atrophy
fine mucosal nodulation and ulceration. The ulcers
are sharply demarcated and have undermined edges
and a layer of gray soft material on the floor. The
ureteral orifice may be deformed by the disease; it
may be distorted and nodular or remain rigidly open:
the characteristic 'golf hole' ureter. The lesions may
extend to the trigone or the rest of the bladder. In the
late stages, there is fibrosis and shrinkage of the blad-
der with reduction of capacity (McClure and Young
1989). Occasionally, there is formation of exuberant
granulation tissue forming polypoid masses which
may mimic neoplasm. The lesions in the ureter and
urinary bladder are always secondary to tuberculosis
of the kidney. The most common site of involvement
in the ureter is ureterovesical orifice or ureteropelvic
junction.

Fig. 10.12. A nephrectomy specimen showing extensive case-

10.4.5 ous material filling the dilated calyxes
Male Genital Tract

Testis and epididymis are involved by metastatic

spread through the blood stream and are usually
associated with tuberculosis in other areas in the
genitourinary system. On gross examination, there
is enlargement, thickening and fibrosis of the testis
and epididymis and there are variable numbers
of necrotic and caseous lesions with discharging
sinuses. (Figs. 10.13, 1O.14) Tuberculosis of the pros-
tatic urethra and penis have rarely been reported.

10.4.6
Female Genital Tract
The most common site for involvement by tubercu-
losis in the genital tract is the fallopian tubes. In the
Fig.IO.B. Epididymo-orchiectomy specimen in which epi-
didymitis is involved by several caseating lesions. The testis is
somewhat atrophic but is not involved
Pathology of Tuberculosis
Fig. 10.14. Cut surface of a testis virtually completely replaced
by amorphous light yellow caseous material
adhesive ulcerative form of tuberculous salpingitis,
there is little distension and enlargement of the tube
is mainly interstitial. The surface may show minute
gray to yellow nodules and there may be adhesions
to adjacent organs and development of fistulae. In the
inflammatory exudative form, the tubes are markedly
distended by exudates, which may be purulent rather
than caseous. The wall is thin and serosa relatively
smooth. There is eversion of the fimbria and the
abdominal ostium is usually patent. Occasionally,
the tube may appear normal at gross inspection. In
the uterine cavity the disease is limited to the endo-
metrium, which may be difficult to recognize at gross
examination. Vulvovaginal disease takes the form of
small shallow ulcers with sinus tracts and fibrosis.
Involvement of the ovaries is uncommon but ovar-
ian tissues may show adhesions and inflammatory
exudates due to granulomatous inflammation.
10.4.7
Central Nervous System
Involvement of the central nervous system is usually
through the blood stream, although the spinal cord
may be involved by extension of the disease from
adjacent lesions in the vertebral bones. Leptomenin-
gitis due to tuberculosis is characterized by a cover-
ing of the surface by a diffuse exudate, which later
develops into small nodules. These changes may be
more pronounced at the base of the brain.
Tuberculous infection may also be present in the
form of a space-occupying lesion in the brain or, rarely,
in the spinal cord. These masses, called tuberculomas,
may be single or multiple and mayor may not be asso-
ciated with concomitant meningitis. They have been
159
Fig. 10.15. A space-occupying lesion (tuberculoma) resected
from the posterior fossa in a 30-year-old man. There are cystic
areas filled with necrotic caseating material
reported throughout the cerebral hemispheres, basal
ganglia, the brain stem and cerebellum. Tuberculoma
is usually a well defined mass separated from the sur-
rounding brain tissue by a zone of fibrosis. The mass
is usually solid with areas of necrosis and may reveal
focal cystic change. (Fig. 10.15) Distinction from a
glioma may not always be possible and correct diag-
nosis usually requires microscopic examination.
10.5
Miliary Tuberculosis
Miliary tuberculosis results from a massive hema-
togenous spread of tubercle bacilli, which may
occur either at the time of primary infection or at a
time remote from primary infection. Lymphohema-
togenous dissemination following primary infection
is the usual mechanism causing miliary tuberculosis
in infants and children. Miliary dissemination may
also occur when an older tuberculous lesion in the
lung or an extrapulmonary site reactivates and dis-
charges massive numbers of tubercle bacilli into the
blood stream.
Miliary tuberculosis may involve many organs
including lung, liver, spleen, bone marrow, kidney,
adrenals, eyes and thyroid (Divinagracia and Harris
1999). The involved organs have numerous small,
gray to reddish brown, punctate rounded lesions of
more or less uniform size. Microscopically, miliary
foci reveal tuberculous granulomas with or without
central caseation. With appropriate stains, acid-fast
bacilli may be found within the macrophages or
epithelioid cell of the granulomas or in the areas of
caseation.
160
10.6
Disseminated Bacille Calmette-Guerin
(BCG)
Vaccination of the newborn and babies with Bacille
Calmette-Guerin has been used widely throughout
the world and has been shown to be a safe method
for providing some protection against tuberculosis.
However, in some cases local ulceration or regional
lymph node enlargement due to granulomatous
disease may be seen. In patients with congenital or
acquired immune deficiency, a more progressive and
disseminated BCG infection may develop, which may
prove fatal unless treated by anti-tuberculous chemo-
therapy. In these cases, a biopsy usually reveals diffuse
proliferation of histiocytes intermixed with variable
numbers of neutrophil without distinct granulomas.
(Fig. 10.16) The histiocytes on Ziehl-Neelsen stain-
ing usually reveal large numbers of intracytoplasmic
acid-fast bacilli (Fig.10.1?) (AI BhlaI2000).
Fig. 10.16. Photomicrograph of a lymph node involved by dis-
seminated BeG infection. The node is replaced by numerous
histiocytes with abundant granular cytoplasm. No granulo-
mas are present. Hematoxylin and eosin stain 300x
Fig. 10.17. Numerous intracytoplasmic acid-fast bacilli pres-
ent within the macrophages. Ziehl-Neelsen stain 400x
M. Akhtar and H. AI Mana
10.7
Role of Fine-Needle Aspiration Biopsy
in the Diagnosis of Tuberculosis
Fine-needle aspiration biopsy is an extremelyeffective
technique for the pathologic diagnosis of tuberculo-
sis. The morphologic pattern encountered in smears
parallels closely with that seen in histologic prepa-
rations. Thus all the components of granulomatous
inflammation including lymphocytes, macrophages,
epithelioid giant cells and other inflammatory cells
are easily recognized. (Fig. 10.18) Caseating necrosis
can also be recognized by the presence of amorphous
material with scattered polymorphonuclear cells and
a few epithelioid cells (Das and Pant 1994; Prasoon
2000; Radhika et al. 1993; Bhattacharya et al. 1998)
(Fig. 10.19). Aspirated material may also be used suc-
cessfully for culturing the tubercle bacilli. In addition,
aspiration smears may be stained by special stains to
identify tubercle bacilli.
Fig.IO.IS. Fine-needle aspiration biopsy smear showing an epi-
thelioid granuloma without caseation. Diff-Quik stain 300x
Fig. 10.19. Aspiration smear featuring an epithelioid granu-
loma associated with a background of necrosis and inflam-
matory cells indicating caseation. Diff-Quik stain 200x
Pathology of Tuberculosis
References
AI Bhlal LA (2000) Pathologic findings for Bacille Calmette-
Guerin infections in immunocompetent and immunocom-
promised patients. Am J Clin Pathol 113:703-708
Allen EA (1995) Tuberculosis and other mycobacterial infec-
tions of the lung. In: Thurlbeck WM, Churg AM (eds)
Pathology of the lung, 2nd edn, pp 299-265
Allison MJ, Zappasodi P, Lurie MB (1962) Host-parasite rela-
tionships in natively resistant and susceptible rabbits on
quantitative inhalation of tubercle bacilli. Am Rev Respir
Dis 85:553-569
Anonymous (1970) Aspergilloma and residual tuberculous
cavities - the results of a resurvey. Tubercle 51:227-245
Barbolini G, Bisetti A, Colizzi V, Damiani G, Migaldi M, Vis-
mara D (1989) Immunohistologic analysis of mycobacte-
rial antigens by monoclonal antibodies in tuberculosis and
mycobacteriosis. Hum Pathol 20:1078-1083
Bhattacharya S, Raghuveer CV, Adhikari P (1998) FNAC diag-
nosis of tuberculosis - an eight years study at Mangalore.
Indian J Med Sci 52:498-506
Dannenberg AM Jr (1993) Immunopathogenesis of pulmo-
nary tuberculosis. Hosp Pract 28:33-40, 51-58
Dannenberg A (1999) Pathogenesis oftuberculosis. In: Schloss-
berg D (ed) Tuberculosis and non-tuberculous mycobacte-
rial infections. Saunders, Philadelphia, pp 17-47
Dannenberg AM Jr, Rook GAW (1994) Pathogenesis of pulmo-
nary tuberculosis: an interplay of tissue-damaging and mac-
rophage-activating immune responses - dual mechanisms
that control bacillary multiplication. In: Bloom BR (ed)
Tuberculosis: pathogenesis, protection, and control. Ameri-
can Society for Microbiology, Washington DC, pp 459-483
Das DK, Pant CS (1994) Fine needle aspiration cytologic diag-
nosis of gastrointestinal tract lesions. A study of 78 cases.
Acta CytoI38:723-729
Divinagracia R, Harris HW (1999) Miliary tuberculosis.
In: Schlossberg D (ed) Tuberculosis and non tubercu-
lous mycobacterial infection Saunders, Philadelphia, pp
271-284
Farrow GM (1997) Diseases of the kidney in the urologic
pathology (Murphy WM ed). Saunders, Philadelphia, pp
440-442
Fayyazi A, Eichmeyer B, Soruri A, Schweyer S, Herms J,
Schwarz P, Radzun HJ (2000) Apoptosis of macrophages
and T cells in tuberculosis associated caseous necrosis. J
Pathol 191:417-425
Jagirdar J, Zagzag D (1996) Pathology and insight into patho-
161
genesis of tuberculosis. In: Rom WN, Gary S (eds) Tuber-
culosis. Little Brown, Boston, pp 467-491
Kobayashi G, Chandrasama P (1999) Non neoplastic lesions of
the colon. In: Chandrasama P (ed) Gastrointestinal pathol-
ogy. Appleton and Lange, Stanford, pp 263-264
McClure J, Young RH (1989) Infectious disease of the urinary
bladder, including malakoplakia. In: Young RH (ed) Pathol-
ogy of urinary bladder Churchill Livingston, New York, pp
351-375
Prasoon D (2000) Acid-fast bacilli in fine needle aspiration smears
from tuberculous lymph nodes. Acta Cytol44:297-300
Radhika S, Rajwanshi A, Kochhar R, Kochhar S, Dey P, Roy P
(1993) Abdominal tuberculosis. Diagnosis by fine needle
aspiration cytology. Acta Cytologica 37:673-678
Ruiz-Serrano MJ, Albadelejo J, Martinez-Sanchez L, Bouza E
(1998) LCx: a diagnostic alternative for the early detection
of mycobacterium tuberculosis complex. Diagn Microbiol
Infect Dis 32:259-264
Scully RE, Mark EJ, Menealy WF, Menealy BU (eds) (1990) Case
45-1990 Case records of Massachusetts General Hospital. N
Engl J Med 323:329-336
Shima K, Dannenberg AM Jr, Ando M et al. (1972) Macro-
phage accumulation, division, maturation, and digestive
and microbicidal capacities in tuberculous lesions. 1. Stud-
ies involving their incorporation of tritiated thymidine
and their content of lysosomal enzymes and bacilli. Am J
PathoI67:159-180
Stevens A, Frances RJ (1996) Micro organisms. In: Bancroft
JD, Stevens A (eds) Theory and practice of histopathologic
techniques, 4th edn. Churchill Livingston, New York, pp
291-308
Thomas JO, Adeyi D, Amanguno H (1999) Fine needle aspira-
tion in the management of peripheral lymphadenopathy in
a developing country. Diagn Cytopathol 21:159-162
Ting YM, Church WR, Ravikrishnan KP (1976) Lung car-
cinoma superimposed on pulmonary tuberculosis 199:
307-312
Vago L, Zervi P, Caldarelli-Stefano R, Cannone M, D'Amico M,
Bonetto S, Barberis M (2000) Polymerase chain reaction for
mycobacterium tuberculosis complex DNA. Use on nega-
tive archival Ziehl-Neelsen cytologic samples. Acta Cytol
44:1023-1028
Yanking M, Chandrasama P (1999) Pathology of small intes-
tine. In: Chandrasama P (ed) Gastrointestinal pathology.
Appleton and Lange, Stanford, pp 199-201
11 Nontuberculous Mycobacteria
JAE-JOON YIM and STEVEN M. HOLLAND

CONTENTS known as one of the nontuberculous mycobacteria,

M. smegma tis, was identified. Currently, 71 species of
11.1 Introduction 163 Mycobacterium are recognized or have been proposed
11.2 Classification of Nontuberculous
Mycobacteriosis 163
(Shinnick and Good 1994). Despite the fact that these
11.3 Epidemiology 164 mycobacteria have occasionally been identified from
11.4 Host Susceptibility Factors 166 clinical specimens (Wolinsky 1979), it is only relatively
11.5 Host Defense Against Mycobacteria 168 recently that serious attention has been paid to them.
11.6 Human Immune Defects 169 A variety of terms have been applied to these
11.6.1 IFNy Receptor 1 Deficiency 169
11.6.2 IFNy Receptor 2 Deficiency 170
nontuberculous mycobacteria, including anonymous,
11.6.3 1L-12 Receptor ~ 1 Deficiency 170 unclassified, unknown, tuberculoid, environmental,
11.6.4 IL-12 p40 Deficiency 170 opportunist, nyrocine, and mycobacteria other than
11.7 NFKB Essential Modulator (NEMO) tuberculosis (MOTT) (Collins et al. 1984; Debrunner
or IKB Kinase Defects 170 et aI.1992). The word 'atypical' has been popular in the
1l.8 Familial Syndrome 171
11.9 Polymorphisms 171
USA, while 'anonymous' and 'opportunist' have been
11.10 Body Morphotype 171 preferred in the UK (Collins et al.1984). However, each
11.11 Clinical Presentation and Diagnosis 172 term has its own limitations: These mycobacteria are
11.11.1 Pulmonary Disease 172 certainly not 'anonymous', there is no reason to regard
11.11.2 Lymphadenitis 173
them as 'atypical', and the term 'opportunist mycobac-
11.11.3 Skin and Soft-Tissue Infection 173
11.11.4 Disseminated NTM Infection 175 teria' could be equally applied to the more virulent
11.12 Treatment 175 Mycobacterium tuberculosis, especially in the setting
11.12.1 Chemotherapy 175 of AIDS (Grange 1989). Although the term nontuber-
11.12.2 Immunotherapy 177 culous mycobacteria (NTM) has been widely used and
11.13 Summary 178
accepted after the American Thoracic Society adopted
References 178
this term in its official statement about these myco-
bacteria (ATS 1990), NTM itself is a kind of misnomer.
These mycobacteria usually do result in the formation
11.1 of tubercles in infected tissue; in other words, they are
Introduction 'tubercle-ous' (Grange 1989). Furthermore, the British
Thoracic Society still uses the term 'opportunist myco-
The Norwegian physician G. H. Armauer Hansen bacteria'in its official guidelines (BTS 2000).
described an organism subsequently known as Myco- In this chapter, we will review NTM disease in
bacterium leprae in the skin of leprosy patients for terms of classification, epidemiology, pathogenesis,
the first time in 1874 (Hansen 1874). Eight years later, diagnosis, and treatment. In addition, the state of
the German physician Robert Koch identified Myco- knowledge of host susceptibility and related cytokine
bacterium tuberculosis. In 1885, 3 years after Koch's therapy will be underscored.
discovery of M. tuberculosis, a smegma bacillus, erro-
neously thought to be connected to syphilis but now
11.2
J.-J. YIM, MD Classification of Nontuberculous
Respiratory and Critical Care Medicine, Seoul National Uni-
Mycobacteriosis
versity College of Medicine, Seoul, South Korea 110744
S. M. HOLLAND, MD
Immunopathogenesis Section, Building 10, Room 11N103, Growth rates, colony morphologies, and biochemi-
10 Center Dr. MSC 1886, Bethesda, MD 20892 1886, USA cal tests have traditionally been used to differentiate

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
164 J.-J. Yim and S. M. Holland

NTM from M. tuberculosis. M. tuberculosis takes
3-6 weeks to grow, has white colonies, and gives a
positive niacin test result; most NTM are niacin-
negative (Wayne 1985). In 1954, Timpe and Runyon
classified NTM for the first time (Timpe and Runyon
1954). Their classification was based on NTM growth
rates, morphology, and colony pigmentation. Accord-
ing to their scheme, if a mycobacterium took?.7 days
to grow, it was a 'slow grower' and fell into groups I,
II, or III. If it took <7 days, it was a 'rapid grower' and
was classified as group IV. Among the slow growers,
if colonies showed yellow pigmentation when grown
in the light, they were photochromogens, type I. If
they were strongly pigmented, yellow to orange
when grown in the dark, they were scotochromogens,
type II. If they showed no or shallow pigmentation
when grown in the dark, they were classified as type
III, nonphotochromogens. In fact, group IV, the rapid
growers, shares many similarities with group III, the
nonphotochromogens, except that the individual colo-
nies grow in <7 days.
Although this Runyon classification system pro-
vided the first basis for a taxonomic understanding
of mycobacteria, it has limitations. It is not particu-
larly relevant to the clinical manifestations of NTM
diseases, and some bacteria cannot be clearly clas-
sified. In fact, Runyon himself proposed that some
mycobacterial species from different categories could
share similar clinical manifestations (Runyon 1974).
Consequently, the classification of NTM according to
their predisposition to involve various organs has
gained popularity (Table ILl).
Over the last decade, the number of new species
added to the NTM has increased dramatically (Wayne
and Sramek 1992). Correct identification of these
newly added NTM species by traditional culture-based
and biochemical methods has become more challeng-
ing and is sometimes inadequate (Springer et al.I993).
The development of new techniques, such as 16S rRNA
gene sequencing (Sansila et al. 1998), PCR-RFLP of a
439 bp segment of the heat shock protein 65 (hsp-65)
gene (Eriks et al. 1996), or high-performance liquid

Table 11.1. Classification of the nontuberculous mycobacteria recovered from humans

Features of the common species

Clinical disease Common etiologic Geography Morphologic features Unusual etiologic

species species

Pulmonary disease M. avium complex Worldwide Usually not pigmented M.simiae

Slow growth (>7 days) M.szulgai
M. kansasii USA, coal mining Pigmented; often large and beaded M. fortuitum
regions, Europe on acid-fast stain M. celatum
M. abscessus Worldwide Rapid growth «7d) M. asiaticum
but mostly USA not pigmented M.shimodii
M. xenopi Europe, Canada Slow growth; pigmented M. haemophilum
M. malmoense UK, northern Europe Slow growth; not pigmented M. smegmatis

Lymphadenitis M. avium complex Worldwide Usually not pigmented M. fortuitum

M. scrofulaceum Worldwide Pigmented M. chelonae
M. malmoense UK, northern Europe Slow growth M. abscessus
(especially Scandinavia)
Cutaneous disease M. marinum Worldwide Photochromogen; requires low M. avium complex
temperatures (28-30°C)
For isolation M. kansasii
M. fortuitum Worldwide, mostly USA Rapid growth; not pigmented M. nonchromogenicum
M. chelonae M. smegmatis
M. abscessus M. haemophilum
M. ulcerans Australia, tropics, Africa, Grows slowly, pigmented
SE Asia
Disseminated disease M. avium complex Worldwide Isolates from patients with AIDS M. abscessus
usually pigmented (80%) M.xenopi
M. malmoense
M. kansasii USA Photochromogen M. genavense
M. chelonae USA Not pigmented M.simiae
M. haemophilum USA, Australia Not pigmented; requires hemin, M. conspicuum
often low temperatures, and CO 2 M. marinum
to grow M. fortuitum

From ATS (1997)

Nontuberculous Mycobacteria 165

chromatography (HPLC) mycolic acid fingerprint- Table 11.2. Newly described or emerging human species of
ing (Butler and Kilburn 1990; Butler et al. 1992) have nontuberculous mycobacteria
provided identifications and are in part responsible Newly described clinically significant human pathogens
for the recent explosion of new species of NTM. Newly
Rapid growing mycobacteria:
described or emerging human species of NTM are M. mucogenicum
listed in Table 11.2. M. goodii
M. immunogenum
M. wolinskyi
M. fortuitum third biovariant complex,
including M. septicum
11.3 Slowly growing mycobacteria:
Epidemiology M. genavense
M. triplex
After the first report of disseminated NTM infection M. lentiflavum
in an AIDS patient in 1982 (Zakowski et al.1982), the M. celatum
M. interjeetum
number of patients with NTM disease dramatically M. intermedium
increased in developed countries (Selik et al. 1987; M. conspicuum
Horsburgh and Selik 1989), even though the preva- M. branderi
lence of NTM disease in AIDS patients in Africa is M. geidelbergense
still low (Okello et al.1990; Morrissey et al.I992). The M. heckeshornense
M. bohemicum
epidemic of AIDS not only increased the number of M. tusciae
cases of NTM disease, it also changed its characteris- Newly described potential human pathogens:
tics. Before the AIDS epidemic, NTM disease was usu- M. confluentis
ally confined to the lungs, lymph nodes, or skin; only M. mageritense
in rare cases was it disseminated (Wolinsky 1979). M. alvei
M. hassiacum
However, NTM disease in AIDS patients is usually M. kubicae
disseminated. In addition, the most common dissem- M. novocastrense
inated NTM in AIDS patients is M. avium, (Zakowski M. brumae
et al. 1982), whereas infection by M. kansasii and M. Old but emerging pathogens:
intracellulare was more frequent in the era before the M. ulcerans
M. haemophilum
AIDS epidemic (Ahn et al. 1979; Falkinham 1996). M. xenopi
Irrespective of the change related to the AIDS M. malmoense
epidemic, the NTM which cause lymphadenitis in
Adapted from Brown-Elliott et al. (2002) with permission
children have also changed: M. scrofulaceum was the
most frequent organism causing cervicallymphadeni-
tis in children (Wolinsky 1979), but M. avium has now
become the most common (Colville 1993; Wolinsky infection of AIDS patients from the hot water system
1995). This observation raises the possibility that the of hospitals (von Reyn et al. 1994). Although MAC is
natural or environmental population of M. avium and one of the major causes of mycobacteriosis in poultry
M. scrofulaceum is changing, that there is a change in and swine, recent serologic and molecular studies sug-
children's behavior, or that other factors (e.g., inciden- gest a low possibility of animal-to-human transmission
tal antibiotic use) may be involved (Falkinham 1996). (Meissner et al. 1977; Ahrens et al. 1995; Guerrero et al.
Water is the major source of infection for numerous 1995). The waterborne M. marinum has been reported
NTM species including M. avium complex (MAC), M. to infect persons through aquarium exposure (Adams
marinum, M. kansasii, M. xenopi, M. simiae, M. fortui- et al. 1970; Barrow and Hewitt 1971), and M. xenopi has
tum, M. chelonei, M. abscessus, and others. MAC grows been suggested to infect humans through an aerosol
well in the natural waters ofGermany (Beerwerth 1973), route (Collins and Yates 1984). In addition, the rapidly
Japan (Ichiyama et al. 1988), Finland (von Reyn et al. growing mycobacteria M. fortuitum, M. chelonae, and
1993), Uganda (von Reyn et al. 1993), Zaire (von Reyn M. abscessus have been reported to cause nosocomial
et al. 1993), the USA (Gruft et al. 1981), and probably outbreaks and pseudo-outbreaks from tap water
many other untested regions. MAC can be aerosolized (Lowry et al. 1988; Maloney et al. 1994), ice prepared
(Meissner and Falkinham 1986), raising the probability from tap water (Laussucq et al. 1988), processed tap
of airborne infection for these mycobacteria. In 1994, water for dialysis (Bolan et al. 1985), and the whirlpool
von Reyn et al. suggested the transmission of MAC footbaths of nail salons (Winthrop et al. 2002).
166 J.-J. Yim and S. M. Holland

Even though many mycobacteria including MAC the rapidly growing NTM M. fortuitum, M. chelonae,
(Paull 1973; Reznikov and Leggo 1974; Reznikov and and M. abscessus (Wolinsky 1979; Griffith et al. 1993;
Dawson 1980; Brooks et al. 1984; Eaton et al. 1995), Hadjiliadis et al. 1999).
M. malmoense (Saito et al. 1994), and M. fortuitum Although disseminated M. avium infection has
(Wolinsky and Rynearson 1968; Paull 1973) have been been common among patients with advanced HIV
isolated from soil, there is no evidence that mycobacte- disease, pulmonary MAC disease has occurred infre-
ria from soil can directly infect human beings. quently in these patients. The frequency of MAC lung
Although the prevalence of NTM disease cannot be disease in patients with disseminated MAC infection
estimated easily, it has been suggested that the preva- is low, between 2.5% andlO% (Kalayjian et al. 1995).
lence of NTM disease is now increasing in the USA This may reflect differences in organism tropism and
(O'Brien et al.1987; Pae et al.1997), Switzerland (Deb- virulence, as bacteremic strains in AIDS patients
runner et al. 1992), and Japan (Sakatani 1994). The tend to be M. avium, whereas M. intracellulare is
reasons for this increase ofNTM disease are unknown, the more common pulmonary pathogen. Horsburgh
but better clinical recognition and improved, increased et al. studied advanced AIDS patients with CD4+ T-
amounts of culturing for NTM are felt to play impor- lymphocyte counts <25/mm 3 and found that the con-
tant roles (ATS 1997). MAC is the most prevalent sumption of hard cheese and pate were associated
species followed by M. kansasii (O'Brien et al. 1987; with recovery of MAC from stool, whereas consump-
Debrunner et al.1992; Sakatani 1994). tion of raw shellfish was associated with recovery of
The NTM are widely thought to be opportunistic. MAC from sputum (Horsburgh et al. 1994).
Those patients most commonly afflicted are ones
with structural lung disease (e.g., bronchiectasis,
pneumoconiosis, chronic obstructive pulmonary
disease [COPDj, prior tuberculosis) (Wolinsky 1979;
O'Brien et al. 1987; Pae et al. 1999). Bronchiectasis 11.4
and MAC infection often coexist, making causality Host Susceptibility Factors
difficult to determine in some cases. Specific mor-
photypes have been associated with pulmonary NTM In 1964, Engbaek reported the first familial outbreak
infection, particularly in women with pectus excava- of fatal MAC infection (Engbaek 1964), suggesting a
tum, thoracic scoliosis, and mitral valve prolapse genetic basis for the development of severe disease
(Iseman et al. 1991). Pulmonary NTM disease has with these relatively nonpathogenic organisms. In
been shown to have a growing female predominance, the 1980s, advanced HIV infection showed the criti-
including a significant number of older women with- cal role of CD4+ T-Iymphocytes as effectors against
out clearly recognized predisposing factors (Prince et mycobacteria: NTM infections typically occurred
al. 1989; Griffith et al. 1993). The majority of elderly only after the CD4+ T-Iymphocyte number fell below
3
male patients have risk factors such as smoking or 50/mm (Horsburgh and Selik 1989; ATS 1997), sug-
COPD (Iseman 1996). In contrast to male smokers gesting that specific T-cell products or activities were
with upper lobe cavitary disease who tend to carry required for mycobacterial resistance. The genetic
the same single strain of MAC indefinitely, female basis of susceptibility to NTM has been characterized
nonsmokers with nodular bronchiectasis tend to through identification of specific mutations involv-
carry 3 or more strains over time. This suggests either ing the interferon-y (IFNy) synthesis and response
polyclonal infection or recurrent infection with dis- pathways (Table 11.3). These include mutations in
tinct strains (Wallace et al.1998). Patients with cystic interferon-y receptor 1 (IFNyR1), interferon-y recep-
fibrosis, in whom there are both structural lung dis- tor 2 (IFNyR2), interleukin-12 receptor ~1 subunit
orders (Olivier et al. 1996) and abnormalities in the (IL12R~1), interleukin-12 p40 (IL12p40), the signal
function of defensins, small antibacterial peptides transducer and activator of transcription 1 (STATl),
secreted by bronchial epithelium (Goldman et al. and NFKb essential modulator (NEMO), an X-linked
1997), form a special risk group for pulmonary NTM trait that affects IL-12 production (Dorman and Hol-
infection. Patients with pulmonary alveolar pro- land 2000; Doffinger et al. 2001; Dupuis et al. 2001).
teinosis are also at higher risk for pulmonary NTM However, despite the growing number of genetic
infections as well as infection with the closely related defects identified in patients with disseminated NTM
Nocardia (Witty et al. 1994). Esophageal motility dis- infections, only about one-third of cases without HIV
orders such as achalasia have been found to be sig- infection have a genetic diagnosis. Therefore, almost
nificantly associated with pulmonary disease due to two-thirds of all disseminated cases to date are still
Nontuberculous Mycobacteria 167

Table 11.3. Human identified immune defects

Defect Location Zygosity Infections
(disseminated mycobacterial disease)

c-IFNJ'RI Sl16X Homozygous M. avium, Salmonella, M. chelonei,

M. fortuitum
131delC Homozygous BCG
22delC Homozygous M. avium
201-2A~G Homozygous M. avium
107ins4; 200+ 1 G~A Heterozygous M. smegmatis, BCG?
561de14; 373+1 G~T Heterozygous BCG, M. avium, M. kansasii,
Listeria monocytogenes
561 del4 Homozygous BCG

c-IFNJ'R2 278delA,G Homozygous M. avium, M. fortuitum

791deiG Homozygous BCG? M. abscessus

AR p-IFNJ'Rl 1871 Homozygous BCG, S. enteritidis, 1. pneumophilia?,

M. tuberculosis?

AR p-IFNJ'R2 R1l4C Homozygous BCG, M. abscessus

AD p-IFNJ'RI 818del4 Heterozygous BCG, M. avium, M. kansas ii, M. spp.,

H. capsulatum
818delT Heterozygous BCG, M. avium
817insA Heterozygous M. avium, M. kansasii, M. chelonei

c-IL-12 p40 p40de14.4 Homozygous BCG, S. enteritidis

c-IL-12RPI de1409-549 Homozygous BCG, S. typhimurium

Q32X Homozygous M. avium, S. paratyphi
Q376X Homozygous M. avium, S. group B
K305X Homozygous BCG, S. enteritidis
783+1 G~C Homozygous BCG
Q214R Homozygous M. avium, S. enteritidis

STATl L706S Homozygous BCG, M. avium

NEMO X420W X-linked M. kansasii, M. sp., S. enteritidis,

S. pneumoniae, P. carinii
1167insC M. avium, P. aeruginosa, CMV
1218insA H. influenzae, S. pneumoniae
D3l1N M.avium
nd S. pneumoniae

R175P S. pneumoniae
L227P
C417F S. pneumoniae
C417R H. influenzae
A288G S. pneumoniae

Polymorphism:
NRAMPI 5'(CA)n 199/other Heterozygous Pulmonary tuberculosis
INT4G~C Heterozygous Pulmonary tuberculosis
D543NG~A Heterozygous Pulmonary tuberculosis
3'UTR TGTG+/del Heterozygous Pulmonary tuberculosis

Abbreviations: c-IFNJ'Rl, complete interferon-y receptor 1 deficiency; c-IFNJ'R2, complete IFNJ'R2 deficiency; AR p-IFNJ'Rl,
autosomal recessive partial interferon-y receptor 1 deficiency; AR p-IFNJ'R2, autosomal recessive partial IFNJ'R2; AD p-IFNJ'Rl,
autosomal dominant partial interferon-y receptor 1; c-IL-12 p40, complete interleukin-12 p40 deficiency; c-IL-12RPl, complete
interleukin-12 receptor ~ 1 deficiency; STATl, signal transduction and activation oftranscription factor 1; NEMO, NFKB essential
modulator; NRAMPl, natural resistance-associated macrophage protein 1; BCG, bacille Calmette-Guerin. Adapted from Guide
and Holland (2002) with permission.
168 J.-J. Yim and S. M. Holland

syndromic (e.g., sarcoid, CD4+ T-Iymphocytopenia) and Scott 1999). Binding results in phosphorylation
or idiopathic. Interestingly, an association has been of the associated cytoplasmic kinases, tyrosine
noted between pulmonary NTM infections and body kinase 2 (Tyk2) and Janus kinase 2 (Jak2), which in
habitus, predominantly in postmenopausal Cau- turn phosphorylate STAT4, leading to transcription
casian women (pectus excavatum, scoliosis, mitral of IL-12-responsive genes, in particular IFNy. IFNy
valve prolapse) (Iseman et al. 1991; Pomerantz et activates neutrophils and macrophages to produce
al. 1996). superoxide and nitric oxide, to increase the sur-
face display of MHC molecules and Fc receptors, to
decrease lysosomal pH, and to increase the intracel-
lular concentration of certain antibiotics (Bermudez
11.5 and Young 1989; Boehm et al. 1997). IFNy signals
Host Defense Against Mycobacteria through a membrane-bound receptor composed
of two chains, IFNyRI and IFNyR2, which are con-
In normal hosts, mycobacteria are initially phago- stitutively bound to their respective Janus kinases,
cytosed by macrophages, which respond with the Jakl and Jak2. IFNy binding initiates aggregation of
production of interleukin-12 (lL-12; see Fig. ILl) IFNyRI and IFNyR2, phosphorylation of Jakl and
(Fulton et al. 1996). IL-12 is a complex heterodimer Jak2, and phosphorylation of the IFNyRI cytoplas-
composed of p35 (chromosome 3) and p40 (chro- mic domain, which enables docking of STATl. Phos-
mosome 5) moieties, which together comprise p70. phorylated STATl (STATl-P) then homodimerizes
IL-12p70 activates T-Iymphocytes and NK cells and translocates as a complex to the nucleus, where
through binding to the IL-12 receptor, composed of it upregulates IFNy-responsive genes such as IL-12,
IL-12R~1 and IL-12R~2 (Gately et al. 1998; Trinchieri major histocompatibility complex genes, and TNF .
The positive feedback loop between IFNyand IL-12 is
pivotal in the immune response to mycobacteria and
other intracellular infections. Defects in any of these
receptors or cytokine genes negatively affect the
production of IFNy and/or IL-12, and consequently
enhance mycobacterial susceptibility (Fig. 11.1)
IL-12 (Dorman and Holland 2000).
In mouse studies, IFNy has been shown to exert a
mycobacteriostatic effect, possibly through acidifica-
tion of mycobacteria-containing vesicles (Appelberg
and Orme 1993). Toxic oxygen radicals do not control
NTM infection in mice (Doherty and Sher 1997), but
they may be important in early tuberculosis infection,
TNFa especially in the lungs (Adams et aI. 1997; Cooper et
al. 2000). Although in vitro data for IFNy control
of NTM using macrophages has been inconsistent
(Douvas et al. 1985; Bermudez and Young 1989; Denis
et al. 1990), the in vivo data have been persuasive.
Fig. ILL Cytokine pathways involved in mycobacterial infec-
Disruption of IFNy production in mice resulted in
tion and host response. Mycobacteria infect macrophages and
stimulate production of IL-12. Subsequently, IL-12 leads to T Widespread tuberculosis with poor granuloma for-
or NK cell production of IFNy. IFNy binds to a specific site on mation and rapid death (Flynn et al. 1993; Cooper
IFNyRl, leading to aggregation of IFNyRl and IFNyRZ. Follow- et aI. 2000). Exogenous administration of IFNy did
ing this aggregation, the associated cytoplasmic Janus kinases, not restore normal resistance in these animals, sug-
Jakl and Jak2, transphosphorylate each other, leading to phos-
phorylation of the intracellular domain of IFNyRl. This phos-
gesting that IFNy may playa critical developmental
phorylation creates a binding site for cytosolic STAT!, which is role, or that injecting systemic IFNy does not activate
in turn phosphorylated. Phospho-STAT! then homodimerizes local defenses in the same manner as local produc-
and is transported to the nucleus, where it upregulates IFNy- tion. Similar results were seen in IFNyR-deficient
responsive genes. IFNy also upregulates TNFa production in mice (Kamijo et al. 1993) and in mice lacking the
the setting of endotoxin (lipopolysaccharide). The mechanisms
IFNy-induced interferon regulatory factor-l (IRF-l)
by which IFNy mediates the killing of mycobacteria are not yet
well characterized. *, denotes genes known to predispose to (Kamijo et al. 1994). These models define a critical
mycobacterial infection when mutated role for IFNy in tuberculosis resistance and control in
Nontuberculous Mycobacteria
the mouse. Similar findings have been reported with
MAC infection in IFNy-deficient mice (Doherty and
Sher 1997). In normal mice infected with MAC, exog-
enously administered IFNy successfully augmented
mycobacterial clearance, in part by suppressing the
production of prostaglandin E2, an inhibitor of mac-
rophage function (Edwards et al. 1986).
11.6
Human Immune Defects
11.6.1
IFNy Receptor1 Deficiency
In 1994, four children of an extended Maltese family
were described with an autosomal recessive pattern
of susceptibility to NTM (Levin et al. 1995). These
related children had severe disseminated disease
with a variety of NTM including M. chelonae,
M. fortuitum, and MAC, in addition to recurrent
infections with Salmonella. Affected children had
impaired TNFa production in response to stimula-
tion by IFNy, while their parents were intermediate
between the patients and normals, suggesting a gene
dosage effect in heterozygotes. Antigen presentation
was also impaired (D'Souza et al.1996). Newport et al.
identified a homozygous point mutation in the extra-
cellular domain of IFNyRI in these patients, result-
ing in a premature stop codon (Newport et al. 1996).
Simultaneously, a Tunisian infant was reported with
disseminated BCG infection and a more proximal
extracellular chain-terminating mutation (131delC)
in the IFNyRI gene (Jouanguy et al. 1996). Since
then, numerous autosomal recessive mutations in the
extracellular domain of IFNyRI have been identified
in patients with disseminated mycobacterial infec-
tions. Identification of compound heterozygotes in
outbred populations indicates that mutations in
IFNyRI are not simply the result of remote founder
effects (Dorman and Holland 2000).
Extracellular mutations leading to disruption of
IFNyRI function lead to failure to activate STATl
in response to IFNy. This leads to an inability to
upregulate IFNy-responsive genes including TNFa.
Since IFNy and IL-12 production are co-dependent,
both show an impaired response to phytohemagglu-
tinin (PHA) stimulation (Holland et al. 1998). How-
ever, complete IFNyRI or IFNyR2 deficiency leads to
loss of IFNy removal from the plasma and elevated
circulating levels of IFNy, despite low levels of IFNy
generation in vitro (Fieschi et al. 2001).
169
Complete IFNyRI deficiency typically presents
as disseminated BCG or NTM disease in infancy or
childhood. Salmonella infections have occurred in
cases from outside North America, probably due
to the greater exposure to that organism in those
parts of the world. Complete IFNyRI deficiency is
associated with a poor prognosis and high mortal-
ity due to incomplete clearance of infection despite
aggressive antibiotic therapy, as well as continued
susceptibility to new infections. Other infections
reported in these patients include Listeria monocy-
togenes, cytomegalovirus (CMV), severe respiratory
syncytial virus (RSV), herpes simplex virus (HSV),
and varicella zoster virus (VZV). Although these are
common viruses prevalent in the general popula-
tion, these infections are often much more severe in
IFNyRI-deficient patients, suggesting that this may
be an aspect of the clinical phenotype (Dorman et
al. 1999). Two Portuguese siblings with disseminated
BCG and presumed tuberculosis were identified with
a homozygous mutation in the extracellular domain
of IFNyRI (I87T) leading to an alteration in a N-
linked glycosylation site (Jouanguy et al. 1997). They
were both cured with antibiotics alone and showed
partial IFNy responsiveness in vitro. The formation
of intact, tuberculoid granulomas by one child with
disseminated BCG further demonstrated residual
IFNyactivity.
Autosomal dominant (AD) mutations in the
IFNyRI gene have been shown to confer only a par-
tialloss of IFNy responsiveness. Jouanguy et al. iden-
tified a four-base deletion (818deI4) in 12 unrelated
people from around the world (Jouanguy et al. 1999),
and another 2 Japanese patients with the same muta-
tion were reported recently (Sasaki et al. 2002). This
latter group also identified less common autosomal
dominant mutations, 818delT and 818insA. These
mutations result in deletions of the Jakl and STATl
intracellular binding sites as well as deletion of the
receptor recycling domain. These mutations impede
signal transduction and receptor removal from the
cell surface while preserving the receptor binding
capacity. Autosomal dominant IFNyRI disease has a
milder phenotype, lower mortality rate (<5%), and is
associated with a later onset of infection and less like-
lihood of dissemination than complete ARIFNyRI
mutations. These ADIFNyRl patients often develop
mycobacterial osteomyelitis; other identified infec-
tions include histoplasmosis, herpesvirus infections,
and Salmonella.
170
11.6.2
IFNy Receptor 2 Deficiency
Complete IFNyR2 deficiency shows similar clini-
cal features, histopathology, and in vitro responses
to complete IFNyRI defects. Dorman and Holland
(l998, 2000) identified a frameshift mutation in
the IFNyR2 gene of a child with disseminated M.
fortuitum and M. avium complex (MAC) infections
resulting in a premature stop codon which caused
IFNy unresponsiveness. A child with an extracellular
domain mutation of IFNyR2 causing a partial defect
leading to reduced IFNy responsiveness but normal
surface IFNyR2 expression was also reported (Doffin-
ger et al. 2000).
11.6.3
IL-12 Receptor ~1 Deficiency
In 1998, mutations in the IL-12R~1 receptor were
identified in several unrelated cohorts (Altare et
al. 1998a; de Tong et al. 1998). These patients were
infected with BCG, NTM, or Salmonella. In contrast
to the patients with disabling mutations in IFNyRI
or IFNyR2, patients with mutations in IL-12R~1
showed intact and mature-appearing granulomas,
later onset of infections, and much higher rates
of survival than did ARIFNyRI-deficient patients.
Most of these patients were treated successfully with
antibiotics and IFNy. However, despite aggressive
therapy, several fatal cases of disseminated BCG in
IL-12R~I-deficient patients have been reported from
Turkey (Sanal et al. 2000). In vitro, despite a complete
lack of expression of IL-12R~I, the cells maintained
a normal proliferation to mitogen or antigen (PPD)
and normal IFNy responsiveness. However, defective
IL-12 signaling was exhibited in terms of NK activ-
ity, proliferation, and poor IFNy production (1-10%
of normal). To date, most of the reported cases have
been homozygous autosomal recessives, even in
the absence of reported consanguinity, presumably
reflecting founder effects in some populations. No
mutations in IL-12R~2 have been found yet, although
they are predicted to be similar in severity and theo-
retically as likely as mutations in IL-12R~1. However,
IL-12R~1 is also one of the critical chains for the IL-
23 receptor, suggesting that IL-12R~1 deficiency also
confers IL-23 unresponsiveness.
I.-I. Yim and S. M. Holland
11.6.4
IL-12 p40 Deficiency
IL-12p40 is the first identified cytokine gene defect
affecting mycobacterial susceptibility (Altare et al.
1998b). Autosomal recessive inheritance of the IL-
12p40 defect was demonstrated in a child with con-
sanguineous Pakistani parents. The patient developed
postvaccine disseminated BCG infection followed by
disseminated Salmonella enteritidis. Both IFNy recep-
tors 1 and 2 were intact along with the IL-12R~1 chain
and a functional IL-12R. Similar to patients with IL-
12R~1 deficiency, this child's cells produced about 10%
of normal IFNy but demonstrated normal TNFa and
IL-8 production. Aggressive antibiotic therapy and two
courses of IFNy (currently still on IFNy) were required
for successful control of the Salmonella infection. The
father had also had recurrent, severe Salmonella infec-
tions in his youth, raising the possibility of haploinsuf-
ficiency in heterozygous carriers.
11.7
NFKB Essential Modulator (NEMO)
or IKB Kinase Defects
Recently, mutations in one of the proteins required
for nuclear factor kappa B (NF-KB) activation have
been shown to cause immunodeficiency associated
with hypohidrotic ectodermal dysplasia (HED-ID)
(Zonana et al. 2000; Doffinger et al. 2001; Orange et
al. 2002). The NFKB essential modulator (NEMO) or
IKKy encodes the regulatory subunit of the IKB kinase
(IKK) complex, which is required for the activation of
the NF-KB signaling pathways and immune defense.
NF-KB regulates the transcription of genes encod-
ing inflammatory proteins and antiapoptotic factors.
Disruptions in NEMO that result in truncation of
the protein or alteration of the zinc finger domain
result in impaired NF-KB activation and a distinct
physical phenotype of conical teeth, poor sweat gland
formation, and abnormal hair growth (hypohidrotic
ectodermal dysplasia) along with recurrent infections.
Gram-positive, gram-negative and mycobacterial
infections, bronchiectasis, and poor febrile responses
are characteristic features of this disease. The defect
in NEMO affects not only the signaling through
the ectodysplasin/dysplasin pathway, explaining the
ectodermal dysplasia phenotype, but also signaling
through the TNFR, IL-IR, IL-18R, CD40, and Toll-like
receptors (TLR), thereby explaining the infection sus-
ceptibility and lack of fever.
Nontuberculous Mycobacteria
11.8
Familial Syndrome
A familial syndrome has been identified in which two
generations of previously healthy women developed
CD4+ T-Iymphocytopenia, disseminated NTM infec-
tion, and subsequent myeloid leukemia (Holland et
al. 1994). Affected individuals displayed a similar
clinical phenotype consisting of marked monocyto-
penia, indolent mycobacterial disease, and progres-
sive lymphocytopenia prior to the development of
myeloproliferative disease. This suggests an autosomal
dominant common genetic defect that controls myco-
bacterial susceptibility and myeloid regulation.
Holland et al. (1994) also reported a family with
disseminated MAC infection in males from two
generations. This family had abnormal regulation of
IL-12, poor IFNy production, and clinical improve-
ment with IFNytreatment (Frucht and Holland 1996;
Frucht et al. 1999). These patients have recently been
shown to have a novel mutation in NEMO (E315A).
11.9
Polymorphisms
Aside from the specific defects described above, there
is a growing interest in the role for genetic polymor-
phisms in mycobacterial susceptibility. In mice,
polymorphisms in the natural resistance-associated
macrophage protein (NRAMPl) gene were found to
influence early, but not late mycobacterial suscepti-
bility (Vidal et al. 1995). NRAMPI is expressed in
activated macrophages and is clearly responsible
for the higher bacterial burdens found 3 weeks after
infection in BCG-susceptible mouse strains. However,
by 6 weeks post infection, both susceptible and resis-
tant strains have cleared the infection. The natural
occurrence of this dichotomous trait in mice is due
to a single amino acid change (Vidal et al. 1995).
NRAMPI encodes an ion transporter that localizes
to the lysosomal membrane during phagocytosis of
mycobacteria as well as other pathogens. It is thought
that NRAMPI may restrict organism replication
through alteration of the phagolysosomal environ-
ment by regulating iron or other divalent cation
transport (Supek et al. 1996; Gruenheid et al. 1997).
The NRAMPI gene clearly plays a critical role in the
early inflammatory response and early control of
mycobacteria but may not influence the course of the
disease later on, as demonstrated by the similar time
to clearance of infection regardless of the NRAMPI
171
status of the strain. A study of the human orthologue
of NRAMPI in a West African population with high
rates of tuberculosis found significant associations
of several polymorphisms with the development of
tuberculosis (Bellamy et al.I998). However, in a small
study of eight older women with pulmonary MAC
infection, none of the described defects or polymor-
phisms in NRAMPI was found (Huang et al. 1998).
A functional polymorphism in an intron of the
IFNy gene has been identified that is correlated with
decreased IFNy production, but this has not yet been
linked to any infection susceptibility (Pravica et al.
1999). However, almost all patients with dissemi-
nated MAC infection produce low levels of IFNy in
response to in vitro cell stimulation by the nonspecific
T-cell mitogen PHA (Holland et al. 1994). No cases of
genetic IFNy deficiency have been identified to date.
Many more mutations and polymorphisms have been
identified in cytokine receptors than in the cytokines
themselves, although the reason for this is unclear.
11.10
Body Morphotype
Many patients with pulmonary NTM infections
appear to have similar clinical characteristics and
body type, including scoliosis, pectus excavatum,
mitral valve prolapse (MVP), and joint hypermobility
(Iseman et al. 1991; Huang et al. 1999). Many of these
physical characteristics are reminiscent of individuals
with heritable collagen-vascular diseases (Bruno et al.
1984; Pyeritz and McKusick 1979; Grahame and Child
1984; Glesby and Pyeritz 1989; Roman et al.1989; Spon-
seller et al. 1995) or the hyper-IgE recurrent infection
syndrome (Job's syndrome) (Grimbacher et al. 1999).
Therefore, patients with pulmonary NTM infections
may possess a subtle connective tissue defect. In this
context, phenotypic abnormalities may be the con-
sequence of a genetic defect that affects both body
morphotype and infection susceptibility.
The relationship between immunity and body mor-
photype has not been well characterized, but certain
associations have been noted. Pulmonary abnormali-
ties including spontaneous pneumothoraces, general-
ized emphysema, cystic lung disease, bronchiectasis,
and increased susceptibility to pneumonia have long
been noted in patients with Marfan syndrome (Guide
and Holland 2002). Patients with cystic fibrosis (CF),
who have been noted to develop NTM infections later
in life (Olivier et al. 1996), showed an increased inci-
dence of spinal curvature as well (Erkkila et al. 1978).
172 J.-J. Yim and S. M. Holland

Patients with Job's syndrome suffer from recurrent more important for the diagnosis of a mycobacterial
pulmonary infections and have recently been shown to disease than for routine bacterial infection, since labo-
have a characteristic phenotype that includes scoliosis, ratories typically do not look for mycobacteria with-
joint hypermobility, and a high arched palate (Grim- out a specific request. In untreated cavitary disease,
bacher et al. 1999). In a comprehensive phenotype sputum smears and cultures are usually positive.
study of Job's patients, Grimbacher et al. found that The placement of mycobacterial skin tests usually
76% of patients had scoliosis, 68% had hyperextensible implies anticipation of M. tuberculosis infection, not
joints, and 57% had recurrent fractures (Grimbacher et nontuberculous infection. In studies from the early
al. 1999). The paradigm of Job's syndrome, in which a 1980s, about 50% of patients with pulmonary NTM
distinct and somewhat similarbody morphotype is rec- infection had positive skin tests with purified protein
ognized in the setting of recurrent infections, suggests derivative (PPD), the antigen filtrate derived from
that an association between phenotype and immunity MTB (O'Brien et al. 1987). Recently, von Reyn and
may also exist in patients with NTM infections. colleagues have revived the use of an antigen derived
from M. avium, the M. avium sensitin (MAS), and
compared it with testing with PPD in patients with
pulmonary MAC or MTB. They looked for dominance
11.11 of one skin test over the other when both MAS and
Clinical Presentation and Diagnosis PPD are placed simultaneously (>5 mm reaction and
>3 mm larger than the heterologous skin test reac-
11.11.1 tion). MAS-dominant skin tests were 97% specific for
Pulmonary Disease discriminating MAC from MTB disease (von Reyn et
al. 1998). MAS is under trial but not yet available in
Clinically, pulmonary NTM infection is usually milder the USA; it is used extensively in Scandinavia.
than active pulmonary TB, although the destructive The radiographic appearance of NTM disease in the
pulmonary changes and chest radiographs can be lung can be similar to TB, with extensive cavity forma-
indistinguishable. Symptoms can persist for months tion, nodules, and airspace disease (Fig.ll.2). Discrimi-
to years before radiographs are taken or mycobacterial nating aspects of pulmonary MAC disease compared
disease emerges in the differential diagnosis. Fatigue with TB are disease limited to the right middle lobe,
is usually the most prominent symptom and one that disease limited to either lower lobe, or combined simul-
is quite troublesome to the patient. Cough productive taneous disease in the right middle lobe and lingula
of mildly purulent sputum may be severe, but it is (Table 1104). Diffuse bronchiectasis involving four or
usually simply annoying, worse in the morning or on more lobes of the lung is also more common for pulmo-
recumbency, and tolerated fairly well. Hemoptysis is nary MAC disease, whereas pleural calcification is more
not usually serious but may occur in the course of dis- suggestive of MTB (Lynch et al.1995). The combination
ease in up to one-third of patients (Griffith et al.1993). of bronchiectasis and lung nodules seen by chest CT
Night sweats and fever eventually occur in the major- scanning had a sensitivity of80% for predicting positive
ity; weight loss is common but not severe. Concurrent sputum cultures for MAC (Swensen et al.1994).
bacterial infection, especially with Pseudomonas aeru- The American Thoracic Society has proposed
ginosa, is not uncommon and may greatly exacerbate formal guidelines for the diagnosis of NTM disease,
cough, sputum production, and fatigue. Various sapro- initially in 1990 (ATS 1990) and then revised in 1997
phytic fungi are frequently cultured from the sputum, (Table U.5) (ATS 1997). Although these are valuable
but their influence on the pace or severity of disease is guidelines for the diagnosis of NTM disease, cautious
undefined. The tempo of the infection is rarely rapid, interpretation of the radiologic or bacteriologic find-
and changes may be seen on radiographs evolving ings in the context of the clinical setting and follow-up
over years if left untreated. Dyspnea is typically seen observation should be underscored. As we recognize
only late in disease when extensive destruction has more cases of pulmonary NTM infection, and the rela-
occurred. The physical examination is not very infor- tive frequency of pulmonary TB in the USA declines,
mative, usually does not reflect the activity or quies- these guidelines may be unnecessarily restrictive in
cence of mycobacterial disease, and is often relatively terms of the criteria required for the diagnosis of an
normal, despite extensive involvement. active NTM infection. In most cases, if a patient has a
The first step toward the diagnosis of pulmonary positive culture for NTM in the context of a compatible
nontuberculous mycobacterial infection is to include history and radiograph, it is likely to be pulmonary
that entity in the differential diagnosis. This is much NTM infection and not simple colonization.
 Nontuberculous Mycobacteria 173

Table 11.4. Radiographic appearance of mycobacterial disease

Feature Nontuberculous Tuberculosis
mycobacteria
Diffuse bronchiectasis' +++
Focal bronchiectasis b + +++
RUL bronchiectasis +++ +++
RML bronchiectasis +++ +
RML and lingular
bronchiectasis +++
RLL bronchiectasis ++ +
Either RLL or LLL +++ +
Nodules +++ +++
Diffuse bronchiectasis
and well-defined nodules ++
a Cavities +++ +++
Airspace disease +++ +++
Pleural calcification + ++
RUL, right upper lobe; RML, right middle lobe; RLL, right
lower lobe; LLL, left lower lobe; + indicates relative prevalence
of radiographic feature in NTM compared with TB, with three
plusses (+++) signifying a very common finding and one plus
(+) a very uncommon one; - indicates absence of that feature
• Diffuse bronchiectasis defined as involving ~4 lobes
b Focal bronchiectasis defined as involving only one lobe.
Adapted from Guide and Holland (2002) with permission

specimen, with or without positive acid-fast bacillus

b
Fig. 11.2a,b. Radiographic findings of pulmonary NTM dis-
ease. a Chest CT of a 69-year-old woman with pulmonary
MAC disease shows characteristic bronchiectasis in the right
middle lobe and lingula. b Chest CT of a 56-year-old woman
with long-standing M. abscessus pulmonary infection showing
extensive left lung destruction with bronchiectasis and conse-
quent chest wall deformity
11.11.2
Lymphadenitis
Lymphadenitis caused by NTM most commonly
involves the submandibular, submaxillary, cervical,
or preauricular lymph nodes of children between the
ages of 1 and 5 years (Margileth et al. 1984; Wolinsky
1995). A unilateral, rapidly enlarging lymph node or
group of nodes is usually the only presenting com-
plaint, with occasional low-grade fever and malaise.
In the USA currently, the majority of infections are
due to MAC (80%) (Lai et al. 1984). This predomi-
nance of MAC is a change from 30 years ago, when M.
scrofulaceum was the most common etiologic myco-
bacteria (Lincoln and Gilbert 1972). The presumptive
diagnosis of NTM lymphadenitis can be made on
the basis of caseating granulomata in the pathologic
smears and cultures, in the absence of a positive PPD
reaction. Culturing NTM from the involved lymph
node can provide a definitive diagnosis, but the yield
of positive culture in cervical lymph nodes is only
50%-85% (Schaad et al.I979).Although the tubercu-
lin skin test is helpful in making a diagnosis, it does
not distinguish tuberculosis from NTM infection in
one-third of cases (Wolinsky 1995). Excisional biopsy
of the involved lymph node is recommend both for
diagnosis and treatment. The role of fine needle aspi-
ration in the diagnosis of NTM lymphadenitis has
been controversial (Bailey et al. 1985; Lau et al. 1991;
Gupta et al. 1993), and incisional biopsy may cause
fistula formation with chronic drainage (Schaad et aI.
1979). In contrast to other forms of NTM infection,
antibiotic therapy is often not needed; excision of
the involved lymph node without chemotherapy is
currently the best treatment option for isolated NTM
lymphadenitis of childhood (Stewart et al. 1994).
11.11.3
Skin and Soft-Tissue Infection
Although almost all species have been described as
occasional pathogens in cutaneous disease (Wolinsky
1979; Falkinham 1996), M. jortuitum, M. abscessus,
Table 11.5. Criteria for the diagnosis of pulmonary nontuberculous mycobacterial disease ......
~
Presumed or confirmed HIV-seropositive potential risk factors Presumed or confirmed HIV-seronegative potential
risk factors
I. Local immune suppression II. General severe immune suppression
Alcoholism (M. avium complex) Leukemia CD4 count <200
Bronchiectasis Lymphoma
Cyanotic heart disease Organ transplantation
Cystic fibrosis Other immunosuppressive therapy
Prior mycobacterial disease
Pulmonary fibrosis
Smokinglchronic obstructive lung disease
None
1. Clinical criteria:
a. Compatible signs/symptoms (cough, fatigue most common; fever, weight loss, hemoptysis, a. Same a. Same
dyspnea may be present, particularly in advanced disease) with documented deterioration
in clinical status if an underlying condition is present and
b. Reasonable exclusion of other disease (e.g., tuberculosis, cancer, histoplasmosis) to explain b. Same b. Same
condition, or adequate treatment of other condition with increasing signs/symptoms
2. Radiographic criteria:
a. Any of the following chest X-ray abnormalities; if baseline films are more thanl year old, a. Same a. Same
should be evidence of progression
• Infiltrates with or without nodules (persistent >2 months or progressive)
• Cavitation
• Nodules alone (multiple)
b. Any of these HRCT abnormalities b. Same b. Same
• Multiple small nodules
• Multifocal bronchiectasis with or without small lung nodules
3. Bacteriologic criteria:
a. At least three available sputum/bronchial wash samples within I year a. Same a. Same
• Three positive cultures with negative AFB smears or
• Two positive cultures and one positive AFB smear or
b. Single available bronchial wash and inability to obtain sputum samples b. Same except b. Same except
• Positive culture with 2+,3+, or 4+ growth or • culture positive with I + • culture positive with I +
• Positive culture with a 2+, 3+, or 4+ AFB smear or or greater growth or greater growth
(excludes M. avium complex) --,
c. Tissue biopsy c. Same c. Same
• Any growth from bronchopulmonary tissue biopsy ~
• Granuloma and/or AFB on lung biopsy with one or more positive cultures ~
'"0..
from sputum/bronchial wash ~
• Any growth from usually sterile extrapulmonary site ~
For a diagnosis of pulmonary disease, all three criteria - (1) clinical, (2) radiographic, and (3) bacteriologic - must be satisfied ~
From ATS (1997) ~
~
0..
Nontuberculous Mycobacteria
M. marinum, and M. ulcerans are the most common
NTM causes of skin and soft-tissue infection (Wolin-
sky and Rynearson 1968). Abscess formation and/or
drainage at the site of puncture wounds, open trau-
matic injuries, or fractures are typical manifesta-
tions of cutaneous NTM infection. Furunculosis in
the lower extremities by M. fortuitum through nail
salon pedicure baths (Winthrop et al. 2002) and noso-
comial infections involving long-term intravenous or
peritoneal catheters have been reported (Wallace et
al. 1983; Hoy et al. 1987). The diagnosis can be made
by culture of NTM from the drainage material or on
tissue biopsy.
M. marinum can cause so-called 'swimming pool
granuloma' or 'fish tank granuloma'. This bacillus
infects abraded skin during the cleaning of fresh-
water fish tanks or from scratches from salt water
fish, shrimp, fins, etc. (Wolinsky 1979). Skin infection
by M. marinum commonly presents as a nodular or
ulcerative sporotrichoid lesion on the elbows, fin-
gers, hands, or knees. These lesions usually resolve
spontaneously within 36 months (Catinella 1978).
However, extensive tendonitis can occur. It is impera-
tive to consider M. marinum early in the differential
diagnosis of any necrotizing tendonitis, as the antibi-
otics effective against it are quite distinct, and failure
to identify this organism often leads to recurrent
operations and procedures with associated tendon
destruction and loss of use.
M. ulcerans usually causes a painless nodule on the
leg without other signs or symptoms in young adults
and children (Marston et al. 1995). It can progress to
a shallow ulcer with a necrotic base. Spontaneous
resolution may occur in 6-9 months, but sometimes
the lesion will spread and cause serious deformity
(Marston et al. 1995). NTM can cause chronic granu-
lomatous infections of tendon sheaths, bursae, joints,
bones, and prostheses inserted during orthopedic pro-
cedures (Kelly et al. 1967). M. marinum and MAC are
prone to causing tenosynovitis of the hands (Wolinsky
1979; Hellinger et al. 1995; Falkinham 1996). On rare
occasions, NTM can infect prosthetic heart valves (M.
chelonae) (Repath et al. 1976) or the genitourinary
tract (MAC, M. chelonae) (Clark et al. 1989).
11.11.4
Disseminated NTM Infection
Disseminated infection due to NTM usually occurs
in the setting of immunosuppression such as AIDS
(Horsburgh and Selik 1989), after solid organ
transplantation (Lichtenstein and MacGregor 1983),
175
leukemia (Ingram et al. 1993), lymphoma (Ingram
et al. 1993), collagen vascular disease (Ingram et
al. 1993), severe combined immunodeficiency syn-
drome (Stone et al. 1992), and others. Patients with
disseminated NTM infection without any apparent
underlying disease have also been reported (Lincoln
and Gilbert 1972). MAC has been the most common
NTM infecting patients with HIV infection (Hors-
burgh and Selik 1989). However, in the other settings
mentioned above, although MAC (Wolinsky 1979;
Horsburgh et al. 1985) is most common, M. kansa-
sii (Lichtenstein and MacGregor 1983), M. chelonae
(Wallace et al. 1983; Cooper et al. 1989; Ingram et al.
1993),M. scrofulaceum (Wolinsky 1979),M. abscessus
(Wolinsky 1979), and M. haemophilum (Kiehn and
White 1994) also cause disease.
In patients with disseminated NTM infection with-
out AIDS, disease can manifest as a fever of unknown
origin (Horsburgh et al.1985) or multiple subcutane-
ous nodules or abscesses (Wolinsky 1979; Cooper et
al.1989; Ingram et al.1993; Kiehn and White 1994). In
AIDS patients, prolonged fever accompanied by night
sweats and weight loss in the setting of <50 CD4+ T-
lymphocytes suggests disseminated NTM infection
(ATS 1997). Diagnosis in both settings is confirmed
by culture of NTM from bone marrow or blood. A
case of disseminated MAC infection without known
risk factors is illustrated in Fig. 11.3.
11.12
Treatment
11.12.1
Chemotherapy
Management of these infections is very much still
evolving and remains challenging even in the most
straightforward cases. Many of the NTM, such as
MAC, are typically resistant to many antimycobacte-
rial agents and frequently relapse after therapy. The
newer macrolides, azithromycin and clarithromycin,
have shown remarkable activity against these infec-
tions. Clear and defined roles for immunotherapy
and surgery remain to be determined.
In the pre-macrolide era, long-term success rates
for the treatment of pulmonary MAC infection
ranged from 44% to 80% and were not aided by the
determination of in vitro susceptibilities. Dutt and
Stead reported on 85 cases of pulmonary and 5 cases
of extrapulmonary MAC infection from Arkansas in
the era before macrolides. Some 65% of their popu-
176
Fig. 11.3. a A 7-year-old-boy with extreme splenomegaly had a draining inguinal lymph node (b). Blood and drainage cultures
grew MAC. c Bone marrow aspirate showed no granulomatous response, and staining of his bone marrow showed numerous
acid-fast bacilli
lation had underlying pulmonary disease, and 31%
were women. Although they found a high rate of drug
resistance in vitro, they were still successful in treat-
ing patients with multiple agents including isoniazid,
ethambutol, and streptomycin at least for 12 months.
Some 63% of their patients cleared their sputa within
3 months, with an overall medical failure rate of 32%.
They found no added benefit to treatment with more
than four drugs (Dutt and Stead 1979). Engbaek et al.
reported a retrospective study of 37 Danish patients
(30% women) with pulmonary MAC disease treated
with combinations of isoniazid, ethambutol, PAS,
streptomycin, cycloserine, and rifampin. There was
a 44% medical success rate despite very high levels
of in vitro resistance of the organisms (Engbaek et al.
1981). Ahn et al. treated 46 patients with pulmonary
MAC, 67% of whom had predisposing lung disease,
and 33% of whom were women. Using a 9-week
inpatient induction regimen of isoniazid, etham-
butol, rifampin, and twice weekly streptomycin for
6 months, they achieved a 91 % sputum conversion
rate. Subsequently, after 18-24 months of therapy, 4
patients relapsed or had recurrent disease (Ahn et
al. 1986). Horsburgh et al. reported on 75 patients,
53% of whom were women, and 77% of whom had
been treated previously. They started patients on an
initial regimen of isoniazid, ethambutol, rifampin,
capreomycin, ethionamide, and cycloserine. The
overall response rate in 4 months was 66%, with
more responders having received more drugs to
which their organism was susceptible in vitro (Hors-
burgh et al. 1987). The existence of multiple case
J.-J. Yim and S. M. Holland
series in the pre-macrolide era successfully treated
without the benefit of in vitro susceptibility testing
suggests that we are still fairly ignorant of the critical
aspects of therapy and may be relying on inessential
components of management currently. However, at
least in the USA, the clinical epidemiology of MAC
lung disease has changed and is now more female
predominant, so the response rates may be changing
for reasons other than treatment.
Macrolides are now the mainstay of NTM therapy.
Wallace et al. gave clarithromycin 500 mg twice daily
for 4 months, followed by multidrug therapy, in an
effort to determine the activity of the macrolide
alone. They showed a 58% apparent cure rate in 20
adults, most of whom were previously untreated, and
40% of whom were women. The drug was fairly well
tolerated; 16% of isolates developed clarithromycin
resistance (Wallace et al. 1994). Azithromycin 600 mg/
day as monotherapy was studied by the same group
in 23 patients. They found similar results except that
toxicities such as gastrointestinal effects and altered
hearing were more common; there was no develop-
ment of macrolide resistance (Griffith et al. 1996).
The combination of high-dose rifabutin (600 mg/day)
with macrolides caused multiple toxicities including
leukopenia, gastrointestinal symptoms, abnormal
liver functions, diffuse polyarthralgias, and anterior
uveitis (Griffith et al. 1995). Huang et al. have treated
27 patients with pulmonary MAC infection (93% were
female, 90% Caucasian). After 12 months of multi-
drug therapy including azithromycin/clarithromycin,
rifampin/rifabutin, and ethambutol, they had a 50%
Nontuberculous Mycobacteria 177

treatment failure rate (Huang et al. 1999). Whether counts rebounded. These results were seen in patients
these changing rates of success with drug therapy with low CD4+ T-lymphocyte counts, suggesting that
reflect changes in the demographics of the patient the IFNyeffect in this setting was not T-cell mediated.
group (increasing numbers of Caucasian women) or We have identified a heterogeneous cohort of
the organism or loss of potency of the multidrug regi- patients with abnormalities in IFNy generation
mens remains to be determined in larger studies. and disseminated MAC infection without HIY. This
The rapidly growing mycobacteria, especially has been associated with abnormal IL-12 produc-
M. abscessus, present a special problem. Pulmonary tion (Frucht and Holland 1996) or idiopathic CD4+
disease due to M. abscessus is more likely to occur T-Iymphocytopenia (Holland et al. 1994) We have
in Caucasian women who are nonsmokers without used subcutaneous IFNy in a total of 23 patients
cavitary disease, more likely to be associated with with disseminated MAC and other NTM infections
previous mycobacterial infection, less likely to be with favorable results (Fig. 11.4, unpublished obser-
medically curable, and more likely to lead to death vations). Treatment has been used for up to 1 year
due to lung destruction than other mycobacterial or longer. These patients have persistent in vitro
lung diseases. This infection is a distinct entity that hypoproduction of IFNy, even long after cure. Three
requires more intravenous therapy than does MAC patients have relapsed after discontinuation of their
infection and may be less responsive (Griffith et al. IFNy, despite continuation of their antimycobacterial
1993). Although surgery is frequently attempted and therapy.
may improve the pulmonary symptoms, M. absces-
sus may cause severe fistulizing disease if there has
been contamination of the operative bed. Treatment
guidelines from the American Thoracic Society and
the British Thoracic Society from 1997 and 1999,
respectively, are briefly summarized in Table 11.6
(ATS 1997; BTS 2000).

11.12.2
Immunotherapy

It was not long after Koch's discovery of M. tuber-

culosis that he developed an extract of the organ-
ism, tuberculin, the forerunner of PPD, as a cure
a
for tuberculosis (Koch 1891). The use of tuberculin
and other extracts as a treatment for tuberculosis
continued well into the middle part of this century.
However, the lack of controlled studies, the require-
ment for individualization of treatment regimens to
avoid complications, and relatively slow results pre-
vented this approach from gaining wide application.
Effective multidrug antituberculous therapy made
tuberculin therapy obsolete.
Animal models and human studies show that IFNy
is a critical factor in the control of mycobacteria.
Squires et al. (1989, 1992) treated 6 MAC-bacteremic
AIDS patients with IFNy with or without concurrent
antimycobacterials (Squires et al.1989, 1992). In those
patients who received IFNy plus antimycobacterials, b
there was a clear decline in mycobacteremia. In con-
trast, when IFNywas used without concurrent antimy- Fig.l1.4a,b. Abdominal CT scan before (a) and after (a)
treatment with subcutaneous IFNg in a 42-year-old man with
cobacterials there was no change in bacterial counts. disseminated MAC infection. Chylous ascites caused by MAC
The treatment trial lasted no more than 6 weeks, at decreased dramatically after 4 months of IFNy treatment
the end of which time IFNy was stopped, and bacterial (Holland et al. 1994)
178
Systemic IFNy has been tried in patients with
refractory pulmonary MAC infections. These cases
usually have severe parenchymal destruction and are
difficult to clear, since penetration of antimycobacteri-
als into diseased lung spaces is compromised. The best
results have been in patients with noncavitary disease,
but still in the 20% range (S.M. Holland, unpublished
observations). Aerosolized IFNy has been proposed
as a method of delivering organ-specific cytokine
therapy, which may reach high levels in diseased air-
ways (Jaffe et al. 1991). Chatte et al. used aerosolized
IFNy 500 flg by nebulizer 3 days/week in a 38-year-old
man with silicosis and severe refractory cavitary MAC
disease (Chatte et al. 1995). His smears converted to
negative, but the cultures remained positive; with the
cessation of aerosolized IFNy, his smears reverted to
positive, his disease progressed, and he died. This was
the first report of aerosolized IFNy in the treatment
of mycobacterial disease, and rapid conversion of
sputum smear in such a long-term refractory patient
was remarkable. However, the failure to convert his
cultures is a concern for future applications. Similar
results of smear but not culture conversion were seen
with aerosolized IFNy therapy for multidrug-resistant
tuberculosis (Condos et al. 1997). Placebo-controlled,
prospective trials are necessary to determine the most
appropriate delivery system and the patient groups
that are most likely to benefit. To date, the adverse
experience profile for IFNy is essentially the same
as that seen in chronic granulomatous disease (The
International Chronic Granulomatous Disease Coop-
erative Study Group 1991).
The therapeutic use of IL-12 in human infectious
diseases is still quite immature. Low dose IL-12
(60 ng/kg subcutaneously twice weekly) in a patient
with pulmonary M. abscessus led to sputum conver-
sion and cure despite failure to improve on IFNy
(Holland and Dorman 1998). IL-12 has been shown
to induce antimycobacterial activity in vitro (Ber-
mudez et a1.1995) and in vivo (Holland and Dorman
1998) beyond the simple increase of IFNy. However,
whether IL-12 has important therapeutic activity
beyond that of IFNy is not yet clear, and whether
there is any role for IL-12 in the setting of IFNy recep-
tor deficiency is unknown. IL-12 toxicities so far have
been greater than those of IFNy but can be controlled
by careful dosing (Leonard et a1. 1997).
Granulocyte macrophage colony-stimulating
factor (GM-CSF) is produced by T-cells, NK cells,
activated macrophages, and neutrophils. It stimu-
lates myelopoiesis and activates phagocytes for
superoxide production and intracellular killing of
bacteria, including mycobacteria (Bermudez and
J.-J. Yim and S. M. Holland
Young 1990). Monocytes and NK cell infection with
MAC in vitro led to GM-CSF detection (Blanchard
et a1. 1991). Exogenous administration of GM-CSF
to MAC-infected beige mice resulted in increased
killing of mycobacteria and an enhanced antimy-
cobacterial effect (Bermudez et al. 1994). Kemper
et a1. treated 8 MAC bacteremic AIDS patients with
high-dose azithromycin alone or azithromycin plus
subcutaneous GM-CSF daily for 6 weeks (Kemper
et a1. 1998). Monocyte activation was enhanced in
the GM-CSF group, and mycobacterial counts in the
blood decreased more in the azithromycin plus GM-
CSF group than in the azithromycin alone group. The
drug was well tolerated. Given the broad experience
with GM-CSF in cancer patients and in vitro data
supporting its efficacy, further evaluation in clinical
trials are warranted.
Steroids have been used extensively in the treat-
ment of tuberculous meningitis, pleurisy, and peri-
carditis in an effort to reduce inflammation and sub-
sequent fibrosis. Controlled studies in pleuritis have
failed to show much benefit with steroids, although
they reduced symptoms faster than medication
alone (Wyser et a1. 1996). Wormser et a1. reported
that 5 AIDS patients with disseminated MAC who
received dexamethasone 1-4 mg daily in addition
to antimycobacterial therapy noted improvement in
symptoms, nutritional state, and laboratory markers
over baseline (Wormser et a1.1994).
11.13
Summary
Precisely because of their low virulence, the NTM
have taught us much about both innate and acquired
immune pathways. They highlight a critical relation-
ship between the physical morphotype and immu-
nity and have been unusually valuable in leading to
the discovery of cytokine and receptor defects predis-
posing to infection. It is hoped that the observations
about cytokine therapy developed in the treatment of
the NTM will find broader application in the treat-
ment of infectious diseases.
References
Adams LB et al (I997) Comparison of the roles of reactive
oxygen and nitrogen intermediates in the host response
to Mycobacterium tuberculosis using transgenic mice.
Tubercle Lung Dis 78:237-246
Nontuberculous Mycobacteria
Adams RM et al (1970) Tropical fish aquariums. A source of
Mycobacterium marinum infections resembling sporotri-
chosis. JAmMed Assoc 211:457-461
Ahn CH et al (1979) A demographic study of disease due
to Mycobacterium kansasii or M intracellulare-avium in
Texas. Chest 75:120-125
Ahn CH et al (1986) A four-drug regimen for initial treatment
of cavitary disease caused by Mycobacterium avium com-
plex. Am Rev Respir Dis 134:438-441
Ahrens P et al (1995) Two markers, IS901-IS902 and p40, iden-
tified by PCR and by using monoclonal antibodies in Myco-
bacterium avium strains. J Clin Microbio133:1049-1053
Altare F et al (1998a) Impairment of mycobacterial immunity
in human interleukin-12 receptor deficiency. Science 280:
1432-1435
Altare F et al (1998b) Inherited interleukin 12 deficiency in a
child with Bacille Calmette- Guerin and Salmonella enter-
itidis disseminated infection. J Clin Invest 102:2035-2040
American Thoracic Society (1990) Diagnosis and treatment of
disease caused by nontuberculous mycobacteria. Am Rev
Respir Dis 142:940-953
American Thoracic Society (1997) Diagnosis and treatment
of disease caused by nontuberculous mycobacteria. This
official statement of the American Thoracic Society was
approved by the Board of Directors, March 1997. Medical
Section of the American Lung Association. Am J Respir Crit
Care Med 156:S1-25
Appelberg R, Orme 1M (1993) Effector mechanisms involved
in cytokine-mediated bacteriostasis of Mycobacterium
avium infections in murine macrophages. Immunology
80:352-359
Bailey TM, Akhtar M, Ali MA (1985) Fine needle aspiration
biopsy in the diagnosis of tuberculosis. Acta Cytol 29:
732-736
Barrow GI, Hewitt M (1971) Skin infection with Mycobac-
terium marinum from a tropical fish tank. Br Med J 2:
505-506
Beerwerth W (1973) Mycobacteria in cattle watering tanks and
in surface waters. Deutsche tierarztliche Wochenschrift80,
398-401
Bellamy R et al (1998) Variations in the NRAMP1 gene and
susceptibility to tuberculosis in West Africans. N Engl J
Med 338:640-644
Bermudez LE, Young LS (1989) Oxidative and non-oxidative
intracellular killing of Mycobacterium avium complex.
Microb Pathog 7:289-298
Bermudez LE, Young LS (1990) Recombinant granulocyte-
macrophage colony-stimulating factor activates human
macrophages to inhibit growth or kill Mycobacterium
avium complex. J Leukoc Bioi 48:67-73
Bermudez LE et al (1994) Recombinant granulocyte-macro-
phage colony-stimulating factor enhances the effects of
antibiotics against Mycobacterium avium complex infec-
tion in the beige mouse model. J Infect Dis 169:575-580
Bermudez LE, Wu M, Young LS (1995) Interleukin-12-stimu-
lated natural killer cells can activate human macrophages
to inhibit growth of Mycobacterium avium. Infect Immun
63:4099-4104
Blanchard DK, Michelini-Norris MB, Djeu JY (1991) Inter-
feron decreases the growth inhibition of Mycobacterium
avium-intracellulare complex by fresh human monocytes
but not by culture-derived macrophages. J Infect Dis 164:
152-157
179
Boehm U et al (1997) Cellular responses to interferon-gamma.
Annu Rev ImmunoI15:749-795
Bolan G et al (1985) Infections with Mycobacterium chelonei
in patients receiving dialysis and using processed hemo-
dialyzers. J Infect Dis 152:1013-1019
British Thoracic Society (2000) Management of opportunist
mycobacterial infections: Joint Tuberculosis Committee
Guidelines 1999. Subcommittee of the Joint Tuberculosis
Committee of the British Thoracic Society. Thorax 55:
210-218
Brooks RW et al (1984) Epidemiology of infection by non-
tuberculous mycobacteria. V. Numbers in eastern United
States soils and correlation with soil characteristics. Am
Rev Respir Dis 130:630-633
Brown-Elliott BA, Griffith DE, Wallace RJ Jr (2002) Newly
described or emerging human species of nontuberculous
mycobacteria. Infect Dis Clin North Am 16:187-220
Bruno L et al (1984) Cardiac, skeletal, and ocular abnor-
malities in patients with Marfan's syndrome and in their
relatives. Comparison with the cardiac abnormalities in
patients with kyphoscoliosis. Br Heart J 51:220-230
Butler WR, Kilburn JO (1990) High-performance liquid
chromatography patterns of mycolic acids as criteria for
identification of Mycobacterium chelonae, Mycobacterium
fortuitum, and Mycobacterium smegmatis. J Clin Microbiol
28:2094-2098
Butler WR, Thibert L, Kilburn JO (1992) Identification of
Mycobacterium avium complex strains and some similar
species by high-performance liquid chromatography. J Clin
Microbiol 30:2698-2704
Catinella PJ (1978) Mycobacterium marinum granuloma (pro-
ceedings). Arch Dermato1114:1564
Chatte G et al (1995) Aerosolized interferon gamma for Myco-
bacterium avium-complex lung disease. Am J Respir Crit
Care Med 152:1094-1096
Clark R et al (1989) Genitourinary infections caused by myco-
bacteria other than Mycobacterium tuberculosis. Tubercle
70:297-300 .
Collins CH, Yates MD (1984) Infection and colonisation by
Mycobacterium kansasii and Mycobacterium xenopi: aero-
sols as a possible source? J Infect 8: 178-179
Collins CH, Yates MD, Grange JM. (1984) Names for mycobac-
teria. Br Med J (Clin Res) 288:1764
Colville A (1993) Retrospective review of culture-positive
mycobacterial lymphadenitis cases in children in Not-
tingham, 1979-1990. Eur J Clin Microbiol Infect Dis 12:
192-195
Condos R, Rom WN, Schluger NW (1997) Treatment of mul-
tidrug-resistant pulmonary tuberculosis with interferon-
gamma via aerosol. Lancet 349:1513-1515
Cooper AM et al (2000) Transient loss of resistance to pulmo-
nary tuberculosis in p47(phox-I-) mice. Infect Immun 68:
1231-1234
Cooper JF et al (1989) Mycobacterium chelonae: a cause of
nodular skin lesions with a proclivity for renal transplant
recipients. Am J Med 86: 173-177
De Jong R et al (1998) Severe mycobacterial and Salmonella
infections in interleukin-12 receptor-deficient patients. Sci-
ence 280: 1435-1438
Debrunner M et al (1992) Epidemiology and clinical signifi-
cance of nontuberculous mycobacteria in patients negative
for human immunodeficiency virus in Switzerland. Clin
Infect Dis 15:330-345
180
Denis M, Gregg EO, Ghandirian E (1990) Cytokine modula-
tion Mycobacterium tuberculosis growth in human mac-
rophages. Int J ImmunopharmacoI12:721-727
Doffinger R et al (2000) Partial interferon-gamma recep-
tor signaling chain deficiency in a patient with Bacille
Calmette-Guerin and Mycobacterium abscessus infection.
J Infect Dis 181:379-384
Doffinger R et al (2001) X-linked anhidrotic ectodermal
dysplasia with immunodeficiency is caused by impaired
NF-kappaB signaling. Nature Genet 27:277-285
Doherty TM, Sher A (1997) Defects in cell-mediated immunity
affect chronic, but not innate, resistance of mice to Myco-
bacterium avium infection. J ImmunoI158:4822-4831
Dorman SE, Holland SM (1998) Mutation in the signal-
transducing chain of the interferon-gamma receptor and
susceptibility to mycobacterial infection. J Clin Invest 101:
2364-2369
Dorman SE, Holland SM (2000) Interferon-gamma and inter-
leukin-12 pathway defects and human disease. Cytokine
Growth Fact Rev 11:321-333
Dorman SE et al (1999) Viral infections in interferon-gamma
receptor deficiency. J Pediatr 135:640-643
Douvas GS et al (1985) Gamma interferon activates human
macrophages to become tumoricidal and leishmanicidal
but enhances replication of macrophage-associated myco-
bacteria. Infect Immun 50: 1-8
D'Souza S et al (1996) Defective antigen processing associ-
ated with familial disseminated mycobacteriosis. Clin Exp
Immunol103:35-39
Dupuis S et al (2001) Impairment of mycobacterial but not
viral immunity by a germline human STATl mutation.
Science 293:300-303
Dutt AK, Stead WW. (1979) Long-term results of medical
treatment in Mycobacterium intracellulare infection. Am
J Med 67:449-453
Eaton T et al (1995) Isolation and characteristics of Myco-
bacterium avium complex from water and soil samples in
Uganda. Tubercle Lung Dis 76:570-574
Edwards CK III et al (1986) Chronic infection due to Mycobac-
terium intracellulare in mice: association with macrophage
release of prostaglandin E2 and reversal by injection of
indomethacin, muramyl dipeptide, or interferon-gamma.
J Immunol136:1820-1827
Engbaek HC (1964) Three cases in the same family of fatal
infection with M. avium. Acta Tuberc Scand 45:105-117
Engbaek HC, Vergmann B, Bentzon MW (1981) Lung disease
caused by Mycobacterium aviumlMycobacterium intracel-
lulare. An analysis of Danish patients during the period
1962-1976. Eur J Respir Dis 62:72-83
Eriks IS et al (1996) Rapid differentiation of Mycobacterium
avium and M. paratuberculosis by PCR and restriction
enzyme analysis. J Clin MicrobioI34:734-737
Erkkila JC, Warwick WJ, Bradford DS (1978) Spine deformities
and cystic fibrosis. Clin Orthop Rel Res 131:146-150
Falkinham JO III (1996) Epidemiology of infection by nontu-
berculous mycobacteria. Clin Microbiol Rev 9:177-215
Fieschi C et al (2001) High levels of interferon gamma in the
plasma of children with complete interferon gamma recep-
tor deficiency. Pediatrics 107:E48
Flynn JL et al (1993) An essential role for interferon gamma
in resistance to Mycobacterium tuberculosis infection. J Exp
Med 178:2249-2254
Frucht DM, Holland SM (1996) Defective monocyte costimula-
J.-J. Yim and S. M. Holland
tion for IFN-gamma production in familial disseminated
Mycobacterium avium complex infection: abnormal IL-12
regulation. J ImmunoI157:411-416
Frucht DM et al (1999) IL-12-Independent costimulation
pathways for interferon-gamma production in familial
disseminated Mycobacterium avium complex infection.
Clin ImmunoI91:234-241
Fulton SA et al (1996) Interleukin-12 production by human
monocytes infected with Mycobacterium tuberculosis: role
of phagocytosis. Infect Immun 64:2523-2531
Gately MK et al (1998) The interleukin-12/interleukin-12-
receptor system: role in normal and pathologic immune
responses. Annu Rev ImmunoI16:495-521
Glesby MJ, Pyeritz RE (1989) Association of mitral valve pro-
lapse and systemic abnormalities of connective tissue. A
phenotypic continuum. JAMA 262:523-528
Goldman MJ et al (1997) Human beta-defensin-l is a salt-sen-
sitive antibiotic in lung that is inactivated in cystic fibrosis.
Cell 88:553-560
Grahame R, Child A (1984) Mitral valve prolapse. Br Med J
(Clin Res) 289:317
Grange JM. (1989) Nomenclature of mycobacterial disease. Am
Rev Respir Dis 140:561
Griffith DE, Girard WM, Wallace RJ Jr (1993) Clinical features
of pulmonary disease caused by rapidly growing mycobac-
teria. An analysis of 154 patients. Am Rev Respir Dis 147:
1271-1278
Griffith DE et al (1995) Adverse events associated with high-
dose rifabutin in macrolide- containing regimens for the
treatment of Mycobacterium avium complex lung disease.
Clin Infect Dis 21:594-598
Griffith DE et al (1996) Azithromycin activity against Myco-
bacterium avium complex lung disease in patients who
were not infected with human immunodeficiency virus.
Clin Infect Dis 23:983-989
Grimbacher B et al (1999) Hyper-IgE syndrome with recurrent
infections - an autosomal dominant multisystem disorder.
N Engl J Med 340:692-702
Gruenheid S et al (1997) Natural resistance to infection with
intracellular pathogens: the Nrampl protein is recruited
to the membrane of the phagosome. J Exp Med 185:
717-730
Gruft H, Falkinham JO III, Parker BC (1981) Recent experience
in the epidemiology of disease caused by atypical mycobac-
teria. Rev Infect Dis 3:990-996
Guerrero C et al (1995) A novel insertion element from Myco-
bacterium avium, IS1245, is a specific target for analysis of
strain relatedness. J Clin Microbiol 33:304-307
Guide SV, Holland SM (2002) Host susceptibility factors in
mycobacterial infection. Genetics and body morphotype.
Infect Dis Clin North Am 16:163-86
Gupta SK et al (1993) Cytodiagnosis of tuberculous lymphad-
enitis. A correlative study with microbiologic examination.
Acta Cytol 37:329-332
Hadjiliadis D, Adlakha A, Prakash UB (1999) Rapidly growing
mycobacterial lung infection in association with esopha-
geal disorders. Mayo Clin Proc 74:45-51
Hansen GA (1874) Undersolgelser angaende spedalskhedens
arsager. Norsk Mag Laegevidenskaben 4:1-88
Hellinger WC et al (1995) Localized soft-tissue infections with
Mycobacterium aviumlMycobacterium intracellulare com-
plex in immunocompetent patients: granulomatous teno-
synovitis of the hand or wrist. Clin Infect Dis 21:65-69
Nontuberculous Mycobacteria
Holland SM, Dorman SE (1998) Interleukin-12 (IL-12) treat-
ment in interferon gamma refractory pulmonary M.
abscessus infection: clinical response of immunologic
reconstitution. J Invest Med 46:218A
Holland SM et al (1994) Treatment of refractory dissemi-
nated nontuberculous mycobacterial infection with inter-
feron gamma. A preliminary report. N Engl J Med 330:
1348-1355
Holland SM et al (1998) Abnormal regulation of interferon-
gamma, interleukin-12, and tumor necrosis factor-alpha
in human interferon-gamma receptor 1 deficiency. J Infect
Dis 178:1095-1104
Horsburgh CR Jr, Selik RM (1989) The epidemiology of dis-
seminated nontuberculous mycobacterial infection in the
acquired immunodeficiency syndrome (AIDS). Am Rev
Respir Dis 139:4-7
Horsburgh CR Jr et al (1985) Disseminated infection with
Mycobacterium avium-intracellulare. A report of 13 cases
and a review of the literature. Medicine (Baltimore) 64:
36-48
Horsburgh CR Jr et al (1987) Response to therapy of pulmo-
nary Mycobacterium avium-intracellulare infection corre-
lates with results of in vitro susceptibility testing. Am Rev
Respir Dis 135:418-421
Horsburgh CR Jr et al (1994) Environmental risk factors for
acquisition of Mycobacterium avium complex in persons
with human immunodeficiency virus infection. J Infect
Dis 170:362-367
Hoy JF et al (1987) Mycobacterium fortuitum bacteremia in
patients with cancer and long-term venous catheters. Am
J Med 83:213-217
Huang JH et al (1998) Analyses of the NRAMPI and IFN-
gammaRI genes in women with Mycobacterium avium-
intracellulare pulmonary disease. Am J Respir Crit Care
Med 157:377-381
Huang JH et al (1999) Mycobacterium avium-intracellulare
pulmonary infection in HIV-negative patients without
preexisting lung disease: diagnostic and management
limitations. Chest 115:1033-1040
Ichiyama S, Shimokata K, Tsukamura M (1988) The isolation
of Mycobacterium avium complex from soil, water, and
dusts. Microbiol ImmunoI32:733-739
Ingram CW et al (1993) Disseminated infection with rapidly
growing mycobacteria. Clin Infect Dis 16:463-471
Iseman MD (1996) Pulmonary disease due to Mycobacterium
avium complex. In: Korvick JA, Benson CA (eds) Myca-
bacterium avium complex infection: progress, research
and treatment. Dekker, New York, pp 45-77
Iseman MD, Buschman DL, Ackerson LM (1991) Pectus exca-
vatum and scoliosis. Thoracic anomalies associated with
pulmonary disease caused by Mycobacterium avium com-
plex. Am Rev Respir Dis 144:914-916
Jaffe HA et al (1991) Organ specific cytokine therapy. Local
activation of mononuclear phagocytes by delivery of an
aerosol of recombinant interferon-gamma to the human
lung. J Clin Invest 88:297-302
Jouanguy E et a1 (1996) Interferon-gamma-receptor deficiency
in an infant with fatal Bacille Calmette-Guerin infection. N
Engl J Med 335:1956-1961
Jouanguy E et al (1997) Partial interferon-gamma receptor 1
deficiency in a child with tuberculoid Bacillus Calmette-
Guerin infection and a sibling with clinical tuberculosis. J
Clin Invest 100:2658-2664
181
Jouanguy E et al (1999) A human IFN-yR1 small deletion hot-
spot associated with dominant susceptibility to mycobacte-
rial infection. Nature Genet 21:370-378
Kalayjian RC et al (1995) Pulmonary disease due to infection
by Mycobacterium avium complex in patients with AIDS.
Clin Infect Dis 20:1186-1194
Kamijo R et al (1993) Mice that lack the interferon-gamma
receptor have profoundly altered responses to infection
with Bacillus Calmette-Guerin and subsequent challenge
with lipopolysaccharide. J Exp Med 178:1435-1440
Kamijo R et al (1994) Requirement for transcription factor
IRF-l in NO synthase induction in macrophages. Science
263:1612-1615
Kelly PJ et al (1967) Infection of synovial tissues by mycobac-
teria other than Mycobacterium tuberculosis. J Bone Joint
Surg Am 49:1521-1530
Kemper CA, Bermudez LE, Deresinski SC (1998) Immuno-
modulatory treatment of Mycobacterium avium complex
bacteremia in patients with AIDS by use of recombinant
granulocyte-macrophage colony- stimulating factor. J
Infect Dis 177:914-920
Kiehn TE, White M (1994) Mycobacterium haemophilum: an
emerging pathogen. Eur J Clin Microbiol Infect Dis 13:
925-931
Koch R (1891) Forsetzung der mittheilung uber ein heilmittel
gegen tuberkulose. Dtsch Med Wochenschr 17:1189
Lai KK et al (1984) Mycobacterial cervical lymphadenopathy.
Relation of etiologic agents to age. J Am Med Assoc 251:
1286-1288
Lau SK et al (1991) Combined use of fine-needle aspiration
cytologic examination and tuberculin skin test in the diag-
nosis of cervical tuberculous lymphadenitis. A prospective
study. Arch Otolaryngol Head Neck Surg 117:87-90
Laussucq S et al (1988) Nosocomial Mycobacterium fortuitum
colonization from a contaminated ice machine. Am Rev
Respir Dis 138:891-894
Leonard JP et al (1997) Effects of single-dose interleukin-12
exposure on interleukin-12-associated toxicity and inter-
feron-gamma production. Blood 90:2541-2548
Levin M et al (1995) Familial disseminated atypical myco-
bacterial infection in childhood: a human mycobacterial
susceptibility gene? Lancet 345:79-83
Lichtenstein IH, MacGregor RR (1983) Mycobacterial infec-
tions in renal transplant recipients: report of five cases and
review of the literature. Rev Infect Dis 5:216-226
Lincoln EM, Gilbert LA (1972) Disease in children due to
mycobacteria other than Mycobacterium tuberculosis. Am
Rev Respir Dis 105:683-714
Lowry PW et al (1988) Mycobacterium chelonae causing otitis
media in an ear-nose-and-throat practice. N Engl J Med
319:978-982
Lynch DA et al (1995) CT features of pulmonary Mycobacte-
rium avium complex infection. J Comput Assist Tomogr
19:353-360
Maloney S et al (1994) Mycobacterium abscessus pseudoin-
fection traced to an automated endoscope washer: utility
of epidemiologic and laboratory investigation. J Infect Dis
169:1166-1169
Margileth AM, Chandra R, Altman RP (1984) Chronic lymph-
adenopathy due to mycobacterial infection. Clinical fea-
tures, diagnosis, histopathology, and management. Am J
Dis Child 138:917-922
Marston BJ et al (1995) Emergence of Buruli ulcer disease in
182
the Daloa region of Cote d'lvoire. Am J Trop Med Hyg 52:
219-224
Meissner G, Anz W (1977) Sources of Mycobacterium avium
complex infection resulting in human diseases. Am Rev
Respir Dis 116:1057-1064
Meissner PS, Falkinham JO III (1986) Plasmid DNA profiles
as epidemiological markers for clinical and environmental
isolates of Mycobacterium avium, Mycobacterium intracel-
lulare, and Mycobacterium scrofulaceum. J Infect Dis 153:
325-331
Morrissey AB et al (1992) Absence of Mycobacterium avium
complex disease in patients with AIDS in Uganda. J Acquir
Immune Defic Syndr 5:477-478
Newport MJ et al (1996) A mutation in the interferon-gamma-
receptor gene and susceptibility to mycobacterial infection.
N Engl J Med 335:1941-1949
O'Brien RJ, Geiter LJ, Snider DE Jr (1987) The epidemiology
of nontuberculous mycobacterial diseases in the United
States. Results from a national survey. Am Rev Respir Dis
135:1007-1014
Okello DO et al (1990) Absence of bacteremia with Mycobacte-
rium avium-intracellulare in Ugandan patients with AIDS.
J Infect Dis 162:208-210
Olivier KN, Yankaskas JR, Knowles MR (1996) Nontuberculous
mycobacterial pulmonary disease in cystic fibrosis. Semin
Respir Infect 11:272-284
Orange JS et al (2002) Deficient natural killer cell cytotoxic-
ity in patients with IKK-gammaINEMO mutations. J Clin
Invest 109:1501-1509
Pae HH et al (1999) Study for clinical chracteristics of nontu-
berculous mycobacterial pulmonary diseases. Tuberculosis
Lung Dis (Korean) 47:735-746
Paull A (1973) An environmental study of the opportunist
mycobacteria. Med Lab TechnoI30:11-19
Pomerantz M et al (1996) Resection of the right middle lobe
and lingula for mycobacterial infection. Ann Thorac Surg
62:990-993
Pravica V et al (1999) In vitro production of IFN-gamma
correlates with CA repeat polymorphism in the human
IFN-gamma gene. Eur J Immunogenet 26:1-3
Prince DS et al (1989) Infection with Mycobacterium avium
complex in patients without predisposing conditions. N
Engl J Med 321:863-868
Pyeritz RE, McKusick VA (1979) The Marfan syndrome: diag-
nosis and management. N Engl J Med 300:772-777
Repath F et al (1976) Prosthetic valve endocarditis due to
Mycobacterium chelonei. South Med J 69:1244-1246
Reznikov M, Dawson DJ (1980) Mycobacteria of the intracel-
lulare-scrofulaceum group in soils from the Adelaide area.
Pathology 12:525-528
Reznikov M, Leggo JH (1974) Examination of soil in the
brisbane area for organisms of the Mycobacterium avium-
intracellulare-scrofulaceum complex. Pathology 6:269-273
Roman MJ et al (1989) Comparison of cardiovascular and skel-
etal features of primary mitral valve prolapse and Marfan
syndrome. Am J CardioI63:317-321
Runyon EH (1974) Ten mycobacterial pathogens. Tubercle 55:
235-240
Saito H et al (1994) Mycobacterium malmoense isolated from
soil. Microbiol ImmunoI38:313-315
Sakatani M (1994) Nontuberculous mycobacteriosis (NTM)
in Japan-epidemiologic and clinical study. Kekkaku 69:
119-124
J.-J. Yim and S. M. Holland
Sanal 0 et al (2000) Isolated cutaneous response to granulo-
cyte-monocyte colony stimulating factor in fatal idiopathic
disseminated Bacillus-Calmette-Guerin infection. Eur J
Pediatr 159:149-152
Sansila A et al (1998) Differentiation between Mycobacterium
tuberculosis and Mycobacterium avium by amplification
of the 16S-23S ribosomal DNA spacer. J Clin Microbiol 36:
2399-2403
Sasaki Y et al (2002) Genetic basis of patients with Bacille
Calmette-Guerin osteomyeliti~ in Japan: identification of
dominant partial interferon-gamma receptor 1 deficiency
as a predominant type. J Infect Dis 185:706-709
Schaad UB et al (1979) Management of atypical mycobacterial
lymphadenitis in childhood: a review based on 380 cases.
J Pediatr 95:356-360
Selik RM, Starcher ET, Curran JW (1987) Opportunistic dis-
eases reported in AIDS patients: frequencies, associations,
and trends. AIDS 1:175-182
Shinnick TM, Good RC (1994) Mycobacterial taxonomy. Eur J
Clin Microbiol Infect Dis 13:884-901
Sponseller PD et al (1995) The thoracolumbar spine in Marfan
syndrome. J Bone Joint Surg Am 77:867-876
Springer B et al (1993) Mycobacterium interjectum, a new spe-
cies isolated from a patient with chronic lymphadenitis. J
Clin MicrobioI31:3083-3089
Squires KE et al (1989) Interferon-gamma and Mycobacterium
avium-intracellulare infection. J Infect Dis 159:599-600
Squires KE et al (1992) Interferon-gamma treatment for
Mycobacterium avium-intracellular complex bacillemia in
patients with AIDS. J Infect Dis 166:686-687
Stewart MG, Starke JR, Coker NJ (1994) Nontuberculous myco-
bacterial infections of the head and neck. Arch Otolaryngol
Head Neck Surg 120:873-876
Stone AB et al (1992) Disseminated Mycobacterium avium
complex in non-human immunodeficiency virus-infected
pediatric patients. Pediatr Infect Dis J 11:960-964
Supek F et al (1996) A yeast manganese transporter related to
the macrophage protein involved in conferring resistance
to mycobacteria. Proc Nat! Acad Sci USA 93:5105-5110
Swensen SJ, Hartman TE, Williams DE. (1994) Computed tomo-
graphic diagnosis of Mycobacterium avium-intracellulare
complex in patients with bronchiectasis. Chest 105:49-52
The International Chronic Granulomatous Disease Coop-
erative Study Group (1991) A controlled trial of interferon
gamma to prevent infection in chronic granulomatous
disease. N Engl J Med 324:509-516
Timpe A, Runyon EH (1954) The relationship of atypical acid-
fast bacteria to human disease: a preliminary report. J Lab
Clin Med 44:202-209
Trinchieri G, Scott P (1999) Interleukin-12: basic principles
and clinical applications. Curr Top Microbiol Immunol
238:57-78
Vidal S et al (1995) The Ity/Lsh/Bcg locus: natural resistance
to infection with intracellular parasites is abrogated by dis-
ruption of the Nrampl gene. J Exp Med 182:655-666
Von Reyn CF et al (1993) Isolation of Mycobacterium avium
complex from water in the United States, Finland, Zaire,
and Kenya. J Clin MicrobioI31:3227-3230
Von Reyn CF et al (1994) Persistent colonisation of potable
water as a source of Mycobacterium avium infection in
AIDS. Lancet 343:1137-1141
Von Reyn CF et al (1998) Dual skin testing with Mycobac-
terium avium sensitin and purified protein derivative to
Nontuberculous Mycobacteria
discriminate pulmonary disease due to M. avium complex
from pulmonary disease due to Mycobacterium tuberculo-
sis. J Infect Dis 177:730-736
Wallace RJ Jr et al (1983) Spectrum of disease due to rapidly
growing mycobacteria. Rev Infect Dis 5:657-679
Wallace RJ Jr et al (1994) Initial clarithromycin monotherapy
for Mycobacterium avium-intracellulare complex lung dis-
ease. Am J Respir Crit Care Med 149:1335-1341
Wallace RJ Jr et al (1998) Polyclonal Mycobacterium avium
complex infections in patients with nodular bronchiectasis.
Am J Respir Crit Care Med 158:1235-1244
Wayne LG (1985) The 'atypical' mycobacteria: recognition and
disease association. Crit Rev MicrobioI12:185-222
Wayne LG, Sramek HA (1992) Agents of newly recognized
or infrequently encountered mycobacterial diseases. Clin
Microbiol Rev 5:1-25
Winthrop KL et al (2002) An outbreak of mycobacterial furun-
culosis associated with footbaths at a nail salon. N Engl J
Med 346:1366-1371
Witty LA, Tapson VF, Piantadosi CA (1994) Isolation of myco-
bacteria in patients with pulmonary alveolar proteinosis.
Medicine (Baltimore) 73:103-109
183
Wolinsky E (1979) Nontuberculous mycobacteria and associ-
ated diseases. Am Rev Respir Dis 119: 107-159
Wolinsky E (1995) Mycobacterial lymphadenitis in children: a
prospective study of 105 nontuberculous cases with long-
term follow-up. Clin Infect Dis 20:954-963
Wolinsky E, Rynearson TK (1968) Mycobacteria in soil and
their relation to disease-associated strains. Am Rev Respir
Dis 97:1032-1037
Wormser GP, Horowitz H, Dworkin B (1994) Low-dose dexa-
methasone as adjunctive therapy for disseminated Myco-
bacterium avium complex infections in AIDS patients.
Antimicrob Agents Chemother 38:2215-2217
Wyser C et al (1996) Corticosteroids in the treatment of
tuberculous pleurisy. A double- blind, placebo-controlled,
randomized study. Chest 110:333-338
Zakowski P et al (1982) Disseminated Mycobacterium avium-
intracellulare infection in homosexual men dying of acquired
immunodeficiency. J Am Med Assoc 248:2980-2982
Zonana J et al (2000) A novel X-linked disorder of immune
deficiency and hypohidrotic ectodermal dysplasia is allelic
to incontinentia pigmenti and due to mutations in IKK-
gamma (NEMO). Am J Hum Genet 67:1555-1562
Diagnosis
12 Serologic Testing for Tuberculosis
CHAKRADHAR KOTARU and EDWARD D. CHAN

CONTENTS 12.1
Introduction
12.1 Introduction 187
12.2 Specific Serological Tests for TB 187
12.2.1 Antigen 5 (38 kDa Antigen) 188 Tuberculosis (TB) is a leading cause of morbidity and
12.2.2 A60 Antigen 189 mortality in the world. The World Health Organiza-
12.2.3 30 kDa Antigen 189 tion (WHO) estimates that one-third of the world's
12.2.4 Antigen p90 189 population is latently infected with Mycobacterium
12.2.5 19 kDa Antigen 189
tuberculosis. From this pool, roughly 10 million
12.2.6 16 kDa Antigen 189
12.2.7 45/47 kDa Antigen Complex 190 active TB cases emerge annually, resulting in 2 to
12.2.8 P32 Antigen 190 3 million deaths. Early identification and proper
12.2.9 Cord Factor 190 treatment of individuals with active TB have a great
12.2.10 Glycolipid Antigens 190 impact on public health. The bacteriologic methods
12.2.11 Lipoarabinomannan 191
are either not sufficiently efficient and specific (acid-
12.2.12 ICT Tuberculosis Test 192
12.2.13 Dermal Response to MPB64 Antigen 192 fast smear) or require an extended turnaround time
12.2.14 Integral Membrane Antigens from the laboratory (culture isolation). The develop-
of M. habana TMC 5135 192 ment of a rapid nonbacteriologic diagnostic test that
12.2.15 Lipopolysaccharide Antigen 192 is both sensitive and specific for active TB continues
12.3 Specific Targeted Populations 192
to be formidable. This difficulty is due, in part, to the
12.3.1 Childhood TB 193
12.3.2 TB in the Elderly 193 inability of many tests to distinguish latent infection
12.3.3 HIV-Positive Individuals 193 from active disease. This chapter focuses on the cur-
12.3.4 Extrapulmonary TB 194 rent serological methods to diagnose TB.
12.3.5 Smear-Negative vs Smear-Positive TB 194
12.4 Conclusion 194
References 195

Edward D. Chan is supported by the NHLBI-
HL-66-112, the Parke-Davis Atorvastatin Research
Award, and the American Lung Association Career
Investigator Award.
C. KOTARU, MD
Department of Medicine and Program in Cell Biology,
National Jewish Medical and Research Center, Division of
Pulmonary Sciences and Critical Care Medicine, University
of Colorado Health Sciences Center, and Denver Veteran
Administration Medical Center, K613e, Goodman Building,
1400 Jackson Street, Denver, CO 80206, USA
E. D. CHAN, MD
Associate Professor of Medicine, K613e Goodman Bldg.,
National Jewish Medical and Research Center, Denver, CO,
1400 Jackson St, Denver, CO 80206, USA
12.2
Specific Serological Tests for T8
Although serological assays may not add to the diag-
nostic yield in those patients in whom sputa exami-
nation is possible an accurate serological test to diag-
nose TB would have considerable advantages in those
patients who are unable to provide adequate sputum
samples (e.g., in children and the very elderly), who
are smear-negative and/or culture-negative, or have
suspected extrapulmonary TB (Steele and Daniel
1991). In populations where the prevalence of tuber-
culin skin test positivity is high, earlier detection
and treatment of active TB by blood screening may
help prevent further transmission of TB. However,
serological testing has been confounded by cross-
reactivity associated with bacillus Calmette-Guerin
(BCG) vaccination or infection with nontuberculous
mycobacteria. For example, using the purified protein

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
188
derivative (PPD) status as a gold standard, the abil-
ity of immunoblot for the A60 antigen complex to
discriminate between TB infection and noninfection
was nearly 100% (Rovatti et al. 1996). However, 90%
of the study group (n=140) tested positive for anti-
A60 antibody after BCG vaccination compared with
only 1.4% prior to vaccination (Rovatti et al. 1996).
Previous seroassays for TB have utilized either a
mixture of M. tuberculosis antigens, such as purified
extracted glycolipids, adsorbed mycobacterial soni-
cates, and PPD, or more distinct mycobacterial anti-
gens (Table 12.1) (Chan et al. 2000). Measurement
of tuberculostearic acid in clinical specimens was
shown to have a high degree of sensitivity and speci-
ficity, but the assay required considerable expertise
to perform (Brooks et al. 1987). Using mycobacte-
rial sonicates in an enzyme-linked immunosorbent
assay (ELISA) on samples of 31 children with clini-
cal TB, Rosen (l990) found a sensitivity of 26% and
specificity of 40%. Khomenko et al. (l996) evalu-
ated an ELISA for M. tuberculosis H37Rv antigenic
preparation and found the sensitivity to be lower at
72% and specificity to be 100%. Conversely, using
affinity-purified H37Rv antibody to detect the anti-
gen, they found the sensitivity and specificity to be
similar, 77% and 90%-93%, respectively. Franco et
al. (2001) assessed the serological response to four
antigenic fractions of H37Rv sonicate by Western
blot analysis. Patients with smear-positive pul-
monary infection and tuberculous pleurisy had a
positivity rate of 50%-60% against all fractions.
Interestingly, of the 16 contacts, 3 had a conver-
sion of their tuberculin skin tests, and the same 3
individuals also had a positive serologic response to
the 19 kDa antigenic fractions, making it potentially
Table 12.1. Antigens used in the serological diagnosis of
tuberculosis
Mycobacterial sonicates
Extracted glycolipids
Purified protein derivative
Antigen 5 (38 kDa antigen)
A60 antigen
45/47 kDa antigen complex
Antigen Kp90
30 kDa antigen
P32 antigen
Cord factor (trehalose dimycolate)
Lipoarabinomannan
16 kDa antigen
Integral membrane antigens of M. habana TMC 5135
Lipopolysaccharide antigen
C. Kotaru and E. D. Chan
useful in identifying patients who are candidates
for the treatment of latent infection. In the follow-
ing sections, we reviewed studies that examined
the presence of specific antibodies directed against
mycobacterial antigens.
12.2.1
Antigen 5 (38 kDa Antigen)
The use of antigen 5, also known as the 38 kDa anti-
gen, in a microtiter plate ELISA has been evaluated
in a number of countries with reasonable sensitivity
and high specificity (AIde et al. 1989; Daniel et al.
1985, 1986; Ma et al. 1986). In one of these studies
from China, the test was 89% sensitive and 94%-
100% specific for patients with active pulmonary
TB (Ma et al. 1986). Generally, the antibody titer
correlated with the extent of pulmonary disease. In
another study, detection of IgG against this 38 kDa
revealed a sensitivity of 64% and a specificity of
81 % (Chiang et al. 1997). Recently, Cole and co-
workers (l996) also used antigen 5 in a serological
test but modified it by using a rapid membrane-
based antibody assay and found the sensitivity to
be 89% (54 of 61) in patients with smear-positive,
culture-positive TB. In a group of 91 patients who
were both smear- and culture-negative, the sensitiv-
ity was 74% even though these patients had been
on antituberculous therapy from 1 to 6 months
at the time of the test. The overall specificity was
93%. In an ELISA using the monoclonal antibody
TBn to the 38 kDa antigen, the sensitivity was
about 80% for patients with either smear-positive/
culture-positive or smear-negative/culture-positive
TB (Bothamley and Rudd 1994). Interestingly, the
sensitivity rate was even higher (-90%) for patients
successfully treated for presumed TB but without
bacteriologic confirmation of disease. Combining
an ELISA to detect antibody to both 38 kDa and
a 34 kDa M. tuberculosis protein, Amicosante and
colleagues (l995) found the overall sensitivity to
be 86%. Importantly, the specificity in PPD-nega-
tive or PPD-positive control patients (n=30 each)
and in 20 patients with nontuberculous bacterial
pneumonia was 100%. Nsanze et al. (l997) reported
the utility of the enzyme immunoassay (EIA) kit
developed by Omega laboratories using the 38 kDa
antigen. Depending on the cut-off points used,
the sensitivities were >95%, while the specificities
were 86%-100%. The test was much less useful for
extrapulmonary TB, with sensitivities ranging from
51% to 71%.
Serologic Testing for Tuberculosis
12.2.2
A60Antigen
The A60 antigen, a thermostable component of PPD,
has also been used in the serodiagnosis ofTB (Charpin
et al. 1990; Cocito 1991; Fadda et al. 1992; Luh et al.
1996; Yu et al.1992; Zou et al.1994). Unfortunately, this
molecule is not specific for mycobacteria because it is
also present in Nocardia and Corynebacterium species.
In 83 patients with smear-negative but culture-posi-
tive pulmonary TB, measurement of both IgM and IgG
in an ELISA revealed a sensitivity of 68% and specific-
ity of 100% (Charpin et al. 1990). Similarly, an ELISA
revealed a sensitivity of 76.2% for patients with active
pulmonary TB and 59% for patients with extrapul-
monary TB (Luh et al. 1996). In 153 individuals with
inactive tuberculous infection, the specificity was only
81.7%, and in over 500 patients with nontuberculous
infection, the specificity was 91.3%. Thus, the positive
predictive value was only 67.9% (Luh et al. 1996). In
a study of 560 Chinese patients with pulmonary and
extrapulmonary TB and over 700 controls, the mea-
surement of IgM appeared to be sensitive (80%) for
active primary TB and specific (100%) for latent TB
(Zou et al. 1994). On the other hand, IgG against the
A60 antigen was more predictive of active postpri-
mary TB (sensitivity of 89%). Among 529 healthy
persons most of whom were vaccinated with BCG,
including 287 who were PPD positive, there was less
than 1% false-positives. Anuradha et al. (2001) evalu-
ated the utility of detecting antibodies to A60 for the
diagnosis of extrapulmonary TB. In 72 patients with
neuro-TB, antibodies were detected in the serum and/
or CSF in almost 80%. The diagnosis of other forms of
extrapulmonary TB by anti-A60 antibodies could be
made in about 60% of the cases.
12.2.3
30 kDa Antigen
Using a 30 kDa antigen purified from culture filtrates
of M. tuberculosis, McDonough and colleagues (1992)
compared a dot EIA with the standard ELISA assay
on patients with active TB using control patients who
had no clinical evidence of TB but whose PPD status
was not known. The specificity for dot EIA and plate
ELISA were 92% and 97%, respectively; the sensitivity
rate for hospitalized patients with TB was 69% and
78%, respectively. Interestingly, the sensitivity rate was
-48% for both methods in ambulatory patients with
active TB, while in 26 HIV-positive patients with TB,
both immunoassays showed extremely poor sensitiv-
189
ity (12% for the plate ELISA and 0% for the dot EIA).
In assessing the lymphocyte proliferative and humoral
responses to patients with active pulmonary TB and
of healthy household contacts, the group with active
disease had a strong humoral response to the 30 kDa
antigen but a poor proliferative response (Torres et al.
1994); the opposite pattern was seen in the exposed
control patients. Thus, a humoral response to tubercu-
lous antigen may not only signify active infection but
may also indicate a nonprotective immune response.
In a study that measured the presence of the 30 kDa
antigen in 25 African patients with smear-positive TB
using a monoclonal antibody, 20 of them tested posi-
tive (sensitivity 80%) (Ng et al. 1995).
12.2.4
Antigen p90
Arikan and co-workers (1997) showed that in 51
patients with active TB, the sensitivity of anti-Kp90
IgA in either sera or body fluids was 82%, while in 71
control patients, the specificity was 90%. In another
study of 88 TB patients, comprised of 32 smear-posi-
tive and 56 smear-negative individuals, the overall
sensitivity was 70% (Alifano et al. 1997). In a control
group of 87 individuals with either nontuberculous
lung disease, healed TB, or healthy volunteers, the
specificity was 92%.
12.2.5
19 kDa Antigen
Although previously promlsmg, antibody titers to
the 19 kDa antigen were found to have poor sensitiv-
ity (8%) in Indian patients with active TB, but it was
higher in patients from the UK (55%-57%) (Bothamley
et al. 1992).
12.2.6
16 kDa Antigen
Using a recombinant 16 kDa antigen of M. tuberculo-
sis, Imaz et al. (2001) compared the seroresponse using
EIA to detect IgG, IgM, and IgA in 74 children with
active TB, in 149 with nonmycobacterial disease, and
in 49 healthy contacts. The average levels of antibodies
were higher in the healthy contacts when compared
with children with nonmycobacterial disease, making
it a potentially useful test for predicting patients who
require prophylaxis against latent infection; however,
190
the low overall sensitivity (34%, 19%,3%, and 43% for
IgG, IgA, IgM, and combined IgG IlgM, respectively)
prevented the test being useful in the diagnosis of
active childhood TB.
12.2.7
45/47 kDa Antigen Complex
In a study that evaluated the IgG response to the 451
47 kDa secreted protein ofM. tuberculosis, the specific-
ity of the test was >98% in control subjects, comprised
of healthy volunteers who were either PPD-positive or
PPD-negative (Diagbouga et al. 1997). However, the
sensitivity of the test was only 40% even in smear-
positive TB patients.
12.2.8
P32Antigen
The serological response to purified protein of M.
bovis BCG (known as P32 antigen) was also evalu-
ated in patients with active TB (Turneer et al. 1988).
Although there was a statistically significant differ-
ence in mean IgG and IgA antibody levels between
active TB patients and healthy control subjects, the
combined sensitivity rate for both immunoglobulin
classes was only 47%. Interestingly, neither naturally
acquired tuberculin hypersensitivity nor BCG vac-
cination affected the positive frequencies in healthy
subjects. This same group of investigators also uti-
lized a dot-blot assay with BCG cellular extract and
found a good correlation between their ELISA assay
and the dot-blot assay (van Vooren et al. 1988).
12.2.9
Cord Factor
The IgG antibody response to M. tuberculosis cord
factor (trehalose-6,6'-dimycolate) was examined in a
group of 99 patients with mycobacterial infection (42
with culture-positive TB, 46 with a clinical diagnosis
of TB, and 11 patients with atypical mycobacterial
infection), 5 patients with lung cancer, and 100
healthy controls. The overall sensitivity of patients
with any bacterial or clinical diagnosis of a myco-
bacterial infection was 83%, and the specificity was
100% (He et al. 1991). The antibody response to M.
tuberculosis cord factor was tested in an ELISA for
patients with active TB and produced a sensitivity of
81 % and a specificity of 96% compared with healthy
C. Kotaru and E. D. Chan
controls who were all skin test-positive (Maekura
et al. 1993). Interestingly, the cord factor antibody
titers declined to normal levels with antituberculous
chemotherapy.
12.2.10
Glycolipid Antigens
Detection of IgG against M. tuberculosis glycolipid
antigen in 57 patients (27 smear-positive and 30
smear-negative) revealed that the overall sensitivity
was 96% and specificity 91 % (Dogan and Aksu 1997).
Escamilla and colleagues (1996) used glycolipids from
M. fortuitum (di- and tri-O-acylated trehaloses) to
detect an antibody response and found the sensitivity
for active TB was >80% and the specificity 98%.
In a large multicenter study from Japan, Maekura et
al. (2001) described the utility of an ELISA for antitu-
berculous glycolipid (anti-TBGL) in the serodiagnosis
of TB. Using a cut-off of anti-TBGL antibody titer
>2 Ulml, the assay had a sensitivity of 81 % and speci-
ficity of 96% in 318 patients with active pulmonary TB
(216 smear-positive andlor culture-positive and 102
smear- and culture-negative, but clinically diagnosed).
Subgroup analysis revealed greater sensitivity (90%)
in patients with smear- and culture-positive TB, as
opposed to culture-positive only (70%) or smear- and
culture-negative, clinically diagnosed patients (73%).
This suggests a correlation between the positivity of
the test and the mycobacterial burden. Another inter-
esting observation that has potential implications in
judging the response to therapy was the decline in
titers in a subgroup of patients with initially high titers
(> 10 U/ml) at the beginning of treatment.
The test performance characteristics of ELISA to
detect antibodies against M. tuberculosis antigenic
glycolipids was evaluated in 142 patients with extra-
pulmonary TB compared with 578 patients with non-
tuberculous disease (Niculescu et al.1999). The sensi-
tivity and specificity were 82 and 92%, respectively.
Simmoney and co-workers (1996) compared the
diagnostic utility of ELISA in detecting antibodies
against glycolipid to those against A60 antigen in 46
HIV-positive and 50 HIV-negative patients with docu-
mented TB. Overall, antibody to A60 was detected in
only 37% of the cases, while 85% of the patients had
a positive antibody response to glycolipid antigens.
When the data were analyzed in the context of HIV
status and the associated level of immunosuppression,
HIV-positive patients with lower CD4 counts were
more likely to have negative antibody responses to
A60 than to glycolipids.
Serologic Testing for Tuberculosis 191

12.2.11 that of M. avium complex (the predominant atypi-

Lipoarabinomannan cal mycobacteria), the fine structural differences
between the LAMs (e.g., degree and configuration
The cell walls of mycobacterial organisms contain of mannose capping) have not been fully elucidated.
complex macromolecules such as lipoarabinoman- Therefore, despite the relatively good specificity rate
nan (LAM) (Fig. 12.1). LAM is a lipoglycan known to reported, the degree of antibody cross-reactivity
have a number of immunomodulatory effects. LAM is against LAM of M. tuberculosis vs that of atypical
comprised of a linear series of ringed mannose sugar mycobacteria remains to be determined.
residues, with occasional branches of single mannose Second, there was a wide range of sensitivity of
residues. At the proximal end of LAM, a phosphati- anti-LAM IgG among HIV-negative patients with
dylinositol group anchors it to the plasma membrane. active TB, while the sensitivity was generally poor
Distal to the mannose residues, a series of arabinose in HIV-positive patients. For example, in HIV-nega-
sugar residues is attached. In M. tuberculosis, these tive patients, the sensitivity ranged from 21.5% to
arabinose residues are further 'capped' with mannose 89% (del Prete et al. 1998; Julian et al. 1997; Lawn
residues. Thus, the LAM of M. tuberculosis is called et al. 1997; Ratanasuwan et al. 1997; Sada et al. 1992,
'ManLAM'. In relatively recent reports, the detection 1999) but was between 7% and 40% in HIV-positive
of anti-LAM antibodies has shown great promise in patients (Boggian et al. 1996; Julian et al. 1997; Lawn
the serological diagnosis of active TB (Boggian et al. et al. 1997; Ratanasuwan et al. 1997; Sada et al. 1992,
1996; del Prete et al. 1998; Julian et al. 1997; Lawn et 1999). In HIV-negative patients, the sensitivity of the
al. 1997; Ratanasuwan et al. 1997; Sada et al. 1990, MycoDot antiLAM IgG test in Tanzania, Ghana, and
1992; Somi et al. 1999). Thailand was 33%,56%, and 63%, respectively (Lawn
In general, three important points can be taken et al. 1997; Ratanasuwan et al. 1997; Somi et al.I999).
from these studies. First, the specificity of the test Third, as with most other types of serological tests,
was excellent, ranging from 84% to 100%. Moreover, the sensitivity of the test fell with negative smears,
Julian and co-workers (1997) tested 14 patients with a finding generally attributed to the lower burden
atypical mycobacteria, and anti-LAM IgG could not of organisms in smear-negative cases and/or to the
be detected in any of them although it is not clear how greater incidence of smear-negativity in patients co-
many of these patients were co-infected with HIV. infected with HIV. We recently tested the accuracy of
Boggian et al. (1996) tested 104 patients with atypical a simplified, visually detectable, colloidal gold-based
mycobacteria for anti-LAM IgG, and only 2 patients serological assay to qualitatively detect IgG directed
were weakly positive; however, all the patients tested against LAM in a population of individuals largely
were HIV-positive. Although the basic structure of comprised of immigrants to the USA from areas with
LAM associated with M. tuberculosis is similar to a high prevalence of TB infection (Chan et al. 2000).

ManLAM Fig. 12.1. Cell wall of M. tuberculosis showing the various mac-
romolecules that comprise the complex structure, including
the plasma membrane, peptidoglycan, lipoprotein ,and vari-
ous Iipoglycans such as mannose-capped Iipoarabinoman-
nan, mycolic acid-arabinogaJactan peptidoglycan complexes
(mAGP), Iipomannan (LM), and phosphatidylinositol diman-
nose (PIM z). The fatty acid chains (acyl groups) are attached
to the plasma membrane

lipoproteins

• 0- mannose-p mycolic phosphatidylinositol

• D-arabinose-f acids (palmitate, tuberculostearate)
• galaetose-f
'
192
In patients with active TB, the sensitivity of anti-LAM
IgG was 85%-93%. Importantly, in five patients with
active TB who were smear-negative, all tested positive
for anti-LAM IgG. The specificity of the test depended
on the presence of tuberculous infection. In a group
of US citizens comprised of young healthy adults and
rheumatology patients, the specificity was 100%. In a
population at risk for TB infection who were either
tuberculin skin test-negative or -positive, the specific-
ity of the test for active TB was 89%. The negative and
positive predictive values of the test were 98% and
52%, respectively. The overall accuracy of the test was
81 %. The high negative predictive value of the test in a
population at risk for tuberculous infection makes it a
potentially valuable screening test for active TE.
12.2.12
ICT Tuberculosis Test
The ICT tuberculosis test is a card test based on the
detection of IgG antibodies directed against five M.
tuberculosis secreted antigens, using an anti-human
IgG labeled with colloidal gold (Rasolofo et al. 2000),
a method similar to the one we employed in the anti-
LAM IgG dot-blot assay (Chan et al. 2000). Rasolofo
and colleagues (2000) showed that the ICT test had
a sensitivity of only 68.2% for smear-positive TB and
65.2% for extrapulmonary TB in a group of patients
from Madagascar. The specificity was 83.3%. Overall,
the ICT assay is not sufficiently predictive for clinical
application.
The choice of a diagnostic test depends on the
pretest probability of a positive response, and there-
fore the prevalence of disease in the population being
tested. Such factors are especially important when
decisions regarding health care policy and screening
of large populations are being considered. McCon-
key et al. (2002) illustrated this point well by testing
the diagnostic yield of a rapid antibody card test in
areas of high prevalence of the disease (Cairo, Egypt)
and of low prevalence(St. Louis, Missouri, USA). In
Egypt, the antibody test was highly sensitive (87%)
for smear-positive and -negative pulmonary TB; the
specificity was 82%. The sensitivity and specificity in
St. Louis were 29% and 79%, respectively.
12.2.13
Dermal Response to MPB64 Antigen
Although not a serological assay, Nakamura and
colleagues (1998) recently reported that the MPB64
C. Kotaru and E. D. Chan
mycobacterial antigen delivered transdermally by the
patch test method was highly accurate for diagnosing
active TB. In this promising study, 52 of 53 patients
(98%) with active TB showed a positive reaction to
MPB64, while none of the 43 PPD-positive controls
were positive. This resulted in an overall sensitivity of
the test for active TB of 98.1%, a specificity of 100%,
and an accuracy of 98.9%.
12.2.14
Integral Membrane Antigens
of M. habana TMC 5135
Chaturvedi and Gupta (2002) evaluated the usefulness
of detecting antimycobacterial antibodies to antigens
that belong to the integral compartment of the plasma
membrane of M. habana in serum and body fluids of
patients with mainly extrapulmonary TB (only 4 of 42
patients had exclusively pulmonary TB). The presence
of antibodies against M. habana integral membrane
antigens (IMAs) was detected in 36 patients, for an
overall positivity rate of 86% (36/42). Anti-M. tubercu-
losis H37RA antibodies were also found in 29 out of 36
anti-M. habana IMA antibody-positive patients.
12.2.15
Lipopolysaccharide Antigen
Meena and co-workers (2002) recently reported the
isolation of an immunodominant lipopolysaccharide
(LPS) antigen from M. tuberculosis H37Rv for the
serodiagnosis of TB. Serum samples obtained from
59 Indian patients (19 patients with active pulmonary
TB, 20 with extrapulmonary TB, and 20 with nontu-
berculous pulmonary disease) and 20 healthy adults
were tested for LPS and three other commercially
available antigen assays. The presence of IgG against
LPS was highly sensitive and specific (84 and 97%,
respectively), similar to the test characteristics of
A60 IgG and superior to the assays for antibodies to
38 kDa or p90 antigens.
12.3
Specific Targeted Populations
In certain individuals, the diagnosis of active TB is
particularly challenging. Such persons include young
children, the elderly, HIV-positive individuals, and
patients with extrapulmonary TE. This is due to the
Serologic Testing for Tuberculosis
greater difficulty in obtaining a sputum sample, the
lower incidence of acid-fast bacilli in respiratory
samples, and/or the indolent or atypical presentation
of TB in such groups of people.
12.3.1
Childhood TB
Although childhood TB represents a small percentage
of all cases, infected children are a reservoir for many
adult cases (Khan and Starke 1995). The bacteriologic
diagnosis of TB is particularly difficult in young chil-
dren because of the greater incidence of disseminated
disease and because children often do not produce an
adequate amount of sputum. Thus, serological meth-
ods to diagnose childhood TB have great potential and
relevance. For children with clinical TB, the detection
of IgG and IgM against A60 antigen is comparable to
that seen in adults. In a study that examined the utility
of detecting specific IgM, IgA, and IgG antibodies to
A60 antigen in childhood TB on the Indian subconti-
nent, the overall sensitivity and specificity of the test
with combined IgA and IgM testing were 76% and
92%, respectively (Gupta et al. 1997). In children who
were acid-fast positive, the sensitivity rose to 95%.
For 150 children with extrapulmonary TB, the overall
sensitivity was 77%. In a study to detect IgG against
an autoclaved suspension of H37Rv in 132 clinical
cases of childhood TB, the sensitivity of the test in
the culture-positive group (n=35) was 69% (Hussey et
al. 1991). Interestingly, there was a positive correlation
between the titer level and increasing age. Prior BCG
did not affect the ELISA result. However, in another
study that examined IgM and/or IgG against A60
antigen in primary TB or TB adenitis in children, the
authors concluded that the test was not diagnostically
useful (Turneer et al.1994). Levels of antiglycolipid IgG
were found to be higher in children with active TB
than in nontuberculous controls, leading the authors
to conclude that the detection of IgG antibodies
against glycolipids was useful as a complementary
technique in the serodiagnosis of pulmonary TB in
children (Simonneyet al. 2000).
12.3.2
TB in the Elderly
Despite the fact that the clinical presentation of TB
in the elderly shares features with disease in younger
people, TB in the aged frequently goes unrecognized
(Fulton and McCallioin 1987). Differing radiographic
193
patterns of primary infection in the elderly compared
with young adults may lower the health care provider's
suspicion ofTB (Morris 1989). In adults, the incidence
of active TB cases is higher in the elderly (Dorken et al.
1987). Miliary TB, often presenting without localizing
signs or symptoms, is also more common in the elderly
person than in young adults, although the diagnosis
is frequently not made until autopsy (Korzeniewska
et al. 1994). Reactivation is associated with co-morbid
diseases that result in waning cell-mediated immunity,
disclosed by reduced skin test positivity to PPD with
increasing age. For example, in men, PPD reactivity
drops from 50% at age 65-74 years to 10% at age 95+,
and in women from 40% to about 5%, respectively
(Dorken et al. 1987). Because of the high rate of con-
version in the absence of symptoms in nursing home
residents, a two-step skin test is recommended for resi-
dents on admission to the nursing home: an initial test
and, if negative or equivocal, a repeat test 1-2 weeks
later. A positive booster effect is defined as an increase
of 6 mm or more from an induration <10 mm on the
first test to 10 mm with the second test. Thus, this
information at baseline can help to distinguish newly
acquired disease from a prior infection.
The bacteriologic diagnosis of TB in the elderly
is generally more problematic due to the greater dif-
ficulty in obtaining sputum and the greater incidence
of disseminated disease. Unfortunately, there has not
been a systematic evaluation of the serologic diagnosis
of active TB in the elderly. Whether age per se impairs
the immune system and thus may affect the antibody
response to bacterial antigens is controversial. For
example, although a decline in the adaptive immunity
has been observed in the elderly, in many cases it is not
clear whether the defect is primary or secondary to an
underlying systemic disorder such as cancer or diabe-
tes.An age-associated decline in humoral immunity has
been observed, as evinced by a lower rate ofseroconver-
sion and decreased antibody titer after vaccination; this
effect may largely be due to loss ofT-helper cell effector
function (Miller 1996). We believe a prospective study is
needed to determine the accuracy of serologic assays in
the diagnosis of pulmonary and extrapulmonary TB in
the elderly. This issue is especially relevant in the context
of nursing home-associated TB, where transmission is
not an uncommon occurrence.
12.3.3
HIV-Positive Individuals
The ability to detect specific antibodies against M.
tuberculosis antigens is substantially diminished in
194
HIV-positive individuals. As noted above, the sensi-
tivity of anti-LAM IgG in HIV-positive patients was
uniformly lower than in HIV-negative patients. In
Ugandan patients with TB who were also HIV-posi-
tive, the sensitivity of IgG antibody (ELISA) to the
30 kDa antigen of M. tuberculosis was only 28% com-
pared with 62% for patients who were HIV-negative
(Daniel et al. 1994). An ELISA and EIA to detect the
30 kDa antigen revealed an extremely low sensitivity
(l1 % and 0%, respectively) in HIV-positive patients
(McDonough et al. 1992). The level of immunosup-
pression may also affect the antibody response as
evidenced by an earlier study, conducted prior to
the inclusion of TB as an AIDS-defining criteria,
which showed that anti-PPD IgG was positive in 8 of
22 patients (36%) co-infected with HIV and TB but
without another criteria for AIDS vs 1 of 20 patients
(5%) with AIDS (Barrera et al.1992). Thus, the gener-
ally poor negative predictive value of serologic assays
for TB in HIV-positive individuals makes them unre-
liable and not cost-effective.
12.3.4
Extrapulmonary 18
An ELISA to detect IgG to a 43 kDa antigen of M.
tuberculosis in various body fluids found it to be
very sensitive (-100%) and highly specific [95.7% for
cerebrospinal fluid (CSF) and 98.1% for pleural and
ascitic fluid] (Wadee et al.1990).An antibody response
to 38 kDa antigen was noted in 73% of patients with
extrapulmonary TB, comparable to the 70% sensitivity
rate for smear-negative pulmonary TB (Wilkins and
Ivanyi 1990). In four groups of patients with various
degrees of confirmed neurologic TB (postmortem
proved, culture proved, clinically suspected, and
tuberculoma), antibody to the 38 kDa antigen was
detected in the CSF in 60%,80%,62.5%, and 0% of the
patients, respectively (Kadival et al. 1994). Detection of
the 30 kDa antigen using a monoclonal antibody was
done on the sera of 51 Mrican patients with clinically
diagnosed tuberculous pericardial effusion, of whom
25 had confirmation by pericardial fluid cultures.
The sensitivity was 61% and specificity 96% (Ng et
al. 1995). Zou et al. (l994) showed that IgG against
A60 antigen was present in 88.5% with reactivation
TB, and 69%-86% of patients with extrapulmonary
TB. Similarly, Lub and co-workers (l996) found the
sensitivity rate for anti-A60 IgG to be 76% for pulmo-
nary TB and 59% for patients with extrapulmonary
TB. Compared with children with pulmonary TB, the
IgG antibody response to glycolipid was significantly
C. Kotaru and E. D. Chan
lower in patients with extrapulmonary TB (Simonney
et al. 2000). Gupta and co-workers (l997) found that the
sensitivity of IgA and IgM directed against A60 antigen
was 72.4% for patients with pulmonary TB and 76.6%
for patients with extrapulmonaryTB. In Indian patients
with pulmonary and extrapulmonary TB, the sensitiv-
ity rate for IgA and IgG antibody directed against A60
was 98.3% for pulmonary TB, 88% for pleural TB, and
86% for extrapulmonary TB (Gupta et al. 1995). Overall,
an assay for antibody against mycobacterial antigens in
either serum or body fluids in patients with extrapul-
monary TB holds great potential.
12.3.5
Smear-Negative vs Smear-Positive 18
If acid-fast smear of sputum or body fluids is posi-
tive, the diagnosis of TB or atypical mycobacteria is
rarely questioned. However, -50% of sputum smear
in patients with bacteriologic-confirmed pulmonary
TB are smear-negative. Thus, an accurate seroassay
may be particularly useful in this group of patients.
Charpin and colleagues (l990) specifically examined
patients with smear-negative, culture-positive TB
using a combined IgG and IgM seroassay for the A60
antigen and demonstrated the sensitivity to be 68%,
specificity 100%, and positive predictive value 100%.
We had previously shown in our immunoblot assay for
anti-LAM IgG that five patients with smear-negative,
culture-positive TB had a positive seroassay (Chan
et al. 2000). Few studies have directly compared the
sensitivity rate of TB patients who are smear-nega-
tive and culture-positive vs those who are smear- and
culture-positive. As shown in Table 12.2, the sensitiv-
ity rate for smear-negative patients is either similar
or lower than for smear-positive cases. The sensitiv-
ity for the MycoDot assay for LAM was especially low
for both groups, likely due to the fact that over 50%
of the subjects were HIV-positive (Somi et al. 1999).
Unfortunately, in those studies that directly compared
smear-negative vs smear-positive cases, the majority
(Table 12.2) showed suboptimal sensitivity and thus
sufficiently low negative predictive value to be unreli-
able at present in smear-negative cases.
12.4
Conclusion
In summary, the appropriate setting for the use of
serological tests to diagnose M. tuberculosis dis-
Serologic Testing for Tuberculosis 195

Table 12.2. Sensitivity of smear-negative vs smear-positive tuberculosis

Antibody detection Smear-positive, Smear-negative, Reference

to antigen 'X' culture-positive culture-positive

Glycolipid antigen 96% 97% Dogan and Aksu (1997)

19kDa 57% 55% Botharnley et al. (1992)
LAM (MycoDot)" 26% 7% Somi et al. (1999)
LAM (ELISA) 88% 67% Sada et al. (1992)
P32 antigen of IgA 33% IgA 24% Turneer et al. (1988)
M. bovis BCG IgG 52% IgG 35%
38kDa 45-89% 16-82% Cole et al. (1996);
Botharnley and Rudd (1994);
Chan et al. (1990)
P90 75% 68% Alifano et al. (1997)
Glycolipid 90% 70% Maekura et al. (2001)

a 51% of the patients were HIV-positive

ease remains to be definitively established. Obvi- References

ous advantages to serologic diagnostic tests are a
greater sensitivity compared with acid-fast smear
and a more rapid turnaround time than culturing.
However, in our opinion, these tests at the current
state of their development cannot be considered as
an alternative to the conventional methods for the
following reasons:
• An overall concern is whether any of these tests
offers any significant advantage in the clinical set-
ting over the gold standard of cultivation, which
is still necessary to obtain the drug-susceptibility
pattern. Importantly, whether the application of
these tests improves the patient outcome has yet
to be determined.
• Suboptimal sensitivity of the serologic assays for
patients with smear-negative TB diminishes the
utility of these tests because this is the group of
patients that is considered potentially to derive
the most benefit from these nonconventional
tests. Similarly, the poor negative predictive value
of serologic assays for TB in HIV-positive indi-
viduals should, in its current state, preclude its use
in AIDS patients.
• The tests are associated with relatively high costs
and greater personnel training requirements,
issues that are of paramount concern especially
in developing countries.
• The relatively low positive predictive value due to
the decreased ability to distinguish between active
and latent TB infection.
• The difficulty serological tests have distinguishing
between M. tuberculosis vs atypical mycobacteria
infection.
Aide SLM et al (1989) Evaluation of an enzyme-linked immu-
nosorbent assay (ELISA) using an IgG antibody to Myco-
bacterium tuberculosis antigen 5 in the diagnosis of active
tuberculosis in children. Am Rev Respir Dis 139:748-751
Alifano M et al (1997) Evaluation of IgA-mediated humoral
immune response against the mycobacterial antigen P-
90 in diagnosis of pulmonary tuberculosis. Chest Ill:
601-605
Amicosante M et al (1995) Evaluation of a novel tuberculosis
complex-specific 34 kDa protein in the serological diagno-
sis of tuberculosis. Eur Respir J 8:2008-2014
Anuradha S et al (2001) Serodiagnosis of extra-pulmonary
tuberculosis using A-60 antigen. J Commun Dis 33:12-16
Arikan S et al (1998) Anti-Kp 90 IgA antibodies in the diagno-
sis of active tuberculosis. Chest 114:1253-1257
Barrera L et al (1992) Humoral response to Mycobacterium
tuberculosis in patients with human immunodeficiency
virus infection. Tuberc Lung Dis 73:187-191
Boggian K, Fierz W, Vernazza PL (1996) Infrequent detection
of lipoarabinomannan antibodies in human immuno-
deficiency virus-associated mycobacterial disease. J Clin
MicrobioI34:1854-1855
Bothamley GH, Rudd RM (1994) Clinical evaluation of a sero-
logical assay using a monoclonal antibody (TBn) to the
38 kDa antigen of Mycobacterium tuberculosis. Eur Respir
J 7:240-246
Bothamley G et al (1992) Serodiagnostic value of the 19 kilo-
dalton antigen of Mycobacterium tuberculosis in Indian
patients. Eur J Clin Microbiol Infect Dis 11:912-915
Brooks JB et al (1987) Selective procedures for detecting fem-
tomole quantities of tuberculostearic acid in serum and
cerebrospinal fluid by frequency-pulsed electron capture
gas-liquid chromatograph. J Clin MicrobioI25:1201-1206
Chan ED et al (2000) Diagnosis of tuberculosis by a visually
detectable immunoassay for lipoarabinomannan. Am J
Respir Crit Care Med 161:1713-1719
Chan SL et al (1990) Serodiagnosis of tuberculosis using an
ELISA with antigen 5 and a hemagglutination assay with
glycolipid antigens. Am Rev Respir Dis 142:385-390
Charpin D et al (1990) Value of ELISA using A60 antigen in
the diagnosis of active pulmonary tuberculosis. Am Rev
Respir Dis 142:380-384
196
Chaturvedi V, Gupta HP (2002) Evaluation of integral mem-
brane antigens of Mycobacterium habana for serodiagno-
sis of extrapulmonary tuberculosis: association between
levels of antibodies and Mycobacterium tuberculosis anti-
gens. FEMS Immunol Med MicrobioI33:1-7
Chiang 1-H et al (1997) Serodiagnosis of tuberculosis: A study
comparing three specific mycobacterial antigens. Am J
Respir Crit Care Med 156:906-911
Cocito CG (1991) Properties of the mycobacterial antigen
complex A60 and its applications to the diagnosis and
prognosis of tuberculosis. Chest 100:1687-1693
Cole RA et al (1996) Clinical evaluation of a rapid immuno-
chromatographic assay based on the 38 kDa antigen of
Mycobacterium tuberculosis on patients with pulmonary
tuberculosis in China. Tuberc Lung Dis 77:363-368
Daniel TM, Debanne SM, van der Kuyp F (1985) Enzyme-
linked immunosorbent assay using Mycobacterium
tuberculosis antigen 5 and PPD for the serodiagnosis of
tuberculosis. Chest 88:388-392
Daniel TM et al (1986) Field evaluation of enzyme-linked
immunosorbent assay for the serodiagnosis of tuberculo-
sis. Am Rev Respir Dis 134:662-665
Daniel TM et al (1994) Reduced sensitivity of tuberculosis
serodiagnosis in patients with AIDS in Uganda. Tuberc
Lung Dis 75:33-37
Del Prete R et al (1998) Detection of anti-lipoarabinomannan
antibodies for the diagnosis of active tuberculosis. Int J
Tuberc Lung Dis 2:160-63
Diagbouga S et al (1997) Immunoblot analysis for serodi-
agnosis of tuberculosis using a 45/47-kilodalton antigen
complex of Mycobacterium tuberculosis. Clin Diag Lab
ImmunoI4:334-338
Dogan UB, Aksu HSZ (1997) Serodiagnostic value of ELISA
in pulmonary tuberculosis in Turkey where tuberculosis is
highly prevalent. Respiration 64:73-75
Dorken E, Grzybowski S, Allen EA (1987) Significance of the
tuberculin test in the elderly. Chest 92:237-40
Escamilla L et al (1996) Mycobacterium fortuitum glycolipids
for the serodiagnosis of pulmonary tuberculosis. Am J
Respir Crit Care Med 154:1864-1867
Fadda G et al (1992) Serodiagnosis and follow up of patients
with pulmonary tuberculosis by enzyme-linked immuno-
sorbent assay. Eur J Epidemiol 8:81-87
Franco J et al (2001) Serological response (Western blot) to
fractions of Mycobacterium tuberculosis sonicate antigen
in tuberculosis patients and contacts. Int J Tuberc Lung
Dis 5:958-962
Fulton JD, McCallioin J (1987) Tuberculosis - diagnostic dif-
ficulty in the elderly. J Clin Exp Gerontol 9:303-11
Gupta S, Bhatia R, Datta KK (1997) Serological diagnosis of
childhood tuberculosis by estimation of mycobacterial
antigen 6O-specific immunoglobulins in the serum. Tuberc
Lung Dis 78:21-27
Gupta S et al (1995) Diagnostic utility of the estimation of
mycobacterial antigen A60 specific immunoglobulins IgM,
IgA, and IgG in the sera of cases of adult human tubercu-
losis. Tuberc Lung Dis 76:418-24
He H et al (1991) Rapid serodiagnosis of human mycobacte-
riosis by ELISA using cord factor (trahalose-6,6'-dimyco-
late) purified from Mycobacterium tuberculosis as antigen.
FEMS Microbiol ImmunoI3:201-204
Hussey G, Kibei M, and Dempster W (1991) The serodiagnosis
of tuberculosis in children: an evaluation of an ELISA test
C. Kotaru and E. D. Chan
using IgG antibodies to M. tuberculosis, strain H37Rv. Ann
Trop Paediatr 11:113-118
Imaz MS et al (2001) Evaluation of the diagnostic value of
measuring IgG, IgM and IgA antibodies to the recom-
binant 16-kilodalton antigen of Mycobacterium tubercu-
losis in childhood tuberculosis. Int J Tuberc Lung Dis 5:
1036-1043
Julian E et al (1997) Detection of lipoarabinomannan anti-
bodies in patients with newly acquired tuberculosis and
patients with relapse tuberculosis. J Clin Microbiol 35:
2663-2664
Kadival GV et al (1994) Detection of antibodies to defined
M. tuberculosis antigen (38 kDa) in cerebrospinal fluids of
patients with tuberculous meningitis. Zentralblatt Bakte-
rioI281:95-101
Khan EA, Starke JR (1995) Diagnosis of tuberculosis in chil-
dren: Increased need for better methods. Emerg Infect Dis
1:115-123
Khomenko AG et al (1996) Serodiagnosis of tuberculosis:
detection of mycobacterial antibodies and antigens. Tuberc
Lung Dis 77:510-515
Korzeniewska M et al (1994) Tuberculosis in young adults and
the elderly. Chest 106:28-32
Lawn SD, Frimpong EH, Nyarko E (1997) Evaluation of a com-
mercial immunodiagnostic kit incorporating lipoarabino-
mannan in the serodiagnosis of pulmonary tuberculosis in
Ghana. Trop Med Int Health 2:978-981
Luh KT et al (1996) Tuberculosis antigen A60 serodiagnosis
in tuberculous infection: application in extrapulmonary
and smear-negative pulmonary tuberculosis. Respirology
1:145-151
Ma Y, Wang YM, Daniel TM (1986) Enzyme-linked immuno-
sorbent assay using Mycobacterium tuberculosis antigen 5
for the diagnosis of pulmonary tuberculosis in China. Am
Rev Respir Dis 134:1273-1275
Maekura R et al (1993) Clinical evaluation of rapid serodiag-
nosis of pulmonary tuberculosis by ELISA with cord factor
(trehalose-6,6'-dimycolate) as antigen purified from Myco-
bacterium tuberculosis. Am Rev Respir Dis 148:997-1001
Maekura R et al (2001) Clinical evaluation of anti-tubercu-
lous glycolipid immunoglobulin G antibody assay for rapid
serodiagnosis of pulmonary tuberculosis. J Clin Microbiol
39:3603-3608
McConkey SJ et al (2002) Evaluation of a rapid-format anti-
body test and the tuberculin skin test for diagnosis of
tuberculosis in two contrasting endemic settings. Int J
Tuberc Lung Dis 6:246-252
McDonough JA et al (1992) Microplate and dot immunoassays
for the serodiagnosis of tuberculosis. J Lab Clin Med 120:
318-322
Meena LS et al (2002) Comparative study of three different
mycobacterial antigens with a novel lipopolysaccharide
antigen for the serodiagnosis of tuberculosis. J Clin Lab
Anal 16:151-155
Miller RA (1996) The aging immune system: primer and pro-
spectus. Science 273:70-74
Morris CDW (1989) The radiology, haematology, and bio-
chemistry of pulmonary tuberculosis in the aged. Q J Med
71:529-535
Nakamura RM et al (1998) MPB64 mycobacterial antigen:
a new skin-test reagent through patch method for rapid
diagnosis of active tuberculosis. Int J Tuberc Lung Dis 2:
541-546
Serologic Testing for Tuberculosis
Ng TT, Strang JI, Wilkins EG (1995) Serodiagnosis of pericar-
dial tuberculosis. Quart J Med 88:317-320
Niculescu D et al (1999) Serodiagnosis of extrapulmonary
tuberculosis by enzyme-linked immunosorbent assay
(ELISA). Roum Arch Microbiol Immunol 58: 111-119
Nsanze H et al (1997) Serodiagnosis of tuberculosis and
leprosy by enzyme immunoassay. Clin Microbiol Infect
3:236-239
Rasolofo V et al (2000) Predictive values of the ICT Tuber-
culosis test for the routine diagnosis of tuberculosis in
Madagascar. Int J Tuberc Lung Dis 4:184-185
Ratanasuwan W et al (1997) Evaluation of the MycoDot test
for the diagnosis of tuberculosis in HIV seropositive and
seronegative patients. Int J Tuberc Lung Dis 1:259-264
Rosen EU (1990) The diagnostic value of an enzyme-linked
immunosorbent assay using adsorbed mycobacterial soni-
cates in children. Tubercle 71:127-130
Rovatti E et al (1996) Evaluation of a western blot serum test
for the diagnosis of Mycobacterium tuberculosis infection.
Eur Respir J 9:288-292
Sada E et al (1990) Evaluation of lipoarabinomannan for the
serological diagnosis of tuberculosis. J Clin Microbiol 28:
2587-2590
Sada E et al (1992) Detection oflipoarabinomannan as a diag-
nostic test for tuberculosis. J Clin Microbiol 30:2415-2418
Simonney N, Bourrillon A, Lagrange PH (2000) Analysis of
circulating immune complexes (CICs) in childhood tuber-
culosis: levels of specific antibodies to glycolipid antigens
and relationship with serum antibodies. Int J Tuberc Lung
Dis 4:152-160
Simonney N et al (1996) Comparison of A60 and three gly-
colipid antigens in an ELISA test for tuberculosis. Clin
Microbiol Infect 2:214-222
Somi GR et al (1999) Evaluation of the MycoDot test in
197
patients with suspected tuberculosis in a field setting in
Tanzania. Int J Tuberc Lung Dis 3:231-238
Steele BA, Daniel TM (1991) Evaluation of the potential role
of serodiagnosis of tuberculosis in a clinic in Bolivia by
decision analysis. Am Rev Respir Dis 143:713-716
Torres M et al (1994) Comparison of the immune response
against Mycobacterium tuberculosis antigens between a
group of patients with active pulmonary tuberculosis and
healthy household contacts. Clin Exp ImmunoI96:75-78
Turneer M et al (1988) Humoral immune response in human
tuberculosis: Immunoglobulins G, A, and M directed against
the purified P32 protein antigen of Mycobacterium bovis
bacillus Calmette-Guerin. J Clin Microbiol 26: 1714-1719
Turneer M et al (1994) Determination of humoral immuno-
globulins M and G directed against mycobacterial antigen
60 failed to diagnose primary tuberculosis and mycobacte-
rial adenitis in children. Am J Respir Crit Care Med 150:
1508-1512
Van Vooren JP et al (1988) A multiblot immunobinding assay
for the serodiagnosis of tuberculosis. J Immunological
Methods 113:45-49
Wadee AA, Boling L, Reddy SG (1990) Antigen capture assay
for detection of a 43-kilodalton Mycobacterium tuberculo-
sis antigen. J Clin MicrobioI28:2786-2791
Wilkins EGL, Ivanyi J (1990) Potential value of serology for
diagnosis of extrapulmonary tuberculosis. Lancet 336:
641-644
Yu CJ et al (1992) Evaluation of the A-60 IgG ELISA serodi-
agnostic test for tuberculosis in Taiwan. J Formosan Med
Assoc 91:614-619
Zou YL et al (1994) Serological analysis of pulmonary and
extrapulmonary tuberculosis with enzyme-linked immu-
nosorbent assays for anti-A60 immunoglobulins. Clin
Infect Dis 19:1084-1091
13 peR and Diagnosis of Tuberculosis
DIANA L. WILLIAMS
CONTENTS
13.1 Polymerase Chain Reaction 199
13.2 PCR of M. tuberculosis 200
13.3 Drug Resistance in M. tuberculosis 202
13.4 PCR Detection of Drug Resistance 204
13.5 Impact of PCR Diagnosis on Patient Care 206
13.6 Summary 207
References 208
The current 'gold standard' for the identification
of M. tuberculosis in clinical specimens is based
on acid-fast microscopy of sputum sediments or
other specimens (AFB smears), followed by culture
confirmation using definitive biochemical or DNA
probe analyses. Although AFB smear results can be
available within 1 day, they are only 50%-70% sensi-
tive and do not distinguish between M. tuberculosis
and other nontuberculosis mycobacteria that may
be present in specimens. Traditional culture-based
methodologies for the detection of M. tuberculosis
require between 1 and 8 weeks to perform and often
have low sensitivity when small numbers of organ-
isms are analyzed (AFB smear-negative). In addition,
culture-based drug-susceptibility testing requires
several additional weeks to identify drug-resistant
M. tuberculosis. These factors make patient diagno-
sis extremely difficult, and therefore, many patients
are routinely started on treatment with minimal
diagnostic information. This potentially results in
postponing appropriate treatment and the failure to
identify rapidly those patients who need to be iso-
lated from the general population because they are
harboring drug-resistant organisms. Therefore, the
rapid and specific diagnosis of tuberculosis is one of
the most pressing needs in the effort to control and
eventually eradicate this disease.
D. 1. WILLIAMS, PhD
Molecular Biology Research Department, Laboratory Research
Branch, National Hansen's Disease Programs at LSU-SVM,
Rm 3517W, Skip Bertman Dr., Baton Rouge, LA 70803, USA
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
Molecular approaches to aid the diagnosis of
human disease have begun to have a major influence
on the current clinical management of various dis-
eases. A major goal of tuberculosis research over the
last decade has been to apply molecular approaches
to develop rapid, reliable procedures that can detect
M. tuberculosis directly from clinical specimens and
thereby avoid the many weeks required for culture
amplification. The result has been the development
of numerous nucleic acid-based assays, including
ones based on strand displacement amplification
(Walker et al. 1992), transcription-mediated ampli-
fication (Jonas et al. 1993), oligonucleotide ligation
amplification (Iovannisci and Winn-Deen 1993),
Q-beta replicase amplification (Shah et al. 1995),
and polymerase chain reaction (PCR) (Mullis and
Faloona 1987) methodologies. Among them, PCR-
based assays have been tested on thousands of clini-
cal specimens and isolates in laboratories around the
world. The purpose of this chapter is to summarize
the progress that has been made in the past few years
in the development and characterization of PCR-
based assays for the rapid diagnosis of tuberculosis
and drug-resistant tuberculosis, to highlight several
critical differences between PCR-based testing and
conventional microbiology, and to discuss the impact
that PCR analysis can potentially have on patient care
and the control of tuberculosis.
13.1
Polymerase Chain Reaction
PCR is a nucleic acid-based methodology, originally
described in 1987, that uses short oligonucleotide
primers to direct the amplification of specific
segments of target DNA sequences (template) by
repeated cycles of denaturation of the template DNA,
primer annealing to this DNA, and primer extension
(Mullis and Faloona 1987). This allows exponential
multiplication of the specified template, so that
beginning with as few as one copy of this sequence,
200 D. L. Williams

as many as one billion copies (amplicons) can be
produced within 2-4 h of specimen processing. In
addition, because of the designed specificity, PCR is
a powerful technique for the detection of small num-
bers of bacterial DNA in crude biological specimens
without the necessity of culture.
PCR requires: (1) specifically selected primers; (2)
an appropriately prepared sample; (3) a thermostable
DNA polymerase, buffer, and dNTPs; (4) appropriate
positive and negative controls; (5) a thermocycler; and
(6) a detection assay. The specificity of PCR lies in the
selection of oligonucleotide primers used in this assay.
Therefore, primers are typically selected from specific
regions of the chromosome not found in other bacte-
rial or eukaryote DNA. Once identified, primers can
be synthetically manufactured and used specifically
to amplify template from crude biological specimens.
In addition, since the primers are incorporated in the
amplicon during the PCR reaction, primers can be
labeled with biotin or specific Buors for subsequent
amplicon detection as discussed below.
procedure kills M. tuberculosis and preserves bacte-
rial DNA. The fixed sediments can be stored at room
temperature indefinitely or can be pelleted, resus-
pended in sterile H20, and placed in a boiling water
bath for 10 min. These crude celllysates serve as an
excellent template for PCR or can be stored at -20°C
for later use.
PCR cycling parameters vary primarily because of
the melting temperature of the primers selected for
amplification. However, a standard protocol involves
an initial denaturation step of 94°C for 5 min followed
by a 30-45 cycle program consisting of: 30 s at 94°C;
30 s at the optimal primer annealing temperature; and
30 s at noc, the optimal extension temperature. This
program is followed by a final 5-10 min hold at n°e.
Once the template is amplified, many methods
are available for the detection of PCR amplicons.
One of the most commonly used and inexpensive
methods is the electrophoretic separation of ampli-
cons on agarose or polyacrylamide slab gels with
subsequent visualization of ampIicons in ethidium

Table 13.1. PCR-based assays for detection of M. tuberculosis in

13.2 clinical specimens
peR of M. tuberculosis Assay Target Reference(s)
'In-house'
Numerous PCR-based assays have been developed for IS6110 Eisenach et aI. (1990, 1991)
the detection of M. tuberculosis, and several of them Brisson-Noel et aI. (1991)
are summarized in Table 13.1. All of these assays use Kent et aI. (1995)
specific procedures to recover DNA efficiently from Mulcahy et aI. (1996)
DesJardin et aI. (1998)
low numbers of M. tuberculosis from clinical speci-
Almeda et aI. (2000)
mens and to remove potential contaminants that may Borun et aI. (2001)
inhibit PCR. Sputum specimens are typically used rpoB Williams et aI. (1 998b)
for the PCR detection of M. tuberculosis, but bron- hsp65 Brisson-Noel et aI. (1991)
chial alveolar washes, skin biopsies, cerebrospinal Shinnick and Jonas (1994)
MPB70 Cousins et aI. (1992)
fluid, blood, gastric lavage fluid, fecal specimens, and
MTB40 DelPortilio et aI. (1991)
tissue biopsy specimens have also been used. 32 kDa antigen Soini et aI. (1992)
The standard protocol for preparing sputum for 65 kDa/IS611O Yeboah-Manu et aI. (2001)
PCR uses N-acetyl-L-cysteine sodium hydroxide (multiplex)
(NALC-NaOH) treatment to digest and decontami- 16S rRNA gene Bottger et aI. (1989)
Boddinghaus et aI. (1990)
nate specimens (Kent and Kubica 1985). The next
Commercial:
step is the efficient lysis of the bacteria. This can be Roche Amplicor 16S rRNA gene www.roche-diagnostics.com
accomplished by sonication, treatment with sodium MTB Bergmann and Woods (1996)
dodecyl sulfate (SDS) plus lysozyme plus heat, pro- Bennedsen et aI. (1996)
tease treatment, and chaeotropic salts followed by Dalovisio et aI. (1996)
Soini et aI. (1996)
a nucleic acid purification step. This insures that
Forbes (1997)
the majority of potential inhibitors of the PCR are Mitarai et aI. (2001)
removed. However, one easy and commonly used Gomez-Pastrana et aI. (2001)
procedure for the preparation of sputum sediments RocheCOBAS 16S rRNA gene www.roche-diagnostics.com
for PCR is the fixation of a portion (100-250 Ill) of Amplicor MTB Eing et aI. (1998)
Bogard et aI. (2001)
these sediments in a final concentration of 70%
Oh et aI. (2001)
ethanol for at least 2 h (Williams et al. 1995). This
peR and Diagnosis of Tuberculosis 201

bromide-stained gels using UV-light transillumina- Because of the large number of copies of this ele-
tion. Documentation of results can be accomplished ment in most strains (1-25) and the random inser-
using a standard or Polaroid camera fitted with the tion of this element in different strains of M. tuber-
appropriate orange filter or by using a computerized culosis, the IS6110 PCR assay has been modified to
image acquisition and analysis instrument, such as a produce a DNA fingerprint of a strain (Thierry et al.
GelDoc 2000 (BioRad Laboratories, Hercules, CA). 1990; Cave et al.1991). This analysis and others based
To increase the sensitivity of amplicon detection, on repeated insertion elements in the genome of M.
amplicons can be transferred from gels to nitrocellu- tuberculosis have been used to type clinical isolates
lose or nylon membranes using a simple capillary dif- for epidemiological studies (Thierry et al.1990; Cave
fusion-based nucleic acid transfer technique or South- et al. 1991; Plikaytis et al. 1993; van Embden et al.
ern blotting (Southern 1975). PCR products can then 1993; van Soolingen et al. 1994; Otal et al. 1997).
be visualized using one of many hybridization-based Other 'in-house' PCR assays target many different
detection techniques with radiolabeled, chemilumines- genes and repeated sequences within the M. tubercu-
cent, or fluorescent-labeled probes. Some protocols use losis genome (Table 13.1). Even though these assays
reverse hybridization of amplicons to probes bound on rely on the amplification of different targets, they
nitrocellulose filters (de Beenhouwer et al. 1995) or to all have similar specimen preparation and specific-
the wells of microtiter plates (Amplicor MTB, Roche ity for smear-positive specimens. However, because
Molecular Systems, Emery, CA, USA). In addition, one the majority of these assays are based on single copy
commercially available assay provides amplification genes, they are not as sensitive as the IS6110 assay.
and detection in a semiautomated system (COBAS In addition to these M. tuberculosis-specific PCR
Amplicor MTB, Roche Molecular Systems). assays, another 'in-house' PCR assay has been devel-
The majority of the assays developed for the detec- oped which identifies the presence of M. tuberculosis
tion of M. tuberculosis are referred to as 'in-house' or other mycobacteria in clinical specimens. This
PCRs because they have been developed and charac- assay uses mycobacterial genus-specific primers
terized by individual laboratories. While these assays which flank a hypervariable region of the 16S rRNA
offer the potential for diagnosing tuberculosis with gene in which species-specific differences have been
a high degree of sensitivity and specificity in a few characterized (Bottger et al. 1989). The PCR ampli-
hours, only a few have been successfully transferred cons are then subjected to DNA sequencing to dis-
to the clinical laboratory. The IS611 0 'in-house' PCR criminate between mycobacterial species (Bottger et
assay, originally developed by Eisenach et al. in 1990, al. 1989; Boddinghaus et al. 1990). This analysis has
targets the multicopy IS611 0 insertion element of M. proven to be a rapid and reliable method for the iden-
tuberculosis. This assay, or variations of it, have been tification of mycobacteria in the clinical laboratory
successfully used in many clinical laboratories to (Kirschner et al. 1993).
detect M. tuberculosis (Table 13.1). Presently, there are only a few commercially avail-
When the IS6110 PCR assay was evaluated for its able nucleic acid amplification assays for the direct
ability to detect M. tuberculosis in sputum specimens detection of tuberculosis from clinical specimens
routinely processed in several clinical laboratories, the (Table 13.1). An example of one of these assays is the
following conclusions were made: (1) this PCR was Amplicor MTB assay, which is based on PCR ampli-
easy to fit into the daily routine of a clinicallabora- fication of the 16S rRNA gene and a colorimetric
tory with a turnaround time of 24-36 h from time of detection system. This assay is recommended for the
specimen collection; (2) existing laboratory personnel detection of M. tuberculosis in liquefied, decontami-
could perform these assays; and (3) false-positives, due nated, and concentrated human respiratory speci-
to amplicon contamination of specimens, were rare mens, including expectorated and induced sputum,
when personnel strictly adhered to specific laboratory bronchial washings, and bronchial alveolar lavages
procedures (Clarridge et al. 1993; Shalwar et al. 1993; (www.roche-diagnostics.com).
Nolte et al. 1993). In addition, when the results of this Clinical trials to evaluate the reliability of this
study were compared with culture-based detection for assay for the diagnosis of pulmonary tuberculosis
the same specimens, there was 100% specificity for M. have concluded that this assay has excellent sensitiv-
tuberculosis and 95% sensitivity with smear-positive, ity, specificity, and positive and negative predictive
culture-positive specimens. However, the PCR-based values for smear-positive, culture-positive specimens
assay had a lower sensitivity for smear-negative, cul- when compared with other 'in-house' assays (Berg-
ture-positive specimens, thus being of lesser value for mann and Woods 1996; Eing et al. 1998). One such
the early detection of disease development. trial evaluated 956 respiratory specimens from 502
202
patients and compared the results of the Amplicor
MTB assay with those of culture. For 135 specimens
that were culture-positive for mycobacteria, the sensi-
tivity, specificity, and positive and negative predictive
values of the Amplicor MTB assay were 79.4%,99.6%,
92.6%, and 98.6%, respectively. For AFB smear-posi-
tive specimens, the sensitivity, specificity, and posi-
tive and negative predictive values of Amplicor MTB
were 97.6%, 100%, 100%, and 90.9%, respectively.
For AFB smear-negative specimens, the sensitivity,
specificity, and positive and negative predictive values
were 40.0%,99.5%,69.2%, and 98.7%, respectively. In
another study, the sensitivity was ~95% and specificity
100% in smear-positive specimens (Woods 2001).
Another commercially available PCR assay is
the Roche COBAS Amplicor MTB assay. This assay
is a semiautomated version of the manually per-
formed Roche Amplicor MTB test (www.roche-
diagnostics.com). When compared with culture and
an IS611O-based 'in-house' PCR protocol, the sensi-
tivity, specificity, and positive and negative predictive
values for this assay were 66.33%, 99.71 %, 94.36%,
and 97.66%, respectively (Eing et al. 1998). The corre-
sponding values for the 'in-house' PCR were 91.08%,
99.85%, 97.87%, and 99.37%, respectively. Another
study showed similar results, giving an overall speci-
ficity of 99.7% and sensitivities ranging from 85.7%
to 90.9% (Bogard et al. 2001).
Therefore, it appears that these commercially
available assays have good sensitivity, specificity, and
predictive values for the detection of M. tuberculosis
in smear-positive and culture-positive specimens
(EI-Hajj et al. 2001). However, many laboratories,
particularly in developing countries, continue to use
methods developed 'in house'. This is primarily due to
a lack of funding required for the expensive reagents
and equipment used to perform these assays. In this
environment, great care should be taken in the vali-
dation of these 'in-house' assays and in maintaining
proper quality control.
In summary, many PCR-based assays have been
developed for the detection ofM. tuberculosis directly
from clinical specimens. Several of these assays have
been shown to be extremely sensitive for the direct
detection of M. tuberculosis from clinical specimens.
13.3
Drug Resistance in M. tuberculosis
Treating tuberculosis with combinations of effec-
tive antibiotics is a proven method for limiting the
D. L. Williams
emergence and spread of existing antibiotic-resistant
strains of M. tuberculosis. In most areas of the world,
current short-course multidrug therapy formulated
for tuberculosis (MDT-TB) includes a 2-month daily
combination of isoniazid (INH), rifampin (RMP), and
pyrazinamide (PZA) and either streptomycin (STM)
or ethambutol (EMB) followed by daily therapy with
INH and one of the other primary drugs for the next
4 months (Stratton and Reed 1986). However, because
of the length of MDT-TB, patient noncompliance often
occurs, particularly when direct observation of therapy
is not instituted (Frieden et al. 1993; Bradford et al.
1996). Since drug-resistant strains often emerge during
intermittent or inadequate chemotherapy (Frieden et
al. 1993; Bradford et al. 1996), clusters of disease caused
by drug-resistant strains are still reported globally
(Cohn et al. 1997; Moore et al. 1997). Therefore, there
is a great need for rapid drug-susceptibility testing to
monitor the evolution of drug resistance and track the
spread of multidrug-resistant tuberculosis (MDR-TB).
Conventional drug-susceptibility testing for tuber-
culosis relies on the cumbersome and time-consuming
culture-based assays which can take up to 2 months to
obtain results (Kent and Kubica 1985; Heifets 1991).
Accordingly, drug-susceptibility testing is not a routine
component of tuberculosis control programs in many
parts of the world. This limits the ability to monitor
the evolution of drug resistance and to establish the
success of a treatment strategy like MDT-TB.
Recently, substantial progress has been made in our
understanding of the molecular basis of M. tuberculo-
sis drug resistance as summarized by Musser (1995),
Blanchard (1996), Ramaswamy and Musser (1998) and
Cockerill (1999). Drug resistance in M. tuberculosis is
primarily attributed to chromosomal mutations in
genes encoding drug targets. These mutations occur
spontaneously as a result of errors in DNA replication.
For example, the frequency of drug-resistant mutants
in a population of M. tuberculosis is estimated at 10-6
for INHR mutants and 10-7 for RMp R mutants. These
frequencies suggest that one organism in a population
of 106 is resistant to INH and one organism in 107 is
resistant to RMP.
Since it has been estimated that some untreated,
smear-positive tuberculosis patients can harbor
>109 viable M. tuberculosis/gram of lung tissue, the
population of M. tuberculosis in such a patient could
contain thousands of drug-resistant organisms. Inap-
propriate therapy for patients with such large bacte-
rial loads selects for these subpopulations of drug-
resistant bacteria, resulting in the development of
drug-resistant tuberculosis and leading to the spread
of one or more resistant phenotypes.
PCR and Diagnosis of Tuberculosis 203

Isoniazid. INH is an effective, synthetic bactericidal
agent for M. tuberculosis that requires the bacterial
catalase peroxidase to convert it to its active form.
Resistance to INH in the majority of isolates corre-
lates with changes in the catalase enzyme as a result
of missense mutations, deletions, or insertions within
the katG which encodes the catalase of M. tubercu-
losis (Zhang et al. 1992; Musser 1995). In addition,
INH and ethionamide (ETH), a structural analogue
of INH, interfere with new cell wall synthesis in M.
tuberculosis (Winder et al. 1971). Resistance to these
drugs also develops as a result of defined mutations
within the inhA locus encoding the enoyl-ACP
reductase and the 3-ketoacyl-ACP reductase involved
with mycolic acid biosynthesis (Banerjee et al. 1994;
Musser et al. 1996; Ramaswamy and Musser 1998).
Additional mechanisms have been described for the
development of INH resistance in M. tuberculosis, but
all targets and modes of action of this drug in M.
tuberculosis are still unknown (Table 13.2).
Rifampin. RMP [3,4-(methylpiperazinyl-iminome-
thylidene-rifamycin)J, a semisynthetic derivative of
rifamycin, has been a key bactericidal component in
most antituberculosis chemotherapeutic regimens
since the early 1970s. RMP appears to exert its potent
bactericidal activity by binding to the DNA-depen-
dent RNA polymerase complex and uncoupling
transcription (Winder 1982). The target for RMP in
M. tuberculosis is the ~-subunit of the DNA-depen-
dent RNA polymerase encoded by rpoB (Telenti et
al. 1993a; Williams et al. 1994, 1998a). Resistance to
RMP in virtually all resistant isolates of M. tubercu-
losis correlates with changes in the structure of the
~-subunit of the DNA-dependent RNA polymerase
as a result of missense mutations, small deletions,
and insertions within an 81-base pair region of the
rpoB referred to as the RMP resistance-determining
region (RRDR) (Telenti et al. 1993a; Williams et al.
1994; Musser 1995; Ramaswamy and Musser 1998).
The resultant ~-subunit is still functional, but it is
refractory to the effects of RMP or requires much
higher levels of the drug to produce bacterial cell
death (Williams et al. 1998a).
Pyrazinamide. PZA is a structural analogue of
nicotinamide that is used as a first-line tuberculosis
drug. The mode of action of this drug is not under-
stood, but it is thought that PZA is converted to the
bioactive form of the drug by pyrazinamidase, an
enzyme produced by M. tuberculosis (Konno et al.
1967). PZAR isolates lose some or all pyrazinamidase

Table 13.2. PCR-based assays for detection of drug-resistant M. tuberculosis

Assay Drug Target Reference(s)

'In-house':
PCR-DNA sequencing INH katG Marttila et aI. (1996); Musser et aI. (1996)
inhA Musser et aI. (1996)
ahpC Sreevatsan et aI. (1997c)
RMP rpoB Telenti et aI. (1993b); Kapur et aI. (1994)
Williams et aI. (1994); Nachamkin et aI. (1997)
Rossau et aI. (1997); Bartfai et aI. (2001)
PZA pncA Sreevatsan et aI. (1997a)
EMB embB Sreevatsan et aI. (1 997b)
FQs gyrA Takiff et aI. (1994)
SM rrs, rpls Sreevatsan et aI. (1996); Nachamkin et al. (1997)
SSCP INH katG Heym et aI. (1995); Rouse et aI. (1995)
Victor et aI. (1996); Telenti et aI. (1993a, 1997b)
Temesgen et aI. (1997)
inhA Rouse et aI. (1995); Telenti et aI. (1997)
ahpC Telenti et aI. (1997a)
RMP rpoB Telenti et al. (1993a,b, 1997a,b)
Felmlee et aI. (1995); Whelen et aI. (1995)
Selvakumar et aI. (1997); Kim et aI. (2001)
EMB embB Sreevatsan et aI. (1 997b)
FQs gyrA Takiff et aI. (1994); Delgado and Telenti (1996)
Heteroduplex RMP rpoB Williams et aI. (1996, 1998b); Thomas et aI. (2001)
Molecular Beacons INH katG Piatek et aI. (2000)
RMP rpoB Piatek et aI. (1998,2000); EI-Hajj et aI. (2001)
Commercial:
Line probe assay (LiPA) RMP rpoB de Beenhouwer et aI. (1995); Rossau et aI. (1997)
Bartfai et aI. (2001)
204
activity (Butler and Kilburn 1983). This occurs as a
result of defined missense mutations within pncA
which encodes this enzyme (Sreevatsan et al. 1997a;
Ramaswamy and Musser 1998; Cockerill 1999).
Streptomycin. STR is an aminocyclitol glycoside
antibiotic that appears to function by binding to
the 16S rRNA of M. tuberculosis to inhibit transla-
tional initiation and detrimentally affects transla-
tion fidelity, which results in protein production
termination (Douglass and Steyn 1993; Finken et al.
1993; Ramaswamy and Musser 1998; Cockerill 1999).
Specific mutations within the 16S rRNA gene (rrs)
and rpsi gene encoding the ribosomal protein S12 in
M. tuberculosis are responsible for the development
of resistance to this drug (Douglass and Steyn 1993;
Finken et al. 1993; Sreevatsan et al. 1996; Ramaswamy
and Musser 1998; Cockerill 1999). However, not all
resistant isolates have mutations in these genes.
Ethambutol. EMB ([ S,S']- 2,2'-[ethylenedimino]di-l-
butanol) is a bactericidal drug that targets M. tubercu-
losis cell wall synthesis by inhibiting incorporation of
mycolic acids into the cell wall (Takayama et al. 1979;
Takayama and Kilburn 1989). This appears to be a func-
tion of three genes organized into an operon designated
as embCAB (Telenti et al.1997b). Mutations in a single
codon of emb are associated with the development of
EMB resistance in M. tuberculosis (Telenti et al. 1997;
Ramaswamy and Musser 1998; Cockerill 1999).
Fluoroquinolones. Ciprofloxacin, ofloxacin, and levo-
floxacin are fluorinated derivatives of quinolones
(FQs) (Hooper and Wolfson 1989) used primarily as
second-line drugs in the treatment of drug-resistant
tuberculosis (Cambau et al. 1994; Sullivan et al. 1995;
Takiff et al.1994; Ramaswamy and Musser 1998). The
target for this drug class is the DNA gyrase, an enzyme
that catalyzes negative supercoiling of DNA (Hooper
and Wolfson 1989). Specific missense mutations within
the quinolone resistance-determining region (QRDR)
of the gyrA, encoding the GyrA subunit of the gyrase,
are responsible for the development of resistance in M.
tuberculosis (Cambau et al. 1994; Sullivan et al. 1995;
Takiff et al. 1994; Ramaswamy and Musser 1998).
13.4
peR Detection of Drug Resistance
Knowledge of the specific mechanisms that result in
the development of antituberculosis drug resistance
D. L. Williams
has been used to develop a variety of rapid, PCR-based
assays for the detection of drug-resistant M. tuberculo-
sis directly from clinical specimens (Table 13.2). These
assays are based on the specific genotypic identifica-
tion of mutations associated with the drug-resistant
phenotype within amplicons derived from drug target
genes. Some of these assays include: direct PCR-DNA
sequencing, PCR single-strand conformation poly-
morphism (SSCP), PCR heteroduplex formation
analysis (PCR-HDA), PCR line probe assay (LiPA), and
molecular beacon sequence analysis (Table 13.2).All of
these assays can be performed on sputum specimens
or clinical isolates within 24 h of sample acquisition,
are highly specific and sensitive, and, therefore, have
the capability to significantly reduce the time needed
for determining drug resistance.
PCR-DNA Sequencing. Direct DNA sequencing of PCR
amplicons is the most definitive method to detect
mutations associated with drug resistance. This analy-
sis can be performed in a well-equipped diagnostic lab-
oratory and requires approximately 1-2 days to obtain
results from clinical specimens. As a result, numerous
PCR-DNA sequencing assays have been developed to
identify M. tuberculosis (Kapur et al. 1995; Kirschner
et al. 1993) and the presence of drug-resistant mutant
strains (Table 13.2). These assays are based on PCR
amplification of the respective target gene directly from
crude celllysates of processed sputum specimens using
oligonucleotide primers which flank these regions of
the chromosome of M. tuberculosis, purification of
the amplicons using commercial reagents, such as
QIAquick PCR Purification Kit (QIAGEN, Valencia,
CA, www.qiagen.com). and dideoxy DNA sequencing
of the amplicons using either manual or automated
DNA sequencing strategies.
Figure 13.1 shows representative DNA electro-
pherograms of a small fragment of the RRDR of
RMp s and RMp R strains of M. tuberculosis obtained
by automated DNA sequencing. In the resistant
strain, a single missense mutation (TCG-"7TTG) in
codon 531 is identified. This results in a substitu-
tion of a serine amino acid residue with a leucine
residue in the ~-subunit of the RNA polymerase of
this mutant and results in the development of high-
level RMP resistance (MIC >64 Ilg/ml) (Williams et
al. 1998a).
PCR-SSCP. PCR single-strand conformation polymor-
phism analysis assays have been developed and used
to detect drug-resistant mutants of M. tuberculosis
directly from patient specimens or clinical isolates
(Telenti et al. 1993a,b; Kim et al. 2001) (Table 13.2).
PCR and Diagnosis of Tuberculosis 205

This is accomplished using the same basic approach affect the mobility of the resultant DNA fragments
taken to produce PCR arnplicons for DNA sequencing (Fig. 13.3). Therefore, when the heteroduplexes are
for gene targets. These double-stranded amplicons are separated by gel electrophoresis on polyacrylamide
then dissociated into single strands by heat denatur- gels and stained with ethidium bromide, unique
ation and separated very slowly by gel electrophoresis heteroduplex profiles are observed for susceptible
under stringently controlled temperature conditions. and resistant genotypes (Williams et al. 1998b). The
Gels are stained to observe ssDNA fragment mobility PCR-UHG requires approximately 6 h to complete
patterns called SSCP profiles. A schematic represen- and uses precast minigels and a nonradioactive
tation of the SSCP assay is shown in Fig. 13.2. DNA detection format.
strands from a RMp R strain demonstrate a unique
mobility pattern compared with RMPS, reflecting the
change in nucleotide composition. Numerous studies
have used this analysis to study drug resistance in M. RRDR rpoB PCR amplieon RRDR rpoB PCR amplieon
RMP,Strain RMP.Strain
tuberculosis (Table 13.2).
111111111111111 1111111111" III
PCR-HDA. PCR heteroduplex assay is another gel j Denature
j
electrophoresis-based assay which has been devel-
--u'-- --u'--
oped to detect simultaneously the presence of M. + Separate by +
tuberculosis and its susceptibility to RMP directly I\..f\...-"~el Electrophoresis ~ I\..f\...-"
from clinical isolates (Williams et al.1994). This assay

§- =
RMP. RMP.

has been modified to reduce the time taken and tech-

nical complexity and can be used directly on clinical
specimens with the use of a universal heteroduplex SSCP
generator (UHG) (Williams et al. 1996, 1998a). The Profiles
UHG contains several base pair mismatches, inser-
tions, and deletions within the RRDR region of Fig. 13.2. Schematic representation of the PCR single-strand
M. tuberculosis rpoB (Williams et al. 1996). When conformation polymorphism (PCR-SSCP) assay for detection
the UHG is heat-denatured and slowly annealed to of rifampin susceptibility of M. tuberculosis
these denatured amplicons, it provides enhanced
mutation detection. Enhanced mutation detection
occurs because large areas of unmatched nucleo- Universal Heteroduplex
peR Am plleons
tides (bubbles) in the newly formed duplexes greatly Mix
Generator
A
n Willy" III &
~ """"""'"
A
Denature c ~ _
CD
B D

GCGCTGGGGCCC
Slowly anneal

4) ~
.S. 5):~
OR
3) c
• II
') c
• =n=
RMPa Strain (Leu531)
Homoduplexes CD Heteroduplexes

GCCGACTG GCGCTGGGGCC(
10% TBE PAGE

,.. .·.. . . 1=-- --I=

S R

Homoduplnu

RM Ps Strain (Ser531)
Fig.13.t. Representative electropherograms of PCR-DNA Fig. 13.3. Schematic representation of the PCR universal het-
sequencing assay for detection of rifampin susceptibility of eroduplex generator (PCR-UHG-Rif) assay for detection of
M. tuberculosis rifampin susceptibility of M. tuberculosis
206
PCR-LiPA. The line probe assay is a PeR-based reverse
hybridization assay that has been developed for the
rapid detection of RMP susceptibility of M. tubercu-
losis (de Beenhouwer et al. 1995; Rossau et al. 1997).
LiPA uses PCR to amplify the RRDR region of rpoB
and, at the same time, label the fragment with biotin.
Denatured PCR amplicons are hybridized with a set
of DNA probes immobilized on a membrane strip at
specific sites (Fig. 13.4). The immobilized probes are
small DNA fragments that are either homologous
with short segments of the RRDR of rpoB from
a RMp s strain or contain mutations in the RRDR
associated with RMp R M. tuberculosis mutants. The
stringency of the hybridization reaction is designed
so that the single-stranded PCR amplicons will bind
only to probes with 100% sequence homology. The
resultant hybrids are detected by immunoenzymatic
staining, and the results are read visually. An example
of LiPA for RMP susceptibility testing is shown in
Fig. 13.4. The genotype of the test organism is deter-
mined by which lines test positive. Therefore, the
presence or absence of a RMp R strain of M. tubercu-
losis can be determined from the pattern obtained.
This assay is commercially available in certain areas
of the world (Rossau et al. 1997).
Molecular Beacons. PCR molecular beacon sequence
analysis has been developed to detect mutations in the
rpoB that are associated with RMP resistance (Piatek et
al. 1998,2000; EI-Hajj et al. 2001) and INH resistance
(Piatek et al. 2000) in M. tuberculosis. Molecular bea-
cons are small hairpin-shaped DNA probes that report
the presence of specific nucleic acids (Fig. 13.5). These
BiotioylatedRRDR rpoD PCR Amplicoo
RMP.
111111""" '"
Denature
Hybridize*
~
RMP. RMP.
r-
I
, :ii
,>,
;i,
II.
•
;~ • Profiles
:.~
lit:: II
*immunoenzymatic stain
D. L. Williams
probes contain homologous sequences to either the
susceptible or resistant genotype. In addition, they
contain a fluorophore on one end and a molecule that
quenches fluorescence when it is in close proximity to
a fluorophore on the other end. When no PCR ampli-
cons are present, the probe remains in its hairpin shape,
and the fluorescence is quenched. However, after these
probes bind to their intended targets, they undergo a
conformational change that restores thefiuorescence
of the internally quenched fluorophore. Therefore, the
more amplicons produced, the more fluorescence. The
specificity for drug resistance detection is in the short
sequence that binds to the amplicon.
In summary, there are several PCR-based muta-
tional detection assays to identify drug-resistant
mutants of M. tuberculosis. These assays can be per-
formed within days of specimen acquisition or isola-
tion of strains from specimens and, therefore, can be
useful for the early detection of drug-resistant tuber-
culosis. Early detection of drug-resistant tuberculo-
sis can potentially improve patient care and reduce
the transmission of drug-resistant tuberculosis by
effective isolation of infected individuals.
13.5
Impact of PCR Diagnosis on Patient Care
Because of the specificity of PCR assays for the detec-
tion of M. tuberculosis, a smear-positive, PCR-posi-
tive specimen would indicate that the acid-fast bacilli
in the sputum of a particular patient are M. tuber-
BiotioylatedRRDR rpoD PCR Amplicoo
RMP.
"'""""''''
LiPA
Fig. 13.4. Schematic representation of the PCR line
probe assay (PCR-LiPA) for detection of rifampin sus-
ceptibility of M. tuberculosis
peR and Diagnosis of Tuberculosis
~+
Qumched Targd DNA
Molecular Beaooo (ss AqJIicoo)
Fig. 13.5. Schematic representation of molecular beacon
sequence analysis. The black circle represent the quenching
molecule. The open circle represents the fluorophore
culosis (Kaul 2001). Therefore, the patient should be
appropriately retained in isolation and treated until
their sputum becomes smear-negative. A PCR-nega-
tive specimen would indicate the presence of a non-
tuberculosis mycobacterial infection, and the patient
could be released from isolation and treated more
appropriately. In addition, because of the availability
of PCR-based assays for the detection of mutations
associated with antituberculosis drug resistance, a
smear-positive, PCR-resistant specimen would indi-
cate that this patient is also harboring drug-resis-
tant tuberculosis. Therefore, the patient should be
appropriately retained in isolation and treated with
additional drug therapy until their sputum becomes
smear-negative. The immediate impact on patient
care, the reduction of transmission of tuberculosis,
and cost of providing care is very apparent.
Because of the high specificity of these assays, they
cannot identify the presence of other mycobacteria in
mixed infections. It is, therefore, recommended that
PCR should be augmented by culture to determine
if other nontuberculosis mycobacteria are present
in specimens. This is especially important in areas
of the world where HIV infection is highly endemic
and other nontuberculosis mycobacteria present
problems for the differential diagnosis of tuberculo-
sis. Recently, newer multiplex PCR assays have been
developed which can be used for the detection of M.
tuberculosis complex and nontuberculous mycobac-
teria. These assays are currently being characterized
for the identification of mixed infections (Ahmed
1999; Yeboah-Manu et al. 2001).
Most M. tuberculosis PCR assays are based on
the detection of DNA or ribosomal RNA targets.
Although beneficial for the initial diagnosis, such
assays have proven unsuitable for rapidly monitor-
ing therapeutic efficacy owing to the persistence of
these nucleic acid targets long after the conversion
of smears and cultures to negative (DesJardin et al.
1998; Hellyer et al. 1999a). It has recently been ascer-
tained that the presence of mRNA is a good indicator
for determining the viability of M. tuberculosis from
sputum specimens. Using mRNA, an assay has been
207
developed to monitor drug therapy (DesJardin et al.
1996; Hellyer et al. 1999a) and has been shown to be
very useful for rapidly determining the efficacy of
therapeutic regimens directly from sputum speci-
mens (DesJardin et al.1996; Hellyer et al. 1999a,b).
13.6
Summary
Methodologies developed over 30 years ago for the
detection of tuberculosis and drug susceptibility test-
ing are cumbersome and require protracted periods
of time to obtain results. The resurgence of tubercu-
losis and drug-resistant tuberculosis has promoted
the development of a number of new options for the
rapid laboratory diagnosis of M. tuberculosis. One
of the most promising and exciting methodologies
has been the introduction of assays employing PCR
amplification technology to detect M. tuberculosis
directly from clinical specimens.
PCR-based assays for the diagnosis of tubercu-
losis and identification of drug-resistant strains are
configured to yield results in a few hours to days
because these assays can detect the presence of M.
tuberculosis and/or mutations present in its genome
that are associated with the development of drug
resistance directly from crude biological specimens.
This represents a major step forward in the ability
to rapidly identify tuberculosis patients and patients
harboring drug-resistant tuberculosis. Therefore, the
implementation of these assays has the potential to
improve patient care and provide substantial sav-
ings in the overall cost of patient care compared with
conventional smear, culture, and speciation alone. In
addition, since rapid detection allows for the rapid
isolation of tuberculosis patients, the nosocomial or
environmental spread of tuberculosis and drug-resis-
tant-tuberculosis to susceptible individuals will most
likely be reduced. However, the major drawbacks for
the implementation of many of these assays in clinical
laboratories in developing countries are the cost of the
reagents used and the need for expensive equipment
and trained individuals to perform the analysis.
This review has attempted to describe many of the
current molecular aspects of the detection of tuber-
culosis using PCR, the modes of action, the resistance
mechanisms of the major drugs used to treat tuber-
culosis, and the current molecular assays for detect-
ing drug resistance in M. tuberculosis. Although
amplification assays hold significant promise to
improve the laboratory diagnosis of tuberculosis,
208
the decision to perform these assays is complicated
because most assays have decreased sensitivity with
specimens that are AFB smear-negative. In addition,
many other factors, such as cost, assay performance,
and technical difficulty must be considered in order
to facilitate the decision-making process about
whether an amplification assay would be appropriate
in a particular laboratory setting.
As nucleic acid-based tests evolve, it is antici-
pated that they will become cheaper and simpler to
perform. This is extremely important in developing
countries, where resources are limited, and tubercu-
losis is highly endemic and often complicated by HIV
infection. Therefore, research continues to be con-
ducted to automate and simplify sample preparation,
amplification, and amplicon detection procedures,
thus making them easier to implement in clinical
laboratories within both developed and developing
countries. This will assist tuberculosis control pro-
grams with their overall goals of providing better
patient care, reducing the transmission of tuberculo-
sis, and providing the capability to monitor trends in
drug resistance at the regional and local levels.
References
Ahmed YH (1999) Detection of Mycobacterium tuberculosis
complex and non-tuberculous mycobacteria by multiplex
polymerase chain reactions. East Mediterr Health J 5:831
Almeda J et al (2000) Clinical evaluation of an in-house IS6110
polymerase chain reaction for diagnosis of tuberculosis.
Eur J Clin Microbiol Infect Dis 19:859-867
Banerjee A et al (1994) InhA, a gene encoding a target for
isoniazid and ethionamide in Mycobacterium tuberculosis.
Science 263:227-230
Bartfai Z et al (2001) Molecular characterization of rifampin-
resistant isolates of Mycobacterium tuberculosis from Hun-
gary by DNA sequencing and the line probe assay. J Clin
MicrobioI39:3736-3739
Bennedsen J et al (1996) Utility of PCR in diagnosing pulmo-
nary tuberculosis. J Clin MicrobioI34:1407-1411
Bergmann JS, Woods GL (1996) Clinical evaluation of the
Roche AMPLICOR@ PCR Mycobacterium tuberculosis test
for detection of M. tuberculosis in respiratory specimens. J
Clin MicrobioI34:1083-1085
Blanchard JS (1996) Molecular mechanisms of drug resis-
tance in Mycobacterium tuberculosis. Annu Rev Biochem
65:215-239
Boddinghaus B et al (1990) Detection and identification of
mycobacteria by amplification of rRNA. J Clin Microbiol
28:1751-1759
Bogard M et al (2001) Multicenter study of a commercial,
automated polymerase chain reaction system for the rapid
detection of Mycobacterium tuberculosis in respiratory
specimens in routine clinical practice. Eur J Clin Microbiol
Infect Dis 20:724-731
D. 1. Williams
Borun M et al (2001) Detection of Mycobacterium tuberculosis
in clinical samples using insertion sequences IS6110 and
IS990. Tuberculosis (Edinb) 81:271-278
Bottger EC (1989) Rapid determination of bacterial ribo-
somal RNA sequences by direct sequencing of enzymati-
cally amplified DNA. FEMS Microbiol Lett 53:171-176
Bradford WZ et al (1996) The changing epidemiology of
acquired drug-resistant tuberculosis in San Francisco,
USA. Lancet 348:928-931
Brisson-Noel A et al (1991) Diagnosis of tuberculosis by DNA
amplification in clinical practice evaluation. Lancet 338:
364-366
Butler WR, Kilburn JO (1983) Susceptibility of Mycobacte-
rium tuberculosis to pyrazinamide and its relationship to
pyrazinamidase activity. Antimicrob Agents Chemother
24:600-601
Cambau E et al (1994) Selection of a gyrA mutant of Mycobac-
terium tuberculosis resistant to fluoroquinolones during
treatment with ofloxacin. J Infect Dis 170: 1351
Cave MD et al (1991) IS611O: conservation of sequence in the
Mycobacterium tuberculosis complex and its utilization in
DNA fingerprinting. Mol Cell Probes 5:73-80
Clarridge JE III et al (1993) Large-scale use of polymerase
chain reaction for detection of Mycobacterium tuberculosis
in a routine mycobacteriology laboratory. J Clin Microbiol
31:2049-2056
Cockerill FR III (1999) Minireview: genetic methods for
assessing antimicrobial resistance. Antimicrob Agents
Chemother 43:199-212
Cohn DL, Bustreo F, Faviglione MC (1997) Drug-resistant
tuberculosis: review of the worldwide situation and the
WHO/IUATLD global surveillance project. International
Union against Tuberculosis and Lung Disease. Clin Infect
Dis 24:S121-S130
Cousins DV et al. (1992) Use of polymerase chain reaction
for the rapid diagnosis of tuberculosis. J Clin Microbiol
30:255-258
Dalovisio JR et al (1996) Comparison of the amplified Myco-
bacterium tuberculosis (MTB) direct test, Amplicor@ MTB
PCR, and IS61l0-PCR for detection of MTB in respiratory
specimens. Clin Infect Dis 23:1099-1108
De Beenhouwer H et al (1995) Rapid detection of rifampicin
resistance in sputum and biopsy specimens from tuber-
culosis patients by PCR and line probe assay. Tuber Lung
Dis 76:425-430
Delgado M, Telenti A (1996) Detection of fluoroquinolone
resistance mutations in Mycobacterium tuberculosis. In:
Persing D (ed) PCR protocols for emerging infectious dis-
eases. ASM Press, Washington DC, pp 138-143
Del Portillo P, Murillo LA, Patarroyo ME (1991) Amplifica-
tion of a species-specific DNA fragment of Mycobacterium
tuberculosis and its possible used in diagnosis. J Clin
MicrobioI29:2163-2168
DesJardin LE et al (1996) Alkaline decontamination of sputum
specimens adversely affects stability of mycobacterial
mRNA. J Clin MicrobioI34:2435-2439
DesJardin LE et al (1998) Comparison of the ABI 7700
(TaqMan) and competitive PCR for quantification of
IS6110 DNA in sputum during treatment of tuberculosis. J
Clin Microbiol 36:1964-1968
Douglass J, Steyn LM (1993) A ribosomal gene mutation in
streptomycin-resistant Mycobacterium tuberculosis iso-
lates. J Infect Dis 167:1505-1506
PCR and Diagnosis of Tuberculosis
Eing BR et al (1998) Comparison of Roche COBAS AMPLI-
CORiZ> Mycobacterium tuberculosis assay with in-house
PCR and culture for detection of M. tuberculosis. J Clin
MicrobioI36:2023-2029
Eisenach KD et al. (1990) Polymerase chain reaction amplifica-
tion of a repetitive DNA sequence specific for Mycobacte-
rium tuberculosis. J Infect Dis 161:977-981
Eisenach KD et al (1991) Detection of Mycobacterium tubercu-
losis in sputum samples using a polymerase chain reaction.
Am Rev Respir Dis 144:1160-1163
El-Hajj HH et al (2001) Detection of rifampin resistance in
Mycobacterium tuberculosis in a single tube with molecular
beacons. J Clin MicrobioI39:4131-4137
Felmlee TA et al (1995) Genotypic detection of Mycobac-
terium tuberculosis rifampin resistance: comparison of
single-strand conformation polymorphism and dideoxy
fingerprinting. J Clin MicrobioI33:1617-1623
Finken M et al (1993) Molecular basis of streptomycin resistance
in Mycobacterium tuberculosis: alterations of the ribosomal
protein S12 gene and point mutations within a functional16S
ribosomal RNA pseudoknot. Mol MicrobioI9:1239-1246
Forbes BA (1997) Critical assessment of gene amplification
approaches on the diagnosis of tuberculosis. Immunol
Invest 26:105-116
Frieden TR et al (1993) The emergence of drug-resistant
tuberculosis in New York City. N Engl J Med 328:521-526
Gomez-Pastrana D et al (2001) Comparison of Amplicor, in-
house polymerase chain reaction, and conventional culture
for the diagnosis of tuberculosis in children. Clin Infect
Dis 32:17-22
Heifets LB (l991) Antituberculosis drugs: antimicrobial activ-
ity in vitro. In: Heifets LB (ed) Drug susceptibility in the
chemotherapy of mycobacterial infections. CRC Press,
Boca Raton, pp 14-57
Hellyer TJ et al (l999a) Quantitative analysis of mRNA as a
marker for viability of Mycobacterium tuberculosis. J Clin
Microbiol 37:290-295
Hellyer TJ et al (l999b) Detection of viable Mycobacterium
tuberculosis by reverse transcriptase-strand displacement
amplification of mRNA. J Clin MicrobioI37:518-523
Heym B et al (l995) Missense mutations in the catalase-per-
oxidase gene, katG, are associated with isoniazid resistance
in Mycobacterium tuberculosis. Mol MicrobioI15:235-245
Hooper DC, Wolfson JS (1989) Mechanism of action and
resistance. In: Hooper DC, Wolfson JS (eds) Quinolone
antimicrobial agents. American Society of Microbiology
Press, Washington DC, pp 5-34
lovannisci DM, Winn-Deen ES (l993) Ligation amplification
and fluorescence detection of Mycobacterium tuberculosis
DNA. Mol Cell Probes 7:35-43
Jonas V et al (1993) Detection and identification of Myco-
bacterium tubercuosis directly from sputum sediments by
amplification of rRNA. J Clin MicrobioI31:2410-2416
Kaul KL (2001) Molecular detection of Mycobacterium tuber-
culosis: impact on patient care. Clin Chern 47:1553-1558
Kapur V et al (1994) Characterization by automated DNA
sequencing of mutations in the gene (rpoB) encoding
the RNA polymerase beta subunit in rifampin-resistant
Mycobacterium tuberculosis strains from New York City
and Texas. J Clin MicrobioI32:1095-1098
Kapur Vet al (l995) Rapid Mycobacterium species assignment
and unambiguous identification of mutations associated
with antimicrobial resistance in Mycobacterium tubercu-
209
losis by automated DNA sequencing. Arch Pathol Lab Med
119:131-138
Kent L et al (1996) Demonstration of homology between
IS6110 of Mycobacterium tuberculosis and DNAs of other
Mycobacterium spp.? J Clin Microbiol 33:2290-2293
Kent P, Kubica G (1985) Public health mycobacteriology:
a guide for the Level III laboratory. US Department of
Health and Human Services, Public Heath Service, CDC,
Atlanta, pp 31-56
Kim BJ et al (2001) Detection of rifampin-resistant Mycobac-
terium tuberculosis in sputa by nested PCR-linked single-
strand conformation polymorphism and DNA sequencing.
J Clin MicrobioI39:2610-2617
Kirschner P et al (1993) Genotypic identification of mycobac-
teria by nucleic acid sequence determination: report of a
2-year experience in a clinical laboratory. J Clin Microbiol
31:2882-2889
Konno K, Feldmann FM, McDermott W (l967) Pyrazinamide
susceptibility and amidase activity of tubercle bacilli. Am
Rev Respir Dis 95:461-469
Marttila HJ et al (1996) katG mutations in isoniazid-resistant
Mycobacterium tuberculosis isolates recovered from Finn-
ish patients. Antimicrob Agents Chemother 40:2187-2189
Mitarai S et al (2001) Potential use of Amplicor PCR kit in
diagnosing pulmonary tuberculosis from gastric aspirate.
J Microbiol Methods 47:339-344
Moore M et al (1997) Trends in drug-resistant tuberculosis in
the United States. J Am Med Assoc 278:833-837
Mulcahy GM et al (1996) IS611O-based PCR methods for the
detection of Mycobacterium tuberculosis. J Clin Microbiol
34:1348-1349
Mullis KB, Faloona FA (l987) Specific synthesis of DNA in
vitro via a polymerase-catalyzed chain reaction. Methods
Enzymol 155:335-350
Musser JM (l995) Antimicrobial agent resistance in myco-
bacteria: molecular genetic insights. Clin Microbiol Rev
8:496-514
Musser JM et al (1996) Characterization of the catalase-per-
oxidase gene (katG) and inhA locus in isoniazid-resistant
and -susceptible strains of Mycobacterium tuberculosis
by automated DNA sequencing: restricted array of muta-
tions associated with drug resistance. J Infect Dis 173:
196-202
Nachamkin I, Kang C, Weinstein MP (l997) Detection of resis-
tance to isoniazid, rifampin, and streptomycin in clinical
isolates of Mycobacterium tuberculosis by molecular meth-
ods. Clin Infect Dis 24:894-900
Nolte FS et al (1993) Direct detection of Mycobacterium tuber-
culosis in sputum by polymerase chain reaction and DNA
hybridization. J Clin MicrobioI31:1777-1782
Oh EJ et al (2001) Improved detection and differentiation of
mycobacteria with combination of Mycobacterium Growth
Indicator Tube and Roche COBAS AMPLICORiZ> System in
conjunction with Duplex PCR. J Microbiol Methods 46:
29-36
Otal I et al (1997) Use of a PCR method based on IS611 0 poly-
morphism for typing Mycobacterium tuberculosis strains
from BACTEC cultures. J Clin Microbiol 35:273-277
Piatek AS et al (1998) Molecular beacon sequence analysis for
detecting drug resistance in Mycobacterium tuberculosis.
Nat BiotechnoI16:359-363
Piatek AS et al (2000) Genotypic analysis of Mycobacterium
tuberculosis in two distinct populations using molecular
210
beacons: implications for rapid susceptibility testing. Anti-
microb Agents Chemother 44:103-110
Plikaytis BB et al (1993) Rapid, amplification-based finger-
printing of Mycobacterium tuberculosis. J Gen Microbiol
139:1537-1542
Poa CC et al (1990) Direct detection and identification of
Mycobacterium tuberculosis by DNA amplification. J Clin
MicrobioI28:1877-1880
Ramaswamy S, Musser JM (1998) Molecular genetic basis of
antimicrobial agent resistance in Mycobacterium tubercu-
losis: 1998 update. Tuberc Lung Dis 79:3-29
Rossau R et al (1997) Evaluation of the INNO-LiPA Rif. TB assay,
a reverse hybridization assay for the simultaneous detection
of Mycobacterium tuberculosis complex and its resistance to
rifampin. Antimicrob Agents Chemother 41:2093-2098
Rouse DA et al (1995) Characterization of the katG and inhA
genes of isoniazid-resistant clinical isolates of Mycobac-
terium tuberculosis. Antimicrob Agents Chemother 39:
2472-2477
Selvakumar N et al (1997) Single strand conformation poly-
morphism profiles with biotinylated PCR products to
detect mutation in rpoB gene of Mycobacterium tubercu-
losis. Curr Sci 73:774-777
Shalwar R et al (1993) Detection of Mycobacterium tubercu-
losis in clinical samples by amplification of DNA. J Clin
MicrobioI29:712-717
Shah J et al (1995) Q-beta replicase-amplified assay for detec-
tion of Mycobacterium tuberculosis directly from clinical
specimens. J Clin MicrobioI33:1435-1441
Shinnick TM, Jonas V (1994) Molecular approaches to the
diagnosis of tuberculosis. In: Bloom B (ed) Tuberculosis:
pathogenesis, protection and control. American Society for
Microbiology Press, Washington DC, pp 517-530
Soini H et al (1992) Detection and identification of mycobacte-
ria by amplification of a segment of the gene coding for the
32-kilodalton protein. J Clin MicrobioI30:2025-2028
Soini H et al (1996) Comparison of AMPLICOR® and 32-kilo-
dalton PCR for detection of Mycobacterium tuberculosis
from sputum specimens. J Clin MicrobioI34:1829-1830
Sougakoff W et al (1997) Nonradioactive single-strand con-
formation polymorphism analysis for detection of fluoro-
quinolone resistance in mycobacteria. Eur J Clin Microbiol
Infect Dis 16:395-398
Southern EM (1975) Detection of specific sequences among
DNA fragments separated by gel electrophoresis. J Mol
Bioi 98:503-517
Sreevatsan S et al (1996) Characterization of rpsL and rrs muta-
tions in streptomycin-resistant Mycobacterium tuberculosis
isolates from diverse geographical localities. Antimicrob
Agents Chemother 40:1024-1 026
Sreevatsan S et al (1997a) Mutations associated by pyrazinamide
resistance in pncA of Mycobacterium tuberculosis complex
organisms. Antimicrob Agents Chemother 41:636-640
Sreevatsan S et al (1997b) Ethambutol resistance in myco-
bacterium tuberculosis: critical role of embB mutations.
Antimicrob Agents Chemother 41:1677-1681
Sreevatsan S et al. (1997c) Analysis of the oxyR-ahpC region in
isoniazid-resistant and -susceptible Mycobacterium tubercu-
losis complex organisms recovered from diseased humans
and animals in diverse localities. Antimicrob Agents Che-
mother 41:600-606
Stratton MA, Reed MT (1986) Short-course drug therapy for
tuberculosis. Clin Pharm 5:977-987
D. 1. Williams
Sullivan EA et al (1995) Emergence of fluoroquinolone-resis-
tant tuberculosis in New York City. Lancet 345:1148-1150
Takayama K, Kilburn JO (1989) Inhibition of synthesis of ara-
binogalactan by ethambutol in Mycobacterium smegmatis.
Antimicrob Agents Chemother 33:1493-1499
Takayama K et al (1979) Inhibition by ethambutol of mycolic
acid transfer into the cell wall of Mycobacterium smegma-
tis. Antimicrob Agents Chemother 16:240-242
Takiff HE et al (1994) Cloning and nucleotide sequence of
Mycobacterium tuberculosis gyrA and gyrB genes and
detection of quinolone resistance mutations. Antimicrob
Agents Chemother 38:773-780
Telenti A et al (1993a) Detection of rifampin-resistance muta-
tions in Mycobacterium tuberculosis. Lancet 341:647-650
Telenti Aet al (1993b) Direct, automated detection of rifampin-
resistant Mycobacterium tuberculosis by polymerase chain
reaction and single-strand conformation polymorphism
analysis. Antimicrob Agents Chemother 37:2054-2058
Telenti A et al (1997a) Genotypic assessment of isoniazid
and rifampin resistance in Mycobacterium tuberculosis: a
blind study at reference laboratory level. J Clin Microbiol
35:719-723
Telenti A et al (1997b) The emb operon, a gene cluster of
Mycobacterium tuberculosis involved in resistance to eth-
ambutol. Nat Med 3:567-570
Temesgen Z et al (1997) Use of polymerase chain reaction
single-strand conformation polymorphism (PCR-SSCP)
analysis to detect a point mutation in the catalase-peroxi-
dase gene (katG) of Mycobacterium tuberculosis. Mol Cell
Probes 11 :59-63
Thierry D et al (1990) IS6110, an IS-like element of Mycobacte-
rium tuberculosis complex. Nucleic Acids Res 18:188
Thomas GA, Williams DL, Soper SA (2001) Capillary elec-
trophoresis-based heteroduplex analysis with a universal
heteroduplex generator for detection of point mutations
associated with rifampin resistance in tuberculosis. Clin
Chern 47:1195-1203
van Embden JD et al (1993) Strain identification of Mycobac-
terium tuberculosis by DNA fingerprinting: recommenda-
tions for a standardized methodology. J Clin Microbiol 31:
406-409
van Soolingen D et al (1994) DNA fingerprinting of Mycobac-
terium tuberculosis. Methods EnzymoI235:196-205
Victor TC et al (1996) katG mutations in isoniazid-resistant
strains of Mycobacterium tuberculosis are not infrequent.
Antimicrob Agents Chemother 40:1572
Walker GT et al (1992) Strand displacement amplification-an
isothermal, in vitro DNA amplification technique. Nucleic
Acids Res 20:1691-1696
Whelen A et al (1995) Direct genotypic detection of Myco-
bacterium tuberculosis rifampin resistance in clinical
specimens by using single-tube heminested PCR. J Clin
Microbiol 33:556-561
Williams DL et al (1994) Characterization of rifampin resis-
tance in pathogenic mycobacteria. Antimicrob Agents
Chemother 38:2380-2386
Williams DL, Gillis TP, Dupree WG (1995) Ethanol fixation of
sputum sediments for DNA-based detection of Mycobacte-
rium tuberculosis. J Clin MicrobioI33:1558-1561
Williams DL et al (1996) PCR-heteroduplex detection of
rifampin-resistant Mycobacterium tuberculosis. In: Persing
D (ed) PCR protocols for emerging infectious diseases.
ASM Press, Washington DC, pp 122-129
PCR and Diagnosis of Tuberculosis
Williams DL et al (l998a) Contribution of rpoR mutations to
development of rifamycin cross-resistance in Mycobacterium
tuberculosis. Antimicrob Agents Chemother 42:1853-1857
Williams DL et al (l998b) Evaluation of a polymerase chain
reaction-based universal heteroduplex generator assay for
direct detection of rifampin-susceptibility of Mycobacte-
rium tuberculosis directly from sputum specimens. Clin
Infect Dis 26:446-450
Winder FG, Collins PB, Whelan D (1971) Effects of ethion-
amide and isoxyl on mycolic acid synthesis in Mycobacte-
rium tuberculosis BCG. J Gen MicrobioI66:379-380
Winder FG (1982) Mode of action of the antimycobacterial
211
agents and associated aspects of the molecular biology of
the mycobacteria. In: Ratledge C, Stanford J (eds) The biol-
ogy of the mycobacteria, vol 1. Academic Press, London, pp
353-438
Woods GL (2001) Molecular techniques in mycobacterial
detection. Arch Pathol Lab Med 125:122-126
Yeboah-Manu D, Yates MD, Wilson SM (2001) Application of a
simple multiplex PCR to aid in routine work of the mycobac-
terium reference laboratory. J Clin MicrobioI39:4166-4168
Zhang Y et al (1992) The catalase-peroxidase gene and iso-
niazid resistance of Mycobacterium tuberculosis. Nature
358:591-593
14 Hematologic Findings in Mycobacterial Infections
Among Immunosuppressed
and Immunocompetent Patients
GORGUN AKPEK

CONTENTS 14.1
Introduction
14.1 Introduction 213
14.1.1 Mycobacterium Tuberculosis 214
14.2 Hematologic Findings in Tuberculosis 215
The relationship between hematologic abnormalities
14.2.1 Anemia 216 and mycobacterial infections dates back to the begin-
14.2.2 White Blood Cell Abnormalities 216 ning of the 20th century. Flinn and his colleagues
14.2.3 Thrombocytopenia 216 suggested at that time that the change in hematologic
14.2.4 Pancytopenia 216 parameters is the earliest finding of tuberculosis (TB)
14.2.5 Possible Mechanisms of Hematologic Findings
in Tuberculosis 217
(Flinn and Finn 1929). The authors said,'In connection
14.2.6 Prognostic Significance of Hematologic with the above studies we have formed the impression
Abnormalities. 219 that an extension of the pathologic process in a tuber-
14.2.7 Tuberculosis-Induced Lymphopenia culosis lung usually manifests itself first in the blood
and Immunosuppression 219
picture, later by physical signs and x-ray, and often last
14.3 Hematologic Abnormalities Associated
with Mycobacterial Infections of all by symptoms: Before embarking upon a general
in Imrnunocompromised Hosts 220 discussion of hematologic manifestations of mycobac-
14.3.1 Hematologic Complications of HIV Infection terial infections, I would like to present two cases that
and AIDS 220 were previously discussed at the Medical Staff Con-
14.3.2 Mycobacterium avium Complex ference in the University of San Francisco, California,
in HIV-Infected Patients 220
14.3.3 Anemia Associated with MAC USA, in 1967. These cases have certain features of the
in HIV-Infected Patients 220 history and clinical findings of TB, in particular, the
14.3.4 Other Cytopenias Associated with MAC striking difference in the hematological manifesta-
in HIV-Infected Patients 222 tions between the two cases. A brief summary about
14.3.5 MAC Infections in Other Immunocompromised
the demographics and clinical presentation of Myco-
Hosts 222
14.3.6 Hemophagocytic Syndrome Associated bacterium TB infection is included thereafter.
with MAC 223
14.3.7 Miliary/Disseminated Tuberculosis in AIDS 223 Case 1. The first patient is a 50-year-old white man
14.3.8 Diagnostic Yield of Bone Marrow Aspiration/Biopsy who came in for evaluation of a fever of about
for Mycobacterial Infections in Patients with HIV
7 months' duration. His symptoms began with mul-
Infection 224
14.3.9 Other Mycobacterial Infections Associated tiple episodes of severe pain in the left upper quad-
with Hematologic Disorders 226 rant of the abdomen and nausea. These symptoms
14.4 Coagulation Abnormalities Associated with Myco- were followed in a few days by fever and generalized
bacterial Infections and Clinical Presentations 226 malaise. According to the patient's description, the
14.4.1 Disseminated Intravascular Coagulation 226
febrile episodes would last for approximately 1 week
14.4.2 Deep Vein Thrombosis Associated
with Tuberculosis 226 and recurred at approximately 4-week intervals with
14.4.3 Thrombotic Thrombocytopenic Purpura Associated temperatures in the range of 38.3°C-40.6°C (lOI°F-
with Mycobacterial Infections 227 105°F). He was noted to have axillary and inguinal
14.5 Conclusion 227 adenopathy 2 months later when seen by a physi-
References 227
cian because of persistent symptoms. Laboratory
data included hemoglobin of 12.2 gm/dl, markedly
elevated erythrocyte sedimentation rate, increased
G.AKPEK,MD
Greenebaum Cancer Center at University of Maryland School
concentration of serum gammaglobulin, and a posi-
of Medicine, 22 South Greene Street, Baltimore MD 21201, tive tuberculin test. Two months later, the patient
USA underwent surgery for the repair of an incarcer-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
214
ated incisional hernia. An exploratory laparotomy
performed at that time revealed no intraabdominal
pathology. Two months later, the patient's fever
recurred and was associated with night sweats and
13 pounds (6 kg) of weight loss since the onset of his
symptoms. At that time no history could be obtained
of cough, pleuritic chest pain, hemoptysis, or produc-
tion of sputum. The family history was negative. The
patient had been a heavy smoker for 20 years.
On physical examination, the only abnormalities
were slight adenopathy in the inguinal and axillary
regions, hepatomegaly, and questionable splenomeg-
aly. Body temperature was 38.3°C (101°F). Hemoglobin
content was 8.7 g/dl, WBC of 3400/mm3, with 30% neu-
trophils, 65% lymphocytes, 3% monocytes, and 2%
basophils. The platelet count was 301,000/mm 3, with
reticulocytes forming 2.1 % of the total. Additional
hematologic data included normal haptoglobulin, leu-
kocyte alkaline phosphatase, coagulation parameters,
and serum iron and iron-binding capacity. The bone
marrow at that time showed generalized hyperplasia,
predominantly of the erythroid series. Skin test was
positive with PPD. A chest X-ray showed fibrocalcific
densities. There was no active pulmonary disease, but
there was evidence of old granulomatous disease.
There were granulomas in the right upper lung field
and some scattered granulomas in the right mid-lung
field. All other work-up for fever of unknown origin
was negative. A scalene lymph node biopsy showed
multiple confluent and epithelioid caseous granulo-
mas. The patient improved on an anti-TB regimen.
Cultures of the scalene lymph node specimen grew
acid-fast organisms.
Case 2. The second patient is a 62-year-old man with a
chief complaint of tenderness in the left upper quad-
rant of the abdomen, with weight loss and weakness
of 5 months' duration. A mass in the left upper quad-
rant was noted on routine physical examination. Four
months later, the patient first noticed anorexia, weight
loss, and weakness. Three months later, he developed
anemia and was admitted to the hospital. The patient
has had regular chest X-rays taken every 6 months for
the past 2 years. These films have shown stable apical
infiltrates. He was also known to have a positive PPD
skin test. He had been a heavy smoker for the past
45 years; he has had chronic nonproductive cough
and a history of acute upper gastrointestinal bleeding,
which responded to medical management.
On physical examination, the patient was pale and
cachectic. Body temperature was 38°C (101 OF). He had
several small lymphadenopathies. Lung expansion was
decreased bilaterally, and the anteroposterior diam-
G.Akpek
eter of the chest was increased. There was dullness and
increased fremitus on the right. Peristaltic movement
was visible beneath a thin abdominal wall. The spleen,
which was large and rock-hard, extended 8 cm below
the left costal margin, and there was a tender nodule,
2x3 cm, at the anterior margin. The liver edge was felt
4 cm below the right costal margin. It was not tender
and firm. Early clubbing of the fingers and severe
muscular wasting of the extremities were noted. Leu-
kocyte numbered 16,000/mm3 , with 60% neutrophils,
many immature myeloid cells, 10% basophils, and
10% eosinophils. Hematocrit was 19%, reticulocyte
count 2.5% of the total, and platelet count 80,000/mm 3•
Normoblasts were present in the peripheral blood. X-
ray examination revealed dense bones. Patchy apical
infiltrates, which had increased since the previous
examination, were noted on chest X-ray. There was
a large right pleural effusion. A skin test with PPD
was positive. Thoracentesis and pleural biopsy were
performed. The latter revealed caseating granulomas.
No intrabronchiallesions were seen at bronchoscopy.
A bone biopsy was performed since three previous
marrow taps had been dry. The bone biopsy revealed
sclerotic bone and myelofibrosis. Triple anti-TB ther-
apy was begun. Subsequently, cultures of both sputum
and pleural fluid grew acid-fast bacilli.
14.1.1
Mycobacterium Tuberculosis
Currently, more than one-third of the world's popula-
tion is infected with Mycobacterium TB: 8 million new
cases and approximately 2 million deaths are reported
each year (Dye et al. 1999). Although the lung is the
primary site of disease in 80%-84% of cases of TB in
the USA (Center for Disease Control and Prevention
2000), extrapulmonary TB has become more common
with the advent of HIV infection, and the risk of TB
increases as the immunosuppression progresses
(Moore et al.1997; Jones et al.1993; Iseman 2000). The
most common extrapulmonary sites of disease are the
lymph nodes, pleura, and bones or joints (Center for
Disease Control and Prevention 2000).
Lymphatic TB is usually seen in children and young
adults; more commonly in women (especially Asian
and Indian women). It presents with unilateral, pain-
less, cervical adenopathy, which is connected to the
skin by sinus tracts late in the course of disease. Exci-
sional biopsy with culture yields the diagnosis of TB
lymphadenitis. PPD is usually positive. It may respond
slowly to medication and rarely may require excision
(Mandel et al. 2000).
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients
Clinical presentations ofTB have changed dramati-
cally since the introduction of anti-TB agents. In the
pre-antibiotic era, late generalized TB was often the
primary disease, occurring mainly in young adults and
frequently associated with pulmonary symptoms. In
the antibiotic era, TB commonly occurs together with
and is frequently obscured by other diseases. It often
afflicts the elderly and is much less frequently accom-
panied by pulmonary symptoms (Slavin et al. 1980).
Disseminated TB involving the lung, liver, spleen, bone
marrow, and lymph nodes can occur occasionally in
an immunocompetent host and is usually associated
with hematologic abnormalities and poor outcome
(Charfi et al.1998). Splenomegaly associated with fever
is another clinical sign of disseminated TB (O'Reilly
1998). The histologic appearance in this rare form
of disseminated hematogenous TB (nonreactive TB)
shows nonspecific necrosis containing disintegrating
polymorphonuclear leukocytes and enormous num-
bers of tubercle bacilli (O'Brien 1954). In a typical case,
granulomas and epithelioid cells are lacking, although
intermediate cases have areas more typical for T8. The
gross pathologic findings are minute soft abscesses
up to 1 cm, which always involve the liver and spleen,
usually the marrow, commonly the lungs and kidneys,
but almost never the meninges. The clinical picture
may be overwhelming sepsis, with splenomegaly and
often an inconspicuous diffuse mottling on the chest
X-ray. Major hematologic abnormalities are common.
Whether the lesions are typical or not, it is necessary
to have bacteriologic confirmation of TB. Near these
lesions, the bone marrow cellularity is often raised or
lowered, reticular fibers are often increased, and some-
times reticulin fibrosis is marked (Tulliez 1976).
In the following sections, common types of myco-
bacterial infections that are associated with various
hematologic findings are discussed.
14.2
Hematologic Findings in Tuberculosis
While most patients with TB do not manifest major
hematologic abnormalities, some patients with late
generalized or chronic hematogenous TB and most
with nonreactive TB have serious hematologic abnor-
malities, including leukopenia, thrombocytopenia,
anemia, leukemoid reactions, myelofibrosis, and
polycythemia (Cameron 1974). Leukemoid reactions
may suggest acute leukemia, although most patients
in whom hematogenous TB coexists with the clinical
picture ofleukemia have both diseases. Slavin and his
215
colleagues reviewed the medical records of a commu-
nity-based university teaching hospital over a lO-year
period (1978 to 1987) to determine the clinical and
laboratory characteristics, diagnostic methods, and
prognostic variables in adults treated for miliary
TB in the rifampicin era. They identified a total of
109 patients including 12 who did not have miliary
nodules on the chest X-ray (all of whom were shown
to have hematogenous dissemination). Hematologic
abnormalities were common: leukopenia (less than
4xlO(9)/I) was present in 16 of 107 patients (15%),
thrombocytopenia (less than 150x10(9)/L) in 24 of
104 (23%), and lymphopenia (less than 1.5X10(9)/L)
in 82 of 94 (87%). Pancytopenia was found in 6
patients, 3 of whom recovered. Disseminated intra-
vascular coagulation occurred in 4 patients, all of
whom died (Slavin et al. 1980).
In a recent study, investigators from India com-
pared peripheral blood and bone marrow findings in
patients with disseminated/miliary TB (DTB/MTB)
as well as pulmonary TB (PTB) (Singh et al. 2001).
They also assessed the effect of anti-TB therapy on the
occurrence of hematologic abnormalities. Thirty-two
patients with DTB/MTB and 23 patients with PTB were
prospectively studied. All patients received standard
anti-TB treatment. They were subjected to a hemo-
gram including a peripheral blood examination, which
was repeated on completion of the anti-TB therapy.
Bone marrow aspiration and biopsy were also done
in all patients before starting the treatment. Normo-
cytic normochromic anemia was the most common
abnormality observed in all groups and subgroups
(DTB/MTB 84%, PTB 86%). Other hematological
abnormalities of the white blood cells include leuko-
penia (DTB/MTB 25%, PTB 0%; p<0.02), neutropenia
(DTB/MTB 22%, PTB 0%; p<0.04), lymphocytopenia,
monocytopenia, leukocytosis, neutrophilia, lympho-
cytosis, and monocytosis. Pancytopenia was observed
only in patients with DTB/MTB (p<0.05). Thrombocy-
topenia was more common in patients with DTB/MTB
(p<0.007), whereas thrombocytosis was more common
in patients with PTB (p<O.04). The DTB/MTB patients
with granulomas in the bone marrow showed certain
significant differences compared with those without
granulomas. These patients showed severe anemia,
peripheral monocytopenia, and bone marrow his-
tiomonocytosis. The hemogram reverted to normal
with anti-TB therapy in these patients. Therefore, it is
important to include TB in the differential diagnosis in
patients with unexplained hematologic abnormalities.
The most frequent hematologic abnormalities
associated with TB are cytopenias. In the following
section, these hematologic findings are discussed.
216 G.Akpek

14.2.1 even be the presenting feature of the latter form

Anemia (Perez et al. 1998; Ghobrial and Albornoz 2001).
When thrombocytopenia occurs in TB, it does so
Severe anemia was quite rare in a large series of most commonly via nonimmunologic mechanisms,
patients with pulmonary TB, occurring in less typically manifesting in the context of pancytopenia
than 20%. It is, however, quite common and occurs that develops secondary to granulomatous infiltra-
in 60%-80% of patients with disseminated TB. tion of the bone marrow. However, thrombocytope-
Therefore, severe anemia is the rule in advanced TB nia and possibly other hematologic complications
(Mandell 2000). Normocytic, normochromic anemia associated with TB can be immune-mediated. In a
(hemoglobin<10 gm/dl) and unexplained fever recent case report, a 49-year-old man who presented
should raise the suspicion of disseminated TB. with immune thrombocytopenic purpura (ITP) and
disseminated TB was described. In that particular
case, the hematological and infectious abnormali-
14.2.2 ties, which did not respond to high-dose intravenous
White Blood Cell Abnormalities corticosteroids and immunoglobulin, resolved after
anti-TB treatment (Ghobrial and Albornoz 2001). An
Leukopenia, leukocytosis, and lymphocytosis are immune basis for TB-induced ITP was also supported
relatively uncommon in TB. The white blood cell by the presence of either platelet antigen-specific
count is usually normal but may be between 10,000 antibodies or platelet surface membrane IgG in two
and 15,000 cells/mm3 in miliary TB (Mandell 2000). studies (Jurak et al. 1983; Boots et al. 1992). In the
Monocytosis is seen in less than 10% of patients. previous report, thrombocytopenia associated with
The presence or absence of monocytosis used to be TB developing concurrently in a mother and son
considered a prognostic marker for the clinical course was described. Antiplatelet antibodies were demon-
and outcome of patients with tuberculosis (Medical strated in the serum of both mother and son (Jurak
Staff Conference 1967). A rising lymphocyte count et al. 1983). Thrombocytopenia and hyponatremia
and falling monocyte count could be signs of a poor associated with miliary TB have also been described
prognosis, and a rising monocyte count a sign of a (Cockcroft et al. 1976).
good prognosis. However, this concept has never been There are other case reports in which patients with
proved. Eosinophilia occurs in about 15% of patients disseminated TB presented with clinical and labora-
and is more common in pulmonary TB than in extra- tory features of ITP (Talbot et al. 1998; Hernandez-
pulmonary TB. Maraver et al. 1996; al-Majed et al. 1995; Yasuda et al.
Leukemoid reactions are extremely rare, occur- 1994; Pavithran and Vijayalakeshmi 1993; Singh et al.
ring in less than 3% of patients with disseminated 1986; Chia and Machin 1979; Levy and Cooper 1964).
TB. The peripheral blood findings and the clinical The majority of reported patients with TB and ITP
features may be exactly the same as those of chronic were women of Middle Eastern and Asian descent.
myelogenous leukemia or, rarely, acute myeloblastic Although TB-related ITP was distributed across all
leukemia. Tuberculous leukemoid reactions and leu- age groups, it occurred most commonly between
kemia may be differentiated by cytogenetic studies, the 3rd and 8th decades of life. Pulmonary TB rep-
which in the case ofleukemia will reveal an abnormal resented the most common clinical presentation,
karyotype pattern. In patients in whom TB is associ- having occurred in 33% of patients, followed by dis-
ated with the hematologic picture of acute leukemia, seminated TB and lymphadenitis at 19% each. Some
it is likely that both diseases are present and that TB 81 % and 35% of patients presented with a platelet
is complicating acute leukemia. The documented count less than 20x109II and 5X109/l, respectively.
instances of TB producing the entire disease picture The highest platelet counts at presentation were seen
of acute leukemia are extremely rare. in 3 of the 4 patients with TB lymphadenitis.

14.2.3 14.2.4
Thrombocytopenia Pancytopenia

Thrombocytopenia is very rare in pulmonary TB. TB may be associated with pancytopenia and aplastic
However, it is a relatively common feature of dis- anemia. Disseminated TB should be considered when
seminated or miliary TB. Thrombocytopenia can pancytopenia is associated with fever and weight loss
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients
or as a cause of other obscure hematologic disorders.
Medd and Hayhoe described a disease called TB mili-
ary necrosis with pancytopenia (Medd and Hayhoe
1955). This disorder had rather distinctive pathologic
features with widespread miliary necrosis and very
little tissue reaction. Clinical symptoms include a
high, spiking fever which is called the 'typhoidal vari-
ant' of TB. Leukopenia and splenomegaly are present.
This is a rare disease, but it is important to be aware
of its existence and to differentiate it from aplastic
anemia of other causes. The clue to the diagnosis is
splenomegaly, which is not consistent with typical
aplastic anemia.
Pancytopenia can be due to hemophagocytic
syndrome, an unusual hematologic manifestation
of TB (Basu et al. 2000). TB-associated hemophago-
cytic syndrome was also reported in a hemodialysis
patient who presented with fever of unknown origin,
anorexia, weight loss, hepatomegaly, and pancytope-
nia (Yang et al. 1996). Bone marrow biopsy revealed
marked hemophagocytosis and granuloma forma-
tion with positive staining for acid-fast bacilli. The
patient expired shortly after making the diagnosis.
Despite its high mortality, disseminated TB in
association with pancytopenia may occasionally
respond to anti-TB chemotherapy with bacteriologic
cure of TB and concomitant resolution of the pan-
cytopenia. Disseminated intravascular coagulopathy
can be accompanied by pancytopenia in some of
these patients (Rosenberg and Rumans 1978).
14.2.5
Possible Mechanisms of Hematologic Findings
in Tuberculosis
There are five possible explanations for hematologic
abnormalities seen during the course of TB:
1. The first possibility is that the association of a blood
disorder and TB may be purely coincidental.
2. The second possibility is that the blood disorder
favors the development of TB, although this is rel-
atively rare. TB is one of the infections commonly
found in patients with hematologic disorders,
particularly in those whose immunologic defense
mechanisms have been further compromised by
chemotherapy.
3. The third explanation is that treatment of blood
disorders results in increased susceptibility to
TB. Patients are treated with corticosteroids and
alkylating agents, which may suppress the normal
immune responses and impair the host defense.
In patients with hematologic disorders, cellular
217
immunity may be suppressed by the disease itself
or by treatment with corticosteroids, immunosup-
pressants, or anticancer agents. Miliary TB devel-
oping in such compromised hosts is cryptic, and
thus its diagnosis is difficult to make.
Missing the diagnosis of miliary TB is a growing
problem and demands special attention. In the
years 1964 to 1974,37 patients with active TB were
misdiagnosed in the Chaim Sheba Medical Center.
The diagnosis was made only after death. Twenty-
one patients were over 60 years old. Eleven had
hematological disorders (Rosenthal et al. 1975).
Miliary TB is fatal if it cannot be detected in time.
Therefore, the possibility of mycobacterial infec-
tion should always be kept in mind when treating
patients with hematologic disorders. Conse-
quently, early suspicion and diagnosis of systemic
TB may be important. Unexplained pyrexia is the
most common symptom of systemic TB (Uetake
et al. 1990). Because of the poor prognosis of
patients with mycobacterial infection and pri-
mary hematologic disorders, prophylactic anti-
TB medication may be considered when treating
patients with hematologic disorders in areas
endemic for TB (Morii and Narita 1998).
4. The fourth possibility is that TB is associated
with the blood disease. The most common hema-
tologic abnormality that is associated with TB is
cytopenia. There are three possible ways in which
TB could cause a reduction in circulating blood
cells:
a. As with any chronic infection, there may be a
suppression of normal hematopoiesis. Chronic
infections are the major causes of anemia of
chronic disease. In association with the latter
there is usually low serum iron and iron-bind-
ing capacity. The plasma iron turnover rate is
rapid, but there is poor utilization of iron for
erythropoiesis. It is likely that suppression of
erythropoiesis in TB is partly due to increased
cytokine release (TNF-a, tumor necrosis factor-
alpha) that is the major mechanism in anemia
of chronic disease (Dallalio et al. 1999).
b. TB can produce hypersplenism, and this might
be a second cause of cytopenia. It essentially
involves increased sequestration because of the
large spleen.
c. There is actual replacement of the marrow with
tubercles and the associated fibrotic reaction.
Occasionally, TB may produce changes in the
blood that are difficult to distinguish from the
underlying hematologic disease. In several large
series of patients with TB, approximately 8%-10%
218
were found to have associated primary blood dis-
orders (Oswald 1963; Glasser et al.1970). Corr and
his colleagues, in reviewing 364 patients with all
types of TB, found that 17% had anemia, 3% had
leukocytosis, and 1% had leukopenia. They found
only one patient with immature granulocytes in
the blood, a patient with metastatic carcinoma.
They also described 6 patients who had primary
hematologic disorders with superimposed TB
(Corr et al. 1964).
In the old days, patients who had leukemic blasts
in the blood and bone marrow were given anti-
TB drugs, and all of them died during the initial
acute illness without improvement of the hema-
tologic picture. But since autopsy failed to reveal
widespread leukemic infiltrates, leukemia was not
believed to be present, and the blasts in the blood
were considered to be a response to TB. However,
subsequent epidemiologic studies failed to prove
this concept. Some of these findings are undoubt-
edly secondary to the TB. In some cases, however,
there is an underlying hematologic disease such
as leukemia and aplastic anemia. Glasser et al.
in their review suggest that anemia, leukopenia,
monocytosis, and granulocytic leukocytosis cer-
tainly represent hematologic responses to TB, for
they disappear following successful treatment
with anti-TB drugs. But the 'leukemic' blood pic-
ture or pancytopenia does not resolve, and these
patients almost invariably die after very brief ill-
nesses (Glasser et al. 1970). Based on these obser-
vations, patients with TB who have blast cells in
the peripheral blood and bone marrow also have
leukemia and, therefore, merit consideration of
anti-leukemic therapy in addition to treatment of
TB. Similarly, there is no strong evidence that TB
causes a reversible pancytopenia. Such patients,
even with anti-TB therapy, invariably die in a
relatively short time.
However, several case reports suggest that treat-
ment of TB rarely restores the blood picture to
normal. This is important in the prognosis of
these diseases. For example, the diagnosis of
refractory anemia with excess blasts, a subtype
of myelodysplastic syndrome, was made based
upon morphological and cytogenetic criteria (del
20q) in a patient with TB lymphadenitis. Hema-
tological parameters remained stable without any
specific treatment for several months, cell counts
even normalized under anti-TB therapy (Ugo et
al. 1996). In another case report, large granular
lymphocytosis associated with pulmonary TB
was described. Anti-TB therapy was started, and
G.Akpek
soon after the clinical symptoms and pancyto-
penia improved, and 1 year later, the hematologi-
cal abnormalities had disappeared (Otsuji et al.
1990).
TB may occur in the setting of all myeloprolif-
erative disorders. More than 100 patients with
pulmonary TB and polycythemia vera have been
reported. The association was sufficiently frequent
that in the old days, the physicians considered that
the lipids of the tubercle bacilli might have eryth-
ropoietic activity. Of all the myeloproliferative
syndromes, it is myelofibrosis and myeloid meta-
plasia that have the most intriguing association
with TB. There are a number of reasonably docu-
mented cases in which the myeloid metaplasia fol-
lowed the onset of TB. The second case presented
at the beginning of this chapter is an example of
this. However, the mechanism of this association
between the two diseases remained speculative.
Karel Pelger described an abnormality of
granulocyte nuclear segmentation in the con-
text of advanced TB. It is now recognized that
the Pelger-Huet nuclear anomaly (PHNA) can
be either hereditary or acquired with systemic
diseases, commonly hematologic dysplasias. A
male patient presenting with cachexia, high fever,
severe hypoproliferative anemia, and acquired
PHNA was described. At autopsy, an overwhelm-
ing TB was discovered in the absence of any other
underlying disease (Shenkenberg et al. 1982).
Based on the above observations, the following
guidelines in the approach to the association of TB
and hematological disorders can be used. First, that
the possibility of TB exists in any case of myelofi-
brosis and myeloid metaplasia, aplastic anemia, or
atypical or poorly characterized blood disorder. In
such a setting, one must make a rigorous search for
acid-fast organisms by cultures of gastric aspirates,
sputum, BAL, bone marrow, blood, and urine. In
addition, the physicians may be forced to consider
biopsy of an enlarged lymph node or, failing this,
a 'blind' biopsy of the scalenus anticus node. This
last procedure yielded the diagnosis in the first case
presented at the beginning of this chapter. Molecu-
lar diagnostic methods such as polymerase chain
reaction analysis should be utilized to diagnose
TB in clinically suspicious cases when the standard
methods fail to identify TB.
5. Finally, anti-TB drugs can cause hematologic
abnormalities. These drugs may all individually
or collectively produce pancytopenia or decrease
individual blood elements. Anti-TB medications
can cause hematologic side-effects, and the inci-
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients
dence of hematologic abnormality appears to be
associated with the degree of immunosuppres-
sion (Kavesh et al. 1989). Leukopenia is the most
common side-effect of anti-TB drugs.
Anti-TB drugs may cause idiosyncratic reactions,
malabsorption, interference with iron metabo-
lism, and hemolysis in patients with red blood
cell enzyme deficiencies. Idiosyncratic reactions
manifested by depression of any or all of the
three cellular blood elements may be caused by
any of the anti-TB drugs. Para-aminosalicylic acid
(PAS) and streptomycin are the drugs most often
responsible (Whitfield 1970). In patients who have
red blood cells deficient in enzymes such as glu-
cose 6-phosphate dehydrogenase, oxidizing agents
such as PAS may cause hemolytic anemia (Whit-
field 1970). Evidence of small-bowel intestinal
malabsorption was found in 36% of 14 patients
studied who were taking PAS for TB (Akhtar et al.
1968). In another study of 68 patients taking PAS,
approximately 90% had subnormal blood folic
acid levels, in contrast to 35% of a control group
of similar patients (Roberts et al. 1966).
Both serum and stainable bone marrow iron levels
are commonly elevated in patients taking isonia-
zid (INH) for 6 months. Sideroblastic anemia has
also been reported in patients receiving INH
therapy. This is likely to be related to interference
with pyridoxine metabolism by INH (Whitfield
1970). Hematologic abnormalities developed in
4 of 814 patients (0.5%) with pulmonary TB who
were treated for 9 months with INH and rifam-
picin, daily for 1 month and twice weekly for
the other 8 months, between January 1976 and
June 1981. Given the infrequency of hematologic
side-effects, only clinical surveillance for toxicity
was recommended (Dutt et al. 1983) for patients
receiving anti-TB medication. Perhaps a complete
blood count should be added to the liver function
studies that are evaluated during the course of TB
therapy.
14.2.6
Prognostic Significance of Hematologic
Abnormalities
Some of the hematologic manifestations of TB are
predictive for the prognosis. In a recent study, the
extent and severity of hematological abnormalities in
380 patients with pulmonary TB were evaluated. Full
blood count, bone marrow aspiration smears, and
bone marrow biopsy were performed. There was a
219
close correlation between the hematologic abnormali-
ties and the severity of clinical findings of pulmonary
TB. This survey revealed that hematologic abnormali-
ties are relatively common in severe pulmonary TB.
Body weight loss, white blood cell count, hemoglobin
level, and erythrocyte sedimentation rate appeared to
be useful indices for the severity of TB. The return of
these indices to a normal level may indicate disease
control correlating with sputum conversion to acid-
fast bacilli negative status (Bozoky et al. 1997). In a
retrospective cohort study, 26 patients (24%) died of
miliary TB a median of 6 days after starting treatment.
Survivors were followed up for a median of 51 weeks.
Stepwise logistic regression identified age greater than
60 years, lymphopenia, thrombocytopenia, hypoalbu-
minemia, elevated transaminase levels, and treat-
ment delay as independent predictors of mortality
(Maartens et al. 1990).
14.2.7
Tuberculosis-Induced Lymphopenia
and Immunosuppression
TB is clearly responsible in some patients for CD4
cell lymphocytopenia that reverses with treatment
(Onwubalili et al. 1987; Turett and Telzak 1994). Many
HIV-negative patients with active TB have CD4 cell
counts much lower than 500 cells/I, which return
toward normal with treatment of TB. Recent clinical
data on 85 HIV-negative patients with TB indicated
that there is a close relationship between the CD4+
T-cell count and the severity of TB. The disease
severity was associated with greater depression of
the total lymphocyte and CD4 cell counts. CD4 cell
counts returned to normal levels in most patients
after 1 month of therapy (Jones et al. 1997).
In TB, Mycobacterium tuberculosis (MTB)-stimu-
lated T-cell responses are depressed transiently,
whereas antibody levels are increased. Lymphopro-
liferative responses of peripheral blood mononuclear
cells (PBMCs) from Pakistani TB patients to both
mycobacterial and candidal antigens were suppressed
by approximately 50% compared with healthy purified
protein derivative (PPD)-positive household contacts.
Production of interferon-gamma (IFN-y) in response
to PPD also was depressed by 78%. Stimulation with
PPD and the 30 kDa alpha antigen of MTB (30 kDa
antigen) induced greater secretion of transforming
growth factor-beta (TGF-~), but not of interleukin
10 (IL-lO) or TNF-a, by PBMCs from TB patients
compared with healthy contacts. The degree of sup-
pression correlated with the duration of treatment:
220
patients treated for less than 1 month had significantly
lower T-cell blastogenesis and IFN-y production and
higher levels of TGF-~ than did patients treated for
longer than 1 month. Neutralizing antibody to TGF-~
normalized the lymphocyte proliferation in response
to PPD, partially restored blastogenesis to candidal
antigen, and significantly increased PPD-stimulated
production ofIFN-yin TB patients but not in contacts.
Neutralizing antibody to IL-lO augmented, but did
not normalize, T-cell responses to both PPD and Can-
dida in TB patients and candidal antigen in contacts.
TGF-~, produced in response to MTB antigens, may
therefore playa prominent role in down-regulating
potentially protective host effector mechanisms and
can be an important mediator of immunosuppression
in TB (Hirsch et al. 1996).
14.3
Hematologic Abnormalities Associated
with Mycobacterial Infections
in Immunocompromised Hosts
14.3.1
Hematologic Complications of HIV Infection
and AIDS
The hematologic manifestations of HIV infection and
AIDS are common and may cause symptoms that are
life-threatening and impair the quality of life of these
patients. These manifestations include morphologic
abnormalities of peripheral blood and bone marrow
changes (Aboulafia and Mitsuyasu 1991). Cytopenias
are the most common hematologic findings in HIV
patients. Anemia and neutropenia are generally caused
by inadequate production because ofsuppression ofthe
bone marrow by the HIV infection through abnormal
cytokine expression and alteration of the bone marrow
microenvironment. Thrombocytopenia is caused by
immune-mediated destruction of the platelets in addi-
tion to inadequate platelet production. The incidence
and severity of cytopenia are generally correlated to the
stage of the HIV infection. Other causes of cytopenia in
these patients include adverse effects of drug therapy,
the secondary effects of opportunistic infections, espe-
cially mycobacterial infections, malignancies or other
preexisting or coexisting medical problems that may be
prevalent in the HIV-infected population. Diagnosis of
the mechanism and cause of the cytopenia may allow
for specific management (Coyle 1997).
In evaluating HIV-infected patients with abnor-
mal blood counts, one must consider the hematologic
G.Akpek
effects of various infectious pathogens or malignan-
cies. Infectious agents such as Mycobacterium avium
complex (MAC) or TB may cause direct bone marrow
suppression or reticuloendothelial system dysfunc-
tion, resulting in depressed blood cell counts (Cas-
tella et al. 1985; O'Hara 1989).
14.3.2
Mycobacter;um av;um Complex in HIV-Infected
Patients
Disseminated MAC infections are common in patients
with HIVI AIDS, occurring in up to 50% of patients
(Ellner et al.1991; Nightingale et al.1992). In contrast
to the immunocompetent host, in which MAC disease
is usually limited to the lungs, bacteremia is by far
the most common syndrome in patients with AIDS.
Frequently, there is widespread dissemination, often
to the liver, bone marrow, lymph nodes (Fig. 14.1),
and spleen. In one study, lymphadenopathy, hepa-
tosplenomegaly, and severe anemia «8.5 g/dl) were
seen in 37%, 24%, and 85% of patients evaluated,
respectively (Benson and Ellner 1993).
MAC infection is one of the most common compli-
cations seen during the course of HIV infection and
is associated with significant suppression of blood
counts (Castella et al. 1985). Bone marrow aspira-
tion and biopsy may be useful in making a specific
diagnosis of several complications of AIDS that result
in cytopenia. In histologic sections of bone marrow,
the diagnosis of MAC infection is suggested by the
finding of granulomas and aggregates of foamy his-
tiocytes (Figs. 14.1b and 14.2a,b) in which the organ-
isms can be seen by acid-fast staining (Fig. 14.2c).
Granuloma formation is frequently lacking or scarce
(Fig. 14.2b), so the pathologist should be alerted that
disseminated MAC is suspected to ensure that acid-
fast staining is performed (French et al. 1997).
14.3.3
Anemia Associated with MAC in HIV-Infected
Patients
Severe anemia (hematocrit <26%) has been reported
to be present in 76% of patients with disseminated
MAC disease (Horsburgh et al. 1994). The presence
of anemia has been correlated with the presence of
disseminated MAC disease in HIV-infected patients
who have absolute CD4+ lymphocyte counts below
100lml (Havlick et al. 1992). The survival of patients
with MAC disease was negatively correlated to the
 Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients 221

a a

b b

Fig.14.la-c. Lymph node histology in an HIV-infected patient
with co-existing MAC infection. a Poorly formed granulomas
(-100). b Poorly formed granulomas composed of numerous
foamy histiocytes (-400). c AFB stain shows numerous intracel-
lular acid-fast mycobacteria (M. avium intracellulare) (-1000).
Courtesy of Metin Ozdemirli, MD, PhD, Assistant Professor in
Hematopathology, Georgetown University, Washington, DC,
USA
Fig.14.2a-c. Bone marrow histology in an HIV-infected patient
with co-existing MAC infection. a Numerous, well formed gran-
ulomas (-100). b A few paratrabecular poorly formed granulo-
mas (-100). c AFB stains show scattered acid-fast mycobacteria
(M. avium intracellulare) (-1000). Courtesy of Metin Ozdemirli,
MD, PhD, Assistant Professor in Hematopathology, Georgetown
University, Washington, DC, USA
222
presence of anemia (Horsburgh et al. 1994). The
presence of anemia, fever, and weight loss in patients
with advanced immunodeficiency should prompt an
evaluation for MAC disease (Coyle 1997).
The anemia in MAC disease is thought to be due
to high levels of inflammatory cytokines as well as
a direct disturbance of the bone marrow microenvi-
ronment by mycobacteria (Coyle 1997). Serum TNF
levels were demonstrated to be markedly elevated in
children with anemia and MAC diseases. In a recent
study, the hematologic manifestations of MAC in 37
HIV-infected infants and children were reviewed.
Anemia was the predominant feature in all patients,
with severe anemia (hemoglobin <6 g/dl) occurring
in 7 of 34 (21 %) patients. This was followed by leuko-
penia (79%), monocytosis (82%), thrombocytopenia
(59%), leukoerythroblastic changes (68%), and neu-
tropenia (41 %). Serum TNF-a was markedly elevated
in all patients with MAC, and there appeared to be an
association between elevated TNF-a and anemia in
these patients (Ellaurie and Rubinstein 1995).
MAC infection may disrupt erythropoiesis in HIV
patients by direct invasion of the bone marrow. Patients
with disseminated MAC appear to have greater impair-
ment of erythropoiesis than those without MAC. This
impairment appears to be mediated by selective sup-
pression of hematopoietic progenitors (Gascon et al.
1993) and production of inhibitory TNF-a (Ellaurie
and Rubinstein 1995). To elucidate the mechanisms of
anemia and other cytopenias in HIV patients, various
cytokine and cytokine receptor concentrations were
measured by ELISA in bone marrow aspirate super-
natants from 19 HIV patients undergoing diagnostic
evaluation and 14 healthy paid volunteer controls.
IL-l ~ and IFN-y were rarely detectable. All cytokines/
receptors detectable in marrow supernatant, except
RANTES, showed mean concentrations 1.6- to 6.2-fold
higher in patients with HIV compared with healthy
controls. Elevated TNF-a and MIP-l ~ was associated
with marrow involvement by lymphoma, Hodgkin
disease, or mycobacterial infection. Concentrations
of all cytokines/receptors measured correlated with
the severity of the anemia. CD8+ lymphocytes were
inversely correlated with concentrations of all cytokines
measured other than MIP-la (Dallalio et al.1999).
14.3.4
Other Cytopenias Associated with MAC
in HIV-Infected Patients
Bone marrow infiltration with MAC may cause
thrombocytopenia or increase the severity of exist-
G.Akpek
ing HIV-associated thrombocytopenia. However,
thrombocytopenia due to bone marrow infiltration
is not nearly as common as anemia of MAC infec-
tion (Coyle 1997). The absence or presence of mega-
karyocytes on bone marrow examination is often
useful in distinguishing thrombocytopenia caused
by decreased production due to marrow infiltration
of mycobacteria from thrombocytopenia caused
by increased destruction during the course of HIV
infection or due to drug side-effect (Coyle 1997).
Bone marrow infiltration with disseminated MAC
infection may cause neutropenia by suppression of
bone marrow progenitors (Coyle 1997).
14.3.5
MAC Infections in Other Immunocompromised
Hosts
Infection with atypical mycobacteria occurs mainly
in patients with a compromised cellular immune
system, in particular in those with a defective T-
cell or monocyte function. The specific immune
response of an adolescent HIV-negative patient with
disseminated M. avium infection and fatal varicella
zoster virus infection was studied in a patient who
presented with dysplastic hematopoesis of all cell lin-
eages and anemia plus leukopenia. No hematologic
malignancy was found. Peripheral lymphopenia and
monocytopenia as well as a lack of natural killer
(NK)-cells and B-cells were noted. Lymphocytes con-
sisted of T-cells (95%), which contained up to 40%
of TCR y8+CD4-CD8- T-cells, few monocytes and
B-cells. Approximately 50% of CD3+ T-cells showed
a CD57+ NK-like phenotype. Functional analysis of
mononuclear cells revealed a good antigen-specific
T-cell function if the antigen-presenting cells were
supplemented from a HLA-matched donor. More-
over, a strong M. avium specific cytotoxicity medi-
ated by TCR a~+ T-cells could be found in vitro and
even ex vivo. In contrast, NK-killing was absent. No
evidence for a defect in IL-12 or IFN-g production
and signaling was found (Wendland et al. 2000).
In a patient with pulmonary MAC disease asso-
ciated with idiopathic CD4+T lymphocytopenia,
hematologic studies showed a low CD4+ cell count
in the absence of any identifiable immunodeficiency,
including HIV infection. With the combination of
chemotherapy and surgery, he had a good clinical
outcome (Ishida et al. 1998). Disseminated atypi-
cal MAC infection in a patient receiving interferon
treatment for hairy-cell leukemia was also described
(Maurice et al. 1988).
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients 223

14.3.6
Hemophagocytic Syndrome Associated
with MAC

This rare manifestation of MAC was recently

described. A 37-year-old man with insulin-depen-
dent diabetes mellitus complicated by end-stage
renal disease who had received a second cadaveric
kidney transplant 4 years earlier presented with
symptomatic splenomegaly, progressive anemia,
prolonged fever, dysphagia, and weight loss. The first
kidney transplant, as well as a pancreatic transplant,
a
had been lost because of acute rejection 7 and 3 years
previously, respectively. Owing to chronic rejection of
his kidney transplant, he required intensive immu-
nosuppressive therapy consisting of tacrolimus,
mycophenolate mofetil, and prednisone. Abdominal
CT performed on admission revealed splenomegaly
with a subcapsular hematoma. Analysis of a bone
marrow aspirate revealed numerous large histiocytes
with striated cytoplasm (onion-skin appearance),
resembling Gaucherie's cells. Acid-fast staining of a
bone marrow biopsy specimen showed that the his-
tiocytic inclusions were actually innumerable acid-
fast bacilli. PCR analysis of the bone marrow identi-
fied the microorganism as MAC. Despite appropriate
antibiotic therapy with azithromycin and ethambu-
tol, respiratory, renal, and liver failure developed, and
the patient ultimately died of profound lactic acidosis
and septic shock due to disseminated MAC infection
(Argiris et al. 1999).

14.3.7
Miliary/Disseminated Tuberculosis in AIDS

HIV-infected patients are predisposed not only to

reactivation of remote TB but also to rapid progres-
sion of recently acquired infection (Daley et al.1992; Di
Perri et al.1989). It is also probable that AIDS increases Fig. 14.3a-c. Lymph node histology in an HIV-infected patient
the susceptibility to acquisition of new infection, and with co-existing tuberculosis. a Granulomas with central
the diagnosis of TB among hospitalized adults aged necrosis (-25). b Granulomas with central necrosis and occa-
sional Langerhans giant cells (-100). c AFB stain shows acid-
15-44 years is also very commonly associated with fast mycobacteria (M. tuberculosis) (-1000). Courtesy of Metin
concurrent HIV infection (40% of patients) (Rosen- Ozdemirli, MD, PhD, Assistant Professor in Hematopathology,
blum et al. 1994). With advanced HIV infection, TB Georgetown University, Washington, DC, USA.
often has an atypical presentation, with extrapulmo-
nary disease a prominent feature. The most common
forms of extrapulmonary TB are lymphadenitis and
disseminated disease (Fig. 14.3). In patients with
advanced HIV, TB may present as progressive primary
infection or disseminated disease. In AIDS patients,
10% with TB and 38% with extrapulmonary TB have
miliary disease (Shafer et al. 1991).
224
Major constitutional symptoms and hectic fevers
are characteristic. The chest X-ray is abnormal in
80% and may include typical miliary mottling. Only
10% are tuberculin-positive (Salzman et al. 1992).
The sputum smear is positive in only 25% (Shafer et
al. 1991), but cultures of many materials are positive,
including blood in 50%-60%. Biopsies show a typi-
cal tuberculous histologic appearance but with more
stainable organisms than in non-HIV miliary TB
(Fig. 14.3). In contrast, the histologic picture is often
nonreactive TB in fatal cases (Salzman et aI. 1992).
Miliary TB in HIV-infected persons may also
cause acute respiratory distress syndrome or tuber-
culous papular skin lesions. Smears of respiratory
secretions are positive in 80%. Fulminant miliary TB
may be associated with severe refractory hypoxemia
(adult respiratory distress syndrome) and dissemi-
nated intravascular coagulation (DIC). In such cases,
adjunctive corticosteroids (60-80 mg of prednisone
daily) are indicated.
14.3.8
Diagnostic Yield of Bone Marrow Aspiration!
Biopsy for Mycobacterial Infections in Patients
with HIV Infection
Hematology consultants are frequently involved in
the care of patients with HIV1AIDS through requests
for diagnostic bone marrow (BM) aspiration and
biopsy, particularly when opportunistic mycobac-
teria or histoplasma infections are suspected. MAC
and TB infections in particular continue to be impor-
tant causes of HIV-related deaths (Selik et aI. 1995;
Sehonanda et aI.1996; Chin et aI.1994). Consequently,
a determination of optimal strategies for the diagno-
sis of these infections continues to be an important
clinical issue. While multiple retrospective studies
have been reported (Barreto et aI. 1993; Northfelt et
aI. 1991; Ciaudo et aI. 1994; Riley et aI.1995; Benito et
aI. 1997; Kilby et aI. 1998), no consensus has emerged
with respect to the specific value of BM examinations
in the diagnosis of MAC or TB infections in immu-
nocompromised HIV-infected patients.
BM examination can be a useful method of
diagnosing opportunistic mycobacterial and fungal
infections in patients with fever, anemia, or neutro-
penia, and underlying HIV infection. Detection of
granuloma with or without positive AFB staining is a
common and valuable histologic clue to opportunis-
tic infection. The diagnostic yield of BM granuloma
formation was evaluated in 6988 BM biopsies taken
from 1973 to 1986. Granulomas were identified in
G.Akpek
72 specimens. The granulomas were associated with
infectious disease (30%), hematologic disorders
(25%), sarcoidosis (11 %), nonhematologic malig-
nancies (10%), drug reaction (5%), other diseases
(6%), and no final diagnosis (6%). Patients with fever
of unknown origin were 15 times more likely to have
marrow granulomas than patients biopsied for other
reasons (Bhargava and Farhi 1988).
In another series of 342 BM examinations from 314
patients with HIV infection, 70 examinations (20%)
detected opportunistic mycobacterial or fungal infec-
tions. Of the 314 patients, III had such infections, and,
hence,63% (70/111) were detected by BM examina-
tion. Special stains for microorganisms detected 16
(32%) of 50 MAC, 10 (22%) of 45 TB infections, 8
(73%) of 11 Histoplasma capsulatum (HC) infections,
and 5 (83%) of 6 Cryptococcus neoformans infections.
BM cultures detected 36 (72%) of the 50 MAC infec-
tions, 13 (29%) of the 45 TB infections, and 63% of
the fungal infections. Marrow examination revealed
infection in only 1 of the 70 specimens (1 %) collected
to evaluate thrombocytopenia alone or hematologic
malignancy, but in 69 (25%) of 274 with fever, neutro-
penia, anemia, or miscellaneous other indications for
marrow examination. Granulomas were detected in
102 (30%) of the biopsy specimens, including 71 (64%)
of those in patients with mycobacterial or fungal infec-
tion. The granulomas showed caseous necrosis in 9,
all in patients with TB, and the 27 patients with TB-
associated granulomas tended to show large, tightly
cohesive granulomas. The presence of granulomas
correlated with opportunistic infection in 82 (80%)
of 102 cases. Without granulomas, special stains were
positive in only 8 (3%) of 240 specimens, suggesting
that special stains alone may not be a cost-efficient way
to diagnose such infections (Nichols et al. 1991).
Although these pathogens are not associated with
specific BM findings, disseminated MAC or TB may
cause increased reticulum, occasional granulomas,
and positive smears for AFB (more abundant in MAC
cases and relatively sparse in TB cases), and rarely
pseudo-Gaucher cells in the BM (Solis et al. 1986;
Argiris et al. 1999). The most common manifesta-
tion of fungal and mycobacterial infection is a diffuse
marrow infiltrate of loose aggregates and clusters of
macrophages (Fig. 14.2b) (Castella et al. 1985; O'Hara
1989). The configuration of acid-fast bacilli in or
around the granuloma also differs between these two
Mycobacterium species. In TB,AFBs resemble a bunch
of cigarettes with a single cigarette dropped next to the
bunch. In MAC, the AFBs tend to be seen as clumps
of bacilli in loose aggregates (Carl J. O'Hara, personal
communication). Histiocytic erythrophagocytosis has
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients 225

Table 14.1. Sensitivity, specificity, and positive and negative predictive values of studies for iden-
tification of mycobacterial infection and histoplasma (Akpek et al. 2001)
Predictive value

Study Sensitivity" Specificitt PositiveC Negatived

BM culture 63% (20/32) 100% (24/24) 100% (20/20) 67% (24/36)

Blood culture 63% (20132) 100% (24124) 100% (20/20) 86% (31/36)
(+) AFB in BM 19% (6132) 100% (24/24) 100% (616) 48% (24/50)
(+) Granuloma 34% (11132) 96% (23124) 91% (11112) 52% (23/44)

a Sensitivity: The percent (or proportion) of patients ultimately diagnosed with MAC/TB or HC
who had positive studies.
b Specificity: The percent (or proportion) of patients who had negative studies who ultimately
did not have MAC/TB or He.
C Positive predictive value: The percent (or proportion) of patients who had positive studies

who were ultimately diagnosed with MAC/TB or He.

d Negative predictive value: The percent (or proportion) of patients who had negative studies
who were ultimately not diagnosed with MAC/TB or HC

also been observed and is a nonspecific finding often
seen in association with infections, including myco-
bacterial infections and lymphomas.
Successful detection of organisms in the BM of
AIDS patients requires the use of special stains and
sensitive culture techniques (O'Hara 1989). The high-
est yield from special stains is obtained when there
is a plentiful infiltrate of macrophages, especially in
aggregates or in poorly formed granulomas (Castella
et al. 1985). Even in the absence of noticeable granu-
lomas or macrophage infiltration, acid-fast organ-
isms and fungi can be diagnosed when present in
high numbers (Cohen et al.1983).
In order to determine the value of BM culture and
BM histology in the diagnosis of opportunistic MAC/
TB and HC infections in immunosuppressed patients
with HIV, we retrospectively reviewed the records of
56 adult patients with HIV who underwent a single
BM aspiration, biopsy, and culture because of unex-
plained fever and/or other clinical features sugges-
tive of MAC/TB or HC infection (Akpek et al. 2001).
Thirty-two patients (57%) were ultimately diagnosed
with MAC/TB (n=30) or HC (n=2) infection by posi-
tive cultures of BM, blood, sputum, or bronchoalveo-
lar lavage (BAL) fluid or by the histologic detection of
organisms in biopsies of BM or other tissues.
The diagnostic sensitivity of blood cultures was
equivalent to that of BM culture (63%) in the diag-
nosis of MAC/TB and HC infections, as reported by
others (Northfelt et al.1991; Kilby et al.1998) and was
greater than that of BM histopathology alone (34%)
(Table 14.1). Blood cultures were found to be nega-
tive in 5 of 20 patients for which BM samples were
culture-positive.
Granuloma and/or histologically apparent organ-
isms were seen in BM biopsy specimens in 11 of 32
individuals (34%) ultimately diagnosed with MAC/TB
or HC infections. Among these 11 patients, both gran-
uloma and acid-fast staining organisms were found in
the BM biopsy specimens of 2 individuals for whom
both BM and blood cultures were negative. The diag-
nosis would have been missed in 6 patients (19%) ifthe
BM had not been examined. The 34% diagnostic sen-
sitivity of BM histopathology observed in our study is
similar to the 20%-30% sensitivities reported in other
studies (Riley et al. 1995; Kilby et al. 1998; Nichols et al.
1991). One-fourth of BM biopsy specimens showing
granuloma were not associated with positive blood
or BM cultures, and this finding suggested a 9% false-
positive rate for this histopathologic feature by itself.
The low diagnostic sensitivity of BM histopathology
alone is not surprising. An autopsy study of 44 AIDS
patients with MAC bacteremia found no histologic
evidence of MAC in 30% (Torriani et al. 1994). Other
studies have also indicated that some individuals
suffer mycobacterial bacteremia without detectable
organ or tissue invasion (Torriani et al. 1996).
A detailed evaluation of clinical and laboratory
parameters observed in our study population indi-
cates that the specific clinical features of high peak
temperature, longer duration of febrile days, and
elevated direct bilirubin at least greater than 1 mg!
dl of the upper limit are predictive of MAC/TB or
HC infections and are therefore useful in selecting
patients most likely to derive diagnostic benefit from
BM examination (Akpek et al. 2001).
With recent evidence supporting the value of MAC
prophylaxis in immunosuppressed HIV-infected
patients (Pierce et al. 1996), it is likely that in the future
an increasing proportion of patients may receive
antimycobacterial prophylaxis at the time of diagnos-
tic evaluation for infection. This change in practice
226
could affect the diagnostic yield of both cultures and
BM histopathology. Furthermore, with the develop-
ment of new molecular methods for the detection
of mycobacterial DNA or ribosomal RNA in clinical
specimens within 24 h, with sensitivities ranging from
40% to 77% and specificities greater than 95% (Havlir
and Barnes 1999; Barnes 1997; Condos et al. 1996), it
is likely that a reluctance to carry out diagnostic BM
examinations will continue, particularly given the
additional considerations of cost, patient discomfort,
risk of needle-stick exposures for individuals involved
in BM sampling, and evidence for responses to empiric
anti-MAC therapy in some patients.
While the diagnostic sensitivity of BM cultures
may not be greater than that of blood cultures in
detecting MAC/TB or HC infections in immunosup-
pressed HIV + patients, histopathologic examination
of BM specimens often results in the relatively rapid
identification of nearly one-third of infected patients
who underwent BM examination and can also iden-
tify infections in some patients who are culture-nega-
tive. Based on the above findings, we should continue
to use BM aspiration, biopsy, and culture for the
diagnosis of opportunistic MAC/TB or HC infections
in immunosuppressed HIV + patients, particularly
when selected clinical features are present.
14.3.9
Other Mycobacterial Infections Associated
with Hematologic Disorders
Mycobacterial disease, especially atypical mycobac-
teria, is relatively often seen in hairy-cell leukemia
(HCL). The clinical picture is usually dominated by
persistent fever. Sweet's syndrome manifested by
fever, skin rash, and HCL has been described (Kram-
ers et al. 1992). Blood cultures grew M. kansasii. The
patient recovered after treatment with recombinant
IFN-a and tuberculostatic drugs. The skin lesions
completely regressed within 1 week after the start of
r-IFN-a. Leukopenia secondary to M.leprae was also
described (Brooks et al. 1990).
G. Akpek
14.4
Coagulation Abnormalities Associated
with Mycobacterial Infections and Clinical
Presentations
14.4.1
Disseminated Intravascular Coagulation
DIC is a very rare complication of pulmonary TB.
However, there have been numerous case reports
published about the association between DIC and TB
(Krishnaswamy 1969; Goldfine et al. 1969; Mavligit
et al. 1972; Jenss and Ostendorf 1980). Clotting fac-
tors might be locally consumed, thus producing the
syndrome of 'multifocal vasculopathic coagulation'
(Krauss and Walker 1979).A patient with cavitary TB
complicated with DIC has been described (Fujita et
aI.1997). Rifampin was considered to treat the clinical
course of DIe. Another patient developed subclinical
DIC due to anti-TB drugs, probably rifampicin. The
patient also developed marked leucocytosis, a flu-like
illness, intravascular hemolysis, and acute renal fail-
ure as part of the drug reaction (Ip et al. 1991).
A review of all 13 patients reported to date shows
that the patients are generally black, middle-aged,
male, alcoholic, and febrile. The form of TB is gener-
ally miliary and is associated with a high mortality.
Eight of these patients had associated ARDS. Only
one patient had an acute tuberculous peritonitis. In
6 patients the coagulopathy began after the start of
therapy; steroids did not appear to affect survival.
The exact pathophysiologic mechanisms involved
in the development of DIC are unknown (Stein and
Libertin 1990).
TB-associated hemophagocytic syndrome (HPS)
has recently been recognized as a benign reactive
histiocytic proliferation with marrow hemophago-
cytosis. A patient with TB complicated by severe BM
failure and DIC was reported. The immunological
disturbances usually occurring in miliary TB may
playa role in the pathogenesis of HPS (Eliopoulos
et al. 1992).
14.4.2
Deep Vein Thrombosis Associated
with Tuberculosis
Deep venous thrombosis (DVT) associated with TB
infection has rarely been reported (Gogna et al.
1999). In a pediatric patient who developed DVT
of the left leg in association with pulmonary TB,
transient protein S deficiency and anticardiolipin
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients
IgG and IgM antibodies were identified (Casanova-
Roman et al. 2002). In a retrospective analysis of
clinically diagnosed and lower limb DVT proven by
contrast venography, DVT complicated admissions
in 46 (3.4%) of 1366 adult patients treated in a TB
hospital during 1986. Analysis of 7542 admissions
during 1978-1986 showed a relative risk of about
5 in patients treated with regimens including
rifampicin compared with other regimens. DVT
was significantly more common in the winter
months and usually occurred within 2 weeks of
treatment being started (White 1989). This prob-
able association between rifampicin and DVT does
not contraindicate use of this drug, but measures to
prevent DVT should be taken for patients receiving
rifampicin.
14.4.3
Thrombotic Thrombocytopenic Purpura
Associated with Mycobacterial Infections
Thrombotic thrombocytopenic purpura (TTP)
during infectious diseases is a known but rare event
(Pene et al. 2001). Various therapeutic approaches
were used with the patient who developed TTP
during the course of primary TB infection: fresh
frozen plasma infusions and plasma exchange, spe-
cific anti-TB therapy, antiplatelet drugs, and steroids.
A complete remission occurred 3 months after the
onset of the acute disease (Toscano et al. 1995). One
of the hypotheses about the pathogenesis of TTP
might be an increased procoagulant activity of IL-l
on endothelial cells. Rifampicin therapy was reported
to be associated with the development of TTP in a
patient with TB (Fahal et al. 1992). A patient with
disseminated MAC infection also developed TTP
(Skopinski et al. 2001).
Excessive bleeding with unclear etiology can occur
rarely in patients with TE. A fatal case of pulmonary
TB masquerading as diffuse alveolar hemorrhage
after autologous stem-cell transplant was recently
reported (Keung et al. 1999).
14.5
Conclusion
The types and mechanisms of hematologic find-
ings seen during the course of mycobacterial
infections were covered in this chapter. Hemato-
logic findings in patients with active mycobacterial
227
infections are usually considered by clinicians a
part of the chronic inflammatory condition, and
thus they are usually overlooked. However, it is
extremely important to recognize the diagnostic
and prognostic value of these hematologic findings
because of the possible impact on the outcome of
these patients. Severe hematologic abnormalities
indicate an advanced mycobacterial infection that
requires immediate attention regarding diagnosis
and treatment, which may alter the course in these
deadly diseases.
References
Aboulafia DM, Mitsuyasu RT (1991) Hematologic abnormali-
ties in AIDS. Hematol Oncol Clin North Am 5:195-214
Akhtar AJ, Crompton GK, Schonell ME (1968) Para-aminosali-
cylic acid as a cause of intestinal malabsorption. Tubercle
49:328-331
Akpek G et al (2001) Bone marrow aspiration, biopsy, and
culture in the evaluation of HIV-infected patients for
invasive mycobacteria and histoplasma infections. Am J
HematoI67:100-106
Al-Majed SA et al (1995) Tuberculosis presenting as immune
thrombocytopenic purpura. Acta HaematoI94:135-138
Argiris A, Maun N, Berliner N (1999) Mycobacterium avium
complex inclusions mimicking Gaucher's cells. N Engl J
Med 340:1372
Barnes PF (1997) Rapid diagnostic tests for tuberculosis:
progress but no gold standard. Am J Respir Crit Care Med
155:1497-1498
Barreto JA et al (1993) Isolation of Mycobacterium avium
complex from bone marrow aspirates of AIDS patients in
Brazil. J Infect Dis 168:777-779
Basu S, Mohan H, Malhotra H (2000) Pancytopenia due to
hemophagocytic syndrome as the presenting manifesta-
tion of tuberculosis. J Assoc Physicians India 48:845-846
Benito N et al (1997) Bone Marrow Biopsy in the diagnosis of
fever of unknown origin in patients with acquired immu-
nodeficiency syndrome. Arch Intern Med 57:1577-1580
Benson CA, Ellner JJ (1993) Mycobacterium avium complex
infection and AIDS: advances in theory and practice. Clin
Infect Dis 17:7
Bhargava V, Farhi DC (1988) Bone marrow granulomas:
clinicopathologic findings in 72 cases and review of the
literature. Hematol PathoI2:43-50
Boots RJ, Roberts AW, McEvoy D (1992) Immune thrombocy-
topenia complicating pulmonary tuberculosis. Case report
and investigation of mechanism. Thorax 47:396-397
Bozoky G et al (1997) Hematologic abnormalities in pulmo-
nary tuberculosis. Orv HetilI38:1053-1056
Brooks BJ Jr,Alvarez S, Yoder L (1990) Leukopenia secondary
to Mycobacterium leprae. J La State Med Soc 142:35-36
Cameron SJ (1974) Tuberculosis and the blood: a special rela-
tionship. Tubercle 5:65-72
Casanova-Roman M et al (2002) Deep venous thrombosis
associated with pulmonary tuberculosis and transient
protein S deficiency. Scand J Infect Dis 34:393-394
228
Castella A et al (1985) The bone marrow in AIDS: a histologic,
hematologic and microbiologic study. Am J Clin Pathol 84:
425
Centers for Disease Control and Prevention (2000) Reported
tuberculosis in the United States, 1999. Centers for Disease
Control and Prevention, Atlanta
Charfi MR et al (1998) Disseminated tuberculosis in non-
immunocompromised host: three case reports. Rev Med
Interne 19:917-920
Chia YC, Machin SJ (1979) Tuberculosis and severe thrombo-
cytopenia. Br J Clin Pract 33:55-56
Chin DP et al (1994) The impact of Mycobacterium avium
complex bacteremia and its treatment on survival of AIDS
patients - a prospective study. J Infect Dis 170:578-584
Ciaudo M et al (1994) Revisited indications for bone marrow
examinations in HIV-infected patients. Eur J Haematol 53:
168-174
Cockcroft DW et al (1976) Miliary tuberculosis presenting
with hyponatremia and thrombocytopenia. Can Med
Assoc J 115:871-873
Cohen RJ et al (1983) Occult infections with M. intracellulare
in bone marrow biopsy specimens from patients with
AIDS. N Engl J Med 308:1475
Condos R et al (1996) Peripheral-blood-based PCR assay
to identify patients with active pulmonary tuberculosis.
Lancet 347:1082-1085
Corr WP, Kyle RA, Bowie EJW (1964) Hematologic changes in
tuberculosis. Am J Med Sci 248:709-714
Coyle TE (1997) Hematologic complications of human
immunodeficiency virus infection and AIDS. In: Gold
JWM, Telzak EE, White DA (eds) Management of the
HIV-infected patient, part II. Infections and malignancies
associated with HIV infection. Saunders, Philadelphia, pp
449-470 (Medical clinics of North America, vol 81)
Daley CL, Small PM, Schecter GF (1992) An outbreak of
tuberculosis with accelerated progression among persons
infected with the human immunodeficiency virus. N Engl
J Med 36:231-235
Dallalio G et al (1999) Cytokine and cytokine receptor concen-
trations in bone marrow supernatant from patients with
HIV: correlation with hematologic parameters. J Invest
Med 47:477-483
Di Perri G, Danzi MC, DeChecchi G (1989) Nosocomial epi-
demic of active tuberculosis among HIV-infected patients.
Lancet 2:1502-1504
Dutt AK, Moers D, Stead WW (1983) Undesirable side effects
of isoniazid and rifampin in largely twice-weekly short-
course chemotherapy for tuberculosis. Am Rev Respir Dis
128:419-424
Dye C et al (1999) Consensus statement: global burden of
tuberculosis: estimated incidence, prevalence, and mortal-
ity by country: WHO Global Surveillance and Monitoring
Project. JAMA 282:677-686
Eliopoulos G et al (1992) Tuberculosis associated hemophago-
cytic syndrome complicated with severe bone marrow fail-
ure and disseminated intravascular coagulation. Nouv Rev
Fr HematoI34:273-276
Ellaurie M, Rubinstein A (1995) Elevated tumor necrosis
factor alpha in association with severe anemia in human
immunodeficiency virus infection and mycobacterium
avium intracellulare infection. Pediatr Hematol Oncol 12:
221-230
Ellner n, Goldberger MJ, Parenti DM (1991) Mycobacterium
G.Akpek
avium infection and AIDS: a therapeutic dilemma in rapid
evolution. J Infect Dis 163:1326-1335
Fahal IH et al (1992) Thrombotic thrombocytopenic purpura
due to rifampicin. Br Med J 304:882
Flinn JW, Finn RS (1929) Annu Rev Tuberculosis 20:347
French AL, Benator DA, Gordin FM (1997) Nontuberculous
mycobacterial infections. In: God JWM, Telzak EE, White
DA (eds) The management of the HIV-infected patient,
part II. Infections and malignancies associated with HIV
infection. Saunders, Philadelphia, pp 361-379 (Medical
clinics of North America, vol 81)
Fujita M et al (1997) Disseminated intravascular coagulation
associated with pulmonary tuberculosis. Intern Med 36:
218-220
Gascon P, Sathe SS, Rameshwar P (1993) Impaired erythropoiesis
in the acquired immunodeficiency syndrome with dissemi-
nated Mycobacterium avium complex. Am J Med 94:41-48
Ghobrial MW, Albornoz MA (2001) Immune thrombocytope-
nia: a rare presenting manifestation of tuberculosis. Am J
HematoI67:139-143
Glasser RM, Walker RI, Herion JC (1970) The significance of
hematologic abnormalities in patients with tuberculosis.
Arch Intern Med 125:691-695
Gogna A et al (1999) Tuberculosis presenting as deep vein
thrombosis. Postgrad Med J 75:104-105
Goldfine ID et al (1969) Consumption coagulopathy in miliary
tuberculosis. Ann Intern Med 71:775-777
Havlick JA et al (1992) Disseminated mycobacterium avium
complex infection: clinical identifiers and epidemiologic
trends. J Infect Dis 165:577-580
Havlir DV, Barnes PF (1999) Tuberculosis in patients with
human immunodeficiency virus infection. N Engl J Med
340:367-373
Hernandez-Maraver D et al (1996) Immune thrombocyto-
penic purpura due to disseminated tuberculosis. Acta
Haematol 96:266
Hirsch CS et al (1996) Cross-modulation by transforming
growth factor beta in human tuberculosis: suppression of
antigen-driven blastogenesis and interferon gamma pro-
duction. Proc Natl Acad Sci USA 93:3193-3198
Horsburgh CR et al (1994) Predictors of survival in patients
with AIDS and disseminated mycobacterium avium com-
plex disease. J Infect Dis 170:573-577
Ip M, Cheng KP, Cheung WC (1991) Disseminated intravas-
cular coagulopathy associated with rifampicin. Tubercle
72:291-293
Iseman MD (2000) Tuberculosis in relation to human immu-
nodeficiency virus and acquired immunodeficiency
syndrome. A clinician's guide to tuberculosis. Lippincott
Williams and Wilkins, Philadelphia, pp 199-252
Ishida T et al (1998) Pulmonary Mycobacterium avium disease
in a young patient with idiopathic CD4+ T lymphocytope-
nia. Intern Med 37:622-624
Jenss H, Ostendorf P (1980) Leukopenia and consump-
tion coagulopathy in miliary tuberculosis. Schweiz Med
Wochenschr 110:1208-1213
Jones BE et al (1993) Relationship of the manifestations of
tuberculosis to CD4 cell counts in patients with human
immunodeficiency virus infection. Am Rev Respir Dis 148:
1292-1297
Jones BE et al (1997) CD4 cell counts in human immuno-
deficiency virus-negative patients with tuberculosis. Clin
Infect Dis 24:988-991
Hematologic Findings in Mycobacterial Infections Among Immunosuppressed and Immunocompetent Patients
Jurak SS, Aster R, Sawaf H (1983) Immune thrombocytope-
nia associated with tuberculosis. Clin Pediatr (Phila) 22:
318-319
Kavesh NG, Holzman RS, Seidlin M (1989) The combined
toxicity of azidothymidine and antimycobacterial agents.
A retrospective study. Am Rev Respir Dis 139:1094-1097
Keung YK et al (1999) Mycobacterium tuberculosis infec-
tion masquerading as diffuse alveolar hemorrhage after
autologous stem cell transplant. Bone Marrow Transplant
23:737-738
Kilby JM et al (1998) The yield of bone marrow biopsy and
culture compared with blood culture in the evaluation of
HIV-infected patients for mycobacterial and fungal infec-
tions. Am J Med 104:123-128
Kramers C et al (1992) Sweet's syndrome as the presenting
symptom of hairy cell leukemia with concomitant infection
by Mycobacterium kansasii. Ann Hematol 65:55-58
Krauss JS, Walker DH (1979) Miliary tuberculosis and con-
sumption of clotting factors by multifocal vasculopathic
coagulation. South Med J 72:1479-1481
Krishnaswamy V (1969) Disseminated tuberculosis presenting
as a bleeding disorder. J Indian Med Assoc 53:85-87
Levy M, Cooper BA (1964) Thrombocytopenic purpura asso-
ciated with tuberculous lymphadenitis. Can Med Assoc J
90:373-374
Maartens G, Willcox PA, Benatar SR (1990) Miliary tuber-
culosis: rapid diagnosis, hematologic abnormalities, and
outcome in 109 treated adults. Am J Med 89:291-296
Mandell GL et al (2000) Principles and practice of infectious
diseases. Mycobacterial infections
Maurice PD et al (1988) Mycobacterium avium-intracellulare
infection associated with hairy-cell leukemia. Arch Derma-
tol 124:1545-1549
Mavligit GM, Binder RA, Crosby WH (1972) Disseminated
intravascular coagulation in miliary tuberculosis. Arch
Intern Med 130:388-389
Medd WE, Hayhoe FGJ (1955) Tuberculous miliary necrosis
with pancytopenia. Quart J Med 24:351
Medical Staff Conference (1967) Hematologic manifestations
of tuberculosis. Calif Med 106:215-221
Moore M, McCray E, Onorato 1M (1999) TB-AIDS versus non-
AIDS TB cases, United States, 1993-1997.1nt J Tuberc Lung
Dis S20-S20 abstract
Morii T, Narita N (1998) Mycobacterial infection in patients
with hematologic disorders. Nippon Rinsho 56:3209-3211
Nichols L et al (1991) Bone marrow examination for the
diagnosis of mycobacterial and fungal infections in the
acquired immunodeficiency syndrome. Arch Pathol Lab
Med 115:1125-1132
Nightingale SD et al (1992) Incidence of Mycobacterium
avium-intracellulare complex bacteremia in human
immunodeficiency virus-positive patients. J Infect Dis
165:1082-1085
Northfelt DW et al (1991) The usefulness of diagnostic bone
marrow examination in patients with human immuno-
deficiency virus (HIV) infection. J Acquir Immune Delic
Syndr 4:659-666
O'Brien JR (1954) Nonreactive tuberculosis. J Clin Pathol 7:
216-225
O'Hara CJ (1989) The lymphoid and hematopoietic systems.
In: Harawi SJ, O'Hara CJ (eds) Pathology and pathophysi-
ology of AIDS and HIV-related diseases. Mosby, St Louis,
pp 150-164
229
Onwubalili JK, Edwards EJ, Palmer L (1987) T4lymphophenia
in human tuberculosis. Tubercle 68:195
O'Reilly RA (1998) Splenomegaly in 2,505 patients in a large
university medical center from 1913 to 1995. West J Med
169:78-87
Oswald NC (1963) Acute tuberculosis and granulocytic disor-
ders. Br Med J 2:1489-1496
Otsuji A et al (1990) Transient large granular lymphocytosis
associated with pulmonary tuberculosis: a case report.
Rinsho Ketsueki 31:1955-1959
Pavithran K, Vijayalekshmi N (1993) Thrombocytopenic
purpura with tuberculous adenitis. Indian J Med Sci 47:
239-240
Pene F et al (2001) Septic shock and thrombotic micro-
angiopathy due to Mycobacterium tuberculosis in a
nonimmunocompromised patient. Arch Intern Med 161:
1347-1348
Perez de Llano LA, Soilan del Cerro JL, Garcia Pais MJ (1998)
Imune thrombocytopenic purpura as presenting form of
miliary tuberculosis. Arch Broncho PneumoI34:411-412
Pierce M et al (1996) A randomized trial of clarithromycin
as prophylaxis as prophylaxis against Mycobacterium
avium complex infection in patients with advanced
acquired immunodeficiency syndrome. N Engl J Med
335:384-391
Riley UB et al (1995) Detection of mycobacteria in bone
marrow biopsy specimens taken to investigate pyrexia of
unknown origin. J Clin PathoI48:706-709
Roberts PD, Hoffbrand AV, Mollin DL (1966) Iron and folate
metabolism in tuberculosis. Br Med J 5507:198-202
Rosenberg MJ, Rumans LW (1978) Survival of a patient with
pancytopenia and disseminated coagulation associated
with miliary tuberculosis. Chest 73:536-539
Rosenblum LS et al (1994) Effect of HIV infection and tuber-
culosis on hospitalizations and cost of care for young
adults in the United States, 1985 to 1990. Ann Intern Med
121:786-792
Rosenthal T, Pitlik S, Michaeli D (1975) Fatal undiagnosed
tuberculosis in hospitalized patients. J Infect Dis 131
[Suppl):S51-S56
Salzman SH et al (1992) The role of bronchoscopy in the diag-
nosis of pulmonary tuberculosis in patients at risk for HIV
infection. Chest 102:143-146
Sehonanda A, Choi YJ, Blum S (1996) Changing patterns of
autopsy findings among persons with acquired immunode-
ficiency syndrome in an inner-city population. Arch Pathol
Lab Med 120:459-464
Selik RM, Chu SY, Ward JW (1995) Trends in infectious dis-
ease and cancers among persons dying of HIV infection
in the United States from 1987 to 1992. Ann Intern Med
123:933-936
Shafer RW et al (1991) Extrapulmonary tuberculosis in
patients with human immunodeficiency virus infection.
Medicine 70:384-397
Shenkenberg TD, Rice L, Waddell CC (1982) Acquired Pelger-
Huet nuclear anomaly with tuberculosis. Arch Intern Med
142:153-154
Singh KJ et al (2001) Significance ofhaematological manifesta-
tions in patients with tuberculosis. J Assoc Physicians India
49:790-794
Singh SP et al (1986) Tubercular lymphadenitis with purpura.
J Indian Med Assoc 84:247-249
Skopinski S et al (2001) Mycobacterium avium complex sep-
230
ticemia: a new cause of thrombocytopenic thrombotic
purpura. Rev Med Interne 22:400-401
Slavin RE, Walsh TJ, Pollack AD (1980) Late generalized tuber-
culosis: a clinical pathologic analysis and comparison of
100 cases in the preantibiotic and antibiotic eras. Medicine
(Baltimore) 59:352-366
Solis DC, Belmonte AH, Ramaswamy G (1986) Pseudo-Gau-
cher cells in Mycobacterium avium-intracellulare infection
in the acquired immunodeficiency syndrome (AIDS). Am
J Clin Pathol 85:233
Stein DS, Libertin CR (1990) Disseminated intravascular
coagulation in association with cavitary tuberculosis.
South Med J 83:60-63
Talbot S et al (1998) Mediastinal nodal tuberculosis presenting
as immune thrombocytopenia. Aust NZ J Med 28:465-466
Torriani PJ et al (1994) Autopsy findings in AIDS patients
with mycobacterium avium complex bacteremia. J Infect
Dis 170:1601-1605
Torriani PJ et al (1996) Disseminated Mycobacterium avium
complex: correlation between blood and tissue burden. J
Infect Dis 173:942-949
Toscano Vet al (1995) Thrombotic thrombocytopenic purpura
associated with primary tuberculosis. Infection 23:58-59
Tulliez M (1976) Bone marrow's histopathology in tuberculosis
G.Akpek
of the hematopoietic organs. Pathol Bioi (Paris) 24:699-711
Turett GS, Telzak EE (1994) Normalization of CD4+ T-Iympho-
cyte depletion in persons without HIV infection treated for
tuberculosis. Chest 105:1335
Uetake T et al (1990) Clinicopathological study of miliary
tuberculosis in patients with hematologic disease. Kek-
kaku 65:273-283
Ugo V et al (1996) Myelodysplasia and tuberculosis. Bull
Cancer 83:548-552
Wendland T et al (2000) Strong alpha beta and gamma delta
TCR response in a patient with disseminated Mycobacte-
rium avium infection and lack of NK cells and monocyto-
penia. Immunol Lett 72:75-82
White NW (1989) Venous thrombosis and rifampicin. Lancet
2:434-435
Whitfield CL (1970) Hematologic abnormalities in tubercu-
lous patients. Arch Intern Med 126:698
Yang CW et al (1996) Tuberculosis-associated hemophago-
cytic syndrome in a hemodialysis patient: case report and
review of the literature. Nephron 72:690-692
Yasuda Y et al (1994) A case of intractable pulmonary
tuberculosis complicated by idiopathic thrombocyto-
penic purpura. Nippon Kyobu Geka Gakkai Zasshi 42:
2301-2305
15 Immunodiagnostics for Latent Tuberculosis Infection
ROHIT K. KATIAL

CONTENTS reactivity with nontuberculous strains, difficulty in

interpretation, and test interference by underlying
15.1 Introduction 231
co-morbid conditions. Because of the resurgence of
15.2 History of TB Skin Testing 231
15.3 TB Immune Response 232 TB infection associated with the AIDS epidemic and
15.4 PPD Strength 232 with increasing urban crowding, more specific and
15.5 PPD Testing Methodology 233 sensitive tests that will detect exposure to Mycobac-
15.6 Mantoux Testing Method of Measuring terium tuberculosis (MTB) are needed. New in vitro
Induration 233
diagnostics are being evaluated as potential replace-
15.7 TB Skin Test Interpretation 234
15.8 Booster 235 ments for the old TB skin test.
15.9 Adverse Reactions to PPD Testing 235
15.10 New Trends in Diagnosing Latent
TB Infection 235
References 237
15.2
History of T8 Skin Testing

TB was an epidemic that first appeared in animals

The opinions or assertions contained herein are
the private views of the author and are not to be
construed as official or as reflecting the views of
the Department of the Army or the Department of
Defense.
15.1
Introduction
Historical accounts date mycobacterial infections
back thousands of years, but it was not until the late
19th century that the causative, agent responsible for
tuberculosis (TB) was discovered. The veterinary
industry in its attempt to eradicate bovine TB led
the effort to diagnose both latent and active TB
infections. The skin testing method successfully
used in animals was soon optimized for human
diagnostics. Screening methods have changed little
in the last 100 years, and the skin test is fraught with
limitations including poor specificity due to cross-
R. K. KATIAL, MD
Associate Professor of Medicine, National Jewish Medical and
Research Center Denver, 1400 Jackson Street, Denver, CO 80206,
USA
in the Paleolithic period long before the disease
occurred in man some 8000 years ago (Bates and
Stead 1993). Four thousand years later, one finds
written reference to TB in the Vedas (Kiple 1993).
In the Indian subcontinent, the Hindus referred to
TB as consumption (Davis 2000), a term used until
quite recently. As urban development increased and
city centers were formed in the early 1600s, the inci-
dence of TB increased sharply. Over the subsequent
200 years, the epidemic spread throughout Western
Europe. At the end of the 18th century, TB was ram-
pant, but the causes were speculative and treatment
ineffective. In 1882, Robert Koch isolated the tubercle
bacillus and thus began the rapid advancement in
understanding both the disease pathophysiology and
diagnostics (Koch 1890, 1891).
Robert Koch developed a MTB filtrate known as
old tuberculin (OT) from heat-sterilized cultures,
which he mistakenly promoted as a therapeutic agent
for TB. Soon thereafter, OT began to be evaluated for
use in early diagnosis, initially in cattle by comparing
skin testing reactivity with autopsy data. The asso-
ciation of subclinical infection with a reactive skin
test was made when asymptomatic cattle with posi-
tive skin tests demonstrated tuberculi on autopsies
(Pearson 1892). Any skin test reactivity at this point
was considered positive for TB because nonspecific

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
232
reactivity from nontuberculous strains had not been
recognized. After widespread testing in animals, the
occasional discrepancy between skin test reactivity
and the absence of organisms on autopsy became
apparent. By the 1930s, it was widely accepted in
the veterinary community that skin test conversion
could occur after exposure to nontuberculous strains
of mycobacteria (Hagan 1931). Similar nonspecific
reactivity was seen in humans when large-scale
testing was done through the 1930s and 1940s. In
addition, it was noted that OT showed batch-to-batch
variability and led to inconsistent skin test results
(Snider 1982). With this observation, work began
on developing a more consistent and standardized
product. In 1934, Florence Seibert prepared the puri-
fied protein derivative (PPD), and in 1941, Seibert
and Glenn prepared lot no,49608, the international
skin test standard reagent PPD-S (Seibert and Glenn
1941). All PPD preparations from that point forward
had to be bioequivalent to PPD-S.
15.3
18 Immune Response
The injurious effects from MTB are a result of the
defensive responses exhibited by the host. The infec-
tious process begins with inhalation of particles in the
respirable range of 1-5 flm. These particles settle in the
alveolus where they are engulfed by alveolar macro-
phages. The immune response to TB is a complex inter-
action between antigen-presenting cells (APC) and a
variety of T-Iymphocytes, with a humoral response
being much less significant. The MTB bacillus is taken
up by a phagosome within the APC that attempts to
fuse with a lysosome and ultimately create a mature
phagolysosome. However, one of the mechanisms of
immune evasion by TB is the prevention of the forma-
tion of the mature phagolysosome (Kaufmann 2001).
The intracellular presence of the bacillus results in
complexing of TB antigens with MHC class II mol-
ecules. This leads to activation of CD4+ T-cells, which
produce IFN-y among other cytokines and ultimately
recruit more macrophages to the site of infection, thus
perpetuating the cycle (Iseman 1999). Antigen presen-
tation also occurs through nonpolymorphic CD1 and
polymorphic MHC class I molecules, thus leading to
activation of CD8+ and double-negative a~T-cells.
Evidence is also mounting that yoT-cells may con-
tribute to the tuberculin response (Kaufmann 2001).
All these T-Iymphocytes have the ability to produce
IFN-y, with CD4+ cells being the primary producer
R. K. Katial
as well as the cell central to the recall delayed-type
hypersensitivity (DTH) response seen after injection
of PPD in individuals previously exposed to MTB.
The contribution of other cytokines such as IL-12,
IL-18, and TNF-~ to the pathogenesis and protective
response of TB infection has been nicely elaborated
in other reviews (Collins and Kaufmann 2001; Iseman
1999; Daniel 1980; Daniel et al.2000; Tsuyuguchi 1996).
IFN-yis the primary cytokine involved with governing
immune protection to TB and has recently been used
in the development of immunodiagnostics.
TB-infected individuals present tuberculin antigens
by various ways and generate effector T-cells, which
then continue to circulate in the blood. The nonsteril-
izing quality of the initial exposure may serve to con-
tinually stimulate the immune system, thus maintain-
ing a reservoir of MTB-specific T-cells. The response
to PPD is a classic DTH reaction, involving vasodila-
tion, edema, and infiltration initially of antigen-spe-
cific T-cells which then secrete cytokines leading to
recruitment of other nonspecific inft.ammatory cells
(Ahmed and Blose 1983). The result of these inft.am-
matory events is induration beneath the skin and
some degree of erythema. The maximum induration
is generally seen 48 h after injection of PPD, and this
is not correlated with the erythema. DTH responses
can persist for years after exposure to the bacillus.
The immune mechanism of the DTH response differs
from the IFN-yresponse mounted for protection from
the disease. Orem and Copper (1999) postulated that
immune protection is governed by cytokine produc-
tion such as IFN-y, while the DTH response may be a
primary result of chemokine production. The clinical
importance of this is that in actively infected individu-
als, the DTH reaction may be depressed initially but
returns within a month of therapy, whereas IFN-y
may remain suppressed for at least 12 months (Ellner
1997; Hirsch et al. 1996, 1999). This clearly suggests
that in vitro measurement of IFN-y production is not
a direct correlate of the less expensive PPD and recall
antigen (e.g., measles, mumps, tetanus) skin tests but
rather that the two reactions are governed by different
immune mechanisms.
15.4
PPD Strength
In the USA, two types of tuberculin are available: OT
and PPD. OT is only used with multipuncture devices,
whereas PPD is used for both multipuncture and the
intracutaneous test. PPD is manufactured by protein
Immunodiagnostics for Latent Tuberculosis Infection 233

precipitation of culture titrate using either tricWoro- Induration at the site of injection is usually read at 48-
acetic acid or ammonium sulfate. Although both prep- 72 h. Howard and Solomon (1988) evaluated readings
arations are standardized by bioequivalency using a at 24, 48, and 72 h and reported 71 % sensitivity and
skin testing methodology (Villarino et al. 1999), the 9% false-positive rate at 24 h. Thus, the authors recom-
antigen content of the two may differ due to the dif- mended reading the test at the standard 48-72 h, while
ferences in the precipitation methods. PPD contains being aware that some reactions may occur earlier and
primarily small proteins (less than 10,000 Da) but also others may persist longer than 72 h.
has polysaccharides and lipids. The small proteins and
lack of adjuvant prevent PPD from being sensitizing
(American Thoracic Society 2000).
The PPD used for intracutaneous testing comes in 15.6
three concentrations: 1 tuberculin unit (TU), 5 TU, Mantoux Testing Method of Measuring
and 250 TV. PPD 5 TU is the only strength that is Induration
bioassayed (standardized in vivo to the PPD-S).
PPD 1 TU was shown to have only 50% sensitivity in The CDC standard for reading PPD induration is
children, while PPD 250 TU identified all specific and palpating along the transverse axis of the forearm
nonspecific reactors (Murtagh 1980; Gryzbowski et (Core Curriculum on Tuberculosis, Centers for
al. 1969). Palmer (1945) studied student nurses who Disease Control 1994). However, several investiga-
were tested with PPD 5 TU and 250 TV. PPD 5 TU tors have compared this palpation method with a
proved to be highly sensitive in identifying students ballpoint method first introduced by Sokal (1975),
with pulmonary calcifications and/or lung infiltrates. who claimed a reduction in the observer variability
On the other hand, reactivity with PPD 250 TU did with the pen method. Correlation between these two
not correlate with pulmonary findings or contact methodologies is high, and differences between read-
history. Therefore, the 1 and 250 TU concentrations ings are small and often not clinically relevant. When
should not be used for testing because of lack of Pouchot et al. (1997) compared PPD readings by pal-
standardization. pation versus the ballpoint method, they found a high
correlation between the two techniques but reported
a 38% broader area of imprecision with palpation.
The skin test readers in the study were not blinded
15.5 to the patients, and only two very experienced read-
PPD Testing Methodology ers performed all the tests. A study by Howard and
Solomon (1988) evaluating 806 volunteers found
The two primary skin-testing methods used are the discordance between the two methodologies in the
multipuncture device and the Mantoux test. Multiple 5-14 mm range. The readers in their study were also
puncture devices such as the tine test are inexpen- not blinded. The authors suggested that the readings
sive, simple to apply, and much more rapid. They obtained with the pen method should be interpreted
introduce antigen through tuberculin-coated prongs with caution in the 5-14 mm range. Jordan et al.
on the applicator device, or by puncturing through (1987) and Bouros et al. (1991) conducted double-
a liquid film placed on the forearm; however, such blind studies comparing the two methods, and both
devices have numerous drawbacks including unre- showed that the pen method is comparable to palpa-
liable dose delivery (Lunn 1980), inconsistent tech- tion in the hands of experienced readers. Bouros and
niques of administration, and significant false-posi- colleagues (1992) later studied the two methods with
tive reactions. A study by the Research Committee of readings from both experienced and inexperienced
the British Thoracic Association (1982) showed up to readers. No significant differences were found in any
23% variability in subjects with duplicate tine tests. range of measurement by the two groups using both
Because of these shortcomings, the current recom- methods. The authors concluded that inexperienced
mendation is that positive tine tests need validation readers can use either methodology accurately. Long-
by the commonly adopted Mantoux method, which field and colleagues (1984) studied interobserver
has become the standard recommended by the variability in 101 patients and found no significant
United States Centers for Disease Control (CDC). difference between the two methods. They found
The Mantoux test is performed by injecting 0.1 ml comparable variability between the two techniques;
of PPD 5 TU intradermally with a 26-27 gauge needle using a 10 mm cut-off, less than 7% of readings would
with bevel upward, creating a pale bleb above the skin. have changed the clinical interpretation.
234 R. K. Katial

At our institution we use the ballpoint method,
measuring along two axes of the forearm. We ret-
rospectively evaluated 5680 skin test records and
analyzed only the positive records (380), comparing
the transverse to the longitudinal axis as well as the
average of the two. We found no statistically signifi-
cant difference (p=0.135) between the transverse
and longitudinal readings; in fact, the total number
of subjects classified positive and negative were the
same in each group, using both a 10 mm and 15 mm
cut-off, although the specific individuals varied
(Table 15.1) (Hershey et al. 2001). This suggests that
the two readings are equivalent. However, when the
larger of the two readings was used to determine the
classification, a slightly higher number of people
(24/380 with a 15 mm cut-off) were classified as
positive than by the single transverse reading. When
the average of the two readings was used, the results
were similar to the standard transverse method but
may reflect enhanced precision because two read-
ings have been made. The intraobserver variability
has been reported to be 3.0 mm greater or less than
the first measurement (Pouchot et al. 1997). In our
data, the variability between the two readings was
3 mm or less in 81 % of the subjects. However, choos-
ing the larger of the two measurements provided
slightly greater sensitivity. It appears that adher-
ence to one technique and proper application and
training are more important than which technique
or axis is used.
15.7
T8 Skin Test Interpretation
In determining the cut-off for a clinically positive
result indicating exposure to MTB, one must under-
stand the reactivity patterns in various populations
as well as the factors contributing to false-positive
and false-negative results. In the 1950s, the World
Health Organization (WHO) collated data of skin
test induration for patients in TB hospitals through-
out the USA, Europe, and Southeast Asia. The high-
est number of reactors had indurations of approxi-
mately 16-17 mm with a normal distribution. In
every country, the frequency curves overlapped. In
some regions, a bimodal distribution was described

Table 15.1. Comparison of skin test readings by reading direction (transverse vs

longitudinal)
No. of Percentage
subjects
(n=380)

Transverse classification at ~1O mm cut-off:

Negative 59 15.5%
Positive 321 84.5%
Longitudinal classification at ~10 mm cut-off:
Negative 59 15.5%
Positive 321 84.5%
Transverse classification at ~15 mm cut-off:
Negative 167 43.9%
Positive 213 56.1%
Longitudinal classification at ~15 mm cut-off:
Negative 167 43.9%
Positive 213 56.1%
Identical readings in mm 108 28.4%
Transverse reading larger than longitudinal 106 27.9%
No. of readings with difference >3 mm 34 8.9%
Longitudinal reading larger than transverse 166 43.7%
No. of readings with difference >3 mm 39 10%
No. of readings reclassified" at 15 mm cut-off 24 6.3%
No. reclassified" at 15 mm cut-off
where reading difference >3 mm 9 2.4%
No. of readings reclassified" at 10 mm cut-off 7 1.8%
No. reclassified" at 10 mm cut-off
where reading difference >3 mm 3 0.7%

" Standard classification is based on transverse reading. Reclassification reflects

change to positive classification when larger longitudinal direction is used
Immunodiagnostics for Latent Tuberculosis Infection
with one peak at 5-9 mm and another at 16-17 mm
(Snider 1982). Based on previous animal data, it
was felt that the smaller peak reflected reactivity to
nontuberculous (NTM) strains, such as M. avium-
intracellulare (MAl). Between 1958 and 1970, a study
evaluating 275,558 Navy recruits tested all subjects
with both PPD-S and PPD-B (MAl strains). The
individuals with large reactions to nontuberculous
PPD tended to have smaller reactions (6-11 mm)
with PPD-S (Edwards et al. 1969). The incidence of
active TB in 0-5 mm PPD reactors was 36/100,000, in
6-11 mm PPD reactors it was 110/100,000, and with
indurations with PPD >12 mm the incidence of dis-
ease was 380/100,000. In another study, Palmer et al.
(1959) evaluated skin-test reactivity in Pakistani chil-
dren and again demonstrated a bimodal distribution
with peaks at 2-5 mm and 17 mm. They concluded
that the first peak was due to NTM, and the larger
size peak represented true TB reactors. The reactions
from NTM tend to be smaller than from MTB. Thus,
increasing the cut-off for a positive reaction increases
the specificity but reduces the sensitivity. In the USA,
this has been the rationale for a 15 mm cut-off, since
the prevalence of NTM is high (particularly in the
southern USA) and of MTB low. A value of 15 mm is
also close to the peak of reactivity described by the
WHO in true reactors. In groups at high risk for TB,
the cut-off is lowered in order to improve the sensitiv-
ity; immunocompromised groups, in particular, may
have smaller size reactions or become anergic due to
immune dysregulation, and thus the cut-off has been
reduced to 5 mm.
15.8
Booster
Although the PPD skin-test reactivity can persist
throughout life, occasionally there is a diminution
or absence in the DTH response to PPD over time.
In such individuals, a return in response may be
seen if tested sequentially over several weeks. This
phenomenon of accentuation of response after repeat
testing is commonly known as the booster effect and
must not be interpreted as a new skin test conversion.
Boostering most often occurs in persons older than
age 55 years in whom cellular immunity may wane.
Health care institutions where repeat TB testing is
performed use an initial screening two-step test (two
PPD skin tests about 3 weeks apart) to establish a
baseline reactivity, in order to avoid falsely classify-
ing a repeat test at a later time as a new conversion.
235
15.9
Adverse Reactions to PPD Testing
Generally, adverse reactions to TB skin testing are
rare; however, immediate hypersensitivity reactions
have been reported and are in no way correlated to
underlying TB infection (Tarlo et al. 1977; Wright et
al. 1989). Individuals may also develop vesicular or
ulcerating lesions at the site of PPD injection. These
severe reactions occur in a small percentage ofpositive
reactors; however, a previous positive does not predict
an increased likelihood of developing a severe cutane-
ous reaction (Reichman and O'Day 1977). Health care
providers use topical corticosteroid creams to treat
these local reactions, but one controlled study using
hydrocortisone did not demonstrate efficacy, although
the use of a more potent steroid may have shown ben-
efit (Hanson and Comstock 1968).
15.10
New Trends in Diagnosing Latent
TB Infection
The immune response to mycobacterial infection is
predominantly cellular (Kaufmann 2001). The PPD
skin test has been a convenient, cost-effective method
for assessing cell-mediated immune responses to
mycobacterial-derived proteins. Although the test
is reasonably priced, there continue to be multiple
factors challenging the accuracy of the PPD skin test
(ST) in different settings. These factors include (but
are not limited to) variability in operator placement
and reading, cross-reactivity among mycobacterial
species (including M. avium and BeG), the need for
the patient to return in 48-72 h for a reading, and
the modulation of the skin response due to underly-
ing illness or immunosuppression. Clearly, there is a
need for another more sensitive and specific test that
would overcome the limitations of the skin test. In
addition, the tools used to diagnose active disease are
currently different from the skin test, which is used
to detect exposure.
The major immunodiagnostic advances to date
have centered on more sensitive and rapid testing to
determine the presence of active disease, but prog-
ress in the area of detecting exposure in healthy indi-
viduals has been slower. Recent work has shifted to
identifying recombinant and purified antigens that
may be used for skin testing and in immunoassays
with greater sensitivity. One of the first tuberculosis
complex-specific antigens discovered was MPT64, a
236 R. K. Katial

24 kDa protein. More recently, other antigens have Table 15.2. Agreement between Q-1FN and SKT for latent TB
been purified, including 38 kDa (antigen b), 10 kDa, infection in various studies
18 kDa, MPT59 (Ag85B), and ESAT-6 (Gennaro 2000). Reference Skin test Agreement between
MPT64 has produced disappointing results when classification Q-1FN test and SKT
used as a skin-test reagent. Only 6% of PPD-reactive (Q-1FN/SKT)a
TB-positive patients responded to MPT64 in compar- Bellete et aI. N (HIV-) 68% (43/63)
ison with a 50% reactivity with MPT59 (Wilcke et al. (2002) P (HIV-) 70% (921131)
1966). Antigen b has been shown to be very antigenic Overall, H1V- 700/0 (135/194)
N (H1V+) 85% (22126)
and, when studied in guinea pigs, elicited a positive
P (H1V+) 40% (ll/28)
skin test in MTB-vaccinated animals but not in those Overall, H1V+ 61% (33/54)
exposed to other strains. The antigen appears to be Converse et aI. N (H1V -) 40% (6115)
more specific than PPD (Haslov et al. 1990). Vorder- (1997) P (H1V - ) 100% (17117)
meier et al. (1992), using an epitope located near the Overall, H1V- 72% (23/32)
N (H1V+) 73% (16122)
38 kDa antigen of MTB, elicited a DTH reaction in
P(H1V+) 92% (ll/12)
PPD-positive individuals but not in negative ones. Overall, H1V+ 79% (27/34)
These antigens may find a role in immunoassays and Desem and Jones N 100% (10/10)
possibly in skin testing by providing the ability to (1998) P 90% (9/10)
detect newly exposed individuals with greater speci- Overall 95% (19/20)
ficity as well as to differentiate those vaccinated from KatiaI et aI. N 80% (16/20)
(2001) P 80% (16/20)
those infected. Overall 80% (32/40)
The immune response to MTB is highly depen- Kimura et aI. N (H1V-) 500/0 (115/229)
dent upon IFN-y production by macrophages and (1998) P (H1V -) 89% (63171)
antigen-specific T-cells; over the past decade, there Overall, HIV - 59% (178/300)
N (H1V+) 85% (128/151)
has been an increasing interest in the development
P (H1V +) 56% (9/16)
and application of in vitro culture assays measuring OveraIl, H1V + 82% (137/167)
IFN-y production in response to tuberculin antigen Mazurek et aI. N 900/0 (738/818)
stimulation as a substitute diagnostic screening test (2001) P 65% (1471227)
for the classic PPD skin test (Lein and von Reyn Overall 85% (885/1045)
Pottumarthy et aI. N (high-risk country) 89% (135/151)
1997). Initially using peripheral blood mononuclear
(1999) P (high-risk country) 64% (55/86)
cells, the methodology evolved to a whole blood cul- N(HCW) 81% (64179)
ture technique that was first validated in Australian P (HCW) 67% (32/48)
cattle. These studies demonstrated that the IFN-y test OveraIl 79% (286/364)
had greater diagnostic sensitivity, lower cost, and Streeton et aI. N 88% (480/545)
rapid results for cattle TB screening (Rothel et al. (1998) P 900/0 (163/182)
Overall 88% (643/727)
1992; Wood et al. 1991; Wood and RotheI1994). This
method was developed for human TB testing using a With ~15% = positive Q-1FN, excluding those with active
human PPD, avian PPD, and the mitogen phytohe- disease or past history of active disease
N =negative skin test defined at various cut-offs depending on
magglutinin (PHA). risk; P =positive skin test defined at various cut-offs depend-
A standardized diagnostic kit with a specifically ing on risk; Hew = health care worker
defined data analysis procedure has been marketed
by CSL in Australia (now Cellestis, Australia): Quan-
tiFERON-TB or Q-IFN. In various studies compar- Converse et al.(1997) used Q-IFN in studies compar-
ing the Q-IFN assay with PPD-ST, the agreement ing the IFN release assay with PPD-ST in popula-
ranges from 40% to 100% when looking at subjects tions at risk for MTB exposure (intravenous drug
without active disease or a history of active disease users with or without HIV infection). They found
(Table 15.2). Agreement has been lower in those with that the Q-IFN assay detected more reactors than
active disease. Streeton et al. (1998) reported a sensi- PPD-ST. Agreement between the two tests was weak.
tivity and specificity of 90.5% and 98%, respectively, We studied 40 patients and found good agreement
for diagnosing MTB exposure in human subjects. between PPD-ST and the Q-IFN kit (kappa=0.73).
In their study, the gold standard for diagnosis was We evaluated the agreement between in vivo and
the PPD-ST, and the study subjects were stratified in vitro tests and did not refer to sensitivity and
according to their skin test induration and risk of specificity because they assume comparison to an
exposure to tuberculosis. Kimura et al. (1998) and adequate gold standard, which we feel does not exist
Immunodiagnostics for Latent Tuberculosis Infection
(Katial et al. 2001). Recently, Mazurek and colleagues
(2001) evaluated 1226 volunteers and found overall
83% agreement between the IFN-y assay and skin test
in individuals at both low and high risk for latent TB
infection. However, only 68% agreement was noted in
the positive PPD-ST group. A similar study by Bellete
et al. (2002) reported an overall agreement in a USA
cohort of 79% (in Baltimore), but lower agreement
(68%) was seen in Ethiopia, an endemic TB area. In
addition, the authors found poor reproducibility in
the in vitro results from one test to another.
One significant drawback of both the skin test and
the Q-IFN test is the nonspecific response to PPD
because of cross-reactivity between tuberculin and
other mycobacterial species. Although the Q-IFN test
does distinguish M. avium from MTB reactivity, it
does not differentiate between responses from MTB
and other mycobacterial species such as M. kansasii,
M. marum, and M. africanum. The false-positivity in
the PPD skin testing due to BCG vaccination is well
documented. However, the in vitro IFN-yproduction
by BCG-vaccinated individuals in response to PPD
may also be influenced by cross-reacting mycobacte-
rial antigens. Streeton et al. (1998) included BCG vac-
cinees in their study but were unable to discern the
effects of BCG on the assay results. Therefore, addi-
tional studies are needed to delineate the diagnostic
value of the Q-IFN kit in this population. The in vitro
diagnostic system affords a distinct advantage over
PPD-ST in that one can test for tuberculin-specific
proteins without unnecessarily exposing the patient.
The low-molecular-weight antigen ESAT-6 (6 kDa)
has been shown to differentiate between MTB and
BCG strains and thus may serve as an additional
stimulant to determine the effect of BCG. Recently,
ESAT-6 was evaluated in the Q-IFN assay and found
to differentiate those infected with TB from controls,
with high sensitivity and specificity (Johnson et al.
1999). We studied low-molecular-weight culture fil-
trate proteins, culture filtrate proteins minus lipoara-
binomannan, soluble cell wall proteins, and cytosolic
proteins in whole blood culture, and none of the sub-
fractions performed any better than the whole PPD
(Katial et al. 2001). However, the crude preparations
used in this study did stimulate IFN-y production in
sensitized individuals and should be further purified
and studied in the whole blood system.
The immune response to TB is complex and
directed to a heterogeneous mixture of antigens rather
than anyone protein. Therefore, studying a cocktail of
native or recombinant antigens may prove to be more
specific and sensitive in diagnosing TB than anyone
immunodominant protein. The advantages of the
237
blood test include the absence of any reading or place-
ment variability and the need for only one office visit.
The current disadvantages are the need for stringent
laboratory requirements dealing with blood handling,
cell culture, and ELISA testing. The areas needing
clarification include defining the performance charac-
teristics and lower cut-off limit for the enzyme immu-
noassay. Data are needed on temporal measurements
to determine if the time of day or different days have
any impact on the whole blood stimulation response.
Finally, the cut-offs differentiating a positive from a
negative response suggested in the Q-IFN kit need
to be verified based on the prevalence of disease in
different populations as has been done with PPD-ST.
These issues should be addressed and followed up by
large-scale trials to assess the true sensitivity, specific-
ity, and positive predictive value of the Q-IFN kit, prior
to Widespread use for clinical MTB testing.
The assessment of IFN-y production by TB-spe-
cific lymphocytes is also being studied by more sen-
sitive techniques such as flow cytometry and Elispot
assays. Tilley and Menon (2000) demonstrated the
correlation between intracellular staining of IFN-y
in TB-specific lymphocytes with the extracellular
secretion of IFN-y as measured by the Elispot assay.
The authors were also able to conclude that the IFN-
y response as measured by the intracellular staining
correlated directly with the Mantoux skin test. As the
data in TB patients using these techniques increase,
the understanding of the underlying TB immune
response will become clear, and some combination
of tests may become the new standard of care for the
diagnosis of latent TB infection.
References
Ahmed AR, Blose DA (1983) Delayed type hypersensitivity
skin testing: a review. Arch DermatoI119:934-945
American Thoracic Society and Centers for Disease Control
and Prevention (2000) Diagnostic standards and classifica-
tion of tuberculosis in adults and children. Am J Crit Care
Med 161:1376-1395
Bates JH, Stead WW (1993) The history of tuberculosis as a
global epidemic. Med Clin North Am 77:1205-1217
Betiete B et al (2002) Evaluation of a whole-blood interferon-y
release assay for the detection of Mycobacterium tubercu-
losis infection in 2 study populations. Clin Infect Dis 34:
1449-1456
Bouros D et al (1991) Palpation vs pen method for the mea-
surement of skin tuberculin reaction (Mantoux test). Chest
99:416
Bouros D et al (1992) The role of inexperience in measuring
tuberculin skin reaction (Mantoux test) by the pen or pal-
pation technique. Respir Med 86:219-223
238
Centers for Disease Control (1994) Core curriculum on tuber-
culosis, 3rd edn. Centers for Disease Control, Atlanta, p 9
Collins H, Kaufmann S (2001) The many faces of host
responses to tuberculosis. Immunology 103:1-9
Converse PJ et al (1997) Comparison of a tuberculin skin
test in high-risk adults of human immunodeficiency virus
infection. J Infect Dis 176:144-150
Daniel T (1980) The immunology tuberculosis. Clin Chest
Med 1:189-201
Daniel TM, Boom WH, Ellner JJ (2000) Immunology of tuber-
culosis. In: Reichman L, Hershfield ES (eds) Tuberculosis:
a comprehensive international approach, 2nd edn. Dekker,
New York, pp 187-214
Davis AL (2000) A historical perspective on tuberculosis and
its control. In: Reichman L, Hershfield ES (eds) Tubercu-
losis: a comprehensive international approach, 2nd edn.
Dekker, New York, pp 3-54
Desem N, Jones SL (1998) Development of a human gamma
interferon enzyme immunoassay and comparison with
tuberculin skin testing for detection of Mycobacterium
tuberculosis infection. Clin Diagn Lab Immunol 5:
531-536
Edwards LB et al (1969) An atlas of sensitivity to tuberculin,
PPD-B and histoplasmin in the United States. Am Rev
Respir Dis 99:1-132
Ellner JJ (1997) The immune response in human tuber-
culosis-implications for tuberculosis control. JID 176:
1351-1359
Gennaro ML (2000) Immunologic diagnosis of tuberculosis.
Clin Infect Dis 30:s243-s246
Gryzbowski S, Brown MT, Stothard D (1969) Infections with
atypical mycobacteria in British Columbia. CMAJ 100:
896-900
Hagan WA (1931) The no lesion case problem in the tubercu-
losis eradication campaign. Cornell Vet 21:163-171
Hanson ML, Comstock GW (1968) Efficacy of hydrocortisone
ointment in the treatment of local reactions to tuberculin.
Am Rev Respir Dis 97:472-473
Haslov K et al (1990) Comparison of the immunological activ-
ity of five defined antigens from mycobacterium tuberculo-
sis in seven inbred guinea pig strains. The 38 kDa antigen is
immunodominant. Scand J ImmunoI31:503-514
Hershey J, Engler RJ, Katial RK (2001) Tuberculin skin tests:
to measure or not to measure in two directions? JACI 107:
s254
Hirsch CS et al (1996) Cross-modulation by transforming
growth factor b in human tuberculosis: suppression of
antigen-driven blastogenesis and interferon g production.
Proc Natl Acad Sci USA 93:3193-3198
Hirsch CS et al (1999) Depressed T-cell interferon-y responses
in pulmonary tuberculosis: analysis of underlying mecha-
nisms and modulation with therapy. JID 180:2069-2073
Howard T, Solomon DA (1988) Reading the tuberculin skin
test: Who, when and how. Arch Intern Med 148:2457-2459
Iseman M (1999) Immunity and pathogenesis. In: A clinician's
guide to tuberculosis. Lippincott Williams and Wilkins,
Philadelphia, pp 63-96
Johnson PDR et al (1999) Tuberculin-purified protein deriva-
tive, MPT-64, and ESAT-6 stimulated gamma interferon
responses in medical students before and after mycobac-
terium bovis BCG vaccination and in patients with tuber-
culosis. Clin Diagn Lab Immunol 6:934-937
Jordon T et al (1987) Tuberculin reaction size measurement by
R. K. Katial
the pen method compared to traditional palpation. Chest
92:234-236
Katial RK et al (2001) Cell mediated immune response to
tuberculosis purified protein derivative: comparison of TB
skin testing to in-vitro g-interferon production in whole
blood culture. Clin Diagn Lab Immunol 8:339-345
Kaufmann SH (2001) How can immunology contribute to the
control of tuberculosis? Nature Rev Immunol1:20-30
Kimura M et al(1998) Comparison between a whole blood
interferon-gamma release assay and tuberculin skin test-
ing for the detection of tuberculosis infection among
patients at risk for tuberculosis exposure. J Infect Dis 179:
1297-1300
Kiple KF (ed) (1993) The Cambridge world history of human
disease. Cambridge University Press, Cambridge
Koch R (1890) Ober bacterilogische Forschung. Dtsch Med
Wochenschr 16:756
Koch R (1891) Fortsetzung der Mitteilungen tiber ein Heilmit-
tel gegen Tuberculose. Dtsch Med Wochenschr 891:17-101
Lein D, von Reyn F (1997) In vitro cellular and cytokine
responses to mycobacterial antigens: application to diag-
nosis of tuberculosis infection and assessment of response
to mycobacterial vaccines. Am J Med Sci 313:364-371
Longfield J et al (1984) Interobserver and method variability in
tuberculin testing. Pediatr Infect Dis 3:323-326
Lunn JA (1980) Reasons for variable response to tine test. Br
Med J 280:223
Mazurek GH et al (2001) Comparison of a whole-blood
interferon gamma assay with tuberculin skin testing for
detecting latent Mycobacterium tuberculosis infection.
JAMA 286:1740-1747
Murtagh K (1980) Unreliability of the Mantoux test using 1
TU PPD in excluding childhood tuberculosis in Papua New
Guinea. Arch Dis Child 55:795-799
Grme I, Cooper A (1999) Cytokine/chemokine cascades in
immunity to tuberculosis. Immunol Today 20:307-312
Palmer CE (1945) Nontuberculous pulmonary calcification and
sensitivity to histoplasmin. Public Health Rep 60,513-520
Palmer CE et al (1959) Experimental and epidemiologic basis
for the interpretation of tuberculin sensitivity. J Pediatr
55:413-428
Pearson L (1892) Tuberculin as a diagnostic agent. Med News
60:358-359
Pouchot J et al (1997) Reliability of tuberculin skin test mea-
surement. Ann Intern Med 126:210-214
Pottumarthy S et al (1999) Evaluation of the tuberculin gamma
interferon assay: potential to replace the Mantoux skin test.
J Clin Microbiol 37:3229-3232
Reichman LB, G'Day R (1977) The influence of a history of a
previous test on the prevalence ands size of reactions to
tuberculin. Am Rev Respir Dis 115:737-741
Research Committee of the British Thoracic Association
(1982) Reproducibility of the tine tuberculin test. Br J Dis
Chest 76:75-78
Rothel JS et al (1992) The gamma-interferon assay for diagno-
sis of bovine tuberculosis in cattle: conditions affecting the
production of gamma-interferon in whole blood culture.
Aust Vet J 69: 1-4
Seibert FB, Glenn JT (1941) Tuberculin purified protein
derivative: preparation and analysis of a large quantity for
standard. Am Rev Tuberc 44:9-25
Snider D (1982) The tuberculin skin test. Am Rev Respir Dis
125:s108-s118
Immunodiagnostics for Latent Tuberculosis Infection
Sokal JE (1975) Measurement of delayed skin test responses.
N Engl J Med 293:501-502
Streeton JA,Desem N, Jones SL (1998) Sensitivity and specific-
ity of a gamma interferon blood test for tuberculosis infec-
tion. Int J Tuberc Lung Dis 2:443-445
Tarlo SM et al (1977) Immediate hypersensitivity to tuberculin
in in-vivo and in-vitro studies. Chest 71:33-37
Tilley PAG, Menon IN (2000) Detection of Mycobacterium-
specific interferon-gamma-producing human T lympho-
cytes by flow cytometry. APMIS 108:57-66
Tsuyuguchi I (1996) Regulation of the human immune
response in tuberculosis. Infect Agents Dis 5:82-97
Villarino ML et al (1999) Comparable specificity of 2 com-
mercial tuberculin reagents in persons at low risk for
tuberculous infection. JAMA 281:169-171
239
Vordermeier HM et al (1992) M. tuberculosis-complex spe-
cific T-cell stimulation and DTH reactions induced with
a peptide from the 38-kDa protein. Scand J Immunol 35:
11-718
Wilcke JT et al (1966) Clinical evaluation of MPT-64 and MPT-
59, two proteins secreted from mycobacterium tuberculo-
sis, for skin test reagents. Tuber Lung Dis 77:250-256
Wood PR, Rothel JS (1994) In vitro immunodiagnostic assay
for bovine tuberculosis. Vet MicrobioI40:125-135
Wood PR et al (1991) Field comparison of the interferon-
gamma assay and the intradermal tuberculin test for the
diagnosis of bovine tuberculosis. Aust Vet J 68:286-290
Wright DN, Ledford DK, Lockey RF (1989) Systemic and local
allergic reactions to the tine test purified protein derivative.
JAMA 262:2999-3000
Organ-related Chapters
16 Childhood Tuberculosis
PETER R. DONALD
CONTENTS
16.1 Introduction 243
16.2 Epidemiology of Childhood Tuberculosis 243
16.3 Pathogenesis and Clinical Manifestations
of Childhood Tuberculosis 245
16.4 Central Nervous System Tuberculosis 248
16.5 Intracranial Tuberculomata 250
16.6 Childhood Tuberculosis and HIV/AIDS 251
16.7 Diagnosis of Tuberculosis in Childhood 251
16.8 Tuberculosis in Infancy 254
16.9 Tuberculosis During Adolescence 255
16.10 Treatment of Tuberculosis in Childhood 256
16.11 Drug-resistant Tuberculosis in Childhood 257
16.12 Prevention and Control of Tuberculosis
in Childhood 258
References 261
16.1
Introduction
Childhood tuberculosis refers to a wide spectrum of
manifestations of tuberculosis seen in children from
birth to adolescence. At one extreme, particularly in
the very young, are life-threatening disseminated
forms of tuberculosis, such as tuberculous menin-
gitis and miliary tuberculosis. At the other extreme,
enlargement of the mediastinal lymph nodes may
pass entirely unnoticed, and the development of a
positive tuberculin test will be the only sign that
tuberculosis infection has occurred. As the children
enter adolescence, "adult-type" pulmonary tubercu-
losis will be seen with increasing frequency, charac-
terized by the development of cavitation involving
mainly the apices of the lungs. Also at this age, large
pleural effusions with straw-coloured fluid become
more common as a manifestation of primary infec-
P. R. DONALD, MD, FRCP (Edin), FCP (SA), DCH (Glasgow),
DTM&H (London)
Department of Paediatrics and Child Health, Faculty of Health
Sciences, University of Stellenbosch, P.O. Box 19063, 7505,
Tygerberg, South Africa
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
tion. It is of considerable interest that the frequency
with which these various features are seen varies with
age and therefore probably reflects changes in the
immune response to tuberculosis infection. Perhaps
because of this wide spectrum of pathology, child-
hood tuberculosis will be looked upon by some as a
relatively benign manifestation of tuberculosis infec-
tion and regarded by others as a major child health
problem, particularly in developing countries.
This chapter will describe the epidemiology of
tuberculosis in childhood, the most important clini-
cal features of childhood tuberculosis, the interaction
of childhood tuberculosis and HIV/AIDS, and the
diagnosis and management of childhood tubercu-
losis.
16.2
Epidemiology of Childhood Tuberculosis
The incidence of childhood tuberculosis will be
determined by the frequency of exposure to a source
of infection, which will nearly always be an adult
with sputum microscopy smear-positive, cavitating
pulmonary tuberculosis. In developed communi-
ties, tuberculosis has become mainly a disease of
the elderly, with a male predominance, occurring
amongst the economically deprived, living on the
fringes of society. Small foci of disease may also be
found amongst disadvantaged groups such as immi-
grants. Children will thus not often be exposed to the
risk of infection and disease. In developing commu-
nities, tuberculosis will be a disease of young adults,
often with a female predominance. Children aged
15 years or less may constitute 40% or more of the
population, and it is therefore not surprising that the
incidence of childhood tuberculosis is far higher in a
developing community than a developed community
and that children make up a far greater portion of
the tuberculosis case load. Thus, not only are there
more young children in these communities, but they
will be exposed to tuberculosis infection at a much
244
younger age than would be the case in a developed
community. Because more serious forms of tuber-
culosis such as miliary tuberculosis and tuberculous
meningitis tend to occur in younger children, these
potentially deadly forms of tuberculosis will occur
more frequently in developing communities.
During the last decade tuberculosis has been
acknowledged as a worldwide threat to health, and
in 1993 the World Health Organization declared the
tuberculosis situation in the world a global emer-
gency. Although considerable effort has since been
expended in promoting the diagnosis of adult pul-
monary tuberculosis by sputum smear microscopy
and curing patients with directly observed short
course treatment (DOTS), the burden of childhood
tuberculosis frequently goes unrecognized. In devel-
oped countries, this is understandable as children
may constitute less than 5% of the tuberculosis case
load; in developing countries, however, this percent-
age may rise to 20% or even 40% in certain high-inci-
dence countries and communities.
Because of the difficulty of diagnosing tubercu-
losis in childhood, estimates of the global burden of
childhood tuberculosis are at best an educated guess.
In an often quoted estimate, the number of cases of
childhood tuberculosis occurring in 1990 was put at
1.3 million, leading to 450,000 deaths (Kochi 1991). In
other estimates, an arbitrary figure of 10% appears to
have been used to calculate the portion of the tuber-
culosis burden attributable to children. Thus, it was
predicted that during 2000 just over 1,000,000 cases
of childhood tuberculosis would occur in the world.
In the African region, however, under the influence
of HIV/AIDS, 447,000 children would make up 20%
of the tuberculosis burden, in contrast to the 1990
estimate of 90,000 cases of childhood tuberculosis
for the region (Dolin et al. 1994). That such a dra-
matic increase has in fact occurred is suggested by
data from several sources. In Malawi, for example, the
number of children 0-14 years of age with tuberculo-
sis managed at the Queen Elizabeth Central Hospital
increased from 64 in 1986 to 507 in 1995. In 1995 and
1996,64% of children being treated for tuberculosis
were mY-seropositive (Harries et al. 1997).
In contrast to the above estimates, which must be
open to considerable doubt, the risk of infection to
which children are exposed can be objectively quan-
tified by determining the annual risk of tuberculosis
infection (ART!). ART! is determined by tuberculin
testing a representative group of children at regular
intervals to detect the incidence of infection and is
considered one of the most objective ways to evaluate
the tuberculosis situation in any community (Bleiker
P. R. Donald
1991). While ART! in most developed countries will
be in the region of 0.01 %, it may rise as high as 2% or
3% in developing countries, implying that between a
third and a half of the population will be infected by
15 years of age. From this figure it then becomes pos-
sible to predict more accurately the number of cases
of childhood tuberculosis that are likely to arise in
a particular community. It is disturbing that sev-
eral recent studies from the developing world have
shown no fall in ART! over long periods. In the Chin-
gleput district of south India, for example, ART! has
remained at 2% since observations started in 1968
(Tuberculosis Research Centre 2001). In Tanzania,
ART! was 1.0% during 1988-1992 and 0.9% during
1993-1998. In this instance, the relatively constant
ART! is reason for a certain amount of optimism,
in that adult tuberculosis notifications have doubled
under the influence of mY/AIDS during this same
period (Tanzania Tuberculin Survey Collaboration
2001). This "success" is ascribed to the early adop-
tion and implementation of the DOTS strategy by
Tanzania.
Figure 16.1 illustrates the age- and sex-related
incidence of tuberculosis up to the age of 20 years
in a developing community. There are several note-
worthy features. Firstly, there is a very high incidence
of tuberculosis in a developing community in early
childhood, with a slight but consistent male predomi-
nance in the first year of life. This susceptibility of
the young infant to tuberculosis is even more strik-
ing when considered as a proportion of the infected
population, and very high rates of disease following
infection of 5,000 to 6,000/100,000 population have
been calculated for infected infants (Rich 1951; Bent-
ley et al. 1954). Not only are the very young more
susceptible to tuberculosis disease, but the disease is
c:
o
~::> 250
Co
8. Tuberculous meningitis
8o 200 Miliary tuberculosis
Lymphobronchial
o
o.... tuberculosis 'Adult type'
tuberculosis
g 150
C1l
"t:l
'0
.5
'" 100
.~
~
2
~ 501....&........- -......- _ - -.....-__.--_
Fig. 16.1. Age- and sex-related incidence of tuberculosis up to
the age of 20 years in a typical developing community
Childhood Tuberculosis
likely to be more serious, disseminated and to take
the form of tuberculous meningitis or miliary tuber-
culosis. With increasing age both the frequency and
severity of disease decline until a nadir is reached
between 5 and 10 years of age, the so-called "safe
school age" of the old German authors. As adoles-
cence is entered, the incidence of disease rises again,
"adult-type" pulmonary tuberculosis with apical
cavitation becomes the norm, and large pleural effu-
sions with a straw-coloured exudate are frequently
encountered.
16.3
Pathogenesis and Clinical Manifestations
of Childhood Tuberculosis
Following infection, the clinical manifestations of
tuberculosis will develop in a certain pattern, which
was described as the "time-table of tuberculosis"
by the great Swedish paediatrician, Arvid Wallgren
(1948). As a generalization it is usually stated that
disease will occur in approximately 10% of individu-
als following infection and that the greatest risk for
most manifestations of disease is in the 3-12 months
following upon infection, although a lifetime risk of
disease does remain. In some communities disease
may follow infection in as many as 15% or 20% of
individuals. The "timetable of tuberculosis" is sum-
marized in Fig. 16.2.
Within 3-8 weeks of infection, the majority of
infected individuals will become tuberculin-positive
Time after
primary infection Clinical Manifestation
2 - 3 months
Fever ofOnset
I
I
Tuberculin Test Positive
3 - 12 months
Primary puhnonary TB
TB Meningitis
Miliary TB
TB Pleural effusion
6 - 24 months
II Osteo-articular TB
> 5 years
II RenalTB
Fig. 16.2. The timetable of tuberculosis
245
and experience an unobtrusive ,fever of onset' (Wallgren
1928). By 1 month later, the features of primary tubercu-
losis will start to be seen, and by the end of one year after
infection, most cases of tuberculous meningitis, miliary
tuberculosis and pleural effusion will have occurred. In
contrast, although many cases of osteoarticular tubercu-
losis will develop within a year ofinfection, they will con-
tinue to occur for several years after infection. Interest-
ingly' although haematuria will often be detected shortly
after infection and M. tuberculosis can be cultured from
the urine, renal tuberculosis will seldom be seen sooner
than 5 years after infection (Munro 1944).
Following upon infection, two events playa major
role in determining the clinical manifestations of
childhood tuberculosis. Firstly, before the establish-
ment of cellular immunity to M. tuberculosis, lympho-
haematogenous dissemination of organisms occurs
and will distribute small numbers of organisms
throughout the body, but particularly to sites with a
good blood supply. In addition to reaching the bone
ends, meninges and kidneys, bacilli also again reach
the lungs and in the form of Simon foci may later give
rise to adult tuberculosis. Should exceptionally large
numbers of organisms be liberated into the blood
by erosion into a blood vessel within a lymph node
or following the establishment of tubercles within a
blood vessel, a tuberculous septicaemia ensues, and
small tubercles will develop throughout the body. In
the lungs, these give rise to small, uniformly sized
nodules the size of millet seeds on chest radiography,
and hence the name miliary tuberculosis (Fig. 16.3).
The second important factor in the pathogenesis
of childhood tuberculosis is the role of the medias-
ttl
oS
f
is
~
isc..
0
§'" I
()
0
.3.
~
~:
r;;'
Fig.16.3. Massive haematogenous dissemination of bacilli
leading to multiple small nodules less than 2 mm in diameter
evenly distributed throughout the lungs
246
tinal nodes which form part of the primary or Ghon
complex in the lungs. Following inhalation of a small
droplet 5-10 flm in diameter containing a very small
number of M. tuberculosis organisms, a tuberculous
focus becomes established in the lungs. From the pri-
mary or Ghon focus organisms spread via the lymph
ducts to the regional lymph nodes. Most often, only
the nodes, either hilar (Fig.16.4) or paratracheal
(Fig. 16.5} or both, are seen on a chest radiograph.
Occasionally, the primary focus is visible or more
commonly it may become visible some 6-9 months
later as calcification develops. In a majority of cases,
there is only a single primary focus, but in approxi-
Fig. 16.4. Enlargement of the hilar nodes in a child 2 years of
age with a Mantoux test giving an induration of 20 mm and the
gastric aspirate positive on culture for M. tuberculosis
Fig. 16.5. Enlargement of the right paratracheal nodes in a 5-
year-old child. Sputum positive on culture for M. tuberculosis
P. R. Donald
mately 20% of individuals two or more primary foci
may be seen, as illustrated in Fig. 16.6. Enlargement
of the hilar nodes with or without ulceration gives
rise to a spectrum of pathological changes including
segmental or lobar collapse or hyperinflation, col-
lapse consolidation, and segmental or lobar expans-
ile consolidation with or without cavitation. With
complete obstruction, collapse of a lobe or segment
will develop (Fig. 16.7}. With partial obstruction and
a ball-valve effect, hyperinflation of a lobe or segment
or a whole lung may develop (Fig. 16.8). Ulceration of
a node and discharge of its contents into the relevant
bronchus may lead to collapse consolidation or in
the absence of obstruction to a lobar or segmental
opacification or the appearance of an expansile
pneumonia. The histology of these lesions may vary
according to the immune response and the numbers
of organisms present in the lesion (Seal and Thomas
1956). At one extreme, a benign ,tuberculin' response
may be seen with a good prognosis and clearing
within 2-3 months. At the other extreme, caseation
with cavity formation may develop and lead to bron-
chiectasis and permanent lung damage (Fig. 16.9}.
Should partial obstruction also be present, a grossly
enlarged tension cyst may be formed. This course of
events is summarized in Fig. 16.10. In the past, these
usually well defined lesions were referred to as epitu-
berculosis, segmental tuberculosis, or endobronchial
tuberculosis. They are perhaps best grouped together
Fig. 16.6. Right-sided paratracheal enlargement with calcifica-
tion appearing in the nodes and in three right-sided primary
foci in a 3-year-old child. Gastric aspirate positive on culture
for M. tuberculosis
Childhood Tuberculosis
Fig. 16.7. Collapse of the right middle lobe in a 2-year-old
child who presented with persistent cough and wheeze. Gas-
tric aspirate positive on culture for M. tuberculosis. A Mantoux
test gave an induration of 18 mm
Fig. 16.8. Enlargement of the paratracheal nodes with hyper-
inflation of the right lung in a 2-year-old child with gastric
aspirate positive on culture for M. tuberculosis
as lymphobronchial tuberculosis, thus emphasizing
the role of both lymph node enlargement and bron-
chial disease in determining the varied features of
these lesions which, none the less, have a common
origin.
Any primary tuberculosis infection will be accom-
panied by a greater or lesser degree of lympho-hae-
247
b
Fig. 16.9. Consolidation of the left upper lobe with several
cavities enlarged by pressure inflation. A fine nodularity is also
present in the right lung, which may indicate bronchogenic
dissemination and a small left-sided pleural reaction. Gastric
aspirate positive on culture for M. tuberculosis
matogenous dissemination of organisms throughout
the body. These have a predilection for certain organ
systems where they may establish themselves and
give rise to disease if not controlled by the individu-
al's immune response. The most important site where
extrapulmonary disease may arise in childhood is the
central nervous system.
248
Benign
"Tuberculin
Response"
Fig.16.10. Lymphobronchial tuberculosis: the spectrum of
possible consequences of tuberculous involvement of the
mediastinal nodes during primary tuberculosis
16.4
Central Nervous System Tuberculosis
Tuberculous meningitis (TBM) is the most serious
complication of childhood tuberculosis and has its
highest incidence between the ages of 2 and 3 years.
At this age it is often associated with miliary tuber-
culosis, and if untreated a majority of children with
miliary tuberculosis will die of TBM. In both devel-
oped and developing countries the diagnosis of TBM
is frequently delayed, with disastrous consequences
for the patient as the prognosis is closely linked to the
stage of the disease at the time of diagnosis. A clas-
sification of the stage of TBM developed by the Brit-
ish Medical Research Council in 1948 is still useful.
Patients with early signs of meningeal irritation but
no other focal signs are at stage I TBM. Patients diag-
nosed at this point should recover fully. When focal
neurological signs appear or the patient is mentally
confused, the disease is considered to be at stage II.
Although the majority of patients presenting at stage
II will survive, they are likely to do so with a physi-
calor intellectual deficit. By stage III, the patient is
comatose, unable to localise pain, and may have a
complete hemiplegia or rarely a quadriplegia. A third
of these patients may die, and the great majority of
survivors will have severe physical and mental handi-
caps (Medical Research Council 1948).
Although TBM may occur at any time following
primary infection, it occurs most often approxi-
mately 3 months after primary infection as result of
the rupture of a parenchymal focus, or less often a
P. R. Donald
meningeal focus into the cerebrospinal fluid (CSF).
This release of antigens into the CSF elicits a severe
inflammatory response, which gives rise to a thick
gelatinous exudate. This exudate envelops the base
of the brain and the adjacent structures. This in turn
leads to cranial nerve palsies, vasculitis, and raised
intracranial pressure due to obstruction to the flow
of CSF. The vasculitis affects the larger blood vessels
such as the middle cerebral artery and the smaller
blood vessels of the basal ganglia and will often lead
to infarctions and irreversible brain damage.
TBM usually has an insidious onset lasting a
week or more marked by increasing lassitude, loss
of appetite and irritability. Occasionally, a more
acute onset may be seen resembling that of other
forms of bacterial meningitis. Many children will
experience a preceding loss of weight. Common pre-
senting complaints include vomiting, cough, fever,
apathy and headache; this last complaint being seen
mainly in children older than 3 years. Because of the
non-specific nature of these features, a high index of
suspicion is necessary when these complaints persist.
A tuberculin test, a chest radiograph or enquiry as
to household contact with an adult with ,smear-posi-
tive' pulmonary tuberculosis may lead to the early
diagnosis of TBM or to the prevention of TBM in
some instances by the early prescription of anti-
tuberculosis treatment. In any case where doubt is
present, a lumbar puncture should be done.
The CSF in TBM will typically be clear with a mean
cell count of approximately 100Xl06/L, predomi-
nantly lymphocytes, a mean protein of 1.5 giL, and a
mean glucose of 1.8 mmollL. It is very important to
note that this is a typical picture. In 5%-10% of cases,
a cell count of >500x106 /L with a predominance of
polymorphonuclear leucocytes may be found; CSF
protein levels may be normal or between 0.45 giL
and 0.8 giL in close to 20% of cases; the CSF glucose
may be >2.2 mmollL in as many as 30%-40% of cases
(Donald et al. 1991). There is thus ample room for
mistakes to be made. In any case of doubt, it is better
to initiate antituberculosis treatment and review the
diagnostic evidence later.
The diagnosis of TBM is confirmed by a culture of
M. tuberculosis from the CSF. This is, however, positive
in only a minority of cases. This is probably due to the
relatively small amounts of CSF that are submitted
for culture from children. Support for the diagnosis is
provided by culture of M. tuberculosis from another
source such as gastric aspirate or lymph node biopsy,
a chest radiograph suggestive of tuberculosis, a posi-
tive Mantoux test, or a history of close contact with
an adult with sputum "smear-positive" tuberculosis.
Childhood Tuberculosis 249

A chest radiograph may show evidence of respira- raised CSF white cell count or protein concentration
tory tuberculosis in up to 70% of children with in such cases would provide sound support for a deci-
TBM, while a Mantoux test should be positive in a sion to continue antituberculosis treatment.
majority of children. It must be emphasized, however, The management of TBM involves three aspects:
that a negative Mantoux test does not exclude TBM. the elimination of viable M. tuberculosis from the
Approximately 50% of children with TBM will have lesions, the control of raised intracranial pressure
a history of contact with an adult with pulmonary and the prevention of further brain damage related
tuberculosis, and in many cases this will be a family to the presence of the basal exudate, and the associ-
member, often the mother or a grandparent. ated hypersensitivity and vasculitis.
Cranial computed tomography or magnetic reso- There are almost no controlled trials evaluating
nance imaging, if available, may also assist and show different regimens for treating TBM. Consequently,
hydrocephalus associated with the basal enhance- a variety of recommendations can be found in the
ment and possibly cerebral infarctions (Fig. 16.11). literature and various official documents. Standard
Other diagnostic aids include CSF polymerase short course regimens given for 6 months and all
chain reaction, the detection of tuberculostearic acid containing rifampicin for the full 6 months have been
in the CSF, CSF adenosine deaminase concentrations, reported to be successful by several groups (Biddulph
CSF lactate or lactate dehydrogenase concentrations, 1990; Jacobs et al. 1992). Our own preference is for a
and the bromide partition test. None of these tests at 6-month regimen of isoniazid, rifampicin, pyrazin-
present has a totally satisfactory sensitivity or speci- amide and ethionamide, with isoniazid and rifampi-
ficity, and while they may be valuable in individual cin given in a dose of 20 mg/kg body weight. In our
patients, the final decision as to whether to treat a region there is an overall incidence of isoniazid resis-
patient remains a clinical one assisted mainly by the tance of 10%. Approximately half of these isolates are
clinical features and the CSF results. resistant at isoniazid concentrations of <5 fig/ml, a
In cases of doubt it can sometimes be valuable concentration easily exceeded even in fast acetyl-
to repeat the lumber puncture a week or two after ators by giving a higher dose of isoniazid (Donald
treatment has been started. It would be most unusual et al. 1992). Rifampicin is a critical sterilizing agent
for the CSF findings to normalize in such a relatively in tuberculosis regimens. It is, however, 80% protein
short space of time in the case of TBM. A persistently bound, and relatively low concentrations enter the
CSF when given in a dosage of 10 mg/kg (Humphries
1992). Ethionamide is notorious for causing gastric
irritation and nausea but obtains a satisfactory entry
into the CSF (Donald and Seifart 1992) and is much
better tolerated by children than adults. We have
used this regimen in over 200 patients to date and
have experienced relatively few problems with drug
toxicity; we documented relapse in only one patient
in whom the dosages of the drugs were reduced after
2 months of treatment.
If the intracranial pressure is monitored, it will be
found to be raised in nearly all children presenting
with stage II or III TBM. In those children who have
a communicating hydrocephalus and the absence
of other intracranial pathology has been confirmed
by computed tomography or magnetic resonance
imaging, raised intracranial pressure can be success-
fully managed by using furosemide (1 mg/kg/day)
and acetazolamide (100 mglkg/day). The presence of
communication between the ventricles and the spinal
Fig. 16.11. Cranial computed tomography with contrast in a canal can be demonstrated by injecting air at the time
2-year-old child with tuberculous meningitis showing acute of lumbar puncture and taking an X-ray of the skull,
hydrocephalus and periventricular oedema. Diffuse enhance-
ment in the territory of the right middle cerebral artery is
which should show air in the ventricles (Fig. 16.12). If
present indicating ischaemia. A mass effect also indicating air is seen at the base of the brain and not in the ventri-
ischaemic changes is seen in the right caudate nucleus cles, a non-communicating hydrocephalus is probably
250
Fig. 16.12. Skull radiograph of a child with tuberculous men-
ingitis. Air was injected at the time of lumbar puncture and
is seen at the base of the brain and in the ventricular system,
indicating that a communicating hydrocephalus is present
present, and a ventriculo-peritoneal shunt is urgently
required to relieve the raised intracranial pressure.
As tuberculin hypersensitivity is thought to play
an important role in the formation of the basal exu-
date in TBM and its associated vasculitis, there is a
long tradition of using corticosteroids to manage this
problem. At first, it was reported that corticosteroids
decreased mortality but did not affect the outcome
amongst survivors (Freiman and Geefhuysen 1970;
Shao et al. 1980). In a later study corticosteroids
improved not only morbidity but also outcome
(Girgis et al.1991). In a more recent, randomised trial,
prednisone was also shown to improve morbidity and
outcome. Interestingly, it did not appear to have any
effect upon raised intracranial pressure (Schoeman
et al. 1996). Tuberculomata, when present, resolved
more quickly, and fewer new tuberculomata devel-
oped in those patients receiving corticosteroids.
16.5
Intracranial Tuberculomata
Tuberculomata are common intracranial tumours in
those communities where tuberculosis is common.
They may be single or multiple in 15%-20% of cases
(Jamieson 1995). They may occur in isolation or in
association with tuberculous meningitis (Fig. 16.13).
P. R. Donald
In a review of 202 cases of TBM seen at our hospi-
tal, associated granulomas were present in 34 (17%)
children (Ravenscroft et al. 2001). Two necrotic
processes occur in these lesions. In gummatous
necrosis the inflammatory granulation tissue itself
becomes necrotic, and the fibrovascular stroma can
be identified on silver reticulin staining. The process
can repeat itself intermittently, giving rise to a lamel-
lated appearance. In the second process there is no
fibrovascular component, and the consistency of the
contents can vary from caseous to purulent liquid.
Gummatous necrosis is isodense or hyperdense on
cranial CT, while caseous necrosis is hypodense or
isodense, reflecting the more fluid character of the
contents (Rutherfoord and Hewlett 1994).
Intracranial tuberculomata may be asymptomatic
and discovered by chance or present with convulsions
or signs or symptoms of raised intracranial pressure.
Unless the lesion is situated at some critical point in
the central nervous system, medical treatment as for
TBM and augmented by corticosteroids will usually
be sufficient. As is the case with tuberculous cervical
lymphadenopathy, it is not unknown for these lesions
to enlarge paradoxically despite successful antitu-
berculosis treatment. Should the lesion give rise to
symptoms, either due to its location or due to raised
intracranial pressure, a course of corticosteroids may
expedite the shrinkage of the lesion.
Fig. 16.13. Computed cranial tomography with contrast in a
child with stage II tuberculous meningitis. There is meningeal
enhancement in the left Sylvian fissure and multiple, homoge-
neously enhancing granulomata
Childhood Tuberculosis
16.6
Childhood Tuberculosis and HIVI AIDS
The last decade has witnessed the continued spread
of HIV/AIDS throughout the world and its pernicious
effect upon the incidence of tuberculosis. Nowhere has
the deleterious interaction of these two formidable
diseases been more dramatic than in the countries of
sub-Saharan Africa where the tide of HIV/AIDS and
tuberculosis is threatening to overwhelm the already
struggling health services. Not surprisingly, this dual
epidemic is also affecting childhood tuberculosis,
and very high rates of HIV seropositivity have been
recorded amongst children with tuberculosis (Chintu
et al. 1993; Harries et al. 1997). Because of the difficulty
ofconfirming the diagnosis of tuberculosis in children,
it is uncertain how many of these children really do
have tuberculosis. In addition to tuberculosis, these
HIV-infected children are also susceptible to a variety
of other bacterial pathogens including S. pneumonia,
H. influenza and S. aureus as well as pathogens such as
Pneumocystis carinii, respiratory syncytial virus and
cytomegalovirus (Graham et al. 2001). Lymphocytic
interstitial pneumonitis frequently presents with a
reticulonodular appearance and hilar adenopathy on
chest radiographs and can easily be confused with
miliary tuberculosis. In a study from Johannesburg,
M. tuberculosis was isolated from 8.5% of 423 HIV-
infected children and 7.6% of 434 uninfected children
(Mahdi et al. 2000). In Cape Town M. tuberculosis was
similarly isolated from 10% of HIV-infected children
and from 14% ofa group of non-HIV-infected children
(Zar et al. 2000). Amongst children with lung disease
persisting more than a month, M. tuberculosis was
isolated from 29% of both HIV-infected and non-HIV-
infected children (Jeena et al. 1998). A post-mortem
study of 32 HIV-infected children from Botswana
confirmed the presence of tuberculosis in 4 (l2.5%)
(Ansari et al. 1999). These findings all suggest that
tuberculosis is definitely a problem in HIV-infected
children and that a high index of suspicion is always
required. In an endemic area, however, it is but one of
a number of other infections to which these children
are subject.
HIV-infected children may present with all of the
features of childhood tuberculosis seen in immuno-
competent children. Several studies from the devel-
oped world have found an increased incidence of
extrapulmonary tuberculosis, as is the case in HIV-
infected adults (Chan et al. 1996; Khouri et al. 1987;
Thomas et al. 2000). This has, however, not been the
case in studies from Africa (Chintu and Zumla 1995).
Despite this, it is necessary to consider tuberculosis in
251
the differential diagnosis of any lesion not respond-
ing to conventional treatment in these children.
16.7
Diagnosis of Tuberculosis in Childhood
The diagnosis of tuberculosis in childhood requires a
high degree of suspicion but at the same time a bal-
anced clinical perspective. Suspicion is required, as
the features of childhood tuberculosis may be subtle
and non-specific. Failure to gain weight properly may
be the only clinical sign that a child is in the active
phase of primary tuberculosis infection. Balance is
required, as the urgency of making a diagnosis of
tuberculosis must be weighed against factors such
as the child's age and the severity of the child's ill-
ness. In a child 8 years of age with a subtle picture of
hilar enlargement on chest radiography and a vague
history of indisposition, one may safely choose to
await further developments and the results of inves-
tigations before making a final diagnosis. In a young
infant with an enlarged liver and spleen and an ill-
defined nodular opacification on chest radiography,
it is a matter of urgency to commence antituberculo-
sis treatment as soon as possible and to dispute later
as to the accuracy of the diagnosis.
In an adult it is unwise to accept a diagnosis of
tuberculosis without microbiological confirma-
tion either by way of culture of M. tuberculosis or
by visualization of acid-fast bacilli on microscopy
of a clinical specimen. Childhood tuberculosis is
paucibacillary, and cavitation of the lung tissue is
relatively unusual. Consequently, as few organisms
are excreted, it would be equally unwise to await
. microbiological confirmation of the diagnosis of
childhood tuberculosis before commencing treat-
ment. Even with intensive investigation of children
with probable tuberculosis in specialist institutions,
microbiological confirmation of the diagnosis will
seldom be obtained in more than 50% of cases
(Schaaf et al. 1995; Smith et al. 1996; Lobato et al.
1998). In the community, where most cases of child-
hood tuberculosis must be diagnosed and managed,
fewer than 10% of cases are likely to be confirmed by
culture or microscopy (Jacobs et al. 1987).
In view of the lack of sensitivity of gastric aspirate
culture and microscopy to detect M. tuberculosis in
children, attempts have been made to use PCR in
the detection of mycobacterial DNA. While satisfac-
tory sensitivity and specificity have been claimed by
some (Pierre et al. 1993; Delacourt et al. 1995), other
252 P. R. Donald

investigators consider the technique to be unreliable
at present (Noordhoek et al. 1994; Smith et al. 1996).
Unfortunately, whatever one's views of this technol-
ogy, it is unlikely to be available in precisely those
communities where most cases of childhood tuber-
culosis occur.
The diagnosis of childhood tuberculosis conse-
quently remains a clinical process drawing on the
patient's history, clinical examination, chest radiog-
raphy, tuberculin testing and lastly culture of gastric
aspirate or sputum in older children. Making use
of this process, the World Health Organization has
recommended a hierarchical approach leading to
children being classified as suffering from suspect,
probable or confirmed tuberculosis. This approach is
summarized in Table 16.1 (World Health Organiza-
tion 1989).
A history of close household contact with an adult
with microscopy smear-positive pulmonary tuber-
culosis should immediately raise the suspicion that
infection may have occurred. From one-third to two-
thirds of children so exposed may be infected (Brai-
ley 1940; Shaw and Wynn-Williams 1963; van Geuns
et al. 1975). It is also recommended that a cough and
wheeze not responding to appropriate treatment,
failure to recover from a preceding illness, or loss
of weight or overt malnutrition should also raise the
suspicion of tuberculosis in a child. The sensitivity
and specificity of these criteria have, however, seldom
been determined.
Amongst a group of 627 children <5 years of
age attending an outpatient department in an area
with a particularly high incidence of tuberculosis,
206 (33%) were found to have one or more of the
above criteria for suspect tuberculosis (Houwert et
al. 1998). The children were evaluated by gastric aspi-
rate, Mantoux testing and chest radiography. After
reviewing the diagnostic criteria 2 months later, 23
children (11 %) were considered to have probable
tuberculosis, and a further 10 (5%) had confirmed
tuberculosis. The most rewarding criterion was close
household tuberculosis contact, and 23 (35%) of such
children had either confirmed or probable tubercu-
losis compared with 13 (22%) of 46 children with
persistent cough and 24 (15%) of 157 children with
loss of weight. Of 11 children who had all 3 criteria,
7 (63%) had probable (4) or confirmed (3) tubercu-
losis. Since this study the spread of HIV/AIDS has
complicated the interpretation of these diagnostic
criteria considerably. Many children with HIV/AIDS
will have lost weight, many will be exposed to house-
hold tuberculosis contact and will be suffering from
recurrent or persistent respiratory infections. As the
resources of many developing countries to carry out
further diagnostic investigations in children are
limited, it seems inevitable that many HIV-infected
children who do not have tuberculosis will be placed
on antituberculosis treatment.
In another study children who presented with
probable tuberculosis were investigated. There was
no difference between the proportion of children
later thought to have probable or confirmed tuber-
culosis and those found not to have tuberculosis in
the presence of clinical features such as weight loss,
cough or wheezing lasting more than 2 weeks, hepa-
tosplenomegaly or peripheral lymphadenopathy.
Furthermore, in this high tuberculosis incidence
population, although 43% and 56%, respectively, of
the children with confirmed and probable tuberculo-
sis had an adult household contact with tuberculosis,
so did 21 % of those later thought not to have tuber-
culosis. To add further diagnostic confusion, 10% of

Table 16.1. World Health Organization (1989) provisional guidelines for the diagnosis of pulmonary tuberculosis in children
A. Suspect tuberculosis
1. An ill child with a history of contact with
a confirmed case of pulmonary tuberculosis
2. Any child:
2.1 Not regaining normal health after
measles or whooping cough
2.2 With loss of weight, cough and
wheeze not responding to antibiotic
therapy for respiratory disease
2.3 With painless swelling in a
group of superficial nodes
B. Probable tuberculosis C. Confirmed tuberculosis
A suspect case and any of the 1. Detection by microscopy or
following: culture of tubercle bacilli
from secretions or tissues or
1. Positive (~10 mm) induration 2. The identification of the tubercle
on tuberculin testing bacilli as Mycobacterium tuberculosis
by culture characteristics
2. Suggestive appearances
on chest radiograph
3. Suggestive histologic appearance
of biopsy material
4. Favourable response to
specific antituberculous therapy
Childhood Tuberculosis 253

the children with tuberculosis confirmed by culture Tuberculin testing is one of the most important
had a chest radiograph which was considered normal investigations supporting the diagnosis of tubercu-
on blind reading by an independent panel (Schaaf et losis in a child. It must, however, be emphasized that
al. 1995). This last finding is not unexpected, as it a positive tuberculin test, however defined, means
is known that when such children are evaluated by only that a child has been infected by M. tuberculosis.
chest computed tomography, hilar nodal enlarge- The younger the child, the more significant a positive
ment will frequently be visible (Delacourt et al.1993). tuberculin test is as infection is likely to have occurred
This finding does emphasize the uncertain division recently, and infection in a young child is more likely
between what clinicians regard as infection only and to proceed to disease than in an older child.
disease in childhood tuberculosis. The tuberculin test must be interpreted in the light
In any area where childhood tuberculosis is of the child's clinical condition, the epidemiological
common, children will also suffer from a multitude of background and whether or not BCG vaccination has
other infectious, allergic and congenital conditions, been administered. There is also a small number of
and will often be malnourished to a greater or lesser individuals who appear unable to respond to tubercu-
degree. It is not surprising that many of these chil- lin, perhaps on a genetic basis (van Eden et al. 1983).
dren will be inappropriately treated for tuberculosis. Recent childhood viral disease, live virus vaccines
Table 16.2 lists some of these conditions which in our such as measles and rubella vaccines, severe malnu-
experience have caused diagnostic problems. When trition, overwhelming tuberculosis disease, cortico-
tuberculosis treatment of a child is commenced on steroid therapy and HIV/AIDS may all be responsible
uncertain grounds, it is essential for the clinician to for a false-negative tuberculin test. Nearly two-thirds
keep the case under review and to consider other of HIV-infected children with tuberculosis may have
diagnostic options. a negative tuberculin test (Chan et al. 1996; Mukadi
Culture and microscopy of gastric aspirate are et al. 1997). Nonetheless, even in communities with a
the means by which tuberculosis will most often be particularly high incidence of tuberculosis and HIV
confirmed in childhood. Although a relatively simple infection, the tuberculin test remains a valuable diag-
procedure, attention to detail will improve the results nostic aid. The majority of children with tuberculo-
(Pomputius et al. 1997). Although hospitalization is sis in these communities will not be HIV-infected,
usually considered necessary for gastric aspirate, this and even amongst those who are HIV-infected, the
is not necessarily so (Lobato et al. 1998). Alternative suppression of tuberculin hypersensitivity will be
means of collecting specimens for culture from chil- related to the stage of AIDS in each child (Schaaf et
dren are nasopharyngeal aspirate, which could also al. 1998).
be used in community work (Franchi et al.1998), and In individuals who have received BCG vaccine, a
sputum induction (Zar et al. 2000), both of which will transverse induration of ~15 mm should be taken
give a comparable yield of positive cultures. When- to indicate M. tuberculosis infection, where BCG has
ever sufficient resources are available, every attempt not been given ~10 mm, and in those who are HIV-
should be made to confirm the diagnosis of tuber- infected ~5 mm (Harries and Maher 1996).
culosis in a child. A successful culture may reveal Chest radiography may provide valuable support-
unexpected drug resistance (Steiner et al. 1985), ing evidence of the presence of unsuspected tubercu-
while RFLP strain typing may explain unexpected losis in a child and is often the means by which the
transmission patterns (Barnes et al. 1997). extent of disease will be evaluated. It is therefore nec-

Table 16.2. Differential diagnosis of childhood tuberculosis

Segmental or bronchopneumonic opacification with or without cavitation on chest radiograph Hilar or paratracheal adenopathy
Asthma with a mucous plug Cystic fibrosis Lymphoma
Foreign body Hypersensitivity syndromes Teratoma
Bronchial compression due to tumours, Loeffler syndrome Neurofibroma
congenital abnormalities
Tropical eosinophilia Other tumours and malignancies
Lobar emphysema Necrotizing bacterial pneumonias Fungal infections
Bronchiectasis Lung abscess Sarcoidosis
Recurrent aspiration Diaphragmatic hernia Bronchogenic or dermoid cyst
Echinococcus cysts
 254 P. R. Donald

a b

Fig. 16.14a,b. Chest radiographs of a 3-year-old child presenting with recurrent cough and wheeze. a Chest radiograph taken
in poor inspiration showing possible hilar enlargement. b A chest radiograph of the same child taken 30 min later showing the
shadow to be the descending branch of the right pulmonary artery

essary that the clinician be aware of the importance of
a good quality chest radiograph. Common technical
faults in children's chest radiographs include inad-
equate inspiration, excessive lordosis, rotation of the
chest, and faulty exposure or processing, leading to a
film that is overexposed or underexposed. One of the
most common problems is poor inspiration, leading
to excessive prominence of the descending branch of
the right pulmonary artery (Grzybowski 1954). To
the inexperienced this can easily assume the appear-
ance ofhilar adenopathy (Fig. 16.14). In babies up to
the age of 18 months, the thymus will frequently be
visible and give rise to confusion, particularly if the
shadow is atypical or the radiograph rotated or taken
in poor inspiration (Fig. 16.15).
of the neonate (Pillay et al. 2001; Ahmed et al. 1999;
Adhikari et al. 1997).
Infection of the very young infant or neonate may
be the result of transplacental infection in utero or
infection may take place during or shortly after birth.
If infection is antenatal and transplacental, the pri-
mary focus of infection will be in the liver, and there
will be enlargement of the regional nodes in the porta
hepatis. These infants are often symptomatic, with
loss of weight, tachypnoea and wheezing or bronchial
breathing audible on auscultation (Schaaf et al.I993).

16.8
Tuberculosis in Infancy

If untreated, tuberculosis has its highest mortality

and morbidity during infancy, and the consequences
of a delay in diagnosis and treatment are potentially
more serious at this age than in older children. In
the experience of Wallgren (1938), 36% of infants
less than 1 year of age at the time of diagnosis died,
while Edith Lincoln (1950) reported that 55% of
infants diagnosed when less than 6 months of age
died. With the spread of HIVI AIDS, tuberculosis
has become a common cause of obstetric mortality
and morbidity, and this has inevitably increased the Fig. 16.15. A typical thymus shadow projecting into the upper
risk of antenatal or postnatal tuberculosis infection right lung field in a 15-month-old child
Childhood Tuberculosis
Hepato-splenomegaly may be present, and tubercu-
losis may be a cause of prolonged neonatal jaundice
(Schaaf and NelI992). Chest radiography will often
show a miliary pattern. Airway narrowing is fre-
quently present and will sometimes be more clearly
seen on a high kilo-volt radiograph. Cavitation may
be present as early as 5 weeks of age (Cunningham
1982). Gastric aspirate or tracheal aspirate will give
a positive culture in a higher percentage of cases
than in older children, and as many as 70%-80% of
infants may have a positive culture for M. tuberculosis
(Schaaf et al.1993; Vallejo et al.1994).
Unfortunately, the tuberculin test will frequently be
negative during the acute phase of the illness but should
not be ignored. Extrathoracic lymph node enlargement
may be palpated as early as 4 weeks of age, and hilar
node enlargement may be seen on chest radiograph
even earlier. Chest radiography and sputum culture of
the mother may be rewarding procedures (Schaaf et al.
1991), and the possibility of maternal urogenital tuber-
culosis should be considered. Other family members
and even hospital personnel should not be forgotten as
the possible source of infection.
Infants born to a mother with tuberculosis are at
considerable risk of infection and disease. Even if the
mother is sputum microscopy ,smear-negative' but
culture positive, a significant risk of infection still
exists because of the closeness of the contact between
the mother and her newborn infant. A number of
studies have shown convincingly that isoniazid will
protect such exposed infants against the risk of infec-
tion and disease (Dormer et al. 1959; Kendig 1960;
Light et al. 1974). If the diagnosis of maternal tuber-
culosis is made shortly before or shortly after birth,
an alternative approach would be to give the infant
isoniazid and rifampicin for as long as the mother is
herself being treated.
16.9
Tuberculosis During Adolescence
Following the onset of puberty there is a rapid rise
in the incidence of tuberculosis, although presumably
the risk of tuberculosis infection remains similar.
The nature of tuberculous disease also undergoes a
marked change and now takes the form of,adult-type'
disease. This is characterized by the development of
apical lesions and the propensity to develop cavitation,
which not only contributes to lung destruction but also
leads to the expectoration of microscopy smear-posi-
tive sputum, so completing the cycle of infection.
255
Tuberculin testing of large groups of adolescents
has found similar rates of tuberculosis infection in
both genders or a slight male predominance in some
cases (Hall 1957). Nonetheless, the risk of develop-
ing tuberculosis during adolescence has been found
to be 2 to 6 times higher in girls (Smith 1967). The
development of tuberculosis is also often associated
with menarche (Comstock et al. 1974). It is therefore
attractive to ascribe this susceptibility to ,adult-type'
tuberculosis to the hormonal and metabolic per-
turbations of puberty, and Arvid Wallgren (1938)
commented, "It is quite likely that the liability to pul-
monary tuberculosis (in adolescence) is intimately
linked up with endocrine activity". It is now known
that the balance between the Thl and Th2 families of
cytokines may be regulated by the adrenal steroids.
Glucocorticoids may promote a Th2-type response,
which would be disadvantageous to the course of
tuberculosis, whereas dehydroepiandosterone will
tend to promote a Thl-type response (Rook et al.
1993). This is also the age at which large straw-
coloured pleural effusions are more often seen fol-
lowing primary infection (Fig. 16.16). It is of inter-
est that despite the higher incidence of adult-type
tuberculosis in girls at this age, these large effusions
are much more frequent in boys (Bentley et al. 1954;
Enarson et al.1982; Weber et al. 2000).
If infection occurs after 7 years of age, the chance
of developing disease during puberty has been put at
8% in American studies (Lincoln et al. 1960). When
tuberculin conversion occurs during adolescence, it
Fig. 16.16. A large, right-sided, pleural effusion in a 12-year-
old adolescent who presented with fever and right-sided chest
pain. The Mantoux test gave an induration of 19 mm, and the
sputum was positive on culture for M. tuberculosis, but nega-
tive for acid-fast bacilli on microscopy
256
is far more likely to be followed by disease. As many
as 15% of such recently infected adolescents may
develop the disease, and disease is most likely to
occur within 1-3 years of infection (Smith 1967).
In view of the various hormonal and social adap-
tations occurring during adolescence, it is not sur-
prising that non-compliance has been identified as
a problem in these children (Nemir 1986; Weber et
al. 2000). Careful attention to potential problems and
community networking is necessary to assist these
patients to successfully complete therapy (Wilcox
and Laufer 1994). The combination of HIV infection
and tuberculosis in adolescence makes compliance
even more difficult to achieve (Hoffman et al. 1996).
16.10
Treatment of Tuberculosis in Childhood
The principles of the treatment of childhood tuber-
culosis are the same as those that apply to adults. A
2-month intensive bactericidal phase of isoniazid,
rifampicin, pyrazinamide and ethambutol or strep-
tomycin is followed by a 4-month sterilizing con-
tinuation phase with isoniazid and rifampicin.. The
aim of the intensive phase is to reduce the number
of rapidly metabolizing active organisms as quickly
as possible. This leads to the symptomatic improve-
ment, a reduction in the risk of infection to others,
while the use of a multidrug regimen will prevent
the development of drug resistance. The most effec-
tive bactericidal drugs are isoniazid, which kills 90%
of active organisms within 48 h (Donald et al. 1997),
ofloxacin and rifampicin (Sirgel et al. 1993).
Most cases of tuberculosis in childhood are pauci-
bacillary and "smear-negative", as children do not
often develop cavitation. Consequently, the risk of
developing drug resistance is probably not as great as
in adults, and the intensive phase can safely be under-
taken with isoniazid, rifampicin and pyrazinamide,
unless there is a history of contact with a case of
drug-resistant tuberculosis. In that case, the regimen
should be appropriately adjusted to take into account
the resistance pattern of the contact's isolates. Where
cavitation is present, it is probably advisable to
assume the presence of a reasonably large bacterial
population even if the smear is negative, and to use
four drugs in the intensive phase as in adults. The 4-
month continuation phase is also undertaken with
isoniazid and rifampicin.
The aim of the 4-month continuation phase is the
elimination of all remaining residual forms of M.
P. R. Donald
tuberculosis to prevent relapse occurring. The most
valuable sterilizing agents are rifampicin and pyra-
zinamide and to a lesser extent isoniazid (Mitchison
2000). The loss of rifampicin with the development
of multidrug resistance necessitates the use of much
longer, complicated and potentially toxic regimens.
In the case of serious forms of extrapulmonary
tuberculosis or disseminated tuberculosis, official
documents often recommend the same treatment
regimens as for new ,smear-positive' adult pulmo-
nary tuberculosis. More conservative clinicians might
prefer to extend the continuation phase to 7 months
of isoniazid and rifampicin (Harries and Maher
1996). Our own practice for children with miliary
tuberculosis, tuberculous meningitis or other cases
where sequestered lesions might be present is to use
a combination of isoniazid, rifampicin, pyrazinamide
and ethionamide, all given for 6 months (Donald et
al. 1998). Isoniazid has excellent pharmacokinetics
and a very large therapeutic margin between the
highest serum concentration that can be achieved
without undue toxicity and the lowest concentration
at which a bactericidal effect can still be measured
(Donald et al. 1997). Pyrazinamide also has good
pharmacokinetics and enters the cerebrospinal fluid
with ease (Donald and Seifart 1988), as does ethion-
amide (Donald and Seifart 1989). Rifampicin, how-
ever, is highly protein bound and at the usual dose
of 12 mg/kg body weight is probably operating at the
lower border of its efficacy. Any further reduction
in dose will increase relapse rates (Long et al. 1979)
and decrease bactericidal activity (Sirgel et al. 1993).
Rifampicin has also been shown to have poor pen-
etration into chronic tuberculous empyema (Elliot et
al. 1995) and cerebrospinal fluid, as is the case with
ethambutol (Ellard et al. 1993). As ethionamide is
better tolerated by children than adults, we advocate
its use in preference to ethambutol in those situa-
tions where isoniazid resistance is a potential threat
or where other drugs, such as rifampicin, may be at
a pharmacokinetic disadvantage. The dosages of the
more commonly used drugs for the management of
tuberculosis in children are given in Table 16.3. In
Table 16.4 the dosages of the alternative drugs are
given. The use of these agents may be necessitated
by the detection or suspicion of drug resistance or
by the development of drug toxicity. These agents are
as a generalization less efficient in the treatment of
tuberculosis and tend to be associated with greater
toxicity. Before embarking on their use, it is wise to
consult those experienced in their usage.
There is good evidence that adult smear-negative
pulmonary tuberculosis can be treated for 4 months
Childhood Tuberculosis 257

Table 16.3. Treatment of tuberculosis in children. Dosage and action of antituberculosis drugs

Action Daily dosage Twice weekly intermittent

Drug Bactericidal Sterilizing mglkg (maximum mg) mglkg (maximum mg)
Isoniazid +++ + 10-20 (300) 20-40 (900)
Rifampicin ++ +++ 10-20 (600) 10-20 (600)
Pyrazinamide - +++ 25-35 (2000) 50-70 (4000)
Ethambutol ++ 15-20 (1250) 50 (1250)
Ofioxacin a ++ ?a 15 (800) ?
Streptomycin + 20 (1000) 25-30 (1500)
aThere is as yet relatively little experience with the use of ofioxacin. It has been shown to have a
high early bactericidal activity and is a valuable drug to have in reserve when hepatotoxicity or
drug resistance precludes the use of other established drugs. There is, as yet, no information as to
its possible sterilizing activity

Table 16.4. Treatment of tuberculosis in children. Dosage of
,second-line' drugs
Drug Daily dosage Maximum
mg/kg dosage
Ethionamide 15-20 1g
Kanamycin 15-20 750 mg
Amikacin 15-20 750 mg
Capreomycin 15-20 750 mg
Cycloserine 15 1g
Para-aminosalicylic acid 300 10 g
Thiacetazonea 4 150 mg
aCommence thiacetazone with 0.5 mglkg for the 1st week,
1 mglkg the 2nd week and 4 mglkg from the 3rd week
with very low relapse rates (Hong Kong Chest Ser-
vice et al. 1989; Dutt et al. 1989). As most childhood
tuberculosis is of limited extent and smear-negative,
it seems probable that the same 4-month regimens
might also apply to children. However, as there are, as
yet, no clinical trials to support this approach, most
official documents still recommend 6 months of
treatment for all forms of childhood tuberculosis.
In contrast to the prospect of, perhaps, shortening
the length of treatment for children with tuberculo-
sis, there is increasing evidence that in the presence
of HIV infection, treatment should be prolonged to
9 months. Of 14 HIV-infected children with culture-
proven tuberculosis, 4 were still culture-positive
4-14 months after commencing treatment (Schaaf et
al. 1998). Recently, it was reported that HIV-infected
adults with pulmonary tuberculosis treated for less
than 9 months had an increased incidence of relapse
within 30 days of stopping treatment as compared
with those treated for longer than 9 months (Driver
et al. 2001). The reasons for this finding are uncertain,
but the poor absorption of antituberculosis agents in
adults with HIVI AIDS has been documented (Pelo-
quin et al. 1996). Others have not found any associa-
tion between tuberculosis recurrence and antituber-
culosis drug concentrations in HIV-infected patients
(Narita et al. 2001). No similar pharmacokinetic data
exist for children.
The management of childhood tuberculosis is not
a priority in most developing communities, where
hard-pressed health service staff have difficulty in
ensuring the compliance of smear-positive adult pul-
monary tuberculosis patients. Under these circum-
stances treatment is often given by family members,
and there must be some doubt about the extent of
compliance that is achieved. Intermittent treatment
of children, given two or three times weekly, is an
acceptable alternative to daily treatment and can lead
to equivalent rates of adherence (te Water Naude et
al. 2000). This would also make supervision of treat-
ment easier for the health services.
A further problem encountered in the manage-
ment of childhood tuberculosis is the lack of for-
mulations specifically designed for children. Several
commercial fixed dose combination preparations are
now available, but in their absence ad hoc prepara-
tions must often be constituted by clinic personnel.
It is very important that such ad hoc preparations
should not be made up before administration and
that they should not be mixed with vitamin solutions
as this can lead to a rapid decline in the drug concen-
trations (Seifart et al. 1991).
16.11
Drug-resistant Tuberculosis in Childhood
Drug-resistant tuberculosis in children is important
from two points of view. Firstly, it is important for
the individual patient that it should be diagnosed
expeditiously and managed appropriately. This is
particularly important in HIV-infected individuals,
258
who may deteriorate rapidly and die if appropriate
treatment is not started as soon as possible (Small et
al. 1993). Secondly, it is important as drug resistance
in children will nearly always be primary resistance,
and the incidence and patterns of resistance in chil-
dren will probably reflect those circulating amongst
adults in a community (Rieder 1993).
Despite early evidence that INH-resistant organ-
isms were less pathogenic for laboratory animals,
there can be no doubt now that INH-resistant organ-
isms can be transmitted and can cause disease (Snider
et al. 1985). In an evaluation of 128 children who were
in close household contact with an adult with multi-
drug-resistant tuberculosis, 66 (52%) were found to
be infected and a further 15 (12%) to have tubercu-
losis disease. RFLP strain typing of M. tuberculosis
cultured from these children found identical drug
susceptibility patterns and strains amongst the chil-
dren and their contacts in a majority of cases (Schaaf
et al. 2000). In nearly 80% of cases, the index case was
a family member. As only a minority of diseased chil-
dren are likely to produce a culture of M. tuberculosis
for drug susceptibility testing, it is important that an
accurate history always be obtained of possible drug-
resistant contacts. This will enable the construction
of the most appropriate treatment regimen as soon
as possible. This is particularly important in those
children with HIV-infection and life-threatening
forms of disease such as tuberculous meningitis or
miliary tuberculosis.
In the Western Cape Province of South Africa, the
rates of initial isoniazid resistance and multidrug
resistance among adults were 3.9% and 1.1 %, respec-
tively, between 1992 and 1993 (Weyer et al. 1995).
From 1995 to 1998, 306 cultures of M. tuberculosis
obtained from children aged 0-13 years of age were
evaluated for drug susceptibility. Among children
0-5 years of age, the incidence of isoniazid resistance
and multidrug resistance was 5.6% and 1%, respec-
tively, which did not differ significantly from that
found earlier in adults. This indicates that the inci-
dence of resistance in children probably reflects the
transmission of drug-resistant strains that is occur-
ring amongst adults. Where cultures can be obtained
from children and tested for drug susceptibility, this
will offer a relatively simple means of monitoring
drug resistance in a community.
In children with a culture of M. tuberculosis resis-
tant to isoniazid only, the use of a standard WHO
retreatment regimen is considered adequate. This
consists of 2 months of isoniazid, rifampicin, pyra-
zinamide, ethambutol and streptomycin followed by
1 month of isoniazid, rifampicin, pyrazinamide and
P.R. Donald
ethambutol and a 6-month continuation phase of
isoniazid, rifampicin and ethambutol (Crofton et al.
1997). The use of isoniazid in the face of resistance
is controversial. In primary isoniazid resistance,
low-level isoniazid resistance (MIC<5Ilg/ml) may
be found in more than half of the patients (Canetti
1965; Tripathy et al. 1969; Schaaf et al. 2000). These
concentrations are easily attainable with a dose of
10-20 mg/kg isoniazid, which will usually be well
tolerated by young children.
In the case of resistance to at least isoniazid and
rifampicin (multidrug resistance), a 5-drug intensive
phase is considered mandatory (Crofton et al. 1997)
and should be given for 3 months or until smear
conversion. The continuation phase should last for
18 months after smear conversion and should con-
tain at least three drugs to which the organism is
sensitive. In most childhood disease a culture of M.
tuberculosis may not be obtained. In these cases it is
advisable to treat the child according to the suscepti-
bility pattern of the adult contact and to use at least
three drugs to which the child has not been exposed
before.
16.12
Prevention and Control of Tuberculosis
in Childhood
Tuberculosis infection in childhood is the conse-
quence of the inhalation of small aerosol droplets
5-10 flm in diameter containing 1-3 M. tuberculosis
bacilli. These are expectorated by adults, who will
usually be suffering from cavitating pulmonary
tuberculosis and whose sputum will be positive on
microscopy for acid-fast bacilli. There is therefore
little doubt that the control of tuberculosis in adults
would rapidly eliminate childhood tuberculosis as a
serious health problem. It is equally clear that there
is little prospect that tuberculosis will be controlled
in developing countries in the foreseeable future
and that the main focus of tuberculosis control
programmes will continue to be the supervision of
directly observed short course treatment for sputum
,smear-positive' adults. Nevertheless, it should
be emphasized that even within the constraints
imposed by financial restrictions and the lack of
resources, opportunities exist to reduce the impact
of tuberculosis on children. In order to make use of
these opportunities, we need to continually remind
ourselves that tuberculosis is an infectious disease
and that in a high incidence community those most
Childhood Tuberculosis 259

often affected will be young adults in their reproduc-

In the very young, disease is particularly likely to
tive years. In many developing communities it is also
take the form of disseminated forms of tuberculosis
women who are bearing the brunt of the HIV/AIDS such as tuberculous meningitis or miliary tubercu-
epidemic and, as indicated above, HIV/AIDS-associ-losis. Mortality rates for those infected at an age of
ated tuberculosis has become an important cause ofunder 1 year and untreated have been calculated to
obstetric mortality and morbidity. These women will
be between 5000 and 6000 per 100,000 population
be in constant close contact with their own children
(Rich 1951; Bentley et al. 1954). The identification of
and those of their extended families. In addition to
young children in close contact with an adult with
the more obvious measures of chemoprophylaxis sputum microscopy "smear-positive" pulmonary
and BCG vaccination, there are several other ways by
tuberculosis therefore offers an opportunity for the
prevention of disease in these children and the early
which the force of the infectious process as it affects
children, but also adults, can be reduced. diagnosis and treatment of disease in those already
It is often forgotten that paediatric outpatient
infected. Depending upon the resources available,
facilities and in particular the overnight holdingsuch children should have a Mantoux test to identify
wards which are commonly encountered in the those infected and a chest radiograph to detect sub-
developing world can serve as a focal point for infec-
clinical disease. As it may take up to 10 -12 weeks for
tious disease transmission. Young children attending
tuberculin hypersensitivity to develop, it is prudent
these facilities will often be accompanied by parents
to commence chemoprophylaxis with isoniazid in
or caregivers, who may themselves be sick. In an these children exposed to infection. After 3 months
evaluation of young children less than 18 months the Mantoux test can be repeated. If this is positive,
of age with suspected tuberculosis attending one chemoprophylaxis should be continued to 6 months,
such facility, a policy of taking a chest radiograph of
and if negative, BCG vaccination can be considered.
the accompanying parent or caregiver revealed pul- A number of controlled clinical trials have dem-
monary tuberculosis in 18% of cases. Amongst the onstrated the efficacy of isoniazid chemoprophylaxis
adults accompanying 17 children less than 6 monthswhen given to children and adults with asymptom-
of age, this was an even more rewarding procedure,atic primary infection (Ferebee 1970). Not only is the
and tuberculosis was found in 6 adults (35%). The incidence of disease reduced, but more serious forms
identification of these adults not only contributed to
of disseminated tuberculosis are prevented. Isonia-
strengthening the diagnosis of tuberculosis in their
zid chemoprophylaxis for 3 months, 6 months and
children, but also helped prevent the further spread
12 months reduced the risk of tuberculosis by 21%,
of tuberculosis in their families and in the outpatient
65% and 75%, respectively. (International Union
department. against Tuberculosis Committee on Prophylaxis
Two other measures, which can be considered in 1982). In an analysis of the cost effectiveness of dif-
busy health care facilities where children and their
ferent durations of prophylaxis, it was then claimed
caregivers congregate, are attention to ventilation
that 6 months of chemoprophylaxis offered the best
and the use of ultraviolet light. Good ventilation can
compromise (Snider et al.1986), and this is currently
significantly reduce the density of bacilli in the air.
the duration of isoniazid chemoprophylaxis recom-
Merely renewing the air 6 times in an hour has been
mended in many official documents. A more recent
calculated to reduce the density of bacilli to 100th of
analysis of data from several studies has however
what it was beforehand (Riley and O'Grady 1961). shown that 9 months of chemoprophylaxis probably
Ultraviolet light can also contribute to the killing
offers the optimal protection against the develop-
of bacilli (Riley et al. 1976). If the provision of arti-
ment of tuberculosis disease (Comstock 1999).
ficial ultraviolet light is thought too expensive, the
A major impediment to the successful implemen-
provision of large windows directed appropriately tation of chemoprophylaxis is non-compliance, and
to make maximum use of available sunlight can con-relatively low percentages of adults and children in
tribute further to reducing the risk of tuberculosis
both developing and developed countries who com-
infection. plete the prescribed course of treatment (Woebeser
et al. 1989; Kopanoff et al. 1978). In developing com-
Chemoprophylaxis. Between 50% and 60% of young munities health care staff are already hard pressed
children in close household contact with an adult to ensure that their adult ,smear-positive' patients
with sputum microscopy ,smear-positive' pulmo- complete their therapy, and it is hardly surprising
nary tuberculosis will become infected. Approxi- that chemoprophylaxis is not accorded a very high
mately 10% of those infected may become diseased. priority. In our own experience of 193 children pre-
260
senting with tuberculous meningitis in the Western
Cape Province of South Africa, 77 (40%) had a close
household contact who had been treated for pul-
monary tuberculosis within the previous 2 years.
Only 17 (22%) of these children had, however, been
prescribed isoniazid chemoprophylaxis, and a mere
7 completed 3 months of prophylaxis (Donald et al.
1995).
In the face of these dismal figures, it is important to
prioritise and to give our attention to those at great-
est risk. These are undoubtedly the children under
2 years of age in household contact with an adult with
sputum ,smear-positive' pulmonary tuberculosis.
These adults will frequently be a parent or grandpar-
ent or another caretaker. As these individuals must
themselves receive supervised treatment, it should
add relatively little to the burden of the health care
personnel to ensure that their children receive pro-
phylactic treatment at the same time.
Although isoniazid for 6 months remains the most
frequently recommended prophylaxis regimen, there
has been considerable interest in shorter regimens
containing rifampicin and pyrazinamide. This inter-
est has been sharpened by our inability to halt the
ongoing spread of HIV-associated tuberculosis and
evidence from Peru that even a model national tuber-
culosis control programme may not be able to reduce
the incidence of tuberculosis successfully even in the
absence of HIV/AIDS (Chaisson 2000). In experi-
ments with M. tuberculosis-infected mice, rifampi-
cin for 3 months or rifampicin and pyrazinamide
for 2 months were more effective than 6 months of
isoniazid in reducing the proportion of mice with
positive spleen cultures (Lecoeur et al. 1989). Several
controlled trials have now found rifampicin and
pyrazinamide for 2 months to give results equivalent
to those of 6 months of isoniazid (Halsey et al. 1998;
Mwinga et al. 1998; Gordin et al. 2000). There are at
present few data relating to children regarding this
approach, but there seems no reason to expect dif-
ferent results.
Bacille Calmette-Guerin Vaccination. Bacille
Calmette-Guerin (BCG) vaccination has been prac-
tised for more than 80 years, but there is still no
consensus as to its value in preventing tuberculosis
in childhood. Analysis of a number of clinical trials in
which BCG was evaluated has shown effects ranging
from significant protection to an increased incidence
of disease. Case-control studies have consistently
shown a protective effect against disseminated forms
of tuberculosis such as TBM and miliary tuberculosis.
It should be pointed out that some case-control stud-
P. R. Donald
ies made use of controls that were not in contact with
a patient with tuberculosis and so would automati-
cally tend to show some degree of protection. Never-
theless, the current consensus is that BCG does offer
a degree of protection of between 60%-80% against
disseminated forms of tuberculosis, and BCG forms
part of the World Health Organization Expanded
Programme of Immunization. BCG will, however,
have only a limited effect upon the epidemiological
situation in a community. It does not break the chain
of infection and protects the individual only against
disease dissemination, but not infection (Styblo and
Meijer 1976). BCG has also been evaluated in a vari-
ety of populations in different parts of the world. In
the light of the above variations in strains, methods
of administration and populations in which BCG
has been evaluated, it is perhaps not surprising that
conflicting results have been obtained.
BCG is derived from a virulent bovine tubercle
bacillus which was attenuated over a 13-year period
of subculturing and 230 transfers from one subcul-
ture to another. Following this, a number of culture
samples were distributed all over the world, and these
were in turn subjected to further subculturing on a
variety of different media. In this manner a number
of daughter strains have been created, possibly by
encouraging the emergence of minority strains
within each population (Osborn 1983). BCG has
been administered to a diversity of populations in a
number of different countries orally, by scarification,
by percutaneous multiple puncture and by intrader-
mal injection. It is therefore perhaps not surprising
that widely differing results have been obtained and
that there is no consensus as to the value of BCG.
Amidst the uncertainty of its action with regard
to tuberculosis, it is of interest to note a number of
other instances in which BCG appears to have had
an undoubted effect upon the immune system. Thus,
BCG has been documented to be instrumental in
helping to control leprosy in Malawi (Karonga Pre-
vention Trial Group 1996) and has a well substanti-
ated role in the management of bladder cancer (Alex-
androff et al. 1999). It is also significant to note that
following the discontinuation of BCG immunization
in Sweden, the incidence of atypical mycobacterial
lymphadenitis increased dramatically (Katila et al.
1987). It has also been speculated that BCG immuni-
zation might be the reason why HIV/AIDS patients
in Sweden have a much lower incidence of Mycobac-
terium avium complex infections than patients in the
USA and that it may therefore be of importance in
reducing the impact of HIV infection (Kallenius et al.
1989). There is also some evidence that BeG vaccina-
Childhood Tuberculosis
tion may have a role in preventing certain forms of
malignancy (Grange and Stanford 1990).
Intradermal BCG vaccination will usually lead to
the formation of a small superficial ulcer approxi-
mately 4 weeks after vaccination. This heals over
8-10 weeks and leaves a flat scar some 50 mm in
diameter. Complications may be seen in 3%-4% of
individuals, particularly if care is not taken with
regard to the technique. The importance of staff
training has been emphasized (Jeena et al. 2001). For
this reason complications tend to be more common
following intradermal vaccination than after per-
cutaneous vaccination. Complications encountered
include axillary lymphadenopathy with suppuration
in some cases, abscess formation at the injection site,
especially if the vaccine is delivered subcutaneously,
and disseminated forms of disease such as osteo-
articular involvement or, more rarely, tuberculous
meningitis. Disseminated BCG disease will usually
be seen exclusively in those with severe immunosup-
pression. Although disseminated BCG was in the past
mainly a disease of infants, it is now increasingly seen
in adults and older children and often in association
with HIV/AIDS (Talbot et al. 1997). At present, HIV
infection is not considered a contraindication to BCG
vaccination, and BCG should continue to be given to
all infants even if HIV-infected provided they have no
symptoms of AIDS.
Other factors which playa role in the occurrence
of complications include the type of vaccine used and
the dose of vaccine given. In Sweden between 1972
and 1974, 29 patients with osteo-articular disease
were reported per 100,000 children vaccinated. When
the BCG strain was changed, no more cases occurred
(Romanus 1987). In several countries the Pasteur
strain has been noted to be more reactogenic than
other strains, and several outbreaks of lymphadenitis
followed a change to the Pasteur strain. In another
instance an outbreak of lymphadenitis followed a
change of vaccine from the Pasteur strain to Tokyo 172
and was found to be due to the injection of an inappro-
priately high dose of vaccine (Kabra et al. 1993).
The management of BCG lymphadenitis is con-
troversial. Most lesions will ultimately heal sponta-
neously. In the case of suppurative lesions, needle
aspiration from a distance of 2-3 cm has been shown
to bring about a more rapid regression of adenopathy
and to prevent sinus formation (Banani and Alborzi
1994). It has also been claimed that streptomycin
administration into the nodes during the non-suppu-
rative phase prevents suppuration and causes earlier
healing than the previously advised use of erythro-
mycin (Kuyucu et al. 1998). Surgery has been advo-
261
cated for large lesions, but it is uncertain whether
the long-term results are indeed cosmetically better
than a more conservative approach. In the case of dis-
seminated forms of BCG disease, chemotherapy must
be given, and it should be remembered that BCG is
resistant to pyrazinamide.
With regard to HIV infection it is currently rec-
ommended that BCG vaccine should not be given
to those who are HIV-infected and have signs and
symptoms of AIDS. In the case of those who are HIV-
infected only, the benefits of BCG vaccination appear
to outweigh the potential risks (O'Brien et al. 1995).
In this context it should be noted that a recent study
amongst adult volunteers demonstrated viable BCG
mycobacteria draining from the vaccination ulcer
2 months after vaccination. These organisms could
theoretically pose a risk to other immunosuppressed
individuals who might be in close contact with the
vaccinated infant (Hoft et al. 1999).
References
Adhikari M, Pillay T, Pillay DV (1997) Tuberculosis in the
newborn: an emerging disease. Pediatr Infect Dis J 16:
1108-1112
Ahmed Y, Mwaba P, Chintu C, Grange J, Ustianowski A, Zurnla
A (1999) A study of maternal mortality at the University
Teaching Hospital, Lusaka, Zambia: the emergence of
tuberculosis as a major non-obstetric cause of maternal
death. Int J Tuberc Lung Dis 3:675-680
Alexandroff AB, Jackson AM, O'Donnel et al (1999) BCG
immunotherapy of bladder cancer: 20 years on. Lancet
353:1689-1694
Ansari NA, Kombe AH, Kenyon TA et al (1999) Mortality and
pulmonary pathology of children with HIV infection in Fran-
cis town, Botswana. Int J Tuberc Lung Dis 3 [Suppl]:S201
Bentley FJ, Grzybowski S, Benjamin R (1954) Pleural effu-
sion. Tuberculosis in childhood and adolescence. National
Association for the Prevention of Tuberculosis, London,
pp 50-54
Bleiker MA (1991) The annual risk of infection rate, the tuber-
culin survey and the tuberculin test. Bull Int Union Against
Tuberc 66:53-56
Brailey M (1940) A study of tuberculous infection and mortality
in children of tuberculous households. Am J Hyg 31:1-43
Canetti G (1965) Present aspects of bacterial resistance in
tuberculosis. Am Rev Respir Dis 92:687-703
Cauthen GM, Pio A, ten Darn HG (1988) Annual risk of infec-
tion. World Health Organization WHO/TB/88 154:1-34
Chaisson RE (2000) New developments in the treatment of
latent tuberculosis. Int J Tuberc Lung Dis 4:S176-S181
Chan SP, Birnbaum J, Rao M, Steiner P (1996) Clinical mani-
festations and outcome of tuberculosis in children with
acquired immunodeficiency syndrome. Pediatr Infect Dis
J 15:443-447
Chintu C, Zumla A (1995) Childhood tuberculosis and infec-
tion with the human immunodeficiency virus. J R Coli Phys
Lond 29:92-95
262
Chintu C, Bhat G, Luo C et al (1993) Sero-prevalence of human
immunodeficiency virus type 1 in Zambian children with
tuberculosis. Pediatr Infect Dis J 12:499-504
Comstock GW (1999) How much isoniazid is needed for the
prevention of tuberculosis among immunocompetent
adults? Int J Tuberc Lung Dis 3:847-850
Comstock GW, Livesay VT, Woolpert SF (1974) The prognosis
of a positive tuberculin reaction in childhood and adoles-
cence. Am J EpidemioI99:131-138
Crofton J, Chaulet P, Maher D (1997) Guidelines for the
management of drug-resistant tuberculosis. World Health
Organization, Geneva, WHO/TB/96.210 (Rev 1)
Delacourt C, Mani TM, Bonnerot V et al (1993) Computed
tomography with normal chest radiograph in tuberculous
infection. Arch Dis Child 69:430-432
Delacourt C, Poveda J-D, Chureau C, Beydon N et al (1995)
Use of polymerase chain reaction for improved diagnosis
of tuberculosis in children. J Pediatr 126:703-709
Dolin PJ, Raviglione MC, Kochi A (1994) Global tuberculosis
incidence and mortality during 1990-2000. Bull WHO 72:
213-220
Donald PR, Seifart HI (1988) Cerebrospinal fluid pyrazin-
amide concentrations in children with tuberculous men-
ingitis. Pediatr Infect Dis J 7:469-471
Donald PR, Seifart HI (1989) Cerebrospinal,fluid concentra-
tions of ethionamide in children with tuberculous menin-
gitis. J Pediatr 115:483-486
Donald PR, van Zyl LE, de Villiers J (1995) BCG vaccination
status of children with tuberculous meningitis and the
use of unsupervised isoniazid prophylaxis. S Afr Med J
85:167-170
Donald PR, Schoeman JF, van Zyl LE, de Villiers IN, Pretorius
M, Springer P (1998) Intensive short course chemotherapy
in the management of tuberculous meningitis. Int J Tuberc
Lung Dis 2:704-711
Dormer BA, Swart JA, Harrison I (1959) Prophylactic isoniazid
protection of infants in a tuberculosis hospital. Lancet 2:
902-904
Driver CR, Munsiff SS, Li J et al (2001) Relapse in persons
treated for drug-susceptible tuberculosis in a population
with a high coinfection with human immunodeficiency
virus in New York City. Clin Infect Dis 33:1762-1769
Dutt AK, Moers D, Stead WW (1989) Smear- and culture-nega-
tive pulmonary tuberculosis: four-month short course che-
motherapy. Am Rev Respir Dis 139:867-870
Ellard GA, Humphries MJ, Allen BW (1993) Cerebrospinal
fluid drug concentrations and the treatment of tuberculous
meningitis. Am Rev Respir Dis 148:650-655
Elliot AM, Berning SE. Iseman MD, Peloquin CA (1995) Fail-
ure of drug penetration and acquisition of drug resistance
in chronic tuberculous empyema. Tubercle Lung Dis 76:
463-467
Franchi LM, Cama RI, Gilman RH, Montenegro-James S, Sheen P
(1998) Detection of Mycobacterium tuberculosis in nasopha-
ryngeal aspirate samples in children. Lancet 352:1681-1682
Gordin FM, Chaisson RE, Matts JP et al (2000) An international
randomized trial of rifampin and pyrazinamide versus
isoniazid for prevention of tuberculosis in HIV-infected
persons. J Am Med Assoc 283:1445-1450
Graham SM, Coulter JBS, Gilks CF (2001) Pulmonary disease in
HIV-infected African children. Int J Tuberc Lung Dis 5:12-23
Grange JM, Stanford JL (1990) BCG vaccination and cancer.
Tubercle 71:61-64
P. R. Donald
Grzybowski S (1954) Radiology in childhood tuberculosis.
Tubercle 35:113-117
Halsey NA, Coberly JS, Desormeaux J et al (1998) Rifampicin
and pyrazinamide vs. isoniazid for prevention of tubercu-
losis in HIV-I infected persons. Lancet 351:786-792
Harries AD, Maher D (1996) TB/HIV a clinical manual. World
Health Organization, Geneva, WHO/TB/96.200
Harries AD, Parry C, Mbewe LN et al (1997) The pattern of
tuberculosis in Queen Elizabeth Central Hospital, Blantyre,
Malawi: 1986-1995. Int J Tuberc Lung Dis 1:346-351
Hoffman ND, Kelly C, Futterman D (1996) Tuberculosis infec-
tion in human immunodeficiency virus-positive adoles-
cents and young adults: a New York City cohort. Pediatrics
97:198-203
Hoft DF, Leonardi C, Milligan T et al (1999) Clinical reacto-
geincity of intradermal Bacille Calmette-Gutfrin vaccina-
tion. Clin Infect Dis 28:785-790
Hong Kong Chest Service/ Tuberculosis Research Centre,
Madras/British Medical Research Council (1989) A con-
trolled trial of 3-month, 4-month and 6-month regimens
of chemotherapy for sputum-smear-negative pulmonary
tuberculosis. Am Rev Respir Dis 139:871-876
Houwert KAF, Borggreven PA, Schaaf HS, Nel E, Donald PR,
Stolk J (1998) Prospective evaluation of World Health
Organization criteria to assist diagnosis of tuberculosis in
children. Eur Respir J 11:1116-1120
Humphries MJ (1992) The management of tuberculous men-
ingitis. Thorax 47:577-581
International Union Against Tuberculosis (1982) Efficacy
of various durations of isoniazid preventive therapy for
tuberculosis: five years of follow up in the IUAT trial. Bull
WHO 60:555-564
Jacobs M, Yach D, Fisher S, Kibel M, Hattingh, Coetzee G
(1987) Mangement of children with tuberculosis in a local
authority of Cape Town. Southern Afr J Epidemiol Infect
2:15-18
Jamieson DH (1995) Imaging intracranial tuberculosis in
childhood. Pediatr Radiol 25: 165-170
Jeena PM, Mitha T, Bamber S, Wesley A, Coutsoudis A, Coova-
dia HM (1996) Effects of human immunodeficiency virus
on tuberculosis in children. Tuberc Lung Dis 77:437-443
Jeena PM, Coovadia HM, Thula SA et al (1998) Persistent and
chronic lung disease in HIV-l-infected African children.
AIDS 12:1185-1193
Jeena PM, Chhagan MK, Topley J et al (2001) Safety of the
intradermal Copenhagen BeG vaccine in neonates in
Durban, South Africa. Bull WHO 79:337-343
Kabra SK, Jain Y, Seth MV (1993) BCG associated adenitis.
Lancet 341:970
KaIlenius G, Hoffner SE, Svenson SB (1989) Does vaccination
with Bacille Calmette-Guerin protect against AIDS? Rev
Infect Dis 11:349-351
Karonga Prevention Group (1996) Randomised controlled
trial of single BCG, repeated BCG, or combined BCG and
killed Mycobacterium leprae vaccine for prevenyion of lep-
rosy and tuberculosis in Malawi. Lancet 348:17-24
Katila ML, Brander E, Backman A (1987) Neonatal BCG vac-
cination and mycobacterial cervical adenitis in childhood.
Tubercle 68:291-296
Kendig EL (1969) Prognosis of infants born to tuberculous
mothers. Pediatrics 26:97-100
Khouri YF, Mastrucci MT, Hutto C, Mitchell CD, Scott GB
(1987) Mycobacterium tuberculosis in children with immu-
Childhood Tuberculosis
nodeficiency virus and routine childhood immunisations.
Lancet 2:669-672
Kochi A (1991) The global tuberculosis situation and the
new control strategy of the World Health Organization.
Tubercle 72: 1-6
Kuyucu N, Kuyucu S, Ocal B et al (1998) Comparison of oral
erythromycin, local administration of streptomycin and
placebo therapy for non-suppurating Bacillus Calmette-
Guerin lymphadenitis. Pediatr Infect Dis J 6:524-525
Lecoeur HF, Truffot-Pernot, Grosset JH (1989) Experimen-
tal short-course preventive therapy of tuberculosis with
rifampin and pyrazinamide. Am Rev Respir Dis 140:
1189-1193
Light IT, Saidleman M, Sutherland JM (1974) Management of
newborns after nursery exposure to tuberculosis. Am Rev
Respir Dis 109:415-418
Lincoln EM, Gilbert LA, Morales SM (1960) Chronic pulmo-
nary tuberculosis in individuals with known previous
primary tuberculosis. Dis Chest 5:473-482
Lobato MN, Loeffler AM, Furst K, Cole B, Hopewell PC (1998)
Detection of Mycobacterium tuberculosis in gastric aspi-
rates collected from children: hospitalization is not neces-
sary. Pediatrics 102, URL: http://www.pediatrics.orglcgi/
content/full/l02/4/e40
Long MW, Snider DE, Farer LS (1979) US Public Health Service
co-operative trial of three rifampicin -isoniazid regimens
in treatment of pulmonary tuberculosis. Am Rev Respir
Dis 119:879-894
Madhi SA, Petersen K, Madhi A, Khoosal M, Klugman KP
(2000) Increased disease burden and antibiotic resistance
of bacteria causing severe community-acquired lower espi-
ratory tract infections in human immunodeficiency virus
type-1 infected children. Clin Infect Dis 31:170-176
Medical Research Council (1948) Streptomycin treatment of
tuberculous meningitis. Lancet 1:582-596
Mitchison DA (2000) Role of individual drugs in the chemo-
therapy of tuberculosis. 1nt J Tuberc Lung Dis 4:796-806
Munro WT (1944) Tubercle bacilluria in the child. Edinb Med
J 51:101-106
Mukadi YD, Wiktor SZ, Coulibaly I-M et al (1997) Impact of
HIV infection on the development, clinical presentation
and outcome of tuberculosis among children in Abidjan.,
Cote d'Ivoire. AIDS 11:1151-1158
Mwinga A, Hosp M, Godfrey-Faussett P et al (1998) Twice
weekly tuberculosis preventive therapy in HIV infection
in Zambia. AIDS 12:2447-2457
Narita M, Hisada M, Thimmappa B et al (2001) Tuberculosis
recurrence:multivariate analysis of serum levels of tuber-
culosis drugs, human immunodeficiency virus status, and
other risk factors. Clin Infect Dis 32:515-517
Nemir RL (1986) Perspectives in adolescent tuberculosis: three
decades of experience. Pediatrics 78:399-405
Noordhoek GT, Kolk AH, Bjune G et al (1994) Sensitivity and
specificity of PCR for detection of Mycobacterium tubercu-
losis: a blind comparison study among seven laboratories.
J Clin MicrobioI32:277-284
O'Brien KL, Ruff AJ, Louis MA et al (1995) Bacillus Calmette-
Guerin complications in children born to HIV-1-infected
women with a review of the literature. Pediatrics 95:
414-418
Peloquin CA, Nitta AT, Burman WJ et al (1996) Low antitu-
berculosis drug concentrations in patients with AIDS. Ann
Pharmacother 30:919-925
263
Pillay Y Kahn M, Moodley Jet al (200!) The increasing burden
of tuberculosis in pregnant women, newborns and infants
under 6 months age in Durban, Kwazulu-Natal. S Afr Med
J 91:983-987
Ravenscroft A, Schoeman JF, Donald PR (2001) Tuberculous
granulomas in childhood tuberculous meningitis: radio-
logical features and course. J Trop Pediatr 47:5-12
Rieder HL (1993) Drug-resistant tuberculosis: issues in epi-
demiology and challenges for public health. Tuberc Lung
Dis 75:321-323
Riley RL, O'Grady F (1961) Airborne infection: transmission
and control. Macmillan, New York
Riley RL, Knight M, Middlebrook G (1976)Ultraviolet suscep-
tibility of BCG and virulent tubercle bacilli. Am Rev Respir
Dis 113:413-418
Rook GAW, Onyebujoh P, Stanford JL (1993) Th1->Th2 switch
and loss of CD4 cells in chronic infections: an immuno-
endocrinological hypothesis not exclusive to HIV, Immu-
nol Today 14:568-569
Romanus V (1987) Bacillus Calmette-Guerin-immunized and
unimmunized children in Sweden: a ten-year evaluation
following the cessation of general Bacillus Calmette Guerin
immunization of the newborn in 1975. Pediatr Infect Dis
J 6:272-280
Rutherfoord GS, Hewlett RH (1994) Atlas of correlative surgi-
cal and neuropathology and imaging (Gresham GA, ed).
Kluwer Academic, Dordrecht, pp 47-74
Schaaf HS, Nel ED (1992) Tuberculosis presenting as choles-
tatic jaundice in early infancy. J Pediatr Gastroenterol Nutr
15:437-439
Schaaf HS, Smith J, Donald PR, Stockland B (1989) Tubercu-
losis presenting in the neonatal period. Clin Pediatr 28:
474-475
SchaafHS, Donald PR, Scott F (1991) Maternal chest radiogra-
phy as supporting evidence for the diagnosis of tuberculo-
sis in childhood. J Trop Pediatr 37:223-225
Schaaf HS, Gie RP, Beyers N, Smuts, Donald PR (1993) Tuber-
culosis in infants less than 3 months of age. Arch Dis Child
69:371-374
Schaaf HS, Geldenhuys A, Gie RP, Cotton MF (1998) Culture-
positive tuberculosis in human immunodeficiency virus
type I-infected children. Pediatr Infect Dis J 17:599-604
Schaaf HS, Vermeulen HAS, Gie RP, Beyers N, Donald PR
(1999) Evaluation of young children in household contact
with adult multidrug-resistant pulmonary tuberculosis
cases. Pediatr Infect Dis J 18:494-500
Schaaf HS, Gie RP, Beyers N, Sirgel FA, de Klerk, Donald PR.
(2000a) Primary drug-resistant tuberculosis in children.
Int J Tuberc Lung Dis 4:1149-1155
SchaafHS,Van Rie A, Gie RP et al (2000b) Transmission of multi-
drug-resistant tuberculosis. Pediatr Infect Dis J 19:695-699
Seifart HI, Parkin DP, Donald PR (1991) The stability of iso-
niazid, rifampicin and pyrazinamide suspensions used for
the treatment of tuberculosis in children. Pediatr Infect
Dis J 10:827-831
Shaw JB, Wynn-Williams N (1954) Infectivity of pulmonary
tuberculosis in relation to sputum status. Am Rev Tuberc
69:724-732
Sirgel FA, Botha FJH, Parkin DP et al (1993) The early bac-
tericidal activity of rifabutin in patients with pulmonary
tuberculosis measured by sputum viable counts - a new
method of drug assessment. J Antimicrob Chemother 32:
867-875
264
Small PM, Schafer RW, Hopewell PC et al (1993) Exogenous
reinfection with multidrug-resistant Mycobacterium tuber-
culosis in patients with advanced HIV infection. N Engl J
Med 328:1137-1144
Smith KC, Starke JR, Eisenach K, Ong LT, Denby M (1996)
Detection of Mycobacterium tuberculosis in clinical speci-
mens from children using a polymerase chain reaction.
Pediatrics 97:155-160
Smith MHD (1967) Tuberculosis in adolescence. Characteris-
tics, recognition, management. Clin Pediatr 6:9-15
Snider DE, Kelly GD, Cauthen GM, Thompson NJ, Kilburn JO
(1985) Infection and disease among contacts of tuberculo-
sis cases with drug-resistant and drug susceptible bacilli.
Am Rev Respir Dis 132:125-132
Snider DE, Caras GJ, Koplan JP (1986) Preventive therapy
with isoniazid. Cost effectiveness of different durations of
therapy. J Am Med Assoc 255:1579-1583
Steiner P, Rao M, Mitchell M, Steiner M (1985) Primary drug resis-
tant tuberculosis in children. Am J Dis Child 139:780-782
Tanzania Tuberculin Survey Collaboration (2001) Tuberculo-
sis control in the era of the HIV epidemic: risk of tubercu-
losis infection in Tanzania, 1983- I998. Int J Tuberc Lung
Dis 5:103-112
Talbot EA, Perkins MD, Fagundes S et al (1997) Disseminated
Bacille Calmette-Guerin disease after vaccination: case
report and review. Clin Infect Dis 24:1139-1146
Te Water Naude JM, Donald PR, Hussey GD et al (2000) Twice
weekly vs.daily chemotherapy for childhood tuberculosis.
Pediatr Infect Dis J 19:405-410
Thomas P, Bornschlegel K, Singh TP et al (2000) Tuberculosis
in human immunodeficiency virus-infected and human
immunodeficiency virus-exposed children in New York
City. Pediatr Infect Dis J 19:700-706
Tripathy SP,Menon NK, Mitchison DA et al (1969) Response to
P. R. Donald
treatment with isoniazid plus PAS of tuberculosis patients
with primary isoniazid resistance. Tubercle 50:257-268
Udani PM (1983) Tuberculosis on children in India. A major
health hazard! Pediatr Clin India 18: 11-42
Vallejo JG, Ong LT, Starke JR (1994) Clinical features, diagnosis,
and treatment of tuberculosis in infants. Pediatrics 94: 1-7
Van Eden W, De Vries RR, Stanford JL et al (1983) HLA-DR3
associated genetic control of response to multiple skin tests
with new tuberculins. J Immunogenet 10:107-114
Van Geuns HA, Meijer J, Styblo K (1975) Results of contact
investigation in Rotterdam, 1967-1969. Bull Int Union
Against Tuberc 50:107-121
Wallgren A (1928) Initial fever in tuberculosis. Am J Dis Child
36:1341-1345
Wallgren A (1938) Pulmonary tuberculosis. Relation of child-
hood infection to disease in adults. Lancet 1:417-420
Wallgren A (1948) The ,time-table' of tuberculosis. Tubercle
29:245-251
Weber HC, Beyers N, Gie RP, Schaaf HS, Fish T, Donald PR
(2000) The clinical and radiological features of tuberculo-
sis in adolescents. Ann Trop Paediatr 20:5-10
Weyer K, Groenewald P, Zwarenstein M, Lombard CJ (1995)
Tuberculosis drug resistance in the Western Cape. South
Afr Med J 85:499-504
Wilcox WD, Laufer S (1994) Tuberculosis in adolescents. Clin
Pediatr 33:258-262
World Health Organization (1989) Provisional guidelines for
the diagnosis and classification of the EPI target diseases
for primary health care, surveillance and special studies.
EPIIGEN/83/4
Zar HJ, Tannenbaum E, Appoles P, Roux P, Hanslo D, Hussey G
(2000) Sputum induction for the diagnosis of pulmonary
tuberculosis in infants and young children in an urban set-
ting in South Africa. Arch Dis Child 82:305-308
17 Primary Tuberculosis in Adults
M. MONIR MADKOUR
CONTENTS
17.1 Epidemiology of Primary Tuberculosis
in Adults 265
17.2 Pathogenesis of Primary Pulmonary
Tuberculosis 266
17.3 Culture Yield of Bacilli from Primary Lesion
and Primary Complex 267
17.4 Groups at Risk of Primary Tuberculosis 267
17.5 Clinical Features and Diagnosis
of Primary Tuberculosis in Adults 268
17.6 Lung Infiltrates and Consolidation 268
17.7 Lymphadenopathy 269
17.8 Pleural Effusion 269
17.9 Miliary Tuberculosis 269
17.10 Microbiological Confirmation 269
17.11 Treatment and Possible Paradoxical
Transient Worsening 270
17.12 Conclusion 270
References 270
Traditionally, primary tuberculosis is a disease of
childhood (Beyers 1979). This remains true in poor,
developing countries where tuberculosis is endemic,
and infections commonly start during childhood (see
Chap. 16). This tradition, however, has been modified
in rich, developed and industrialized countries that
adopted successful tuberculosis control programs
starting in the nineteenth century. Developed industri-
alized countries at present harbor only 5% of the world
burden of tuberculosis, and the population group most
infected is over 65 years of age (Alexander et al. 1979)
(see Chap. 3). In these developed countries, exposure to
the infection occurs mostly in the elderly population,
leading to primary pulmonary tuberculosis.
During the current epidemic of HIV, co-infection
of adults with primary tuberculosis has been increas-
ingly reported, leading to its dissemination (Pitchenik
and Rubinson 1985; Wasser et al. 1988). This shift
towards the presentation of primary tuberculosis in
the elderly and among HIV-infected patients appears
M. M. MADKOUR, MD, DM, FRCP
Consultant, Department of Medicine, Riyadh Armed Forces
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
to be related to decreasing childhood exposure in the
developed, industrialized countries (Khan et al. 1977;
Miller and MacGregor 1978; Woodring et al. 1986;
Choyke et al. 1983; Hulnick et al.I983).
As a consequence of this shift of primary tuber-
culosis to the adult population, some authors from
developed countries suggested that the term 'child-
hood tuberculosis' and 'adult tuberculosis' be aban-
doned (Berger and Granada 1974; Choyke et al.I983).
We reviewed the medical records of 10 adult patients
with primary tuberculosis who attended our hospital
from 1990 through 2000. Their clinical features will
be discussed in this chapter. The imaging features of
these patients, including the follow-up radiograph
findings during treatment and after its completion,
are clearly described in our separate chapter specially
devoted to the radiology of pulmonary tuberculosis.
Imaging figures on primary pulmonary tuberculosis
are numbers: 23.3,23.4,23.8,23.9,23.10,23.12,23.16,
23.17 and 23.18.
17.1
Epidemiology of Primary Tuberculosis
in Adults
Since the shift of patient care from long-term san-
atorium management into general hospitals, the
clinicians' awareness of primary tuberculosis has
decreased. This is because tuberculous patients are
no longer managed by clinicians with a special inter-
est in the disease, along with the decline in the inci-
dence of the disease in the populations of developed
countries (Ellersten 1959; Khan et al. 1977; Varkey
and Politis 1981; Choyke et al. 1983). The incidence
of primary tuberculosis has been reported to range
between 10% and 34% (Stead et al. 1968; Khan et al.
1977; Miller and MacGregor 1978; Woodring et al.
1986; Buckner et al. 1990; Miller et al. 1993; Miller
1994; McAdams et al. 1995). Most of the epidemi-
ological data on primary tuberculosis in adults are
reported from the USA and Europe.
266
The criteria adopted to determine the incidence of
adult primary tuberculosis in various reported series
vary slightly among different authors. The criteria for
diagnosis basically rested upon the conversion of PPD
skin testing as well as imaging features (Chiba and
Kurihara 1979; Choyke et al. 1983). Choyke and col-
leagues from Duke and Yale Universities reviewed 103
patients with adult-onset primary pulmonary tuber-
culosis seen over a period of 6 years. Their criteria
included recent PPD skin test conversion with adenop-
athy or pleural effusion in a patient with no history of
tuberculosis and previously normal chest radiographs.
Black male predominance was noted in approximately
60%, and 73.8% was aged 40 years or more.
These authors reported that the incidence of
active primary tuberculosis in adult is approximately
10%-20%.
Miller and Miller (1993) considered a recent con-
version of the PPD skin test with isolated pleural
effusion, lymphadenopathy, isolated middle and
lower lobe infiltration, and most cases of miliary
tuberculosis as diagnostic of primary tuberculosis.
These authors estimated the incidence of primary
tuberculosis in adults as ranging from 23% to 34%.
In Norway, Heldal and colleagues (2000) reviewed the
cases of Norwegian-born patients with pulmonary
tuberculosis. They found that most cases in their
series were due to endogenous reactivation of the
disease. The incidence of adult primary pulmonary
tuberculosis in these patients was reported to be 10%
in 1975,19% in 1985, and 16% in 1995.
17.2
Pathogenesis of Primary Pulmonary
Tuberculosis
Primary pulmonary tuberculosis is transmitted by
inhalation of air-borne Mycobacterium tuberculosis
present in the cough particles from the sputum
of a smear-positive patient. The inhaled bacilli
usually deposit in the middle or lower lung field.
The exact mechanisms of the protective immunity
response against the development of the disease
in humans have not been totally clarified (Ellner
1997). M. tuberculosis lipoproteins and glycolipids
induce activation of macrophage, T-cell, and cyto-
kine expression (Wallis and Johnson 2001). Alveolar
macrophages phagocytose the bacilli, and activation
of the cell-mediated immune response (CMI) plays
an essential role in combating infection (Reddy and
Hayworth 2002; Garcia et al. 2002). The bacilli may
M. M. Madkour
survive, especially within nonactivated monocytel
macrophages that enter the alveoli from the blood-
stream (Dannenberg 2001).
Early in the primary infection (within a few hours),
the bacilli are transported by macrophages to the hilar
and/or paratracheal lymph nodes (Milburn 2001).
Tissue-damaging delayed hypersensitivity (DTH)
develops, and the bacilli-laden macrophages are killed.
Key factors in this process include monocyte-derived
interleukin (IL)-12 and tumor necrosis factor (TNF)-oc
as well as T-cell-derived IL-2 and interferon (IFN)-y.
These cytokines playa crucial role in the induction
of macrophage-mediated elimination of mycobacte-
ria (Vanham et al. 1997; Garcia et al. 2002). During
active pulmonary tuberculosis, signs of both immune
depression and immune activation are concomitantly
present (Vanham et al.1997). The balance of cytokines
produced by lymphocytes in response to infection
is believed to have a profound effect on the clinical
outcome. Depression of peripheral blood T-cell and
cytokine production has been demonstrated during
active tuberculosis disease (Garcia et al. 2002). As the
infection progresses, local areas of inflammation with
granuloma formation and cellular infiltrates occur in
the middle or lower zones of the lung. This is the pul-
monary component of the primary Ghon focus and, in
combination with hilar and/or paratracheallymphad-
enitis, forms the primary complex.
Tuberculin conversion may occur 3-6 weeks after
the initial infection (Milburn 2001). The outcome of
the infection depends on the interaction between
the host immune system response and the bacilli. In
patients with strong and competent cellular immunity,
the Ghon focus and regional lymphadenitis heal with
fibrosis and calcifications within weeks or months.
However, the primary complex lesion may progress
into lung parenchymal or lymph node infection or
both in what is known as progressive primary tuber-
culosis. Many tuberculosis specialists adopted the
use of the term 'progressive primary tuberculosis' if
it occurs within 3-8 months after tuberculin conver-
sion. About 10% of all cases of primary tuberculosis
progress to a chronic form indistinguishable from
reactivation disease (Goppert and Left 1979). Within
the first year of primary pulmonary infection, the
incidence of clinically significant disease is approxi-
mately 1.5%, and the cumulative risk during the first
5 years is 5%-10% (Chiba and Kurihara 1979).
Balasubramanian and colleagues (1994) from Wis-
consin, USA, extensively reviewed the published work
of other investigators which dealt with the patho-
genesis of tuberculosis. Particular emphasis was laid
on the infective dose, the yield of bacilli from the pri-
Primary Tuberculosis in Adults
mary lesion, predominant locations, and the argument
regarding the endogenous versus exogenous pathway
leading to post-primary tuberculosis.
The following paragraphs include some personal
abstracts as well as the investigators' views and
schools of thought on the issues of the pathogenesis
of primary pulmonary tuberculosis. The size of the
infective dose has been debated among investiga-
tors in a review article. On the one hand, some
investigators indicated that a large infectious dose
occurred more frequently in household contacts of
sputum-positive cases. The infective dose includes
the number of bacilli per infectious particles inhaled
during anyone period of exposure. Other investiga-
tors indicated that the infectious dose in tubercu-
losis was very small. The size of a droplet nucleus
disseminated during a cough by a sputum-positive
tuberculosis patient is 5.0 ~m in diameter. It is esti-
mated that the number of bacilli in a droplet nucleus
is 1-10. Accordingly, these observations suggest that
the infectious dose in tuberculosis is very low.
The location of the primary complex was noted
by Medlar (1948) in his series of 105 individuals
found during autopsies performed on 1225 bodies.
The single primary complex was located within 1 cm
of the pleural surface in 85%, and in the lower half
of the lung field in 66%, while only 12% were supra-
clavicular; he ascribed this pattern of lung lesion
distribution to the direction of the airflow.
The primary focus is the initial lesion produced
by tubercle bacilli in any tissue. The primary or
Ghon complex includes both the primary focus and
homologous lesions in the draining lymph nodes.
The primary complex most commonly occurs in the
lung, but may also develop in the skin, intestines, gen-
ital tract, and tonsils. A mucous membrane primary
focus may not incite a typical tuberculous healing
reaction, so that only fibrocalcific lymph nodes may
remain as indicators of a primary infection in these
tissues. Therefore, fibrocalcific lesions in the lung or
lymph nodes of a patient with organ and miliary
tuberculosis may be taken to represent evidence of a
prior primary infection (Slavin et al. 1980).
17.3
Culture Yield of Bacilli from Primary
Lesion and Primary Complex
Canetti (1972) suggested 'that tubercle bacilli do
not survive indefinitely in tuberculous foci'. Work
on guinea-pigs with inoculations of the component
267
of the primary complex (before the era of chemo-
therapy) indicated that no live bacilli could be cul-
tured in 85% of calcified lesions and in 50% in the
encapsulated stage.
This review explored as well the work and school
of thought of other investigators on how a calcified
primary focus can be a source of reactivation of the
disease when live bacilli no longer exist. Theories of
alternative locations created during the primary dis-
ease by the progeny of bacilli from the primary lesion
which had been disseminated hematogenously have
been postulated.
Other aspects of pathogenesis including the post-
primary tuberculosis and the specified pathways
(endogenous versus exogenous) by which the disease
may occur later in life have already been elaborated.
17.4
Groups at Risk of Primary Tuberculosis
Adults and the elderly in developed countries with
diseases that affect the integrity of their immune
system will increase the chances for the develop-
ment of primary tuberculosis infection (Alexander
et al. 1979; Ikejoe et al. 1992). Young adults with pre-
existing HIV infection are particularly susceptible to
co-infection with primary pulmonary tuberculosis.
Patients on chemotherapy or steroids for cancer and
other diseases are particularly vulnerable to the devel-
opment of primary tuberculosis. Drug abusers, the
homeless, and the elderly residing in nursing homes
are susceptible to the development of adult primary
pulmonary tuberculosis. Traveling to endemic areas
or a long airplane flight has been reported as risk
factors for the transmission of tuberculosis. Kenyon
and colleagues (1996) from the USA reported the
transmission of tuberculosis during a long flight
(8.75 hours) from a 32-year-old Korean woman tour-
ist to four passengers on the same flight in April 1994.
The patient was traveling from Chicago to Honolulu,
and she had been on antituberculous treatment for
the past 2 years. She was coughing during the flight
and had fever. Eventually, she became very sick and
was admitted to hospital and died 1 week later with
massive hemoptysis. The sputum smear was highly
positive, and culture was positive for M. tuberculo-
sis. The CDC was involved in the investigation that
included notifications to the passengers of that flight.
Careful and thorough epidemiological screening was
conducted including all risk factors and previous
PPD skin test history. Four patients who were known
268
tuberculin skin test-negative had initial testing after
exposure which was negative and had a final test
repeated 12 weeks later that showed conversion to
positive results. In the Netherlands, Cobelens et al.
(2000) investigated Dutch long-term travelers to
countries of high tuberculosis endemicity. PPD skin
testing was done before traveling and 2-4 months
after their return. The results of 656 individuals were
reported: there were 12 with M. tuberculosis infection,
and 2 had active disease (1.8%). The overall incidence
rate was estimated as 3.5 per 1000 person-months of
travel and 2.8 per 1000 person-months of travel after
the exclusion of health-care workers. The inclusion
criteria for this study were over 15 years of age, not
yet received BCG vaccination, and intended to travel
for 3-12 months to these highly endemic countries.
Exclusion criteria were history of tuberculosis or
positive skin test, diabetes mellitus, HIV, or history
of immunosuppressive therapy.
17.5
Clinical Features and Diagnosis
of Primary Tuberculosis in Adults
Primary pulmonary tuberculosis in adults may pres-
ent with constitutional symptoms as well as cough,
fever, hemoptysis, or weight loss that are often mis-
diagnosed initially as bacterial pneumonia. It is only
because of the atypical clinical features in association
with the so-called 'unusual' radiographic findings
(Segarra et al. 1963; Parmar 1967; Dodd et al. 1978;
Palmer 1979; Hadlock et al. 1980; Stead 1981; Lee et
al. 1993) or the adoption of 'routine' microbiological
examination of the sputum by smear staining and
culture for acid-fast bacilli that an accurate diagnosis
can be achieved (Choyke et al. 1983). A high index of
suspicion by the attending clinician is essential for an
early diagnosis. Failure in considering tuberculosis as
a possible cause in these patients will delay the initia-
tion of appropriate chemotherapy. The disease may
progress, and serious dissemination to other body
organs or systems may occur.
Guidelines and criteria were used by several
authors to identify and classify patients as having
adult primary pulmonary tuberculosis (Stead et al.
1968; Khan et al. 1977; Choyke et al. 1983). Khan and
colleagues (1977) classified adult patients with tuber-
culosis in the lower lung field as having primary pul-
monary tuberculosis if one or more of the following
criteria were also present: (1) known recent tuber-
culin conversion, (2) tuberculous pleural effusions
M. M. Madkour
without roentgenographic evidence of post-primary
disease, (3) hilar adenopathy with or without paren-
chymal infiltration. These authors found 12 patients
with adult primary pulmonary tuberculosis in their
large series of 88 adult patients with newly proven
diagnosis of active pulmonary tuberculosis who ful-
filled these criteria. They were seen over a 12-month
period and attended two hospitals in Boston, USA.
Ten patients were young adults (19-38 years old),
and 2 were in their 60s. Choyke and colleagues (1983)
from the USA adopted the same criteria and reported
their own series of 103 adult patients with primary
pulmonary tuberculosis with positive M. tuberculosis
cultures from at least one source.
In our own series of 10 adult patients with primary
pulmonary tuberculosis, the criteria for the classi-
fication depended mainly on radiographic findings.
Recent PPD skin test conversion or the status of a
previous tuberculin test is unknown in many of our
patients. Choyke and colleagues (1983) from the USA
stated: 'The diagnosis of primary tuberculosis rests
upon tuberculin skin test (PPD) conversion.' How-
ever, they reported that there was inadequate docu-
mentation of the PPD skin test in their own series.
Symptoms lasting a few days or weeks at the time
of presentation are commonly reported in 84% by
Khan et al. (1977) and Choyke et al. (1983). These
included constitutional symptoms, fever, cough, and
hemoptysis. The initial diagnosis of tuberculosis on
admission was only made in 10%, and 70% were ini-
tially diagnosed to have 'bacterial pneumonia', and
diagnosis was delayed for up to 6 weeks as reported
in Choyke's series of 103 patients.
17.6
Lung Infiltrates and Consolidation
Plain radiographs of the chest may look normal,
and lung infiltrates may only be depicted on CT.
Segmental or lobar consolidation of the right middle
or lower lobe is the most common pattern of distri-
bution in primary tuberculosis in adults as we noted
in 6 of our patients (60%) (Figs. 23.3,23.9,23.16, and
23.18). Consolidation in the upper lobe may also
occur, but it is rare. Associated hilar or right para-
tracheal lymphadenopathy with consolidation was
noted in these 6 patients as well. Cavitation (Dahl
1952) in primary tuberculous consolidation was
noted in 1 of our patients (Fig. 23.4).
In the series of 12 patients reported by Khan et al.
(1977), consolidations in the lower and middle lobes
Primary Tuberculosis in Adults
were noted in 6 patients (50%) and in the upper lobe,
in 1 patient. In Choyke and colleagues' (1983) series
of 103 patients, they reported consolidations in the
lower and middle lobes in 77 patients (75%) and in
the upper lobe in 36 patients (35%), bilateral in 13
(13%), cavitations in 8 (8%), and normal radiographs
in 10 patients (10%). In Belgium, van den Brande and
colleagues (1998) reported the incidence of adult
pulmonary tuberculosis in a low prevalence area.
They reviewed the chest radiographs of patients who
had bacteriologically proven pulmonary tuberculosis
over two periods (1981-1985 and 1986-1990). They
found 114 and 105 patients, respectively. Approxi-
mately 50% of these patients were aged 65 years and
over. Radiological features were classified as 'usual' or
'unusual' presentation.'Unusual' radiological features
were defined as solitary pleural effusion, isolated hilar
or mediastinal lymphadenopathies, normal chest x-
ray, lower lung field tuberculosis, nodular lesions, dif-
fuse infiltration, and atelectasis. The usual imaging
features of pulmonary tuberculosis were present in
76% and 68%, respectively. Unusual imaging features
of pulmonary tuberculosis were present in 24% and
35%, respectively. The authors noted the shift in the
incidence of tuberculosis to elderly subjects who had
never been infected before. They also noted that their
'unusual' imaging features resemble those observed
in childhood primary pulmonary tuberculosis. They
concluded that there is a recent trend towards a more
'unusual' presentation of primary pulmonary tuber-
culosis in the elderly.
17.7
Lymphadenopathy
Unilateral or bilateral hilar lymphadenopathy and/or
right paratracheal lymph node enlargement are the
hallmarks of primary tuberculosis. They were noted
in 9 patients with adult primary pulmonary tuber-
culosis (90%) in our own series, and their imaging
features are illustrated in our chapter on radiography
of pulmonary tuberculosis (Figs. 23.3, 23.4, 23.6, 23.7,
23.8,23.9,23.12,23.16, and 23.18). Hilar and/or para-
tracheal tuberculous lymphadenopathy in primary
tuberculosis is more commonly reported in children
than adults (McAdams et al. 1995; Woodring et al.
1986; Leung et al. 1992). Caseation and necrosis
of tuberculous hilar or paratracheal lymph nodes
may occur (Fig.23.8). Enlarged lymph nodes may
obstruct the airway and may lead to atelectasis. It
may invade the surrounding structures, particularly
269
the esophagus, phrenic nerve, recurrent laryngeal
nerve, superior vena cava, or pericardium (Figs. 23.10
and 23.11). Other authors reported the incidence of
tuberculous hilar lymphadenopathy as 15%-50% in
their series (Khan et al. 1977; Choyke et al. 1983).
17.8
Pleural Effusion
Pleural effusion in primary tuberculosis in an adult
may occur in 29%-70% of patients (Roper and War-
ing 1955; Stead et al. 1968; Miller and MacGregor
1978; Khan et al. 1977; Choyke et al. 1983; Epstein
et al. 1987).
It is often unilateral, but bilateral effusion may
rarely occur (Fig. 23.12). Associated parenchymal
lesions and hilar lymphadenopathy may be present
but not depicted on plain radiographs. Computed
tomography is more sensitive in depicting such an
association. Pleural effusion due to primary tuber-
culosis in adults was noted in one patient from our
series (Fig. 23.12).
17.9
Miliary Tuberculosis
Miliary tuberculosis may occur in both primary and
reactivation tuberculosis (Munt 1972; Gelb et al.1973;
Sahn and Neff 1974; Grieco and Chmel 1974). It is
increasingly reported in immunosuppressed adults,
particularly with HIV co-infection, in up to 13%
(Choyke et al. 1983; Lee and 1m 1995). The plain
radiograph may initially look normal, and CT is more
sensitive in depicting the miliary lung shadowing.
17.10
Microbiological Confirmation
Microbiological confirmation of the diagnosis in
adult primary pulmonary tuberculosis may be dif-
ficult. The sputum smear and culture with positive
yield were reported in as many as 60%-70% (Khan et
al. 1977; Choyke et al. 1983). Bronchoscopy to obtain
samples was performed in 34 patients, and in 27, the
cultures were positive as reported by Choyke et al.
(1983). Pleural fluid and biopsies were positive with
yield of the bacilli in 10 out of 26 cases reported by the
270
same authors. Other sources reported about patho-
logical specimens obtained by open lung biopsies and
thoracotomy in difficult-to-diagnose cases.
17.11
Treatment and
Possible Paradoxical Transient Worsening
The treatment of primary tuberculosis in adults does
not differ from the treatment of reactivation disease.
Directly observed therapy (DOT) is the most appro-
priate therapeutic regimen for primary pulmonary
tuberculosis in adults. It is usually given on an outpa-
tient basis for 6 months. In the initia12-month phase,
four drugs are used (rifampicin, isoniazid, pyrazin-
amide, and either ethambutol or streptomycin) fol-
lowed by a 4-month continuation phase of rifampicin
and isoniazid. Treatment is usually given daily but
can be given thrice weekly (see Chap. 44). During the
initial 3 months of chemotherapy, paradoxical tran-
sient worsening of the original radiographic findings
or development of new lesions may occur, as reported
by many authors (Weber et al. 1968; Matthay et al.
1974; Campbell and Dyson 1977; Amodio et al. 1986;
Lamont et al. 1986; Akira et al. 2000). The mecha-
nisms and pathogenesis of this paradoxical response
remain unclear. Enhanced immune responses to a
sudden destruction and release of tubercular anti-
gens or even a drug reaction has been postulated by
several authors (Onwubalili et al. 1986; Marshall and
Chambers 1988; Akira et al. 2000). Paradoxical tran-
sient worsening should be differentiated from true
worsening of tuberculous lesions due to co-infection
with HIY, other bacteria, or fungi. In our series, wors-
ening of the original parenchymal shadowing as well
as the development of new ipsilateral or contralateral
consolidations were noted in two patients (Fig. 23.16
and 23.17). Radiological evaluation should be done
at 2-3 months after initiation of chemotherapy (Bass
et al. 1986).
17.12
Conclusion
Primary pulmonary tuberculosis in adults is rarely
reported in endemic developing countries, where it
is mostly reported in children. The recent PPD skin
test conversions, considered as important criteria for
the diagnosis in developed countries, are difficult to
M. M. Madkour
document in most patients in endemic, resource-
poor countries. In developed, industrialized coun-
tries where the tuberculosis control program has
been applied for many years, with a low incidence
of the disease in the population, primary pulmonary
tuberculosis has shifted to adults and the elderly
population.
The criteria for classification of primary pul-
monary tuberculosis in adults, particularly, the
recent PPD skin test conversions, can be applicable.
The clinical and imaging features of adult primary
pulmonary tuberculosis are no longer considered as
'usual' because of the shift of the disease to adults in
developed countries. The clinical, imaging, and other
investigations to confirm the diagnosis are discussed.
Our own experiences with 10 patients with adult pri-
mary pulmonary tuberculosis have been discussed,
and the imaging features of these patients can be
seen in our separate chapter, Imaging of Pulmonary
Tuberculosis, Figs. 23.3, 23.4, 23.8, 23.9, 23.10, 23.12,
23.16,23.17, and 23.18. Paradoxical transient worsen-
ing of radiology features during the early phase of
chemotherapy is also documented.
References
Akira M, Sakatani M, Ishikawa H (2000) Transient radiographic
progression during initial treatment of pulmonary tuber-
culosis: CT findings. J Comput Assist Tomogr 24:426-43 I
Alexander WJ, Avent CK et al (1979) Simple primary tuber-
culosis in an elderly woman. J Am Geriatr Soc 27:123-125
Amodio J, Abramson S, Berdon W (1986) Primary pulmo-
nary tuberculosis in infancy: a resurgent disease in urban
United States. Paediatr RadioI16:185-189
Balasubramanian V et al (1994) Pathogenesis of tuberculo-
sis: pathway to apical localization. Tuberc Lung Dis 75:
168-178
Bass JB, Farer LS, Hopewell PC et al (1986) Treatment of tuber-
culosis and tuberculosis infection in adults and children.
Am Rev Respir Dis 134:355-363
Berger HW, Granada MG (1974) Lower lung field tuberculosis.
Chest 65:522-526
Beyers JA (1979) The radiological features of primary pul-
monary tuberculosis. S Afr Med J 55:994-997
Buckner CB, Walker CW et al (1990) Radiologic manifestations
of adult tuberculosis. J Thorac Imaging 5:28-37
Campbell lA, Dyson AJ (1977) Lymphnode tuberculosis: a
comparison of various methods of treatment. Tuberculosis
58:171-179
Canetti G (1972) Endogenous reactivation and exogenous
reinfection. Their relative importance with regard to the
development of non-primary tuberculosis. Bull Int Union
Tuberc 47:116-122
Chiba Y, Kurihara T (1979) Development of pulmonary tuber-
culosis with special reference to the time interval after
tuberculin conversion. Bull Int Union Tuberc 54:263-264
Primary Tuberculosis in Adults
Choyke PL,Sostman HD et al (1983) Adult onset of pulmonary
tuberculosis. Radiology 148:357-362
Cobelens FGJ, Deutekom HV et al (2000) Risk of infection with
Mycobacterium tuberculosis in travelers to areas of high
tuberculosis endemicity. Lancet 356:461-465
Dahl RV (1952) The first appearance of a pulmonary cavity
after primary infection with relation to time and age. Acta
Tuberc Scand 27:140
Dannenberg Am Jr (2001) Pathogenesis of pulmonary Myco-
bacterium bovis infection: basic principles established by
the rabbit model. Tuberculosis 81:87-96
Dodd GD, Capitanio MA et al (1978) Case 11 (primary
tuberculosis). In: Film interpretation session. Radiologi-
cal Society of North America, 64th scientific assembly and
annual meeting, Nov 1978. Summary of discussion and
diagnoses. Radiology 129:831
Ellertsen E (1959) Epidemics of primary tuberculosis and
their significance. Acta Tuberc Scand 37:203
Ellner JJ (1997) Regulation of the human immune response
during tuberculosis. J Lab Clin Med 130:469-475
Epstein DM, Kline LR et al (1987) Tuberculous pleural effu-
sions. Chest 91: 106
Garcia M, Vargas A, Castej6n R et al (2002) Flow-cytometric
assessment of lymphocyte cytokine production in tuber-
culosis. Tuberculosis 82:37-41
Gelb AF, Lefler C et al (1973) Miliary tuberculosis. Am Rev
Respir Dis 108:1327
Goppert EF, Left A (1979) The pathogenesis of pulmonary and
military tuberculosis. Arch Intern Med 139: 1381-1383
Grieco MH, Chmel H (1974) Acute disseminated tuberculosis
as a diagnostic problem. Am Rev Respir Dis 109:554
Hadlock FP, Park SK et al (1980) Unusual radiographic find-
ings in adult pulmonary tuberculosis. AJR 134:1015-1018
Heldal E et al (2000) Pulmonary tuberculosis in Norwegian
patients. The role of reactivation, re-infection and primary
infection is assessed by previous mass screening data and
restriction fragment length polymprphism analysis. Int J
Tuberc Lung Dis 4:300-307
Hulnick DH, Naidich DP, McCauley Dr (1983) Pleural tuber-
culosis evaluated by computed tomography. Radiology 149:
759-765
Ikejoe J, Kakeuchi N et al (1992) CT appearance of pulmo-
nary tuberculosis in diabetic and immunocompromised
patients: comparison with patients who had no underly-
ing disease. AJR 159:1175-1179
Kenyon TA, Valway SE et al (1996) Transmission of multi-
drug-resistant Mycobacterium tuberculosis during a long
airplane flight. N Engl J Med 334:933-938
Khan MA, Kovnat DM et al (1977) Clinical and roentgeno-
graphic spectrum of pulmonary tuberculosis in the adult.
Am J Med 62:31-38
Lamont AC, Cremin BJ, Pelleret RM (1986) Radiological pat-
terns of pulmonary tuberculosis in paediatric age group.
Pediatr RadioI16:2-7
Lee KS, 1m JG (1995) CT in adults with tuberculosis of the
chest: characteristic findings and role in management AJR
164:1361-1367
Lee KS, Song KS et al (1993) Adult-onset pulmonary tuber-
culosis: findings on chest radiographs and CT scans. AJR
160:753-758
Leung AN, Muller NL, Pineda PR et al (1992) Primary tuber-
culosis in childhood. Radiographic manifestations. Radi-
ology 182:87-91
271
Marshall BG, Chambers MA (1988) Central nervous system
tuberculosis: the paradox of host immune response. J Infect
36:3-4
Matthay RA, Neff TA, Iseman MD (1974) Tuberculous pleural
effusion developing during chemotherapy for pulmonary
tuberculosis. Am Rev Respir Dis 109:469-472
McAdams HP, Erasmus J, Winter JA (1995) Radiologic man-
ifestations of pulmonary tuberculosis. Radiol Clin North
Am 33:655-678
Medlar EM (1948) The pathogenesis of minimal pulmonary
tuberculosis: a study of 1225 necropsies in cases of sudden
and unexpected death. Am Rev Tuberc 58:583-611
Milburn HJ (2001) Primary tuberculosis. Curr Opin Pul Med
7:133-141
Miller WT (1994) Tuberculosis in the 1990s. Radiol Clin North
Am 32:649-661
Miller WT, MacGregor RR (1978) Tuberculosis: the frequency
of unusual radiographic findings. AJR Am J Roentgenol
130:867-875
Miller WT, Miller WT Jr (1993) Tuberculosis in the normal
host: Radiological findings. Semin RoentgenoI28:109-118
Munt RW (1972) Miliary tuberculosis in the chemotherapy
era: with a clinical review in 69 American adults. Medicine
(Baltimore) 51:139
Onwubalili JK, Scott GM, Smith H (1986) Acute respiratory
distress related to chemotherapy of advanced pulmonary
tuberculosis. A study of two cases and review of the lit-
erature. OJ Med 59:599-610
Palmer PES (1979) Pulmonary tuberculosis - usual and unusual
radiographic presentations. Semin RoentgenoI14:204-243
Parmar MS (1967) Lower lung field tuberculosis. Am Rev
Respir Dis 96:310
Pitchenik AE, Rubinson HA (1985) The radiographic
appearance of tuberculosis in patients with the acquired
immunodeficiency syndrome (AIDS) and pre-AIDS. Am
Rev Respir Dis 131:393-396
Reddy VM, Hayworth DA (2002) Interaction of Mycobacterium
tuberculosis with human respiratory epithelial cells (Hep-2).
Tuberculosis 82:31-36
Roper WH, Waring JJ (1955) Primary serofibrinous pleural
effusion in military personnel. Am Rev Tuberc 71:616
Sahn S, Neff TA (1974) Miliary tuberculosis. Am J Med 56:495
Segarra F, Sherman DS et al (1963) Lower lung field tuber-
culosis. Am Rev Respir Dis 87:37-40
Slavin RE, Walsh TJ et al (1980) Late generalized tuberculo-
sis: a clinical pathologic analysis and comparison of 100
cases in the preantibiotic and antibiotic eras. Medicine 59:
352-366
Stead WW, Kerby GR et al (1968) The clinical spectrum of
primary tuberculosis in adults. Confusion with reinfection
in the pathogenesis of chronic tuberculosis. Ann Intern
Med 68:731-745
Stead WW (1981) Tuberculosis among elderly persons: an
outbreak in a nursing horne. Ann Intern Med 94:606
Van den Brande PV, Dockx S et al (1998) Pulmonary tuberculosis
in the adult in a low prevalence area: is the radiological pre-
sentation changing? Int J Tuberc Lung Dis 2:904-908
Vanham G, Toossi Z, Hirsch CS et al (1997) Examining
a paradox in the pathogenesis of human pulmonary
tuberculosis: immune activation and suppression/anergy.
Tuberc Lung Dis 78:145-158
Varkey B, Politis J (1981) Pulmonary tuberculosis. A multi-
faceted disease. Postgrad Med 69:117-121, 124-126, 129
272
Wallis RS, Johnson JL (2001) Adult tuberculosis in the 21st
century: pathogenesis, clinical features and management.
CUff Opin Pul Med 7:124-132
Wasser LS et al (1988) Endobronchial tuberculosis in the
acquired immunodeficiency syndrome. Chest 94:1240-1244
M. M. Madkour
Weber AL, Bird KT, Janower ML (1968). Primary tuberculosis
in children with particular emphasis on changes affecting
the tracheobronchial tree: AJR 103:123-132
Woodring JH, Vandiviere H et al (1986) The radiographic fea-
tures of pulmonary tuberculosis. AJR 146:497-506
18 Miliary/Disseminated Tuberculosis
M. MONIR MADKOUR

CONTENTS mary tuberculosis, immediately after the post-primary

period, or at a time far remote from the post-primary
18.1 Introduction 273 period as late generalized tuberculosis (LGT) (Slavin
18.2 Epidemiology of MiliarylDisseminated
Tuberculosis 274 et al. 1980). A wide spectrum of pathological features
18.3 Pathology and Pathogenesis occurs as a result of hematogenous dissemination of
of Miliary/Disseminated Tuberculosis 275 the bacilli. This spectrum of pathology is determined
18.4 Clinical Features of Miliary/Disseminated by the size of the bacillary inoculum load, the viru-
Tuberculosis 277 lence of the bacilli, and the status of the host immune
18.5 Meningitis and Tuberculomas
in Miliary/Disseminated Tuberculosis 277 response. It ranges from a rare but acute fulminating
18.5.1 Case Illustration 1 279 form due to the release of massive myriads of caseous
18.5.2 Case Illustration 2 283 and necrotic tubercles into the blood with a nonreac-
18.5.3 Case Illustration 3 288 tive and an anergic response, very low count of CD4+
18.5.4 Case Illustration 4 290 T-cells with scanty or absent granuloma formation;
18.5.5 Case Illustration 5 292
18.6 Adult Respiratory Distress Syndrome that form of the disease is only diagnosed at autopsy.
and Disseminated Intravascular Coagulation If, on the other hand, tuberculous bacteremia is slight
in Miliary Tuberculosis 292 with the formation of few tubercles, this discrete type of
18.7 Pneumothorax in Miliary/Disseminated generalized dissemination is usually without immedi-
Tuberculosis 293 ate clinical significance, although these tubercles serve
18.8 Diagnosis of MiliarylDisseminated
Tuberculosis 293 as 'seed beds' for the later development of organ tuber-
18.9 High Resolution CT 296 culosis or LGT (Slavin et al. 1980). Between these two
18.10 Laboratory Diagnosis of Miliary pathological and clinical extremes lies a spectrum of
Tuberculosis 296 pathology that varies in severity. The more common
18.10.1 Polymerase Chain Reaction 297 classic miliary tuberculosis resembles the fulminating
18.11 Treatment of MiliarylDisseminated
Tuberculosis 297 form (Prout et al. 1980). Miliary tuberculosis is defined
References 298 as a hematogenous dissemination of the bacilli result-
ing in widespread, active, visceral, caseous tubercle
formations measuring 1-3 mm in diameter (the size of
millet seeds) with radiologic or pathologic evidence of
pulmonary micronodules (Sahn and Neff 1974; Slavin
18.1 et al. 1980; Penner et al. 1995). Disseminated tubercu-
Introduction losis is defined as a hematogenous transmission of the
bacilli with active caseous tubercle formation in two
Miliary/disseminated tuberculosis signifies the wide- or more extrapulmonary sites and with no pulmonary
spread occurrence of caseating visceral tuberculosis miliary nodular shadowing on chest radiography
that occurs by hematogenous dissemination of the (Penner et al. 1995; Sahn and Neff 1974). Proudfoot
bacilli from an active caseous focus or foci located in the et al. (1969) described the fulminating dissemination
lung or extrapulmonary sites. Hematogenous dissemi- form of tuberculosis in patients with other co-existing
nation of the bacilli may occur during the course of pri- underlying disease that may lead to a clinical presen-
tation with atypical features due to an impaired cell-
mediated immune response and the absence of miliary
M. M. MADKOUR, MD, DM, FRCP
lung shadowing as 'cryptic disseminated tuberculosis',
Consultant, Department of Medicine, Riyadh Armed Forces which is usually diagnosed at autopsy. Disseminated
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia tuberculosis is a diagnostic challenge even in endemic

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
274
areas when the attending clinician has a high index
of suspicion. Chest radiography is usually the initial
diagnostic investigation for miliary tuberculosis. The
sensitivity of chest radiography was found to range
from 59% to 69% in a study based on population and
group control subject (Kwong et al. 1996). ARDS may
occur as a complication of miliary tuberculosis, and
chest radiography may be difficult to interpret since the
miliary nodules are superimposed on a more diffuse
ground (Armstrong et al. 1995). Fatal consequences if
undiagnosed and untreated at an early stage may affect
21 %-64% of patients with disseminated tuberculosis
(Monie et al. 1983; Al-Jahdali et al. 2000; Prout et al.
1980; Bobrowitz 1982). Miliary/disseminated tubercu-
losis has increased in incidence in the USA, particularly
among HIV co-infected patients (Rieder et al. 1991; Hill
1991; FitzGerald et al. 1991; Korzeniewska-Kosela et al.
1992).
Iatrogenic miliary/disseminated tuberculosis due
to Mycobacterium bovis, induced by intravesical BCG
immunotherapy for the treatment of urinary bladder
cancer, has been increasingly reported (Foster 1997;
McParland et al. 1992; Palayew et al. 1993).
18.2
Epidemiology of Miliary/Disseminated
Tuberculosis
The incidence of miliary/disseminated tuberculo-
sis is difficult to determine in either developing or
developed countries (Vijayan 2000). Many patients
were undiagnosed in life, and it is only at autopsy
that the disease may be found (JuuI1977; Bobrowitz
1982; Slavin et al. 1980; Prout et al. 1980).
In the USA, Slavin et al. (1980) from Johns
Hopkins reviewed the records of 120 autopsies per-
formed between May 1937 and December 1959. They
excluded 20 cases due to missed records. The remain-
ing 100 autopsies had microbiologically proven TB,
histopathological features, and evidence of hematog-
enous spread involving the spleen, liver, lung, bone
marrow, kidney, adrenal glands, and other organs.
The total number of autopsies performed during
these 22 years was 14,224, and LGT was found in 120
(0.8%). Sixty of the 100 patients lived in the pre-anti-
biotic era (mid-1948) and 40 in the antibiotic period.
Male predominance was noted in 61 (61 %), of whom
44% were black. There were 39 women, 28% of whom
were black. The total number of black patients was 72
and white patients, 28. The authors estimated that 1
case of LGT was found in every 142 autopsies.
M. M. Madkour
In Canada, Long et al. (1997) reviewed the records
of 2013 patients with active tuberculosis between
January 1, 1979, and December 31,1993, to determine
the clinical-pathologic-radiologic correlation in dis-
seminated tuberculosis with and without a miliary
pattern on radiography. All patients came from Man-
itoba, with a low seroprevalence of HIV in that area.
They adopted a rigorously defined; radiographic or
histopathologic evidence of hematogenous dissemi-
nation and found 56 patients (3%) (42 miliary and
14 nonmiliary). Women were 2.4 times more likely to
be affected than men in this series. The mean age for
both sexes was 50±26 years. Fourteen patients died,
of whom 5 (12%) had miliary and 9 (64%) had non-
miliary disseminated tuberculosis.
In Glasgow, Scotland, Monie et al. (1983) reviewed
the records of 1000 patients with tuberculosis
reported during a 3-year period (1976-1978) and
found 28 (2.8%) patients with miliary tuberculosis.
Seventeen (61%) had miliary shadowing, 7 (25%)
had other radiological changes consistent with
active or previous pulmonary tuberculosis, and 4
(14%) had no radiological changes suggesting pul-
monary tuberculosis. About 50% of patients were
over 60 years of age, and 23% were non-European in
origin. These authors reported fatalities in 9 (32%) of
the 28 patients with miliary tuberculosis.
Ormerod and Horsfield (1995) from Blackburn,
UK, reviewed notified cases of tuberculosis in the
Blackburn, Hyndburn and Ribble Valley District
Health Authority (DHA), an area of high prevalence
for tuberculosis. They found 39 cases of miliary tuber-
culosis over 16 years (1978-1993 inclusive). Fourteen
patients had miliary TB, and 25 had cryptic miliary
TB. The authors noted ethnic differences with regard
to the number of cases and the age at onset. The
mean age among Caucasian patients was 51.3 years,
while among Indian subcontinent (ISC) patients, it
was 37 years. The incidence of miliary tuberculosis
in Blackburn was approximately 0.25/100,000 in the
Caucasian ethnic group and 7.99/100,000 in the ISC
ethnic group. These authors reported the death rate
due to miliary tuberculosis as 10%.
In Denmark, Juul (1977) reported on clinically
undiagnosed, active tuberculosis in 895 autopsies.
They found active tuberculosis undiagnosed in life
in 86 (0.1 %) patients. Miliary tuberculosis found in 38
(44.2%). The mean age of these patients was 70.7 years,
with a female predominance.
In NewYork,Bobrowitz (1982) reported 21 patients
in whom tuberculosis was not diagnosed in life. At
autopsy, 10 had suffered miliary tuberculosis and
11, pulmonary tuberculosis. Fifteen were male and
MiliarylDisseminated Tuberculosis
6 female, and 13 were white, 6 black, and 2 Hispanic.
Over 70% of the patients were more than 60 years
of age. The main diagnosis of the 10 patients with
miliary tuberculosis on admission was pneumonia,
malignancy, or sepsis. The authors explained the fail-
ure to discover tuberculosis in these patients as due
to combinations of other diseases and conditions. The
incidence of miliary tuberculosis in developing coun-
tries is more difficult to determine (Prout et al. 1980).
Lack of notification of miliary tuberculosis is due to
failure to make the diagnosis while alive in countries
like South Africa, which has a high incidence of tuber-
culosis (Prout et al. 1980). These authors reviewed the
computerized records of medical diagnoses over a 6-
year period and found 64 cases of miliary or dissemi-
nated tuberculosis. Thirty-seven cases fulfilled the
inclusion criteria: miliary lung shadowing, detection
of the bacilli by stain or culture, caseating granulomas
from liver or bone marrow biopsies, and involvement
of two or more organs. These authors also reviewed
the autopsies of 2,200 patients who died between
1973 and 1978 and found a further 34 cases (of them
9 fell into both groups and were excluded). A total
number of 62 cases (live and autopsies) was reviewed.
Eight (12.9%) were white, 20 (32.3%) colored, and 34
(54.8%) black. Thirty-six (58.1%) were male and 26
(41.9%) female. Over 72% of men and 42% of women
were over 40 years of age. Alcoholism, malnutrition,
diabetes, cytotoxicity or radiotherapy for malignan-
cies were concomitantly present in 25 patients (40.3%).
Forty patients (64%) died, and 31 deaths were directly
attributed to disseminated tuberculosis.
In Saudi Arabia, AI-Jahdali et al. (2000) reported
780 cases of tuberculosis seen over a period of 7 years.
Miliary tuberculosis was found in 47 (6%) patients,
and 68% were over 60 years of age. A male pre-
dominance was noted (n=30), and the mean age was
61 years for both sexes. Risk factors such as diabetes,
prior tuberculosis, chronic liver disease, renal failure,
immunosuppressive drugs, and malignancies were
present in over 50% of these patients. Ten patients
died (21 %), all were due to miliary tuberculosis.
In our own hospital in Riyadh, computerized
records of medical diagnoses for patients admitted
during the period from 1983-2000 inclusive (using
ICD-9 code description) revealed a total number of
2,484 tuberculosis patients, with miliary tuberculosis
reported in 79 patients (3.18%).
Since the HIV epidemic in 1980s, the incidence of
miliary/disseminated tuberculosis in HIV co-infected
patients in the USA has increased. The incidence
of miliary/disseminated tuberculosis has therefore
increased from 1.3% of all reported cases of tuber-
275
culosis to 8%-10% among HIV co-infected patients
(Rieder et al. 1991; Hill 1991; Schaefer et alI991).
18.3
Pathology and Pathogenesis
of Miliary/Disseminated Tuberculosis
Acute miliary/cryptic disseminated tuberculosis is
part of a broad spectrum of hematogenous spread
of Mycobacterium tuberculosis with various and vari-
able degrees of clinical manifestation and severity.
It is most commonly encountered in children and
young adults in developing countries. It may occur
during primary disease or within 6-8 months of
primary disease (progressive primary). In endemic
areas, hematogenous spread may occur in neonates,
infants, and children or in immunocompromised
patients, leading to primary miliary or disseminated
tuberculosis. In developed countries, adults particu-
larly those co-infected with HIV may develop primary
miliary/disseminated tuberculosis via hematogenous
transmission of the infection.
Miliary/disseminated tuberculosis may occur
in post-primary disease, either due to endogenous
reactivation or exogenous re-infection (Auerbach
1944,1959). Hematogenous spread of the bacilli from
a focus or foci located in the lung or extrapulmonary
sites is the source of infection. Transmission of the
organisms via the lymphatics to the blood circula-
tion or direct invasion of the blood vessels by lique-
fied caseous material may occur due to a rupture or
erosion of a contiguous focus or foci (Geppert and
Leff 1979; Yu et al. 1986; Murray et al. 1978; Prout et
al. 1980). Slavin et al. (1980) used the term late gen-
eralized tuberculosis (LGT) to describe generalized
hematogenous spread occurring at a time far remote
from the post-primary period. If tuberculous bactere-
mia is slight, a discrete type of generalization is usu-
ally without immediate clinical significance, although
these 'seed beds' may serve later in the development of
organ tuberculosis and LGT. However, if the general-
ization of the tuberculous bacteremia is massive, nec-
rotizing tubercles are formed in various organs, and
this is known as acute miliary tuberculosis. In post-
primary disease, it is often believed that a single focus
at the pulmonary or extrapulmonary site is consid-
ered the source of hematogenous spread of the bacilli.
Clinical and pathological studies reported by several
authors indicated that more than one focus located at
different unrelated anatomical sites in various body
organs may be reactivated simultaneously, discharg-
276
ing an enormous quantity of mycobacterial inoculum
into the bloodstream. Slavin et al. (1980) stated that
'Late Generalized Tuberculosis (LGT) derived from
a single extrapulmonary focus was very infrequent
(7%); whereas in more than half of the cases, the
origin of dissemination appears to have been derived
from large foci present in a variety of organs'. This
may occur intermittently, episodically, or continu-
ously in 54% of cases (Pagel et al. 1964; Saye 1936;
Slavin et al. 1980; Jacques and Sloan 1970). Evidence
of intermittent or episodic hematogenous spread of
infection was noted at autopsy. Soft exudative tuber-
culous lesions were mixed with other lesions showing
partial or complete healing with fibrosis, calcifica-
tions, and ossifications (Slavin et al. 1980; Saye 1936;
Jacques and Sloan 1970). It has been debated by some
authors whether the lung is the main source of focal
active caseous spread via the hematogenous route to
other body organs; they believe that an extrapulmo-
nary focus or foci are the most common source of
hematogenous spread of infection. Slavin et al. (1980)
reported that 'in the antibiotic era, chronic pulmonary
tuberculosis and late generalized tuberculosis (LGT)
coexisted far less commonly than they did prior to the
advent of chemotherapy'. They reviewed the clinical
and autopsy findings of 100 patients with LGT in two
eras (pre-antibiotic and antibiotic periods) and con-
cluded,'thus, both clinical and autopsy findings in the
antibiotic period support Biehl's statement - that only
a minority of cases of miliary tuberculosis arise from
patients with pulmonary tuberculosis'.
Hematogenous release of large quantities of
necrotic caseous material of uniform size spread via
the blood will cause embolization of the capillaries
of most organs, particularly those with the richest
vascular supply such as the liver, spleen, bone marrow,
brain, and lungs. It is generally fatal if undetected and
not treated early. At the other end of the spectrum
of the hematogenous bacteremia, few bacilli may
be discretely seeded into these body organs without
immediate clinical manifestation. At a later stage,
these hematogenous 'seed beds' may be reactivated
in an organ or in all organs, with the development of
miliary tuberculosis. They may also remain nonpro-
gressive and well localized and cause few or no clini-
cal symptoms. The course of the disease may take the
form of 'chronic miliary tuberculosis', which is slowly
progressive and not fatal (Pagel et al. 1964; Saye 1936).
Fatal cases due to late generalized miliary tubercu-
losis do occur, but some clinicians consider these as
representative of the terminal event of long-standing
pulmonary or extrapulmonary disease (Yu et al.1986;
Slavin et al. 1980). Factors determining the pattern of
M. M. Madkour
the disease presentation will depend on the size of the
bacilli inoculum, its virulence, and the host immune
response. In its severe acute fulminating form, septi-
cemia may occur with extensive tissue damage and
necrosis with very little evidence of a cell-mediated
immune response. The diagnosis of this form is mostly
missed in life because of the absence of the typical
miliary lung shadowing seen on chest radiography,
and the diagnosis is only made at autopsy (Chapman
and Whorton 1946; Biehl 1958; Treip and Meyers 1959;
Proudfoot et al.1969; Munt 1972; Prout et al.1980).
Proudfoot et al. (1969) described this form of
the disease as 'cryptic disseminated tuberculosis',
Patients with cryptic disseminated tuberculosis may
have a co-existing underlying disease that may pres-
ent with atypical features and also cause impairment
of the cell-mediated immunity such as alcoholism,
cirrhosis, malnutrition, diabetes mellitus, connec-
tive tissue diseases, HIV, or may be taking immu-
nosuppressive chemotherapy, and the diagnosis is
then missed in life, with an absence of miliary lung
shadowing, and is usually made at autopsy (Millar
and Horn 1979; Bobrowitz 1982; Yu et al. 1986;
Crump 1998). Patients in the late stage of AIDS with
disseminated tuberculosis have a very low count of
CD4+ T-cells and an anergic response with scanty or
absent granuloma formation as reported by Zumla
and Grange (see Chap. 29). In immunocompetent
patients, the cell-mediated immune response (CM!)
and tissue-damaging delayed hypersensitivity (DTH)
responses will be activated. The bacilli cell wall anti-
gens will activate the macrophages, CD4+ T cell,
increased production of cytokines, TNF-a, IFN -y,
IL-12 to combat the infection. The bacilli may survive,
and during these responses concomitant immune
depression and immune activation occur at the same
time (Dannenbergh 2001; Vanham et al. 1997; Ellner
1997). The infection may overcome the immune
system responses, and the disease progresses. At sites
of the capillary bed seeded with bacilli, granuloma
formation with central caseation and necrosis may
occur simultaneously in all affected body organs.
Adult respiratory distress syndrome (ARDS) is
known as a complication of miliary/disseminated
tuberculosis; it requires mechanical ventilation and
has a high mortality rate ranging from 40% to 60%
(Mohan et al. 1996; Dyer and Potgieter 1984; Murray
et al. 1978; Dyer et al. 1985; Dee et al. 1980). The exact
pathogenetic mechanisms by which miliary tuberculo-
sis causes ARDS is not clearly understood. Penner et aI.
(1995) suggested that the immunological aspects and
pathogenesis of ARDS caused by miliary tuberculosis
are similar to those caused by Gram-negative sepsis. In
MiliarylDisseminated Tuberculosis 277

miliary tuberculosis, a massive release of mycobacteria including fever, sweating, loss of appetite, weight
lipoarabino-mannan cell wall antigens causes wide- loss, and weakness. Respiratory symptoms may
spread perifocal inflammatory responses, injury to the include dry or productive cough, hemoptysis, short-
alveolocapillary membrane, interstitial granulomatous ness of breath, and pleuritic chest pain. Neurological
infiltration, obliterative endarteritis, increased capillary symptoms such as headaches, confusion, disorienta-
endothelial cell susceptibility to the toxic effect of the tion, cranial nerve palsies, or even coma may be the
released TNF-a and ICAM-l (Penner et al.1995; Sutton presenting features. A history of tuberculosis may be
et al. 1974; Petty and Ashburgh 1971; Massaro and Katz present in 10%-20% of patients (Slavin et al. 1980).
1964; Ashburgh et al.1967; McClement et al.1951). Fever is the most common symptom and may be
The alveoli will be filled with inflammatory exu- the only presenting feature (pyrexia of unknown
dative edema due to increased local vascularity, with origin). Other symptoms of tuberculosis related to
vasculitis leading to pulmonary edema which is the other body systems or organs such as musculosk-
hallmark of ARDS (So and Yu 1981; Penner et al. eletal, genitourinary tract, or the abdomen are less
1995; Dee et al. 1980). The alveolar exudate mate- frequently encountered at the time of presentation.
rial includes mostly polymorphonuclear leukocytes, Concomitant chronic illnesses such as diabetes mel-
erythrocytes, fibrin, as well as caseating granulomas. litus, alcoholism, connective tissue diseases, chronic
This will lead to mismatching of ventilation and per- renal failure, silicosis, HIV, or immunosuppressive
fusion and an impaired diffusion capacity (Mohan et chemotherapy may be found in patients with miliary/
al.1996; William and Yoo 1973). disseminated tuberculosis. The physical signs may be
Acute miliary/disseminated tuberculosis may few, including tachypnea, wasting, lymphadenopathy,
also be caused by Mycobacterium bovis. Miliary M. hepatic and splenic enlargement, cutaneous lesions,
bovis induced by intravesical BCG immunotherapy or features of meningitis (Ray et al. 2002; Sharma
used for the treatment of bladder cancer has been et al. 1981; Pasculle et al. 1980; Bateman et al. 1980;
reported (Deresiewicz et al. 1990; Rawls et al. 1990; K6ylii et al.1997; Shibolet et al.1979; Prout et al.1980;
Kesten et al. 1990; Gupta et al. 1988; McParland et al. Evans et al. 1998; Long et al. 1997). The frequency of
1992; Iantorno et al. 1998; Palayew et al. 1993; Foster symptoms and signs of miliary/disseminated tuber-
1997; Jasmer et al. 1996; Rabe et al. 1999; Soloway culosis as reported by some authors are shown in
1988). The mechanism by which BCG induces mili- Tables 18.1 and 18.2.
ary tuberculosis has been debated among authors
(Lotte et al. 1984; de Hertogh et al. 1989; Orihvela et
al. 1987). Some authors suggested that it is a hyper-
sensitivity reaction in response to BCG, particularly 18.5
when they failed to grow the organisms. Other Meningitis and Tuberculomas
authors explained the widespread lung shadowing in Miliary/Disseminated Tuberculosis
as due to hematogenous dissemination of the bacilli
which were isolated from various pathological fluid Miliary/disseminated tuberculosis may involve the
and tissue biopsy specimens. Other nontuberculous central nervous system with evidence of meningitis
Mycobacterium species may cause miliary and dis- or tuberculomas in up to 30% of patients (Slavin et
seminated diseases similar to those caused by M. al.1980; Biehl 1958; Maartens et al.1990). Headaches,
tuberculosis. These nontuberculous mycobacteria are hemiplegia, seizures, and coma may indicate involve-
particularly found in patients with HIV. Mycobacte- ment by meningitis and/or tuberculoma, as noted in
rium avium-intracellulare, M. kansasii, and M. fortui- our case no. 3 (Fig. 18.3f,g). However, the patients may
tum are the ones most frequently reported (Bone and have no symptoms that are related to the central ner-
Stableforth 1981; Longdale et al. 1992). vous system, and cerebral tuberculoma may only be
depicted by MRI. The cerebrospinal fluid (CSF) may
be either clear or turbid with lymphocytic pheocy-
tosis, raised protein concentration, with a low level of
18.4 glucose (Gelb et al. 1973). Cerebrospinal fluid smear
Clinical Features of Miliary/Disseminated is rarely positive for acid-fast bacilli, and culture may
Tuberculosis yield the bacilli in up to 20% of patients (Hill 1991).
The prognosis may be poor if undiagnosed and not
The clinical symptoms at the time of presentation of treated early, with the development of hydrocephalus,
miliary/disseminated TB may only be constitutional coma, and death.
Table 18.1. Most common symptoms of miliary/disseminated tuberculosis
I~
Reference No. of Fever Cough Expectorant Dyspnea Chest pain Hemoptysis Weight loss Headaches Others
cases (%) (%) (%) (%) (%) (%) (%) (%)

Al-Jahdali et al. (2000) 47 72 60 30 34 - - 30 - Fatigue - 49

Sweating - 45
Long et al. (1997) 54 82 59 26 30 11 9 62 - Sweating - 37
Weakness - 70
Maartens et al. (1990) 109 92 72 72 72 72 - 92 25 Abdominal pain - 21
Musculoskeletal - 13
Bobrowitz (1982) 21 95 43 19 33 - 14 33 - Disorientation and confusion - 52
Stenius-Aarniala (1979) 26 88 15 - 31 - - 38 12 Abdominal pain - 35
Fatigue - 35
Nausea & vomiting - 23
Prout et al. (1980) 59 44 63 37 37 15 8.5 59 17 Anorexia - 51
Weakness - 42
Abdominal pain - 30
Confusion - 17
Munt (1972) 67 84 65 - 1 - 6 95 10 Abdominal pain - 7
Weakness - 93
Anorexia - 91
Biehl (1958) 62 62 63 - 23 - 6 61 18 Weakness - 65
Anorexia - 42
Abdominal pain - 10
Chapman and Whorton (1946) 63 79 82 - 64 - 15 85 7 Anorexia - 90

Table 18.2. Most common signs of miliary/disseminated tuberculosis

Reference No. of Fever Tachypnea Wasting Lympha- Chest signs Hepato- Spleno- Ascites CNS Others
cases denopathy megaly megaly
Al-Jahdali et al. (2000) 47 72 - - 4 - 36 13
Long et al. (1997) 54 82 - 62 20
Maartens et al. (1990) 109 96 - - 21 72 52 15 38 20
Prout et al. (1980) 62 44 33 - 28 53 62 12 16 - Epididymo-orchitis, subcutaneous
abscess arthritis 13 ~
Pleural effusion 8 ~
Munt (1972) 69 84 - 67 30 46 36 13 - Neck stiffness 26 Pericardial effusion 8 s::
ll>
Choroidal TB 7
Biehl (1958) 63 62 60 62 - 57 35 12 - - - ~
l:
...
MiliarylDisseminated Tuberculosis
18.5.1
Case Illustration 1
A 39-year-old male clerk working in our hospital pre-
sented to me on 8 September 1999 with a I-month his-
tory of right knee pain which was mild and associated
with swelling. Symptoms were progressive. He was in
good health before presentation and denied a history of
trauma. He was always overweight and did not notice
any recent weight loss, and had no fever or sweating.
Musculoskeletal system investigation was otherwise
normal. He was using analgesics as self-medication,
which were controlling his knee pain. There was no
low-back pain or other systemic symptoms that could
be related to connective tissue diseases. He was treated
in the ophthalmology clinic for dryness of the eyes but
that lasted only for a few months before resolving. He
denied a history of drinking raw milk or consumption
of raw liver or meat or contact with animals. Other
systemic enquiries were normal. On physical exami-
nation he looked well, apyrexial, not in pain and not
pale. The right knee was swollen with effusion but not
tender or warm and had a full range of movement.
There was no thigh or leg muscle wasting. Other joint
examinations were normal. Other systematic physical
examination showed no lymphadenopathy, and other
systems were normal. Synovial fluid aspiration was
obtained, and other investigations included hemo-
gram, ESR, biochemical profiles including serum
urate, blood culture, and Brucella agglutinins, plain
radiography of the knees and chest. He was referred to
our orthopedic surgeon for arthroscopy, but he refused
this investigation. The hemoglobin was 13.2 gldl with
normal white cell count, but the ESR was raised
50 mm/h. Biochemical parameters were normal apart
from slightly raised serum urate at 467 f.lmolll (normal
range 210-430 f.lmolll). Chest and knee radiography
results were normal (Fig. 18.1a,b). Synovial fluid white
cell count was normal, and culture was negative. Plain
radiography and MRI of the right knee showed two
small defects in the joint's surface on the medial
condyle of the femur, one measured 7 mm and the
second 3-4 mm in diameter. This was interpreted as
osteonecrosis but could also be due to an inflamma-
tory process (Fig. 18.1c-f, j). Arthroscopy was refused
again by the patient. He was taking Diclofenac retard
100 mg once daily orally and was controlling his mild
knee pain, and the swelling resolved. Four months later
he presented to the emergency room with a 1O-day his-
tory of fever, rigor, night sweating, and 4-day history
of cough with whitish sputum, left pleuritic chest pain,
and shortness of breath. He also had a history of diar-
rhea for 3 days before presentation. He was anorexic
279
with occasional vomiting, and although he was on a
weight-reducing diet, he did not notice faster weight
reduction. On physical examination in the emergency
room, he looked unwell, tachypnea, feverish 40°C,
pale, and had lost 5 kg of weight since last weighed
4 months earlier. There was no lymphadenopathy, no
positive abnormal sounds heard over the lung, and
no pleural rub. Abdominal examination revealed an
enlarged, tender liver extending 4 cm below the costal
margin and enlarged spleen extending 3 cm below
the costal margin, but there was no ascites or other
palpable masses. The right knee was swollen with mild
knee effusion. Chest radiograph looked normal, but
the hemoglobin decreased to ILl gldl with normal
blood film, and there was evidence of malaria, while
the white cell count was normal. Blood culture was
done, and the blood biochemical parameters were
normal. When I reviewed the patient the following
day, a softer chest radiograph was done which showed
faint reticular lung shadowing at both lung bases
(Fig. 18.lg). Miliary tuberculosis was the presumptive
admission diagnosis, and he was started on four-drug
chemotherapy. Other investigations were carried out
during admission. The sedimentation rate (ESR) was
90 mm/h, negative blood and sputum cultures. Bru-
cella serology was repeated, autoimmune screenings
were all negative. Liver enzymes were abnormal; ALT
79 U/l (normal range 2-40 U/I); alkaline phosphatase
1151 U/I (normal range 98-279 U/I), albumin 33 gil
(normal range 38-51 gil), urate 467 f.lmolll (normal
range 210-430 Ilmolll), and total bilirubin was
normal 12 f.lmolll (normal range 2-22 f.lillolll). The
tuberculin test was negative; it had been done twice,
once at the initial visit 4 months earlier and during
the current admission. MRI of the chest showed dif-
fuse confluent miliary lung shadowing widely spread
over both lung fields with small areas of nonhomo-
geneous consolidation in the lateral segment of the
right middle lobe (Fig. 18.1h,i). Ultrasound of the
right knee showed significant synovial thickening but
no effusion. Ultrasound of the abdomen showed an
enlarged spleen with a span of 16 cm but without focal
lesions in it; the liver was enlarged, but no focal lesions
were depicted; and there was no definite para-aortic
lymphadenopathy. The other abdominal organs were
normal. The nonsteroidal anti-inflammatory medica-
tion was discontinued, and the patient remained on
antituberculous medication. Bronchoscopy was done
which showed an inflamed bronchial mucosa but no
endobronchial lesions, and bronchoalveolar lavage
(BAL) was performed. BAL was sent for cytology,
AFB smear, and culture - post-bronchoscopy sputum
specimens were sent for microbiological examination.
 280 M. M. Madkour

a b

c [>
Fig. 18.1 a Frontal radiograph of the right knee at the initial presentation with monoarthritis. This was interpreted as showing
no bony abnormalities. b Chest radiograph at the initial presentation with right knee monoarthritis, showing no abnormal
imaging features. c, d MRI of the right knee: sagittal Tl-weighted images demonstrate presence of effusion (arrow) and two
osteochondral lesions (arrowheads) at the medial femoral condyle appearing as central areas of increased signal intensity sur-
rounded by signal void rim. Appearances are typical of osteochondritis dissecans, but an infective pathology is also possible.
e, f MRI of the right knee: sagittal T2-weighted images at the same level demonstrates joint effusion better (arrow) and the
osteochondral defect as high signal intensity indicating activity (arrowheads). g Chest radiograph during acute presentation
demonstrates numerous fine discrete nodules bilaterally (arrowheads) consistent with miliary tuberculosis. h Axial CT scan
with lung windowing demonstrates generalized increase in parenchymal density with ill-defined fine nodular pattern in a
random distribution (arrowheads). i Axial CT scan at a caudal level demonstrates air-space consolidation in the left lower lobe
(arrowheads). j Frontal radiograph of the right knee during the acute miliary tuberculosis presentation demonstrates evidence
of periarticular osteoporosis (arrowheads) and subchondral erosions (arrows). Note preservation of joint space (*). k Isotope
bone scan (99mTc) demonstrates marked tracer uptake in the right knee (arrowheads). I Isotope bone scan demonstrates tracer
uptake in 010 vertebral body (arrowheads). m Axial CT of the lung with mediastinal windowing demonstrates destruction of
the lower dorsal vertebral body with associated paravertebral abscess (arrows). n Sagittal gadolinium-enhanced Tl-weighted
MR image of the spine demonstrates destruction of the vertebral body (arrows), paraspinal abscess (large arrowheads) with sub-
ligamentous spread (small arrowheads) causing compression of the spinal cord (*).0 Axial gadolinium-enhanced Tl-weighted
MR image with fat saturation demonstrates paraspinal abscess with wall enhancement (arrows)
 Miliary/Disseminated Tuberculosis 281

e f

[>
 282 M. M. Madkour

m n

o Fig. 18.1 (Continued) k-o

MiliarylDisseminated Tuberculosis 283

The patient agreed to arthroscopy, which showed a cough, expectoration, loss of appetite, and weight
hemorrhagic hypertrophic synoviurn with grade II loss. The swelling in the forehead was painless, and
degenerative changes at the medial femoral condyle. she denied any history of head trauma. She had low
Debridement, synovial tissue biopsies, and wash-out back pain and amenorrhea for the same duration. She
were sent for histopathological and microbiological was in perfect health before the present illness. Other
examination. Bone marrow biopsy was performed. systemic reviews were noncontributory, but she gave
The patient responded to the antituberculous treat- a history of contact with a housemaid who had pul-
ment, and the fever and dyspnea began to improve monary tuberculosis 6 years before the onset of the
within 1 week of hospitalization. However, he com- current illness. She had no family history of tubercu-
plained of back pain, and there was tenderness over losis. She was a high school student and single. She
T9-1O. He later admitted that he had had occasional was initially diagnosed to have tuberculosis a few
back pain before, but it was so mild that it was relieved days before her presentation to our hospital.
with the nonsteroidal anti-inflammatory treatment he She initially presented to our satellite military hos-
used for his knee. Isotope bone scan and MRI of the pital in Al-Kharj (80 km south of Riyadh) and was
spine and other joints were performed. MRI showed diagnosed as having tuberculosis and was referred to
destruction of T9 with collapse of the vertebral body, our hospital a few days after initiating her on antitu-
gibbus deformity, and para-spinal abscess with exten- berculous antibiotics. On physical examination, the
sion into the epidural space (Fig. 18.1m-o). Isotope patient was unwell, pale, underweight (38 kg) with
bone scan showed increased uptake at T9-1O and in forehead cystic swelling measuring 3.0x2.0 cm in
the right knee (Fig. 18.1k,1). Histopathological results diameter. It was not tender or hot or red, and there
of the bone marrow showed multiple areas of granulo- was no discharge from it. She had fever with a tem-
matous lesions with central caseation, but bacilli were perature of 39SC and tachycardia of llO/min. She
not seen, and the culture was later reported as negative. had no peripheral lymphadenopathy. Cardiovascular,
The histopathological findings of the synovial tissue respiratory, abdominal, and central nervous system
and debridement specimen showed multiple granulo- examinations showed no abnormalities. Joint exami-
matous synovitis with central caseation. BAL culture nations were normal; she had localized tenderness
and post-bronchoscopy sputum grew M. tuberculosis on percussion over multiple areas of the spine, but
that was sensitive to isoniazid, rifampicin, pyrazin- there were no deformities or kyphosis. The hemo-
amide, and ethambutol. The temperature settled after gram showed WBC 5.3X109, hemoglobin 8.1 g/dl, and
9 days of antituberculous medication. The difficulties ESR 90 mm/h. Clinical chemistry parameters were
in the diagnosis of this patient were due to several normal. Chest radiograph showed normal cardiac
factors: the mild nature of the right knee symptoms size and clear lung fields. There were two ill-defined,
with absence of constitutional symptoms, the refusal soft-tissue mediastinal masses on the right lateral
of exploratory arthroscopy as an important diagnos- border and upper left border of the cardiac shadow
tic means of investigation, the use of nonsteroidal with no calcification. Frontal and lateral dorsolumbar
anti-inflammatory medication that masked back pain spine radiographs showed loss of height of L2 and
despite the enquiry into spinal symptoms, the initially reduced bone density ofL! and L3 (Fig. 18.2a). Sagit-
normal-appearing chest radiograph was done in the tal CT showed destruction of L2 (Fig. 18.2b). MRI of
emergency room when he presented with acute chest the lumbar spine showed destruction of the body of
and constitutional symptoms, the multiplicity of skel- L2 with compression of the dural space by a para-
etal foci that were the most likely source of hematog- spinal abscess (Fig. 18.2c-e). Postcontrast CT of the
enous dissemination rather than a single focus. The abdomen showed anterior subligamentous abscess
patient responded well and continued antituberculous with a large right psoas muscle abscess (Fig. 18.2f,g).
medication for 12 months. He has been followed up MRI of the abdomen at the same level showed a
for over 2 years with no evidence of recurrence or any large right psoas abscess displacing the right kidney
neurological or other complications. (Fig. 18.2h). Drainage of the right psoas abscess was
performed in the prone position (Fig. 18.2i).
The dorsal spine MRI showed another lesion
18.5.2 affecting the mid-dorsal region (Fig. 18.2j,k). Chest
Case Illustration 2 CT with contrast of lung and mediastinal windows
depicted a small lung parenchymal granuloma
A 16-year-old Kuwaiti girl presented on 29 July 2000 and caseating lymph nodes in the retrosternal
with a 6-month history of forehead swelling, fever, carina, subcarinal region, and prevascular groups
 284 M. M. Madkour

a b

c d[>
Fig. IS.2a-s. A 16-year-old girl presented with painless, soft-tissue swelling of the forehead, fever, cough, and low-back pain.
a Frontal and lateral radiographs of the dorsolumbar spine demonstrate loss of height of L2 vertebral body (arrow) due to
destructive lesion demonstrated better on the lateral view (arrowheads). There is also reduced bone density of the vertebral
body of L1 and L3 and narrowing of disc spaces (D). b Sagittal reformatted CT image demonstrates destruction of the body
of L2 with sclerosis of the remaining part (arrowhead). c MRI of the lumbar spine: sagittal post-enhanced Tl-weighted image
demonstrates destruction of the body of L2 with large tuberculous cavity (arrow). Note posterior comparison on the dural
space (open arrow). d Right para-sagittal image demonstrates anterior subligamentous tuberculous abscess (arrows) which had
a tract to form right paraspinal abscess (A) in a. e Sagittal T2-weighted image with fat suppression demonstrates the paraspinal
abscess as high signal intensity mass (A), the subligamentous abscess (long arrows), diffused increased signal intensity of the
body of L2 and to lesser extent the body of L1 (open arrowheads). Note abscess cavity in L2 (arrow). f Post-enhanced CT of
the abdomen at the level of L2 demonstrates destruction of the vertebral body (black arrow), anterior subligamentous abscess
(arrowhead), and a large right psoas muscle abscess. g Post-enhanced axial CT at a caudal level demonstrates further destruction
of L2 vertebral body (black arrow), tracking of the anterior subligamentous abscess (open black arrow), into a large right psoas
muscle abscess (A). h MRI of the abdomen at same level as f: post-enhanced Tl-weighted image demonstrates destruction of
 MiliarylDisseminated Tuberculosis 285

e f

g h l>
the body of L2 (black arrow) and a large psoas muscle abscess (A), note the enhancing thick wall of the abscess (small arrows).
The abscess is displacing the right kidney (thick black arrow). i Axial CT of the abdomen demonstrates percutaneous insertion
of drainage catheter at the site of psoas abscess (arrow) (the examination is performed in the prone position). j Sagittal MRI
of the dorsal spine: T2-weighted image with fat suppression demonstrates large anterolateral abscess appearing as diffuse high
signal intensity anterior and to the right of mid-dorsal region (arrows). No associated bone or disc lesion is seen. k, I MRI of the
same patient. Post-enhanced II-weighted sagittal (k) and axial (I) images demonstrate the abscess (A) as low signal intensity
surrounded by thick enhancing wall (arrows). m, n CT of the chest: post-enhanced CT at the level of aortic arch. m Lung window
demonstrates narrow lung parenchyma separate from small granuloma (arrow). n Mediastinal window demonstrates caseat-
ing lymph node with characteristic ring enhancement in the retrosternal and prevascular group (arrows). 0, p Post-enhanced
axial CT at the level above the carina (0) and subcarinal (p) demonstrates appearance characteristic of subcarinal region. Note
ring enhancement of the wall (arrow). q, r Frontal and lateral skull radiographs demonstrate focal area of bone destruction in
the frontal bone. Note loss of the outer table of the skull bone on the lateral radiograph (white arrows) and focal ill-defined
radiolucency due to bone destruction (black arrows). s Axial CT of the brain demonstrates bone destruction of the frontal bone
(curved arrow) and associated soft-tissue mass due to abscess formation (white arrow)
286 M. M. Madkour

Fig. 18.2 (Continued) i-m m[>

 MiliarylDisseminated Tuberculosis 287

n o

p q

r Fig. 18.2 (Continued) n-s

288 M. M. Madkour

(Fig. 18.2m-p). Plain radiograph of the skull depicted 18.5.3

a focal area of bone destruction in the frontal bone
(Fig. 18.2q,r). Axial CT of the brain showed a soft-
tissue abscess associated with frontal bone destruc-
tion (Fig. 18.2s). Ultrasound of the abdomen showed
no evidence of hepatic or splenic involvement, and
CT showed no intra-abdominal free fluid or other
abdominal organ involvement.
Culture of psoas abscess drainage yielded M. tuber-
culosis after 14 days incubation and was sensitive to
all antituberculous agents. The forehead abscess was
drained by our neurosurgeon, and tissue biopsies
were obtained. Histopathological examination of
the forehead abscess showed caseating granulomas,
and culture yielded M. tuberculosis. The patient was
treated with isoniazid 300 mg, rifampicin 450 mg,
pyrazinamide IG, ethambutol 800 mg, and pyridoxine
25 mg as single oral daily dosing for 2 months and
continued on isoniazid and rifampicin for a further
10 months. She responded well to treatment while
in hospital, and her temperature settled with weight
gain. She remained in hospital for 2 months and was
discharged in September 2000. She stayed in Riyadh
with members of her family and was followed up in
the outpatient clinic regularly. She made a full recovery
with no sequelae. This patient had disseminated tuber-
culosis with spinal disease at multiple levels, mediasti-
nallymph node abscesses, psoas abscess, and forehead
abscess. Although there was no lung parenchymal
active disease depicted on plain chest radiography,
CT showed small parenchymal lung granuloma. This
patient did not have HIV or other diseases and was not
on any medication that could have contributed to such
severe disseminated tuberculosis.
Case Illustration 3
A 27-year-old man transferred to our infectious
disease unit isolation room with advanced AIDS,
multidrug-resistant pulmonary tuberculosis, left
hemiplegia due to tuberculous abscess of the brain,
and recurrent seizures. This patient was diagnosed
in another hospital in Riyadh with pulmonary tuber-
culosis and HIV co-infection in 1997. His pulmonary
tuberculosis relapsed in 1999 due to poor compliance
with medication. He was moved from one hospital
to another and finally came to our hospital on 29
April 2001 until his death on 3 November 2001 due
to disseminated multidrug-resistant tuberculosis.
On arrival at our hospital, he was cachectic, pale,
dyspneic, had oral thrush, and looked very ill. The
hemogram showed low WBC 2.8109, and CD4
count ranged between 14 and 30/111, hemoglobin was
8.2 g/dl, and ESR was 84 mm/h. The HIV viral load
was 75,000 COP/ml in 1999. Sputum culture yielded
M. tuberculosis resistant to rifampicin and isoniazid.
He was already on AZT, lamuvidine, rifampicin, clar-
ithromycin, co-trimoxazole, fluconazole, phenytoin,
and pyridoxine. Chest radiograph showed paratra-
cheal and hilar lymphadenopathy and fibronodular
infiltrative left upper lobe lesions with thin-walled
cavities (Fig. 18.3a). Axial CT of the chest showed evi-
dence of endobronchial tuberculosis with caseating
hilar lymphadenopathy (Fig. 18.3b-d). In June 2001,
the patient developed abdominal pain, vomiting,
abdominal distension, and constipation for 2 days.
The abdomen was tender all over with absent bowel
sounds. Abdominal radiograph and axial CT showed

Fig. 18.3a. Plain chest radiograph demonstrates paratracheal and

hilar adenopathy (arrows) and fibronodular infiltrative changes in
the left upper lobe (curved arrow) with thin-walled cavities (arrow-
heads). b Axial CT of the chest with lung windowing demonstrates
nodular and branching opacities of endobronchial tuberculosis
(arrowheads) in both upper lobes, dilated bronchi (arrows). c Axial
CT at caudal level, lung windowing demonstrates multiple thin
cavities (arrows). d Axial post-enhanced CT of the chest at the level
of carina demonstrates enlarged lymph nodes with typical ring
enhancement and caseating center (arrows). e Axial post-enhanced
CT of the abdomen demonstrates significantly dilated small-bowel
loops (B) and caseating mesenteric lymph node (curved arrow).
There was no ascites (dry type). At surgery, 20 cc of pus were aspi-
rated from the mesenteric abscess. f MRI of the brain: post-enhanced
TI-weighted image demonstrates tuberculoma and tuberculous
abscess in the left parietal lobe. Note the enhancement of the
tuberculoma (arrowhead) and the irregular enhancing wall of the
tuberculous abscess (arrow). g MRI post-enhanced fast spin-echo
inversion recovery image (FSEIR) at the same level demonstrates
surrounding edema (arrowheads) a l>
 Miliary/Disseminated Tuberculosis 289

dilated small-bowel loops and caseating mesenteric
lymphadenopathy. Exploratory laparotomy after
taking all precautionary measures in the theatre was
performed. A massively distended small bowel was
noted with mild changes in color. There was a large
inflammatory mass in the root of the mesentery of
the small bowel. Thick pus was freely present in the
abdominal cavity, and 20 cc was aspirated and sent
to the Microbiology Department. Bowel adhesions
were noted and released. It was decided to close
with interrupted sutures, and the patient tolerated
the procedure. In October 2001, he went into a coma,

b c

d e

f g
 290 M. M. Madkour

and brain MRI showed enlargement of the tubercu-
loma of the brain with development of left parietal
lobe brain abscess (Fig. 18.3f,g). The patient died
1 month later.
18.5.4
Case Illustration 4
A 21-year-old married female patient presented
to the surgical department with a right iliopsoas
mass. She had a history of abdominal pain last-
ing 6 months associated with fever, profuse night
sweating, loss of appetite, and weight loss. She was
well before that time and had no previous illnesses.
She noted a swelling in the right inguinal region
which was associated with a gradual but progres-
sive pain that increased with flexion of the right
hip and was relieved by extension of the leg. The
pain became worse 2 months before presentation,
and the swelling became larger, and she noted a loss
of 13 kg in weight. She had an associated history of
dysuria and dribbling of urine during the 2 months
before presentation. She had no hematuria but had
loin pain on both sides. Other systematic reviews
were noncontributory. She was using pain-killers
to relieve her pain during that time which had
been provided by a private clinic. She was also
given several courses of antibiotics for a period
of 2-3 months before presentation to our hospital.
There was no past history of any illnesses. She had
married at the age of 18 years and had one healthy
boy. She had a history of vaginal discharge which
was yellowish after IUCD insertion 1 year before
presentation, and it was subsequently removed.
On physical examination, she was underweight,
and blood pressure was 92/66 mmHg. There was
a cystic swelling at the right inguinal area. Chest
examination showed dextrocardia and situs inver-
sus. There were positive clinical findings in the
lungs including crepitations and bronchial breath-
ing in the left upper zone. The heart was otherwise
normal. Abdominal examination revealed soft lax
abdomen with no masses or organomegaly. The
right iliac fossa was tender with vague fullness but
no definite masses felt. She was pale and feverish,
with a temperature of 38.5°C. The patient felt pain
during examinations of her hip joint and resisted
manipulation. Other systemic examinations were
normal. The hemogram showed normal white
cell count, hemoglobin 8.8 gldl, and ESR 90 mml
h. Biochemical parameters were normal. Chest
radiograph showed dextrocardia and pneumonic
consolidation in the left mid-zone (Fig. 18.4a).
Axial CT of the chest showed airspace consolida-
tion with air bronchogram (Fig. 18.4b).
Axial CT of the chest postcontrast showed an
abscess in the left erector spinae muscle (Fig. 18.4c,d).

a
Fig. 18.4a-h. A 21-year-old woman with situs inversus and dextrocardia presented with rigor, fever, night sweating, loss of
weight, and right hip pain. She was found to have a right iliopsoas mass. a Chest radiograph demonstrates pneumonic consolida-
tion in the left mid-zone (arrow) (figure should read situs inversus). b Axial CT of the chest with lung windowing demonstrates
air space consolidation with air bronchogram (arrowheads). There is minimal para-pneumonic effusion (open arrow). c, d Axial
post-enhanced CT of the patient at the thoraco-abdominal region demonstrates abscess cavity in the left erector spinae muscle
(arrows). Note the situs inversus with the liver on the left (arrowhead) and dextrocardia. e, fAxial CT of the pelvis: nonenhanced
(e) and post-contrast-enhanced (0 axial images at the same level demonstrate multiple pelvic abscesses (A). Note enhancement
of the walls of the abscesses (arrows; B). g, h Axial CT at the level of the hip joints: nonenhanced (g) and post-contrast-enhanced
(h) axial images demonstrate right inguinal abscess
 Miliary/Disseminated Tuberculosis 291

Axial CT of the pelvis pre- and post-enhancement
showed large multiple pelvic abscesses with wall
enhancement (Fig. 18.4e,f). Axial CT at the level of
the hip joints showed right inguinal abscess with ring
enhancement after contrast (Fig. 18.4g,h). Percutane-
ous drainage of the right inguinal abscess and other
pelvic abscesses was performed under CT guidance.
Other sputum cultures and cultures of the abscess
contents yielded M. tuberculosis that was sensitive
to all antituberculous drugs. Histopathology of the
inguinal abscess wall showed granulomas. The patient
had negative testing for HIV and no evidence of any
underlying diseases. She responded well to antituber-
culous treatment without any sequelae.

e f

g h
292
18.5.5
Case Illustration 5
A 26-year-old man was referred from Al-Kharj (80 kIn
south of Riyadh) with a 6-month history of fever,
sweating, dry cough, left-sided chest pain, dyspnea on
mild exertion, weakness, loss of appetite, weight loss,
and back pain. He had a strong family history of tuber-
culosis. He was initially treated with antibiotics and
analgesics, but his symptoms gradually progressed. He
was healthy previously and was not on any medication
prior to the presenting illness. Other systemic reviews
were normal. On physical examination, he looked
unwell, pale, underweight, and his temperature was
38.5°C. There was no palpable peripherallymphade-
nopathy. Chest examination revealed features of left-
sided pleural effusion. Cardiovascular system exami-
nation was normal. The abdomen was soft, but there
was a splenomegaly with tenderness on palpation.
There was no other organ enlargement and no ascites.
The spine showed no deformities, but tenderness was
elicited at different areas on percussion, mostly at the
mid-thoracic and lumbar spine. There was no evidence
of neurological deficit, and the fundi were normal.
The hemogram showed a normal white cell
count, hemoglobin 8.3 g/dl, and ESR 94 mm/h. All
biochemical parameters were normal. Chest radio-
graph showed left-sided, encysted pleural effusion,
and a paraspinal shadow was noted, consistent with
a paraspinal abscess (Fig. 18.5a). Plain radiograph
of the thoracolumbar spine depicted a paraspinal
abscess with narrowing of the disc space at TlO-ll
{Fig. 18.5d). Ultrasound of the abdomen showed
a poorly defined, hypoechoic lesion in the spleen
consistent with a splenic abscess (Fig. 18.5j). Axial
CT of the chest showed a large, encysted, left pleural
effusion with pleural thickening {Fig. 18.5b). MRI
of the dorsolumbar spine showed several lesions
affecting the vertebral bodies at multiple levels in
the thoracic and lumbar vertebrae with paraspinal
abscess formation {Fig. 18.5c,f,h,i). Pleural aspira-
tion and biopsy were performed for histopathologi-
cal and microbiological investigations. Pleural biopsy
showed multiple areas of granulomas consistent with
tuberculosis. The patient was seen by the spinal sur-
geon, and surgery was planned after initiating anti-
tuberculous treatment. Paraspinal abscess drainage
and bone fusion for stabilization of the spine were
performed {Fig. 18.5e). Culture of the granulation
tissue and abscess content were positive for M. tuber-
culosis which was sensitive to all antituberculous
drugs. Histopathology of spinal granulation tissue
showed caseating granulomas. Serological testing
M. M. Madkour
for HIV was negative. The patient responded well
to treatment. His antituberculous treatment was
extended to 16 months because of the multiplicity of
spinal lesions. He made a full recovery and had no
spinal deformities or late neurological deficits.
18.6
Adult Respiratory Distress Syndrome
and Disseminated Intravascular
Coagulation in Miliary Tuberculosis
ARDS is an uncommon but serious complication of
miliary tuberculosis with a high hospital mortality
rate of up to 69%, similar to that of other etiologies
such as Gram-negative sepsis (Penner et al. 1995).
It has no pathognomonic clinical or imaging fea-
tures that could differentiate a tuberculous from a
nontuberculous etiology. Goldfine et al. (1969) were
the first to describe respiratory failure in patients
with miliary tuberculosis, and Agarwal et al. (1977)
were the first to use mechanical ventilation for such
patients. The criteria for diagnosing ARDS is based
on the American-European Consensus Conference
(Bernard et al. 1994) which includes:
1. The ratio of partial pressure of arterial oxygen
(Pa02) to the fraction of inspired oxygen (Fi0 2)
is ::;200.
2. Chest radiographic features of bilateral lung infil-
trates.
3. A pulmonary artery occlusion pressure ::;18 mmHg.
4. Or absence of clinical or imaging evidence of
raised left atrial pressure.
The onset of ARDS due to miliary tuberculosis
is often preceded by fever and chest symptoms for
a few weeks, but may occur acutely over a few days
with intense dyspnea (Mohan et al.1996; Murray et al.
1978; So and Yu 1981; Dyer et al. 1985). A high index
of suspicion by the treating clinician and willingness
to consider tuberculosis as a possible cause of ARDS
of unknown origin are essential and may reduce
the mortality. Tachypnea and bilateral inspiratory
crackles over both lungs may be found. Other physi-
cal examination results may detect splenomegaly,
hepatomegaly, and signs of meningitis. Some authors
have suggested that ARDS due to miliary tubercu-
losis may have a longer duration of symptoms than
1 week, while Gram-negative septicemia and viral
pneumonia produce symptoms much faster. Hypo-
tension is uncommon (7%) in ARDS due to miliary
tuberculosis, while it is more frequent in ARDS due
Miliary/Disseminated Tuberculosis 293

to sepsis (Dyer et al. 1985). Disseminated intravascu- TB is not clear, but disseminated emphysematous
lar coagulation (DIC), acute renal failure, hepatitis, tuberculosis, rupture of a pleural bleb, or rupture of
and intercurrent infections are common features a subpleural caseous miliary nodule may be the cause
(up to 87.5%) in ARDS due to miliary tuberculosis (Mert et al. 2001). There are only a few single case
and Gram-negative septicemia (Murray et al. 1978; reports on such complications in the literature. Mert
Piqueras et al. 1987; Dyer et al. 1985; Mohan et al. el al. (2001) found eight cases in the English literature
1996). Death may occur with 1-7 days of the onset between 1974 and 1999 and reported one case of their
of ARDS (Murray et al. 1978), and early diagnosis own. These authors reported their own series of 38
and initiation of antituberculous chemotherapy may patients with miliary tuberculosis, only 1 of whom
reduce the mortality (Mohan et al.1996). had pneumothorax while on antituberculous treat-
The etiological diagnosis of ARDS due to miliary ment. The size of the pneumothorax as noted by other
tuberculosis may be achieved in up to 35% of patients authors ranged from small up to 80%. Pneumothorax
by sputum smear and in 20% by urine sediment smear in these rare cases was mostly unilateral. Graf-Deuel
showing the bacilli. The chest radiograph may be and Knoblauch (1994) described 12 patients with bilat-
strongly suggestive of miliary tuberculosis in 15% eral pneumothorax due to various causes and reviewed
of patients (Dyer et al. 1985). Other investigations the literature and found 56 published cases including
in smear-negative patients should include bron- 3 patients due to miliary tuberculosis. Chandra et al.
choscopy with brushing, transbronchial lung biopsy, (1988) described an 18-year-old woman with miliary
bone marrow trephine, and liver biopsies as well as tuberculosis. She was given antituberculous chemo-
lumbar puncture if features of meningitis are present. therapy, but on the 20th day of admission, she devel-
The hemogram often shows normal white cell count, oped bilateral simultaneous pneumothorax that was
anemia, or pancytopenia. Patients with ARDS due to aspirated by needle and tube drainage and improved.
miliary tuberculosis may develop disseminated intra- Two weeks later, recurrent right-sided pneumothorax
vascular coagulation (Die). The pathogenesis and was followed by left-sided pneumothorax and a bron-
factors contributing to DIC occurrence are not clear. chopleural fistula that required surgery and responded
Tuberculous vasculitis may possibly be a contributing well to treatment. Severe chest pain and breathlessness
factor to the initiation oflocal consumption of clotting may occur at the onset of pneumothorax, even while
factors leading to Die. Subclinical coagulopathy may the patient is on chemotherapy. It may respond to
possibly occur as indicated by the presence of micro- needle aspiration, tube drainage, or may require tho-
thrombi in the biopsy specimen from these patients. racotomy if a bronchopleural fistula develops.
The development of DIC may be noted with clinically
overt bleeding or the results of laboratory tests con-
sistent with its occurrence (Murray et al. 1980; Dyers
et al. 1985; Maartens et al. 1990). The development 18.8
of DIC in such cases is often fatal (Rosenberg and Diagnosis of Miliary/Disseminated
Rumans 1978). In endemic areas or when clinical and Tuberculosis .
imaging features are suggestive of miliary tuberculosis
as a possible cause of ARDS, empirical antituberculous The chest radiograph is the most important initial
treatment should be commenced while the diagnosis diagnostic tool for patients with miliary tuberculosis
is actively pursued. (Berger and Samortin 1970). However, most authors
reported a wide range in the incidence of depicting
miliary lung nodularity: from 30% to 93% of patients.
The diagnostic accuracy of the chest radiograph was
18.7 determined with a high specificity and good interob-
Pneumothorax in Miliary/Disseminated server agreement by Kwong and colleagues (1996)
Tuberculosis from Vancouver, Canada. These authors reviewed all
cases with miliary TB diagnosed over a lO-year period
Pneumothorax as a complication of miliary tuberculo- (1982-1992), with inclusion of a group of control sub-
sis is extremely rare (Mert et al. 2001; Graf-Deuel and jects and interpretation by three independent blinded
Knoblauch 1994; Chandra et al. 1988). Pneumothorax observers. They identified miliary nodular lung shad-
in cavitary pulmonary tuberculosis may occur due to owing with a sensitivity ranging from 59% to 69%,
rupture of the subpleural cavity into the pleural space. higher in those with HIV (71 %-85%), and lower when
The possible mechanism of pneumothorax in miliary the diagnosis was made only at autopsy (30%-60%).
 294 M. M. Madkour

e !:.id.... f [>
 MiliarylDisseminated Tuberculosis 295

g h

Fig.18.5a-j. A 26-year-old male patient presented with chest pain, fever, rigor, abdominal and back pain. a Chest X-ray demon-
strates large, left-sided, encysted, pleural effusion (arrowhead). Note paraspinal shadow consistent with abscess formation (open
arrows). b Axial post-contrast-enhanced CT demonstrates collapsed left upper lobe (black arrowhead) and a large pleural effu-
sion (open arrow). Note thickening of the parietal pleura (small arrows). c MRI of the spine: post-enhanced II-weighted image
demonstrating lesion in the vertebral body of Dll (large white arrow); note dural enhancement (small arrowheads). Note also
enhancement of the thickened visceral and parietal pleura (small white arrows) around the pleural effusion (large arrowhead).
d Frontal X-ray of the dorsolumbar spine demonstrates paraspinal abscess (curved arrow). Narrowing of the disc at DlO-ll
level (open arrows). e Postsurgical frontal radiograph of dorsolumbar spine demonstrates resolution of the paraspinal abscess
(drained surgically). Bone fusion for stabilization has been performed (long arrows). f Sagittal post-enhanced II-weighted
image demonstrates multiple spinal involvement by tuberculous foci (arrows). Note large intraspinal abscess involving two
adjacent vertebral bodies and crossing the intervening disc space (open arrow). Abscess is also tracking anteriorly along the
anterior spinal ligament (long arrows). Note enhancement of the dura (arrowheads). g MRI of the spine after surgery: sagittal
proton-density and T2-weighted image demonstrate marked improvement with resolution of the osteomyelitis and paraspinal
abscess. Bone graft (arrow) has been performed for stabilization. h Sagittal post-enhanced II-weighted image demonstrates
enhancement of the last two lumbar vertebral bodies (arrows) and a small intravertebral abscess (arrowhead). i Sagittal proton
and T2-weighted images at the same level demonstrates the abscess cavity at L5 (arrows). j Same patient, ultrasound examination
of the spleen demonstrates poorly defined, hypoechoic lesion in the spleen consistent with splenic abscess (arrows)
296 M. M. Madkour

The interobserver agreement of the three independent
radiologists was 90%. The chest radiography features
of miliary tuberculosis characteristically consist of
widespread, bilateral, nodular lung opacities in the
upper and lower zones often measuring 1-3 mm in
diameter in 90% of patients and greater than 3 mm
in 10% (Kwong et al. 1996). Areas of consolidation
may be noted in up to 30% of patients (Fig. 18.1i) and
are depicted better by CT of the chest. Parenchymal
lung cavities may be seen in 3%-11.9% of patients
with miliary tuberculosis (Kwong et al. 1996; Long et
al. 1997; Hong et al. 1998). An associated mediastinal
and/or hilar lymphadenopathy may be noted in about
15% of patients (Kwong et al. 1996). Other additional
radiographic findings may include pleural effusion
and calcified granulomas (Fig. 23.14a, b).
ing were found in 44% each. The authors also reported
other HRCT findings including pre-existing TB lesions
in 44%, lymphadenopathy in 32%, pleural effusion in
16%, and bronchogenic spread in 16%.
Diverse diseases may cause a similar miliary lung
shadowing including hematogenous dissemination of
infectious diseases, sarcoidosis, metastatic malignancy,
histoplasmosis, pneumoconiosis, and interstitial fibro-
sis (Curull et al.1985; Willcox et al.1986). Centrilobular
(core) and paraseptal (peripheral) nodules may be
identified with HRCT. Similar additional findings in
chest radiography such as pleural effusion, lung cavi-
ties, and lymphadenopathy may also be noted (Hong et
al. 1998). The imaging features of miliary tuberculosis
are not pathognomonic of the disease (McGuinness et
al.1992; Chugh and Agarwal 1997; Miyake et al. 1997).

18.9 18.10
High Resolution (T Laboratory Diagnosis of Miliary
Tuberculosis
HRCT of the chest findings in patients with miliary
tuberculosis has been reported in only a few articles A rapid and specific laboratory diagnosis is the most
with a relatively small number of patients (Oh et al. pressing need in confirming the diagnosis of miliary
1994; Optican et al. 1992; McGuinness et al.1992). These tuberculosis. The many weeks that are required for
reports suggested that HRCT is more specific and culture yield in patients with miliary tuberculosis
observed the following imaging findings in patients cannot be used as a means of confirming the diag-
with miliary tuberculosis, miliary nodules, ground- nosis before starting chemotherapy. A rapid confir-
glass opacities (GGOs), reticulation, and interlobular matory laboratory investigation may not always be
septal thickening. Oh et al. (1994) described the GGOs easy to achieve. A chest radiograph showing miliary
as multiple microscopic granulomas with acid-fast nodular lung shadowing in the presence of clinical
bacilli after obtaining transbronchiallung biopsies of
these areas. The GGOs appear as random patchy opaci-
Table 18.3. Rapid diagnosis by Ziehl-Neelsen stain: number of
ties and are the second most common (miliary nodules
patients and percentage of positive AFB
being the first) HRCT feature of miliary tuberculosis.
A large retrospective study on HRCT findings of Body fluid Prout Maartens
et aI. (1980) et aI. (1990)
miliary tuberculosis was reported by Hong et al. (1998)
from Seoul, Korea. They reviewed 25 patients with Sputum 39 (31%) 64 (33%)
microbiologically andlor histopathologically proven Urine 17 (18%) 22 (14%)
CSF 31 (3%) 24 (8%)
miliary tuberculosis who had undergone HRCT. Mili-
Bronchial lavage 3 (100%) 51 (27%)
ary nodules that are uniformly distributed throughout
both lung fields, mostly ranging from 1-3 mm in
diameter and sharply defined, were the most common
findings in 24 of the 25 patients (96%). Larger nodules
Table 18.4. Rapid diagnosis by tissue biopsies: number of
up to 5 mm in diameter due to coalescence or enlarge-
patients and percentage of positive granulomata
ment of granulomas were noted in 5%-10% of all nod-
ules. GGOs occurred in 23 (92%) of these patients and Biopsy site Al-Jahdali Maartens Prout
et al. (2000) et aI. (1990) et al. (1980)
are considered the second most common CT finding.
These authors observed that in two of their patients Liver 16 (88%) 11 (100%) 12 (92%)
with severe dyspnea and impending ARDS due to Bone marrow 11 (73%) 22 (82%) 15 (71%)
TransbronchiaI 10 (70%) 48 (63%) 2 (100%)
miliary TB, extensive GGOs were depicted by HRCT.
Lymph nodes 2 (100%) 9 (100%)
Interlobar reticulation and interlobar septal thicken-
Miliary/Disseminated Tuberculosis
features that suggest miliary tuberculosis may be
the only available clues for the attending clinician.
Maartens et al. (1990) reported on 109 patients with
miliary tuberculosis, and in 7 patients, the diagnosis
was based on clinical grounds and classical miliary
nodules on chest radiography; they were treated with
chemotherapy with a good response.
Rapid confirmation of the diagnosis of miliary
tuberculosis may be achieved in up to 83% of patients
by means of ZieW-Neelsen stain and by the finding
of granulomata in tissue biopsies or by fine-needle
aspiration (AI BWa12001; Maartens et al. 1990; Prout
et al. 1980; Al-Jahdali et al. 2000). The most common
body fluids used for Ziehl-Neelsen stain include
sputum, gastric fluid aspirate, urine, CSF, bronchial
lavage and brushings, and pleural fluid (Sharma et
al. 1988) (Table 18.3). Common sites of tissue biop-
sies to detect granulomata in patients with miliary
TB include transbronchial lung tissue, bone marrow
trephine, liver, pleura and lymph nodes (Stallworth et
al. 1980; Steiner et al.1976) (Table 18.4).
Sputum culture in patients with miliary tubercu-
losis may be positive in 30%-60% (Prout et al. 1980;
Monie et al. 1983; Maartens et al. 1990; Al-Jahdali et
al. 2000).
18.10.1
Polymerase Chain Reaction
PCR assays are a rapid, specific, and noninvasive
method of confirming the diagnosis of tuberculosis.
PCR using the IS6110 method can be used in iden-
tifying the DNA fingerprinting of M. tuberculosis in
the sputum, urine, bronchial lavage, gastric aspirate,
blood, and tissue biopsies. It has a specificity of 100%
and sensitivity of 95% with a turnaround time of 24-
36 h (Clarridge et al. 1993; Shalwar et al. 1993; Nolte
et al. 1993). This rapid and specific diagnostic tool is
very useful in confirming the diagnosis of miliary
tuberculosis when it is available. Penner et al. (1995)
used PCR to determine the strain of Mycobacterium.
Tuberculosis was found in 6 of their 13 patients (46%)
with miliary tuberculosis, pulmonary and dissemi-
nated disease with respiratory failure that required
mechanical ventilation. In disseminated tuberculosis,
extrapulmonary systems and organs affected can be
diagnosed with other investigations appropriate to
the site of involvement. These may include various
imaging modalities with guided tissue biopsies, sur- .
gical drainage of abscesses as demonstrated in the
five case illustrations from our own patients. Other
investigations should include serological testing for
297
HIV and PPD skin test. The hemogram may show
pancytopenia or 10wWBC (Cassim et al.1993). Tuber-
culin skin test may be positive in 25%-75% of patients
with miliary/disseminated tuberculosis, as reported in
various series, but is low among those co-infected with
HIV (Hill 1991; Maartens et al. 1990; Gelb et al. 1973;
Grieco and Chmel1974; Munt 1972).
18.11
Treatment of Miliary/Disseminated
Tuberculosis
The treatment of miliary/disseminated tuberculosis is
similar to that used for pulmonary and extrapulmo-
nary disease. In miliary tuberculosis, the four-drug
regimen includes isoniazid, rifampicin, pyrazinamide,
and ethambutol which is used for 2 months and is
followed by isoniazid and rifampicin for a further
4-6 months. Drug susceptibility and resistance should
be determined. In patients co-infected with HIY, the
duration of treatment should be extended to at least
9 months. In disseminated tuberculosis, the duration
of treatment may be extended to at least 12 months,
particularly when the bone is involved. Surgical inter-
vention may be required to stabilize the spine, release
abdominal adhesions if intestinal obstruction arises,
drain abdominal, paraspinal, inguinal abscesses, or
treat other pulmonary or extrapulmonary complica-
tion (see case illustrations).
The use of corticosteroids as a modulator of the
inflammatory process in miliary tuberculosis with
ARDS, meningitis, or pericarditis has been reported
by many authors and debated by others (Dyer and
Potgieter 1984; So and Yu 1981). Penner et al. (1995)
used steroids in 8 of their 13 patients (61.5%) with
miliary and pulmonary tuberculosis with ARDS
and respiratory failure; they required mechanical
ventilation. The median time interval between hos-
pitalization and commencement of treatment was
2 days (range 1-36 days) (Maartens et al.1990). Fever
resolved in approximately 7 days (range 1-55 days).
Death occurred in 25% of patients within 1-2 weeks
of receiving treatment. The most important single
factor in the mortality of patients with miliary
tuberculosis is a delay in the diagnosis (Maartens et
al. 1990). Other factors that contribute to mortality
include old age, lymphopenia, hypoalbuminemia,
elevated liver transaminases, and delay in early com-
mencement of treatment. Other underlying chronic
illnesses may contribute to mortality. The develop-
ment of complications such as ARDS, DIC, Addison's
298
disease, acute renal or multiorgan failure also con-
tributes to the raised rate of mortality. The mortal-
ity rate among patients with miliary tuberculosis is
approximately 70%, and 77% among those requiring
mechanical ventilation due to ARDS, DIe leading to
respiratory failure. Mortality rates reported in large
series ranged between 20% and 80% of patients
(Maartens et al. 1990; Prout et al. 1980; AI-Jahdali et
a1.2000; Monie et al.1983; Stenius-Aarniala and Tuki-
ainen 1979; Dahmash et al. 1995; Evans et al. 1998;
Mohan et al.1996; Ormerod and Horsfield 1995).
References
Agarwal MK, Muthuswamy PP et al (1977) Respiratory failure
in pulmonary tuberculosis. Chest 72:605-609
Al Bhlal LA (2001) Fine-needle aspiration cytology of post-
vaccinial disseminated bacillus Calmette-Guerin infection.
Diagn CytopathoI24:333-335
Al-Jahdali H,Al-Zahrani K et al (2000) Clinical aspects of mili-
ary tuberculosis in Saudi adults. Int J Tuberc 4:252-255
Armstrong P, Wilson AG, Dee P et al (1995) Imaging of disease
of the chest, 2nd edn. Mosby Year Book, Chicago, p 183
Ashburg DG, Bigelow DB et al (1967) Acute respiratory distress
in adults. Lancet 2:319
Auerbach 0 (1944) Acute generalized miliary tuberculosis. Am
J Pathol 20: 121-136
Auerbach 0 (1959) The natural history of the tuberculous
pulmonary lesion. Med Clin North Am 43:239-251
Bateman DE, Makepeace W et al (1980) Miliary tuberculosis
in association with chronic cutaneous tuberculosis. Br J
Dermatol103:557-560
Berger HW, Samortin TG (1970) Miliary tuberculosis: diag-
nostic methods with emphasis on the chest roentgeno-
gram. Chest 58:586-589
Bernard GR, Artigas A et al (1994) The American-European
Consensus conference on ARDS: definitions, mechanisms,
relevant outcomes, and clinical trial coordination. Am J
Respir Crit Care Med 149:818-824
BieW JP (1958) Miliary tuberculosis: a review of 68 adult
patients admitted to a municipal general hospital. Am Rev
Tuberc Pul Dis 77:605-622
Bobrowitz ID (1982) Active tuberculosis undiagnosed until
autopsy. Am J Med 72:650-658
Bone M, Stableforth D (1981) Miliary infection to Mycobacte-
rium aYium-intracellulare. Tubercle 62:211-213
Cassim KM, Gathiram Vet al (1993) Pancytopenia associated
with disseminated tuberculosis, reactive histiocytic hae-
mophagocytic syndrome and tuberculous hypersplenism.
Tubercle Lung Dis 74:208-210
Chandra KS, Prasad AS et al (1988) Recurrent pneumothora-
ces in miliary tuberculosis. Trop Geogr Med 40:347-349
Chapman CB, Whorton CM (1946) Acute generalized mili-
ary tuberculosis in adults. A clinicopathological study
based on 63 cases diagnosed at autopsy. N Engl J Med
235:239-248
Chugh 1M, Agarwal AK (1997) Bilateral miliary pattern in
sarcoidosis. Indian J Chest Dis Allied Sci 39:245-249
M. M. Madkour
Clarridge J et al (1993) Large-scale use of polymerase chain
reaction for detection of Mycobacterium tuberculosis in a
routine mycobacteriology laboratory. J Clin Microbiol 31:
2049-2056
Crump JA (1998) Miliary tuberculosis with paradoxical expan-
sion of intracranial tuberculomas complicating human
immunodeficiency virus infection in a patient receiving
highly active antiretroviral therapy. Clin Infect Dis 26:
1008-1009
Curull V, Morell F et al (1985) Dyspnea, fever and miliary pat-
tern. Chest 88:285-286
Dahmash NS, Fayed DF et al (1995) Diagnostic challenge of
tuberculosis of the elderly in hospital: experience at a uni-
versity hospital in Saudi Arabia. J Infect 31:93-97
Dannenberg AM Jr (2001) Pathogenesis of pulmonary Myco-
bacterium boYis infection: basic principles established by
the rabbit model. Tuberculosis 81:87-96
Dee P, Teja K et al (1980) Miliary tuberculosis resulting in
adult respiratory distress syndrome: a surviving case. Am
J Roentgenol134:569-572
De Hertogh Fierer E et al (1989) Hypersensitivity reaction to
bacillus Calmette Guerin treated with plasmapheresis. Am
J Med 86:343-344
Deresiewicz RL, Stone RM, Aster JC et al (1990) Fatal dis-
seminated mycobacterial infection following intravesical
bacillus Calmette Guerin. J UroI144:1331-1334
Dyer RA, Potgieter PD (1984) The adult respiratory distress
syndrome and bronchogenic pulmonary tuberculosis.
Thorax 39:383-387
Dyer RA, Chappell WA et al (1985) Adult respiratory distress
syndrome associated with miliary tuberculosis. Crit Care
Med 13:12-15
Ellner JJ (1997) Regulation of the human immune response
during tuberculosis. J Lab Clin Med 130:469-475
Evans RH, Evans M et al (1998) Massive hepatosplenomegaly,
jaundice and pancytopenia in miliary tuberculosis. J Infect
36:236-239
FitzGerald JM, Grzyrowski S et al (1991) The impact of
immunodeficiency virus infection on tuberculosis and its
control. Chest 100: 191-200
Foster DR (1997) Miliary tuberculosis following intravesical
BCG treatment. Br J Radiol 70:429-431
Gelb AF, Leffler C et al (1973) Miliary tuberculosis. Am Rev
Respir Dis 108:1327-1333
Geppert EF, Leff A (1979) The pathogenesis of pulmonary and
miliary tuberculosis. Arch Intern Med 139:1381-1383
Goldfine ID,Schachter H et al (1969) Consumption coagulopa-
thy in miliary tuberculosis. Ann Intern Med 71:775-777
Graf-Deuel E, Knoblauch A (1994) Simultaneous bilateral
spontaneous pneumothorax. Chest 105:1142-1146
Grieco MH, Chmel H (1974) Acute disseminated tuberculosis
as a diagnostic problem. A clinical study based on 28 cases.
Am Rev Respir Dis 109:554-560
Gupta RC, Lavengood R et al (1988) Miliary tuberculosis due
to intravesical bacillus Calmette-Guerin therapy. Chest 94:
1296-1298
Hill AR (1991) Disseminated tuberculosis in the acquired
immunodeficiency syndrome era. Am Rev Respir Dis 144:
1164-1170
Hong SH, 1m JG et al (1998) High resolution CT findings
of miliary tuberculosis. J Comput Assist Tomogr 22:
220-224
Iantorno R, Nicolai M et al (1998) Miliary tuberculosis of the
Miliary/Disseminated Tuberculosis
lung in a patient treated with bacillus Calmette-Guerin for
superficial bladder cancer. J Urol 159: 1639-1640
Jacques J, Sloan JM (1970) The changing pattern of miliary
tuberculosis. Thorax 25:237-240
Jasmer RM, McCowin MJ et al (1996) Miliary lung disease after
intravesical bacillus Calmette-Guerin immunotherapy.
Radiology 201:43-44
Juul A (1977) Clinically undiagnosed active tuberculosis. Acta
Med Scand 202:225-229
Kesten S, Title L et al (1990) Pulmonary disease following
intravesical BCG treatment. Thorax 45:709-710
Korzeniewska-Kosela M, FitzGerald JM et al (1992) Spec-
trum of tuberculosis in HIV infected patients in British
Columbia: a report of 40 cases. Can Med Assoc J 146:
1927-1934
Koylii R, Tozkoparan E et al (1997) Unusual miliary tubercu-
losis presenting with generalized lymphadenopathy and
abdominal involvement. Int J Tuberc Lung Dis 1:474-476
Kwong JS, Carignan S et al (1996) Diagnostic accuracy of chest
radiography. Chest 110:339-342
Langdale LA, Meissner M et al (1992) Tuberculosis and the
surgeon. Am J Surg 163:505-509
Long R, O'Connor R et al (1997) Disseminated tuberculosis
with and without a miliary pattern on chest radiograph:
a clinical-pathologic-radiologic correlation. Int J Tuberc
Lung Dis 1:52-58
Lotte A, Wasz-Hockert 0 et al (1984) BCG complications. Esti-
mates of the risks among vaccinated subjects and statistical
analysis of their main characteristics. Adv Tuberc Res 21:
107-193
Maartens G, Willcox PA et al (1990) Miliary tuberculosis: rapid
diagnosis, haematologic abnormalities, and outcome in 109
treated adults. Am J Med 89:291-296
Massaro D, Katz S (1964) Rapid clearing in hematogenous
pulmonary tuberculosis. Arch Intern Med 113:573
McClement JH, Renzetti AD, Carroll D et al (1951) Cardiopul-
monary function in hematogenous pulmonary tubercu-
losis in patients receiving streptomycin therapy. Am Rev
Tuberc 64:583
McGuinness G, Naidich DP et al (1992) High resolution CT
findings in miliary lung disease. J Comput Assist Tomogr
16:384-390
McParland C, Cotton DJ et al (1992) Miliary Mycobacterium
bovis induced by intravesical bacilli Calmette-Guerin
immunotherapy. Am Rev Respir Dis 146:1330-1333
Mert A, Bilir M et al (2001) Spontaneous pneumothorax: a
rare complication of miliary tuberculosis. Ann Thorac
Cardiovasc Surg 7:45-48
Millar JW, Horne NW (1979) Tuberculosis in immunosup-
pressed patients. Lancet 1:1176-1178
Miyake S, Yoshizawa Yet al (1997) Non-Hodgkin's lymphoma
with pulmonary infiltrates mimicking miliary tuberculosis.
Intern Med 36:420-423
Mohan A, Sharma SK et al (1996) Acute respiratory distress
syndrome (ARDS) in miliary tuberculosis: a twelve year
experience. Indian J Chest Dis Allied Sci 38:157-162
Monie RDH, Hunter AM et al (1983) Retrospective survey of
the management of miliary tuberculosis in South and West
Wales, 1976-8. Thorax 38:369-372
Munt PW (1972) Miliary tuberculosis in the chemotherapy
era: with a clinical review in 69 American patients. Medi-
cine 51:139-155
Murray HW, Tuazon CU et al (1978) The adult respiratory
299
distress syndrome associated with miliary tuberculosis.
Chest 73:37-43
Myers JA (1970) Miliary tuberculosis in the elderly. Br Med
J 1:565
Nolte FS et al (1993) Direct detection of Mycobacterium tuber-
culosis in sputum by polymerase chain reaction and DNA
hybridization. J Clin MicrobioI31:1777-1782
Oh YU, Kim YH et al (1994) High-resolution CT appearance of
miliary tuberculosis. J Comput Assist Tomogr 18:862-866
Optican RJ, Ost A et al (1992) High-resolution computed
tomography in the diagnosis of miliary tuberculosis.
Chest 102:941-943
Orihuela E, Herr HW, Pinsky CM et al (1987) Toxicity of
intravesical BCG and its management in patients with
superficial bladder tumors. Cancer 60:326-333
Ormerod LP, Horsfield N (1995) Miliary tuberculosis in a high
prevalence area of the UK: Blackburn 1978-1993. Resp Med
89:555-557
Pagel W, Simmonds FAH et al (1964) Pulmonary tuberculosis,
4th edn. Oxford University Press, London .
Palayew M, Briedis D et al (1993) Disseminated infection after
intravesical BCG immunotherapy. Chest 104:307-309
Pasculle AW, Kapadia SB et al (1980) Tuberculous bacillemia,
hyperpyrexia, and rapid death. Arch Intern Med 140:426-427
Penner C, Roberts D et al (1995) Tuberculosis as a primary
cause of respiratory failure requiring mechanical ventila-
tion. Am J Respir Crit Care Med 151:867-872
Petty TL, Ashburg DG (1971) The adult respiratory distress
syndrome: clinical features, factors influencing prognosis
and principles of management. Chest 60:233-239
Piqueras AR, Marruecos L et al (1987) Miliary tuberculosis
and adult respiratory distress syndrome. Intensive Care
Med 13:175-182
Pitchenik AE, Fertel D, Bloch AB (1988) Mycobacterial disease:
epidemiology, diagnosis, treatment,and prevention. Clin
Chest Med 9:425-441
Proudfoot AT, Asktan A, Douglas AC et al (1969) Miliary tuber-
culosis in adults. Br Med J 2:273-276
Prout S, Benatar SR et al (1980) Disseminated tuberculosis. S
Afr Med J 58:835-842
Rabe J, Neff KW et al (1999) Miliary tuberculosis after intra-
vesical bacilli Calmette-Guerin immunotherapy for carci-
noma of the bladder. Am J Roentgenoll72:748-750
Rawls WH, Lamm DL et al (1990) Fatal sepsis following
intravesical bacillus Calmette Guerin administration for
bladder cancer. J UroI144:1328-1330
Ray M, Kataria S et al (2002) Unusual presentation of dissemi-
nated tuberculosis. Indian Pediatr 39:88-91
Rieder HL, Kelly GD, Bloch AB et al (1991) Tuberculosis diag-
nosed at death in the United States. Chest 100:678-681
Rosenberg MJ, Rumans LW (1978) Survival of a patient with
pancytopenia and disseminated coagulation associated
with miliary tuberculosis. Chest 73:536-539
Sahn SA, Neff TA (1974) Miliary tuberculosis. Am J Med 56:
495-505
Saye L (1936) Tuberculosis miliar cronica. Rev Med Barcelona
25:387-423
Shafer RW, Kim DS et al (1991) Extrapulmonary tuberculosis
in patients with human immunodeficiency virus infection.
Medicine 70:384-397
Shalwar R et al (1993) Detection of Mycobacterium tubercu-
losis in clinical samples by amplification of DNA. J Clin
MicrobioI29:712-717
300
Sharma SK, Shamim SQ et al (1981) Disseminated tuberculo-
sis presenting as massive hepatosplenomegaly and hepatic
failure. Am J Gastroenterol 76:153-156
Sharma SK,Pande IN et al (1988) Bronchoalveolar lavage (Bal)
in miliary tuberculosis. Tubercle 69:175-178
Sharma SK, Pande IN et al (1992) Pulmonary function and
immunologic abnormalities in miliary tuberculosis. Am
Rev Respir Dis 145:1167-1171
Shibolet A, Dan M et al (1979) Recurrent miliary tuberculosis sec-
ondary to infected ventriculoatrial shunt. Chest 76:328-330
Slavin RE, Walsh TJ et al (1980) Late generalized tuberculosis: a
clinical pathologic analysis and comparison of 100 cases in
the preantibiotic and antibiotic eras. Medicine 59:352-366
So SY, Yu D (1981) The adult respiratory distress syndrome
associated with miliary tuberculosis. Tubercle 62:49-53
Soloway MS (1988) Intravesical therapy for bladder cancer.
Urol Clin North Am 15:661-669
Stallworth JR, Brasfield DM et al (1980) Congenital miliary
tuberculosis proved by open lung biopsy specimen and
successfully treated. Am J Dis Child 134:320-321
Steiner P,Rao M et al (1976) Miliary tuberculosis in two infants
M. M. Madkour
after nursery exposure: epidemiologic, clinical, and labora-
tory findings. Am Rev Respir Dis 113:267-271
Stenius-Aarniala BS, Tukiainen P (1979) Miliary tuberculosis.
Acta Med Scand 206:417-422
Sutton FD, Hudson LD et al (1974) Recognition and management
of the adult respiratory distress syndrome. Chest 66:34S-36S
Treip C, Meyers D (1959) Fatal tuberculosis in a general hos-
pital. A diagnostic problem. Lancet 1:164-1'67
Vanham G, Toossi Z, Hirsch CS et al (1997) Examining a para-
dox in the pathogenesis of human pulmonary tuberculosis:
immune activation and suppression/anergy. Tubercle Lung
Dis 78:145-158
Vijayan VK (2000) Disseminated tuberculosis. J Indian Med
Assoc 98:107-109
Willcox PA, Potgieter PD et al (1986) Rapid diagnosis of
sputum negative miliary tuberculosis using the flexible
fibreoptic bronchoscope. Thorax 41:681-684
Williams MH, Yoo OH (1973) Pulmonary function in miliary
tuberculosis. Am Rev Respir Dis 107:858
Yu YL, Chow WH et al (1986) Cryptic miliary tuberculosis. Q
J Med 59:421-428
19 Tuberculosis and Pregnancy
M. MONIR MADKOUR

CONTENTS risk factor for the development of tuberculosis has

never been proven, as suggested in the past (Snider
19.1 Introduction 301 1984; Espinal et al. 1996). Pregnancy has no effect
19.2 Epidemiology 301
19.3 Pathogenesis 302
on the incidence or prognosis of tuberculosis. The
19.4 Pregnancy-Tuberculosis: Interrelations 303 effect of tuberculosis on the outcome of pregnancy,
19.5 Clinical Features 303 the fetus, and the neonate has lost most of its
19.6 Personal Series 303 importance due to the existence of effective anti-
19.6.1 Case Illustration 1 304 tuberculous agents. However, a delay in diagnosing
19.6.2 Case Illustration 2 304
19.6.3 Case Illustration 3 306 active tuberculosis poses a real risk to the pregnant
19.7 Diagnosis 306 woman, fetus, and infant. Tuberculosis in pregnant
19.8 PPD Skin Test 306 women may be asymptomatic, particularly when
19.9 Chest Radiography 307 it invades extrapulmonary sites in the body, and
19.10 Microbiology and Other
therefore presents a great risk of neonatal morbid-
Advanced Laboratory Tests 307
19.11 Breast-Feeding and Its Contraindications 307 ity and mortality. Symptoms related to tuberculosis
19.12 Treatment of Active TB During Pregnancy 308 may be mild and mimic those caused by normal
19.13 Treatment of Multidrug-Resistant TB pregnancy such as dyspnea and fatigue (Gogus et al.
During Pregnancy 309 1993; Machin et al. 1992). The disease may first pres-
19.14 Preventive Treatment During Pregnancy 309
ent in the neonate, and acute tuberculosis may later
References 311
be discovered in the mother. Hageman et al. (1980)
reported that tuberculosis was initially presented
and diagnosed in neonates in 15 of 26 cases, and
subsequent investigations of their mothers detected
19.1 active disease. Congenital and neonatal tuberculosis
Introduction associated with undiagnosed and untreated moth-
ers carries an infant mortality rate of 50% (Nemir
The prevalence of tuberculosis in women of child- and O'Hare 1985). Antenatal screening of pregnant
bearing age and during pregnancy is increasing women in endemic areas should include plain chest
both in developing countries and among ethnic radiography if pulmonary tuberculosis is suspected,
immigrants in urban areas of developed countries with proper shielding of the abdomen.
(Centers for Disease Control and Prevention 1994, The outcome of pregnancy for the tuberculous
1995; Ahmed et al. 1999; Hageman 1998). The mor- woman, fetus, and neonate who are treated promptly
bidity and mortality among pregnant women with with antituberculous drugs is similar to that for preg-
tuberculosis, the fetus, and the neonate have also nant women without tuberculosis.
increased, particularly among those co-infected with
HIV (Adhikari et al.1997).A high index of suspicion
by attending clinicians and effective measures during
the prenatal period are essential for reducing these 19.2
risks (Starke 1997). Pregnancy as a predisposing or Epidemiology

Specific epidemiological data on the incidence of

M. M. MADKOUR, MD, DM, FRCP tuberculosis during pregnancy at present is unknown
Consultant, Department of Medicine, Riyadh Armed Forces even in developed countries such as the USA. Avail-
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia able data from the 1970s, based on hospitalized

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
302
pregnant women with tuberculosis in a New York
hospital, was estimated by Schaefer et al. (l975) as
ranging from 0.6% to 1% and was higher (3.2%)
among those with pulmonary disease. These data
were equivalent to those from nonpregnant women
with tuberculosis of comparable age. In the late 1980s
and 1990s, the incidence of tuberculosis in the USA
has increased in young adults and children, which
implies that tuberculosis during pregnancy may be
more prevalent.
Such a rise was attributed to the HIV epidemic
as the most important risk factor, immigrants of
different ethic origin coming from countries where
tuberculosis is endemic, decline in public health
services, and increased transmission in congregate
settings. Tuberculosis among foreign-born persons
has increased from 22% in 1986 to 35% in 1995 in the
USA (McKenna et al.1995).
The epidemiology of tuberculosis in England and
Wales has changed its pattern since 1988 (Ormerod
2001). During the 1980s, white ethnic women over
the age of 50 contributed 50% of the cases, and only a
minority of patients were of child-bearing age (Med-
ical Research Council 1992). In 1998, over 56% of
cases were foreign-born immigrants from the Indian
subcontinent and of African ethnic origin. The
median age of these immigrants was under 30 years,
while among the white ethnic women, tuberculosis
declined to 37% with a minority of child-bearing age
(Rose et al. 2001).
In developing countries, poverty compounded
by HIV infection form closely associated risk fac-
tors and contributed to an estimated 70% mortality
due to tuberculosis in the age group 15-40 years
including women of child-bearing age (Connolly and
Nunn 1996).
In Africa, there are no data on the number of
women of child-bearing age co-infected with HIV
and tuberculosis. The World Health Organization
(WHO) estimated that about 50% of women of child-
bearing age living in sub-Saharan Africa has been
infected with M. tuberculosis (Connolly and Nunn
1996). In South Africa, the incidence of tuberculosis
is 311 per 100,000 population, and the prevalence rate
of HIV infection at an antenatal clinic was reported
as 7.6% (Epidemiology Comments, Department of
Health, Pretoria 1996). In developing countries, the
maternal mortality attributed to tuberculosis is high.
Ahmed and colleagues (l999) from Zambia reported
hospital-based, nonobstetric maternal mortality
among pregnant women with tuberculosis. These
authors analyzed 251 maternal deaths over a 2-year
period (1996-1997) in Lusaka University Hospital.
M. M. Madkour
A total of 106 (42%) maternal deaths was due to
direct obstetric causes, and 145 (58%) were due to
nonobstetric causes. Tuberculosis, a nonobstetric
cause of maternal death, was associated with 25%
of pregnant women. Other reports from Zambia
indicated that 25% of antenatal women were infected
with HIV, and a rate of 1 woman in 8 was presumed
to be co-infected with HIV and TB. The risk of devel-
oping overt tuberculosis is expected at 8% annually
(Fylkesnes et al. 1997; Dolin et al. 1994).
Neonatal tuberculosis is rare, and there are no
reports of a recent increase in incidence from coun-
tries with a high incidence of HIV infection. Adhikari
et al. (l997) from South Africa reported a hospital-
based incidence of neonatal tuberculosis. Eleven
neonates with culture-confirmed tuberculosis were
reported from King Edward VIII Hospital in Natal,
seen in a I-year period (l996-1997). The mothers of
six of these neonates were co-infected with TB and
HIV. One neonate and two mothers died within the
first 3 months. These data suggest that the increase
in the co-infections of pregnant women in endemic
areas for both diseases was associated with an
increased incidence of neonatal TB.
19.3
Pathogenesis
The pathogenesis of tuberculosis during pregnancy is
similar to that in nonpregnant women. Mycobacterium
tuberculosis enters the body commonly by inhalation.
The bacilli are ingested by the alveolar macrophages.
A few weeks later two responses take place: tissue-
damage response (delayed-type hypersensitivity-spe-
cific humoral response) and macrophages-activating
response (cell-mediated immunity).
Cytokines produced by alveolar macrophages and
T-Iymphocytes (CD4+) take an active role in the host
defense. Subsequently, granuloma formation occurs
in the lungs. Granulomas may heal with fibrosis and
scarring, but some bacilli may remain viable and
reactivate if the patient's immunity is suppressed.
Hematogenous spread of the bacilli to the placenta
from pulmonary or extrapulmonary disease is the
most common route of transmission. Rarely, trans-
mission by direct spread and extension of the disease
from pre-existing genital tuberculosis may occur
(Hallum and Thomas 1995; Kaplan et al. 1980; Bazaz-
Malik et al. 1983). The organisms are transmitted to
the placenta and may infect the chorionic villi, amni-
otic fluid, and decidua. The bacilli may be transmit-
Tuberculosis and Pregnancy
ted to the fetus via the umbilical vein, infecting the
liver first, and may pass through the main circulation,
leading to lung disease. Fetal ingestion or aspiration
of infected amniotic fluid may also be another mode
of transmission of tuberculosis to the fetus in utero
or the time of birth (Vallejo and Starke 1992).
Transmission of infection in the neonatal period
may also occur via inhalation if the mother is not
diagnosed or treated.
19.4
Pregnancy-Tuberculosis: Interrelations
"For the virgin, no marriage;
For the married, no pregnancy;
For the pregnant, no confinement;
For the mother, no suckling."
In an excellent comprehensive review on the inter-
relationships between tuberculosis and pregnancy,
Miller and Miller (1996) abstracted these statements
from archaic reports reflecting the opinions and
advice of scientists and clinicians in various medical
eras. The influence of pregnancy as a risk factor for
the development of tuberculosis, altering the course
of the disease and prognosis, has intrigued clinicians
for centuries since the days of Hippocrates through
the early 1920s.
The pendulum of opinion swung widely from the
beneficial value of pregnancy on tuberculosis to the
detrimental effect on the patient as it worsened the
course of the disease, and abortion was even advised.
Even before the discovery of antituberculous treat-
ment, pregnancy was reported not to worsen the
course of the disease (Hill 1928, cited by Miller
and Miller 1996). Shortly after the introduction of
antituberculous drugs, many reports in the early
1950s, 1960s, and 1970s confirmed that there was
no significant difference between the patients who
improved and those whose tuberculosis progressed
during pregnancy or after delivery (Cohen et al.1952;
Hedvalll953; Schaefer et al.1975; Espinal et al.1996).
There is no evidence that pregnancy increases the
risk of tuberculosis in the postpartum period in HIV-
infected or uninfected women (Espinal et al. 1996).
The effect of tuberculosis on the course of pregnancy,
fetus, and neonate is similar to that in nontubercu-
lous pregnant women.
Untreated active tuberculosis poses a real great
risk to the pregnant woman, fetus, and neonate.
Women with active tuberculosis have a significantly
303
higher incidence of spontaneous abortion, congenital
malformations of the fetus, and neonatal disease
(Bjerkedal et al. 1975).
19.5
Clinical Features
Tuberculosis may present initially with symptoms
mimicking those physiologically related to normal
pregnancy such as dyspnea, fatigue, and lack of
energy. The attending clinician has to have a high
index of suspicion and make diligent efforts to diag-
nose the disease in order to prevent fetal and neonatal
infection, particularly in endemic areas and among
groups at risk (HIV, alcoholics, drug addicts, the
homeless). Pregnancy does not alter the mode of
onset or the presentation of initial symptoms. How-
ever, one-half to one-third of pregnant women with
tuberculosis are asymptomatic and unaware of the
disease, and the diagnosis may be missed (Wilson
et al. 1973; Schaefer et al. 1975; Carter and Mates
1994). However, the clinical features of tuberculosis
during pregnancy are similar to those in nonpreg-
nant women.
19.6
Personal Series
In our series, review of the medical records of preg-
nant women with active tuberculosis who attended
the Military Hospital in Riyadh were analyzed.
Between 1985 and 2000, there were 39 patients with
active tuberculosis during their pregnancy. Their age
ranged between 16 and 42 years (average 19.7 years).
We found a past history of tuberculosis (treated or
partially treated) in 13 patients (33.3%). History of
contact and a family history of tuberculosis were
found in 9 patients (23%). Pulmonary tuberculosis
was the most common site of infection and found
in 23 of these pregnant women (59%). Extrapulmo-
nary localizations of active tuberculosis were found
in 16 (41 %). Cough, fever, easy fatigability, shortness
of breath, and weight loss were the most common
presenting features and found in 23 patients (59%).
Accidental discovery of cervical lymph adenopathy
during antenatal screening was noted in 9 patients
(23%).
Four women were aware of the cervical lymph-
adenopathy, while the other 5 women were unaware of
304
its presence. None of the pregnant women with tuber-
culous lymphadenitis were symptomatic at the time of
the antenatal screening. The 4 women who were aware
of lymph node enlargement did not notice any recent
changes in size. Two of these women with TB lymph-
adenitis presented with pancytopenia or thrombocy-
topenic purpura (see case illustrations).
The mean gestational age at the time of clinical
presentation or suspicion during antenatal screen-
ing was 24.6 weeks, with a range of 1-35 weeks.
Tuberculous pleural effusion was found in 3 patients
(7.7%). One patient had a vulval swelling that was
biopsied and found to be tuberculous by culture
and histopathology findings (2.6%). One patient
had military tuberculosis (2.6%). One patient had
tuberculosis of the knee (2.6%). One patient had
cold abscess in the right thigh (see case illustration)
(2.6%). All 39 patients had positive tuberculin skin
tests. Pulmonary tuberculosis was diagnosed in 23
pregnant women by plain chest radiography find-
ings and confirmed by cultures from sputum, early
morning gastric aspiration, and urine, identifying
M. tuberculosis. None of the 39 patients in our series
had HIV co-infection, drug-resistant tuberculosis, or
nontuberculous mycobacterial disease.
In New York, the incidence of tuberculosis during
pregnancy has increased in association with the
HIV epidemic. Margono et al. (1994) reviewed the
computerized records of 16 pregnant women with
active tuberculosis attending two hospitals. In a 6-
year period (1985-1991) they found 5 cases, while
they found 11 cases in the following 2-year period
(1991-1992). A HIV test was positive in 7 among 11
tested pregnant women with tuberculosis.
Pulmonary tuberculosis was the most common
site (10 cases, 62.5%), and extrapulmonary sites were
found in 6 cases (37.5). Tuberculous lymphadenitis
was the least common site of extrapulmonary local-
ization reported by these authors in contrast to our
own series where the lymph node was the most com-
mon location.
Early reports on tuberculosis during pregnancy
in the USA mostly involved pulmonary localizations
(Good et al. 1981). In Great Britain, pregnant women
with tuberculosis mostly had an extrapulmonary site
of the disease. In a prospective study carried out in
Northwick Park District General Hospital, Llewelyn
et al. (2000) reported 13 pregnant women with active
tuberculosis seen prospectively over a 30-month
period (1995-1998). Extrapulmonary sites of local-
ization of tuberculosis were present in 9 patients
(69.2%). All patients in this recent study were immi-
grants to Britain.
M. M. Madkour
19.6.1
Case Illustration 1
A 36-year-old Saudi woman in the 24th week of
pregnancy presented with a swelling in the middle
third of the right thigh. The swelling started as a
small area but increased in size over the following
3 months before presentation, when it became pain-
ful. There was no history of trauma or discharge. The
patient had fever, sweating, and loss of appetite over
the same period. She had never been ill before, and
other systematic enquiries were negative. The patient
was not acutely ill or in distress. A large soft-tissue
swelling extending through the anterolateral aspect
of the proximal third of the right thigh was seen. The
skin over it was normal, and it was slightly tender. The
uterus size corresponded with 24 weeks' gestation.
The hemogram, coagulation profile, and bio-
chemical values were normal. ESR was 40 mm/h.
Ultrasound of the right thigh depicted a large,
soft-tissue, cystic swelling with thick mobile fluid,
which could be an abscess or hematoma. It measured
16 cm in length and contained about 800 ml of fluid
(Fig. 19.1a).
MRI of the right thigh was done, axial T1 and
T2 slices obtained through both upper femora
and the lower pelvis. The post-gadolinium images
were not diagnostic due to the patient's movement.
MRI depicted a huge, well-defined, elliptical mass
in the right rectus femoris muscle extending from
its attachment down to the distal thigh (Fig.1b,c).
Similar masses were also noted in the right iliopsoas
muscle and right pectineus muscle. These masses
showed low T1 and high T2 signal intensity, most
probably indicating pus caused by tuberculous
abscesses. No bone involvement was seen, and MRI
of the thoracolumbar spine showed no evidence of
spinal involvement (Fig. 19.1b,c).
Aspiration of 200 ml of yellowish purulent pus was
obtained and sent for microbiological examination
and culture for tuberculosis. The culture yielded M.
tuberculosis that was sensitive to all antituberculous
drugs. Rifampicin, INH, and ethambutol were given
to this patient, and she responded well to treat-
ment. A normal, spontaneous, full-term delivery of a
healthy infant was the result in this patient.
19.6.2
Case Illustration 2
A 33-year-old woman presented to my clinic with a
2-month history of fever, sweating, and drug cough.
 Tuberculosis and Pregnancy 305

Fig.19.1a-d. A 36-year-old pregnant woman presented with a swelling of her right thigh. a Ultrasound shows a well-defined
cystic mass with relatively thick wall (curved arrow) consistent with localized abscess. TI-weighted (b) and T2-weighted (c) MRI
of the same patient show the abscess in the right vastus lateralis muscle as low signal intensity lesion on TI and as homogeneous
high signal intensity on T2 (arrowheads). d MRI of lower pelvis demonstrated a gravid uterus with gestational sac (arrowheads).
This was done in an attempt to trace the source of this tuberculous abscess and avoid missing abdominal or pelvic pathology
(the safety of MRI in pregnancy is not yet clear). Patient completed the treatment and delivered a normal healthy boy

She had lost 21 kg in weight over that period. She cough, and weight loss. She looked pale, feverish, but
was 95 kg in weight before the illness and 74 kg at there was no change in the size of the left supracla-
the time of presentation. She had a past history of vicular lymph nodes. There were no other clinical
two swellings on the left side of the neck for the past findings. There were purpura and petechiae over the
17 years with no recent changes in their size. She had lower limbs. Repeat chest radiograph, with shield
five healthy children. The patient looked pale, her tem- over the abdomen, remained normal. The hemo-
perature was 38.3°C. There were two enlarged lymph gram showed pancytopenia with Hb 10.9 G/dl, WBC
nodes in the left supraclavicular region but no other 2.400/mm3, platelets 31,OOO/mm 3, and the blood film
clinical findings. The hemogram showed Hb 1O.9G/dl, showed microcytic hypochromic features. Her serum
WBC 9000/mm 3 , platelets 211,OOO/mm 3, and the blood iron and other biochemical parameters were other-
film showed hypochromic microcytic anemia. wise normal. Coombs test was negative. Bone marrow
Biochemical parameters were normal, but the aspiration and biopsy showed hypercellularity in the
Mantoux test was positive. The patient refused lymph myeloid and erythroid series with normal matura-
node biopsy and discharged herself against medical tion and increased number of megakaryocytes.
advice. Two months later she presented to me with a The marrow iron store was normal. Ultrasound of
3-day history of hematuria and skin petechiae over the abdomen was normal. Lymph node biopsy was
her legs. She was pregnant with gestational age of performed and showed granulomas with caseation.
9 weeks. She continued to have fever, sweating, dry The biopsy culture yielded M. tuberculosis 4 weeks
306
later. She responded well to antituberculous treat-
ment, and the hemogram returned to normal after
3 months. She subsequently had a normal, spontane-
ous delivery of a healthy baby.
19.6.3
Case Illustration 3
A 30-year-old woman from AI-Qurayyat, northern
Saudi Arabia, was referred to our hematologist
because of thrombocytopenia, anemia, and convul-
sions 5 days after a normal delivery. The patient was
well during her antenatal follow-up until the last
month of her pregnancy when she presented with
epistaxis and melena. At that time, her hemoglobin
was only 5.8 Gldl, low platelet of 12,OOO/mm3 with
normal white cell count. The bleeding time was
prolonged, and bone marrow aspiration could not
be done at the referring hospital. She was suspected
to have idiopathic thrombocytopenic purpura and
was started on prednisolone 60 mg oral daily. She
was given 5 units of blood and 4 units of platelets
and vitamin K. Five days prior to her transfer to our
hospital, she delivered a healthy girl. After delivery,
she developed headache and recurrent attacks of
convulsions. Brain CT scan with contrast was nor-
mal. Her fits were controlled with phenytoin. She was
admitted on arrival to our general intensive care unit
(GICU). She gave a history of easy bruising for the
past few years and had not been investigated for this
complaint. She had delivered a baby boy 4 years ago
with fetal abnormalities, the exact nature of which
was not known, who died 3 months after birth. Her
history included pulmonary tuberculosis while preg-
nant with that congenitally deformed boy 4 years ago.
She discontinued her antituberculous medication
after 2 months when she discovered that she was
pregnant.
She decided herself to discontinue taking the
antituberculous drugs for fear of taking medication
during pregnancy. The patient looked ill, emaci-
ated, with bruises all over her body, particularly
in the upper limbs. She was not jaundiced and had
no lymphadenopathy. Fine apical crepitations were
heard over the right lung. Other systemic examina-
tions were normal. The hemogram showed features
of severe iron deficiency anemia with low plate-
lets of 18,OOO/mm3 and normal white cell count.
Biochemical parameters were normal. Plain chest
radiography showed multiple cavitating lesions in
the right upper lobe compatible with active tubercu-
losis. Abdominal cT showed no abnormalities. Bone
M. M. Madkour
marrow aspiration and trephine showed marked
hypercellular marrow with hyperactive myeloid and
erythroid series with normal maturation and an
increased number of megakaryocytes. The iron store
was normal. EEG and MRI of the brain were normal.
Lumber puncture showed normal CSF. She was
seen by our pulmonologist and started on four anti-
tuberculous antibiotics. The prednisolone dosage
was gradually reduced and stopped. The patient's
platelets and hemoglobin returned to normal. She
had no further convulsions or bruises. Her follow-up
at the chest and hematology clinic was satisfactory,
and the patient had a full recovery.
19.7
Diagnosis
Prenatal health care screening may help in identi-
fying pregnant women with active pulmonary or
extrapulmonary tuberculosis. Early detection of
active disease may help in preventing serious con-
sequences to the mother, fetus, and neonate. A high
index of suspicion by the attending clinician is essen-
tial' particularly in endemic areas and among women
at high risk for tuberculosis infection. Detailed his-
tory and clinical examination and screening investi-
gations are required to achieve this goal. Investiga-
tions may include microbiological methods, plain
chest radiography, tuberculin skin testing, and other
laboratory tests as appropriate to each patient.
19.8
PPD Skin Test
Earlier concepts suggested that a false-negative
tuberculin skin testing might occur as a result of
the suppression of cell-mediated immunity due to
pregnancy (Finn et al. 1972). Pregnancy is found by
controlled studies to have no demonstrable effect on
tuberculin skin testing, and the test is valid at any
time during pregnancy (Snider 1984; Present and
Comstock 1975; Huebner et al. 1993).
In endemic areas, a positive tuberculin skin test
identifies individuals with previous infection by M.
tuberculosis but does not indicate the state of disease
activity, and a negative response may be noted in
patients with active disease. HIV-infected pregnant
women react to tuberculin skin testing in a similar
way to nonpregnant women with HIV infection with
Tuberculosis and Pregnancy
similar rates of anergy. Anergy is not more frequent
as a result of pregnancy among women with HIV
infection (Mofenson et al.1995).
In the USA, the recommendations of the Advisory
Committee at the Center for Disease Control (CDC
1990) with regard to performing the tuberculin test
for screening individuals at high risk are as follows:
(a) pregnant woman with clinical features suggestive
of tuberculosis, (b) women with known or suspected
exposure to tuberculosis, (c) women at high risk of
developing tuberculosis. A skin reaction of 15 mm
in low-risk patient and 5-10 mm in immunocomp-
romised patients is considered positive and should
lead to the inclusion of chest radiography as part of
the screening (Division of Tuberculosis Elimination,
CDC 1991).
The interpretation of the tuberculin skin test and
the rationale of using various induration size mea-
surements as a positive indication among different
populations at variable risk have been the subject
of debates in many recent publications. In popula-
tions at high risk of developing tuberculosis, contact
with active tuberculosis patients, those with clini-
cal or imaging evidence of tuberculosis, or patients
with HIV or immunocompromised individuals,
induration of at least 5 mm is considered positive
and indicating infection with M. tuberculosis. For
children in contact with high-risk adults, immune-
suppressed patients due to other medical illness,
and foreign-born persons, an induration of at least
10 mm is considered positive. In populations at low
risk for tuberculosis, an induration of at least 15 mm
is considered positive.
There has been no classification scheme study of
tuberculin induration measurements for pregnant
women with or without HIV infection, nevertheless,
the tuberculin skin test reaction during pregnancy
is not different from that for nonpregnant women
(American Academy of Pediatrics 1992; American
Thoracic Society 1993; Centers for Disease Control
and Prevention 1995).
19.9
Chest Radiography
In the past, routine chest radiography was performed
as part of screening investigations for all pregnant
women (Freth 1952). This routine was found to be
potentially hazardous due to irradiation of the fetus,
particularly during the first trimester (Bonebrake et
al. 1978). However, chest radiography can be per-
307
formed for pregnant women if active pulmonary
tuberculosis is suspected particularly in endemic
areas and among women at high risk. In nonendemic
areas, a positive tuberculin skin test in pregnant
women is an indication for chest radiography.
Appropriate shielding of the abdomen will reduce
irradiation to 50 millirads, which is a less hazardous
dose for the fetus, and chest radiography is preferably
done after the first trimester. Women with chest
symptoms even with a negative tuberculin skin test
should have a chest radiograph done (Swartz and
Reichling 1978; Brent 1989).
19.10
Microbiology and Other
Advanced Laboratory Tests
The diagnosis of active tuberculosis during preg-
nancy may be difficult, particularly when conven-
tional sputum smear staining is negative. Specimens
from sputum, gastric or bronchial aspirates, urine
and other body fluids, or tissue biopsies should be
sent for culturing. Recent advances in automated
culture systems such as BACTEC (Becton Dickson
Diagnostic Instrument Systems, Towson, MD, USA)
are now widely used in many laboratories. Immu-
nological tests and PCR are useful tools that can be
used, when available, for the diagnosis of tuberculosis
during pregnancy.
19.11
Breast-Feeding and Its Contraindications
Breast-feeding by mothers with newly discovered
active tuberculosis, before initiating treatment, has
been a subject of debate among authors in devel-
oped and developing countries. Authors from devel-
oped countries, where alternative feeding of infants
is readily available, have their own views. Authors
from poor, developing countries where feeding sub-
stitutes are not easily available have different views
as well. To make matters worse, co-infection with
HIV complicates the issue of breast-feeding in these
poor countries. Transmission of HIV via breast milk
to infants is a real threat, and starvation and infant
death follow if breast-feeding is contraindicated in
such circumstances. This issue is clearly defined in
developed countries as a contraindication to breast-
feeding (Oxtoby 1988).
308
Nursing mothers with active tuberculosis who
are receiving antituberculous treatment should be
encouraged to breast-feed their infants. Concerns
about the side-effects of antituberculous drugs
secreted in the mother's milk have been investigated
by many authors. Rifampicin, INH, and streptomycin
concentrations in breast milk have been measured
after the administration of 600 mg, 300 mg, and 1 g
of these drugs, respectively. Peak milk concentrations
ranged between 10 and 30 mg!l when rifampicin was
given in a dose of 600 mg. A concentration level of
16.6 mg!l was found 3 h after INH injection at a dose
of 300 mg, while streptomycin 30 min after intra-
muscular injection of 1 g dose reached a concentra-
tion of 1.3 mg!I. These concentrations of antituber-
culous drugs in breast milk have no toxic effect on
the infant (Lawrence and Lawrence 2001; Berlin and
Lee 1979; Vorherr 1974; Fugimoritt and Imais 1957;
Snider and Powell 1984). These antituberculous med-
ications have been used directly and safely in infants
(Snider 1984).
Tuberculosis infection and active tuberculosis are
considered two separate issues with regard to breast-
feeding. Breast-feeding is not contraindicated in
women with previous tuberculosis infection. In women
with active tuberculosis, the American Academy Com-
mittee has indicated that respiratory contact puts
these infants in jeopardy. Breast milk in such patients
does not contain tubercle bacilli and can be pumped
into bottles and fed to these infants until medication
has commenced, and the mother is no longer con-
sidered infective. The committee also recommended
that active tuberculosis is an indication for temporary
isolation of the mother until she has received about
2 weeks of antituberculous therapy. Mothers with TB
mastitis should have the milk pumped and discarded
until the breast lesion has completely healed (Ameri-
can Academy of Pediatrics 1992).
The presence of concomitant HIV and active
tuberculosis in mothers living in developed countries
is an absolute contraindication to breast-feeding. The
rate of HIV transmission to infants via breast milk is
estimated to range between 5% and 20% or higher
(Lawrence 1997; Lawrence and Lawrence 2001;
Oxtoby 1988).
Data on freezing or heating breast milk to destroy
HIV type I are sparse and insufficient (Ando et aI.
1989). In developing countries where alternative
affordable infant feeding, sanitary measures, and
medical resources are not available to prevent the
transmission of HIV to infants, the World Health
Organization, UNICEF, and the United Nations AIDS
program recommend the following measures: (l) If
M. M. Madkour
an adequate and safe milk substitute is available, then
breast-feeding should not be used. (2) If substitutes
are not available, then women who are HIV-positive
should be offered the choice of what is appropriate
for their circumstances and supported in their choice
(WHO 1998).
19.12
Treatment of Active T8 During Pregnancy
Pregnant women with active pulmonary or extra-
pulmonary tuberculosis should be treated without
delay. Delay in treatment poses a real threat to the
pregnant women, her fetus, and the contact. Treat-
ment should not be interrupted in women with active
tuberculosis if pregnancy is discovered while she is
on antituberculous medication. The only exception is
streptomycin, which may cause fetal ototoxicity and
should be replaced unless no alternative drugs are
available (Hamadeh and Glassroth 1992). Concerns
about the teratogenic effect of antituberculous drugs
have been extensively studied by many authors in
recent years. Isoniazid, rifampicin, ethambutol, and
pyrazinamide are the most widely used and care-
fully evaluated drugs with no evidence of increased
teratogenicity (Steen and Seainton-Ellis 1977; Snider
et al. 1980; Starke 1997; Czeizel et aI. 2001).
The human teratogenic potential of isoniazid,
rifampicin, ethambutol, pyrazinamide, ethionamide,
prothionamide, and cycloserine during pregnancy
was carefully studied by Czeizel from Hungary and
his colleagues from Denmark (Czeizel et al. 2001).
These authors surveyed case-controlled, population-
based, congenital abnormalities from the Hungar-
ian National Birth Registry between 1980 and 1996,
regarding the safety of 7 oral antituberculous drugs
used during pregnancy. These drugs were used dur-
ing the second and third months of gestation to treat
pregnant women with active tuberculosis (critical
period for most major congenital abnormalities). The
authors reported their findings as shown in their table
(Table 19.1). They found no detectable teratogenic risk
to the fetus; however, the number of pregnant women
who were treated with these drugs during the critical
period was limited (6 cases vs 21 controls).
The guidelines and recommendations with regard
to the choice of the antituberculous drug regimen are
subject to revisions and updating depending on drug
sensitivity and resistance tests. The most commonly
used regimen at present involves the use of two to
four or more oral drugs including isoniazid, rifam-
Tuberculosis and Pregnancy 309

Table 19.1 Characteristics of the antituberculosis drugs studied, in addition to the number of cases and controls
during the study period between 1980 and 1996 (reproduced with permission from Czeizel et al. 2001)

Antituberculosis Table dose Recommended daily Cases Controls for all CAs Crude POR
drugs (mg) treatment (mg) (n=22865) (n=38151)
n (0/0) n (0/0) POR (950/0CI)
Isoniazid 50 50-300 6 (0.03) 17 (0.04) 0.6 (0.2-1.5)
Rifampicin 150,300 450-600 0(0.00) 1 (0.00) 0.6 (0.0-13.7)
Ethambutol 250 15-15 mglbwkg 4 (0.02) 6 (0.02) 1.1 (0.3-13.7)
Pyrazinamide 500 35 mglbw kg (1000-2000) 0(0.00) 3 (0.01) 0.2 (0.0-4.6)
Ethionamide 250 750-1000 0(0.00) 1 (0.00) 0.6 (0/0-13.7)
Prothionamide 250 500-1000 0(0.00) 1 (0.00) 0.6 (0.0-13.7)
Cycloserine 250 750-1000 1 (0.00) 0(0.00) 5.0 (0.2-122.9)
Total 11 (0.05) 29 (0.08) 0.6 (0.3-1.3)
POR, prevalence odds ratio; CA, congenital abnormality; CI, confidence interval; bw, body weight

picin, ethambutol, and pyrazinamide. The duration
of treatment may range between 6 and 9 months.
Ethambutol should be added in the first 2 months
to INH and rifampicin. Alternatively, pyrazinamide
can be used in the first 2 months with INH and
rifampicin. Pyridoxine 50 mg daily should be added
to prevent seizure in the newborn when INH is used.
The dosages of these drugs are similar to those used
in nonpregnant women. Streptomycin should be
avoided unless no alternative antituberculous drug
is available. Kanamycin and capreomycin share the
same potential for producing ototoxicity as strepto-
mycin. Successful treatment of active tuberculosis
during pregnancy depends on the identification of
the organisms by culture and sensitivity, the use of
multiple drugs to which the organisms are sensitive,
and close follow-up of the patient for treatment
compliance.
19.13
Treatment of
Multidrug-Resistant T8 During Pregnancy
Single or multiple drug-resistant tuberculosis dur-
ing pregnancy is increasing. Isoniazid-resistant
tuberculosis during pregnancy can be treated with
rifampicin and ethambutol for 18 months (Vallejo
and Starke 1992). However, the management of
pregnant women with multidrug-resistant tubercu-
losis is a real dilemma. Iseman recommended the use
of 5-6 drugs in areas of known multidrug-resistant
tuberculosis until drug sensitivity patterns are iden-
tified. This makes the treatment of pregnant women
with multidrug-resistant tuberculosis extremely
difficult, because some of these medications may
be contraindicated or of unknown teratogenic effect
(Iseman 1993). There are no clear recommendations
or guidelines to this problem because of possible seri-
ous consequences to the fetus. Good and colleagues
reported their views and stated: 'it is reasonable to
consider therapeutic abortion in this setting' (Good
et al. 1981).
Nitta and Milligan (1999) from the MDR tuber-
culosis unit in Los Angeles reported their experi-
ence of management of four pregnant women with
multidrug-resistant pulmonary tuberculosis. MDR
tuberculosis was acquired due to their nonadherence
to previous regimens of antituberculous treatment.
One of them chose termination of her pregnancy. In
one, treatment was withheld, and only rifampicin and
INH were given to prevent TB progression; another
patient was allowed to continue self-administered
medication during her pregnancy. Yet another
patient was not treated during pregnancy. All three
women who carried their gestation to full term had
healthy infants. Retreatment of all four women was
done using 5-6 drugs including ethambutol, cyclo-
serine, para-aminosalicylic acid (PAS), capreomycin,
ofloxacin, and clofazimine given as daily DOT for
17-24 months. The authors concluded that indi-
vidualized treatment is considered according to each
patient's medical and psychosocial needs. They also
demonstrated that MDR tuberculosis during preg-
nancy can be managed safely and successfully.
19.14
Preventive Treatment During Pregnancy
Pregnant women infected with tuberculosis with a posi-
tive tuberculin skin test but who are asymptomatic form
310
a controversial issue (Vallejo and Starke 1992). Should
they receive preventive treatment, and what are the
guidelines in developed and developing countries?
Authors from developed countries do not adopt
the same policy or guidelines with regard to preven-
tative therapy during pregnancy. Not all infected,
pregnant women in developed countries have the
same chance of developing active tuberculosis. In
the USA, the incidence of tuberculosis infection (not
disease) in the general population is approximately
7% (MedchillI999). Individuals who have a positive
tuberculin test have a 5% chance of developing active
tuberculosis within the first 1-2 years after exposure
and can gain an additional 5% chance over the rest of
their lives (Barnes and Barrows 1993).
As pregnancy has no effect on the development
of active tuberculosis, some clinicians may delay
initiating preventive therapy until after delivery
(Hamadeh et al. 1992). Starke (1997) has debated
the value of initiating isoniazid treatment during
pregnancy for such individuals particularly those
with poor compliance to treatment. However, he
structured an algorithm to serve as a guideline for
clinicians in the USA to help identify those who will
need treatment and when to start it (Fig. 19.2).
Positive Mantoux Sj Test During Pregnancy
Chest Radiograph (abdominal shield)
I
Nonnal Abnonnal
Recent High risk "Older" infection Calcified or
Exposure for likely fibrotic lesion active
y
progression (not active) and tuberculosis
to disease no symptoms
. I.
Start Isom8Z1'd
After delivery
Start isoniazid Start isoniazid
and monitor
carefully
after delivery
if repeat chest
radiograph is
nonnal
M. M. Madkour
Starke recommended the initiation of preventive
therapy to those co-infected with HIV. The cost-effec-
tiveness of preventive therapy was reported in the
USA by Medchill from Phoenix in 1999. His study was
based on the data about the chance of development of
active TB among those with a positive tuberculin test
and found preventive therapy to be cost-effective.
Isoniazid is given as a single daily dose of 300 mg
for 6-12 months. Hepatotoxicity should be monitored
at least monthly (Hamadeh and Glassroth 1992).
In Great Britain, preventive therapy is less widely
used in asymptomatic women with a positive tuber-
culin test but may be recommended for refugees and
new immigrants (Ormerod 2001).
In developing countries, preventive treatment
during pregnancy is not an adopted policy as tuber-
culosis is highly endemic, and the prevalence of
a positive tuberculin test is high. In a large study
from Chile, 840 pregnant women in the third tri-
mester were assessed by tuberculin test. Over 50% of
them had a positive tuberculin test, none developed
symptoms of tuberculosis during pregnancy, and
93% were followed up for 1 year with no evidence of
active disease. The Chilean Ministry of Health policy
is 'not to perform tuberculin testing as a means of
Consistent with
or symptoms
I
Three sputum AFB
smears and
cultures
I
Start multidrug
therapy Fig. 19.2. Evaluation and treatment
(isoniazid, of a pregnant woman with a positive
rifampin, Mantoux skin test. AFB, acid-fast bacilli
ethambutol, (Starke's algorithm, Starke 1997, from
usually Clinics in Perinatology; with permis-
pyrazinamide) sion from Dr. Stark)
Tuberculosis and Pregnancy
identifying infected women and not to treat asymp-
tomatic tuberculin-positive individuals' (Sepulveda
et al. 1995).
References
Adhikari M et al (1997) Tuberculosis in the newborn: an
emerging disease. Pediatr Infect Dis J 16:1108-1112
Ahmed Yet al (1999) A study of maternal mortality at the uni-
versity teaching hospital, Lusaka, Zambia: the emergence
of tuberculosis as a major non-obstetric cause of maternal
death. Int J Tuberc Lung Dis 3:675-680
American Academy of Paediatrics (1992) Committee on
infectious diseases: screening for tuberculosis in infants
and children. Pediatrics 93:313
American Thoracic Society (1993) Control of tuberculosis in
the United States. Am Rev Respir Dis 146:1623
Ando Y et al (1989) Effect of freeze thawing breast milk on
vertical HTLV-l transmission from seropositive mothers to
children. Jpn J Cancer Res 80:405
Barnes PF, Barrows SA (1993) Tuberculosis in the 1990's. Ann
Intern Med 19:400-410
Bazaz-Malik G et al (1983) Tuberculosis endometritis. A clini-
copathological study of 1000 cases. Br J Obstet Gynaecol
90:84-88
Berlin C, Lee C (1979) Isonizid and acetylisoniazid disposition
in human milk, saliva and plasma. Fed Proc 38:426
Bjerkedal T et al (1975) Course and outcome of pregnancy in
women with pulmonary tuberculosis. Scand J Respir Dis
56:245-250
Bonebrake CR et al (1978) Routine chest roentgenography in
pregnancy. JAMA 240:2747-2748
Brent RL (1989) The effect of embryonic and fetal exposure to
x-ray, microwaves and ultrasound: counseling the pregnant
and non-pregnant patient about these risks. Semin Oncol
16:346-368
Carter EJ, Mates S (1994) Tuberculosis during pregnancy:
the Rhode Island experience, 1987 to 1991. Chest 106:
1466-1470
Centers for Disease Control (1990) Screening for tuberculosis
and tuberculosis infection in high-risk populations. Rec-
ommendations of the Advisory Committee for Elimination
of Tuberculosis. MMWR Morb Mortal Wkly Rep 39:1-7
Centers for Disease Control and Prevention (1994) COR Cur-
riculum on tuberculosis, what the clinician should know,
3rd edn. Department of Health and Human Services,
Atlanta, pp 6-7
Centers for Disease Control and Prevention (1995) Screening
for tuberculosis and tuberculosis infection in high-risk
populations. MMWR Morb Mortal Weekly Rep 44:19
Cohen JD et al (1952) The tuberculous mother. Am Rev Respir
Dis 65:1-23
Connolly M, Nunn P (1996) Women and tuberculosis. World
Health Statist Quart 49:115-119
Czeizel AE et al (2001) A population-based case-control study
of the safety of oral anti-tuberculous drug treatment dur-
ing pregnancy. Int J Tuberc Lung Dis 5:564-568
Division of Tuberculosis Elimination, Center for Prevention
Services (1991) Center for Disease Control and American
311
Thoracic Society: core corriculum on tuberculosis. US
Public Health Services, New York
Dolin PJ et al (1994) Global tuberculosis incidence and mor-
tality during 1990-2000. Bull WHO 72:213-220
Epidemiology Comments (1996) Report of the review of the
tuberculosis control programme of South Africa, June 10
to 25, 1996: Department of Health, Pretoria, South Africa,
vol 23, pp 2-20
Espinal MA et al (1996) The effect of pregnancy on the risk of
developing active tuberculosis. J Infect Dis 173:488-491
Finn R, Hill CA et al (1972) Immunological responses in
pregnancy and survival of foetal hemograft. Br Med J 3:
150-152
Freth A (1952) Routine x-ray examination of the chest at an
antenatal clinic. Lancet 1:287-288
Fujimori H, Imais (1957) Studies on dihydrostreptomycin
administered to the pregnant and transferred to their
fetuses. Jpn Obstet Gynecol Soc 4:133-149
Fylkesnes K et al (1997) The HIV epidemic in Zambia: socio-
demographic prevalence patterns and indications of trends
among childbearing women. AIDS 11:339-345
Gogus S et al (1993) Neonatal tuberculosis. Pediatr Pathol13:
299-304
Good JT et al (1981) Tuberculosis in association with preg-
nancy. Am J Obstet GynecoI140:492-498
Hageman J (1998) Congenital and perinatal tuberculosis:
discussion of difficult issues in diagnosis of difficult in
diagnosis and management. J Perinatal 18:389-394
Hageman J et al (1980) Congenital tuberculosis: critical
reappraisal of clinical findings and diagnostic procedures.
Paediatrics 66:980-984
Hallum JL, Thomas HE (1995) Full-term pregnancy after
proved endometrial tuberculosis. J Obstet Gyaecol Br Emp
62:548
Hamadeh MA, Glassroth J (1992) Tuberculosis and pregnancy.
Chest 101:1114-1120
Hedvall E (1953) Pregnancy and tuberculosis. Acta Med Scand
147:1-101
Huebner RE et al (1993) The tuberculin skin test. Clin Infect
Dis 17:968-975
Iseman MD (1993) Treatment of multi-drug-resistant tuber-
culosis. N Engl JMed 329:784-791
Kaplan C et al (1980) Placental tuberculosis in early and late
pregnancy. Am J Obstet Gynecol137:858-860
Lawrence RA (1997) A review of the medical benefits and
contraindications to breast-feeding in the United States.
Maternal and Child Health Technical Information Bul-
letin. National Center for Education in Maternal and Child
Health, Arlington
Lawrence RM, Lawrence RA (2001) Given the benefits of
breast-feeding, what contraindications exist? Pediatr Clin
North Am 48:235-251
Llewelyn M et al (2000) Tuberculosis diagnosed during
pregnancy: a prospective study from London. Thorax 55:
129-132
Machin GA et al (1992) Perinatally acquired neonatal tuber-
culosis: report of two cases. Peadiatr PathoI12:707-716
Margono F, Mroueh J, Garely A et al (1994) Resurgence of
active tuberculosis among pregnant women. Obstet Gyne-
col 83:911-914
McKenna MT et al (1995) The epidemiology of tuberculosis
among foreign-born persons in the United States, 1986 to
1993. N Engl JMed 332:1071
312
Medchill MT (1999) Prenatal purified protein derivative skin
testing in a teaching clinic with a large Hispanic popu-
lation. Am J Obstet GynecoI180:1579-1583
Medical Research Council Cardiothoracic Epidemiology
Group (1992) National Survey of Notifications of Tuber-
culosis in England and Wales in 1988. Thorax 47:770-775
Miller KS, Miller JM Jr (1996) Tuberculosis in pregnancy:
Interactions, diagnosis and management. Clin Obstet
GynecoI39:120-142
Mofenson et al (1995) Mycobacterium tuberculosis infection
in pregnant and non-pregnant women infected with HIV
in the women and infants transmission study. Arch Intern
Med 155:1066-1072
Nemir RL, O'Hare D (1985) Congenital tuberculosis. Am J Dis
Child 139:284-287
Nitta AT, Milligan D (1999) Management of four pregnant
women with multi-drug-resistant tuberculosis. Clin Infect
Dis 28:1298-1304
Ormerod P (2001) Tuberculosis in pregnancy and puer-
perium. Thorax 56:494-499
Oxtoby MJ (1988) Human immunodeficiency virus and other
viroses in human milk: placing the issue in broader per-
spective. Pediatr Infect Dis J 7:825-835
Present PA, Comstock GW (1975) Tuberculosis sensitivity in
pregnancy. Am Rev Respir Dis 112:413-416
Rose AMC et al (2001) Tuberculosis at the end of the 20th
century in England and Wales: results of a national survey
in1998. Thorax 56:173-179
M. M. Madkour
Schaefer G et al (1975) Pregnancy and pulmonary tuber-
culosis. Obstet GynaecoI46:706-715
Sepulveda RL et al (1995) The influence of BCG immunization
on tuberculin reactivity in healthy Chilean women in the
third trimester of pregnancy. Tuberc Lung Dis 76:28-34
Snider D (1984) Pregnancy and tuberculosis. Chest 86:
105-135
Snider DE, Powell K (1984) Should women taking anti-tuber-
culosis drugs breastfeed? Arch Intern Med 144:589-590
Snider DE et al (1980) Treatment of tuberculosis during preg-
nancy. Am Rev Respir Dis 122:65-79
Starke JR (1997) Tuberculosis, an old disease but a new
threat to the mother, fetus and neonate. Clin Perinatol 24:
107-127
Steen JSM, Seainton-Ellis DM (1977) Rafampicin in pregnancy.
Lancet ii:604-605
Swartz HM, Reichling BA (1978) Hazards of radiation exposure
for pregnant women. JAMA 239:1907-1908
Vallejo JG, Starke JR (1992) Tuberculosis and pregnancy. Clin
Chest Med 13:693-707
Vorherr H (1974). Drug excretion in breast milk. Postgrad
Med J 56:97-104
Wilson EA et al (1973) Tuberculosis complicated by pregnancy.
Am J Obstet GynecoI115:526-529
World Health Organization (1998) HIV and infant feeding:
guidelines for healthcare managers and supervisors. World
Health Organization, Geneva. Publication WHO/FRH/NUT
98.2, UNAIDS/98.4, UNICEF/PD/NUT/(J) 98.2
20 Post-primary Pulmonary Tuberculosis
M. MONIR MADKOUR, Y. ABusABAAH, ALI BEN MOUSA, ALI AL MASOUD

CONTENTS Post-primary pulmonary tuberculosis is a chronic

disease commonly caused by either endogenous reac-
20.1 Pathogenesis and Pathology 313
tivation of a latent infection or exogenous re-infection
20.2 Clinical Features and Complications 315
20.2.1 Hemoptysis in Pulmonary Tuberculosis 316 by Mycobacterium tuberculosis. It has other synonyms
20.2.2 Massive Hemoptysis in Tuberculosis 316 derived mostly from the route of transmission of the
20.2.3 Rasmussen Aneurysm 317 infection or from the age of the patient at the onset
20.2.4 Endobronchial Tuberculosis 318 of the disease, including endogenous reactivation pri-
20.2.5 Bronchiectasis as a Complication
mary tuberculosis, exogenous re-infection pulmonary
of Post-primary Pulmonary TB 318
20.2.6 Pneumothorax as Complication tuberculosis, or adult-onset pulmonary tuberculosis.
of Post-primary Pulmonary TB 318 The term being used here is 'post-primary pulmonary
20.2.7 ARDS and Tuberculosis 319 tuberculosis'to include both re-infection and reacti-
20.2.8 Tuberculous Bronchopleural Fistula 319 vation forms. The clinical features of the disease are
20.2.9 Aspergilloma (Mycetoma) 319
not specific, and the imaging features are suggestive
20.2.10 Tuberculomas 320
20.2.11 Tuberculous Pulmonary Gangrene 320 but can simulate other diseases. The definitive diag-
20.2.12 Tuberculosis, Lung Cancer nosis depends on the identification of M. tuberculosis
and Other Neoplasia 320 bacilli, using conventional microbiological methods
20.3 Diagnosis of Post-primary Pulmonary TB 320 of sputum smear and culture or radiometric culture
20.3.1 Imaging Features of Post-primary
methods such as BACTEC or DNA probe PCR-based
Pulmonary TB 320
20.3.2 Microbiology of the Sputum assays which can identify drug-resistant strains as well.
and Bronchial Aspirate 321 Sputum smear and culture remain the most important
20.3.3 Bronchoscopic Diagnosis 321 investigative methods. Smear-negative sputum may
20.3.4 Tuberculin Skin Test 322 delay the diagnosis for 4-8 weeks or longer if the cul-
20.3.5 Other Diagnostic Investigations 322
ture is also negative. A presumptive diagnosis based on
20.4 Treatment of Post-primary Pulmonary TB 322
20.4.1 Response to Treatment 324 the clinical and radiographic features should be made
References 324 with initiation of treatment after the exclusion of other
possible causes of the radiographic findings.
Infection with HIV is a great risk for the develop-
ment of either endogenous reactivation or exogenous
re-infection pulmonary disease (Millar and Horne
1979; Barnes et al. 1991; Heyderman et al. 1998). The
development of drug-resistant pulmonary tubercu-
losis is a real global concern that indicates a failure
M. M. MADKouR, MD, DM, FRCP of the tuberculosis control program. The current
Consultant, Department of Medicine, Riyadh Armed Forces treatment is both toxic and expensive, and new drug
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
development is sparse at present.
Y. ABusABAAH, ABIM
Fellow, Respiratory Medicine, Department of Medicine,
Riyadh Armed Forces Hospital, P.O. Box 7897, Riyadh 11159,
Saudi Arabia
A. BEN MousA, MD 20.1
Department of Medicine, Riyadh Armed Forces Hospital,
Pathogenesis and Pathology
P.O. Box 7897, Riyadh 11159, Saudi Arabia
A. AL MAsouD, MD
Department of Medicine, Riyadh Armed Forces Hospital, The development of post-primary pulmonary tuber-
P.O. Box 7897, Riyadh 11159, Saudi Arabia culosis as a result of endogenous reactivation or exog-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
314
enous re-infection in low-risk and high-risk areas has
been debated for many years among authors (Glynn
et al. 2001). Balasubramanian and colleagues reviewed
the world literature on these issues and added their
own views (Balasubramanian et al. 1994). Schools
of thought adopted by those researchers based on
microbiological studies of tissue specimens from
the lung and lymph nodes of the primary complex.
Some authors quoted by reviewers suggested that the
primary complex is sterile within 5 years, while other
authors suggest that virulent bacilli lie dormant in a
metastatic site seeded hematogenously within the vul-
nerable region. Cultures of apical lung lesions yielded
viable bacilli in 25%-76% as reported by some of the
quoted authors, and transmission via the bloodstream
was suggested.
Other researchers examined the sputum and urine
cultures in patients living in Bangalore, India, and com-
pared the bacilli virulence, INH sensitivity, and phage
typing of the isolates and drew conclusions about the
biological evidence of exogenous re-infection in that
study. Epidemiologists have noted that high- or low-
risk incidence rates of infection in different areas of
the world played an important role (in exogenous re-
infection or endogenous reactivation). In areas with a
risk greater than 1% of developing countries, with the
likelihood of repeated episodes of droplet infection
via the airway, the disease is more likely to be due to
exogenous re-infection. In areas of low-risk (less than
0.05%), as in developed countries, endogenous reac-
tivation is more likely to be the cause of the disease.
Some other authors gave evidence of the contributions
of both exogenous re-infection as well as endogenous
reactivation. Balasubramanian et al. hypothesized that
the implantation of the vulnerable region is directly
transmitted via the airway (exogenous re-infection).
They also concluded that more studies using genetic
fingerprinting of Mycobacterium bacilli isolates would
help in the future in disease control in endemic areas
and in the development of new vaccines. Immune sup-
pression resulting from HIV infection leads to higher
rates of co-infection with tuberculosis.
Recent molecular epidemiological studies have
indicated that up to 40% of newly diagnosed tuber-
culous patients and over 70% of recurrences may
be due to exogenous re-infection (Stead and Bates
2000; Bates et al. 2001). Among the immunosup-
pressed patients due to HIV, exogenous re-infec-
tion was the leading mechanism. Endogenous
reactivation contributed to only 16% of patients in
a low-risk area in developed countries and more
than 70% in a high-risk area in developing coun-
tries (Alland et al. 1994; Brande et al. 1998; Small et
M. M. Madkour et aI.
al. 1994; van Rie et al. 1999; McDonough et al. 2000;
Rook and Zumla 2001; Caminero et al. 2001; Ban-
dera et al. 2001; Wallis and Johnson 2001).
Caminero and colleagues (2001) reported 912
patients with culture-proved pulmonary tuberculo-
sis between 1991 and 1996 on a Spanish island with
a moderate risk rate of incidence. They were treated
and followed up for at least 12 months after complet-
ing chemotherapy and were culture-negative. Twenty-
three patients (2.5%) became culture positive again.
DNA fingerprinting results were available from 18
patients with recurrence of the disease. DNA finger-
printing was available of pretreatment and recurrent
isolates and was reviewed. In 8 patients (44%), the
genotype of the recurrent isolate showed a different
pattern to the pretreatment isolate. These authors
concluded that in 8 patients, 2 of them HIV-positive,
exogenous re-infection was the cause of the recur-
rent pulmonary tuberculosis. Immune suppression or
reduction of immune responses may also occur in cer-
tain diseases such as silicosis, diabetes mellitus, or with
corticosteroid and other immunosuppressive drugs
used for malignancies or connective tissue diseases.
Regardless of the mode of infection (exogenous or
endogenous), the apex of the lung (apical-posterior
segments) is the most common site of post-primary
pulmonary tuberculosis. Specific factors permitting
the progression of tuberculosis in most cases are not
yet known (Wallis and Johnson 2001).
The host immune response and the role of the cell-
mediated immunity of activated macrophages and
T-cells and the expression of cytokines (Garcia et al.
2002) in response to M. tuberculosis glycolipids and
lipoproteins have been discussed in the pathogen-
esis of primary pulmonary tuberculosis in Chap. 17.
Post-primary pulmonary tuberculosis commonly
affects the apical and posterior segments of the
upper lobe or the superior segment of a lower lobe.
After its localization, inflammatory granulomatous
nodular formations with cellular infiltrates, fibrosis,
central necrosis, and caseation may take place. Hilar
and paratracheal lymphadenitis (Woodring 1986)
in post-primary tuberculosis is a rare occurrence
and reported in approximately 5% of patients (see
Figs. 23.18, 23.20, 23.21, 23.22, 23.39, 23.40b). Heal-
ing with fibrosis and calcification (Fig. z.25) of lung
parenchymal lesion may occur, leading to traction
of the trachea (Fig. 23.21, 23.22, 23.23), and in late
advanced stages when the lung is destroyed and
replaced by extensive fibrosis, displacement of the
mediastinum may occur (Fig. 23.24).
The initial upper lobe infiltration may form a
pneumonic consolidation, and cavitation often
Post-primary Pulmonary Tuberculosis 315

occurs in 40%-80% of patients (Fig. 23.20, 23.26a, b) sheep with udder tuberculosis. It can also be transmit-
(Rohenberg and Shaw 1996). Tuberculous consolida- ted to humans from cows as an airborne infection
tion with cavity formation may expand and form a (Geppert and Leff 1979; Liesegang and Cameron 1980;
lung abscess discharging large numbers of the bacilli Sauret et al. 1992). The incidence ofM. bovis in humans
in the sputum as was noted in one of our patients is higher in rural areas with infected herds (Moda et al.
(Fig. 23.26d). Tuberculous cavities with areas of 1996). Ingestion of contaminated milk, particularly by
necrosis and caseation may rupture into the pleura, young children living in developing countries, leads to
leading to empyema (Fig. 23.34b) or a bronchopleu- establishment in the cervical and less frequently in the
raj fistula (Fig. 23.36). Rupture of a tuberculous cavity axillary lymph nodes (scrofula). The bacilli may also
into the trachea or bronchi is a common complica- affect the intestine, kidney, bones, and central nervous
tion and occurs in up to 40% of patients (Rohenberg system in endemic areas with a similar pathogenicity
and Shaw 1996). Bronchogenic spread of necrotic and to M. tuberculosis (Moda et al. 1996). HIV patients
caseous tissue loaded with the bacilli to other parts of co-infected with M. bovis have been reported in San
the lung fields on the same side or opposite lung may Diego, USA, near the border with Mexico (Dankner
occur (Figs. 23.19,23.20, 23.27b,c, 23.28b,c). Infection et al.1993). Nontuberculous mycobacterial organisms
starts at the newly seeded sites with nodular infiltrate may also cause a similar pulmonary infection (see
bronchopneumonia or similar tissue destruction and Chap. 11).
fibrosis to the lung parenchyma (Figs. 23.25, 23.26a,
z.27b, c, 23.28b). A severe form of transbronchial
spread of infection takes the form of dissemination
in both lungs (Fig. 23.28c). 20.2
Endobronchial involvement of the bronchial wall is Clinical Features and Complications
common, leading to scarring and luminal narrowing
and post-stenotic emphysema (Figs. 23.22b, 23.30a, b, Post-primary tuberculosis often presents with a
23.25, 23.31b), atelectasis, and cystic or tubular bron- gradual onset of symptoms that may be tolerated by
chiectasis due to traction or endobronchial fibrosis as the patient. The duration of symptoms before pre-
a common complication of the tuberculous disease sentation may vary widely, from 3 days to 23 months
process. It is often located in the upper lobes but can (Dahmash et al. 1995; Maartens and Beyers 2002).
occur at any other site (Figs. 23.30a, b, 23.31a, b). Post- The presenting features may initially be related to the
tuberculous bronchiectasis, although often asymp- respiratory system or present as constitutional symp-
tomatic, may cause hemoptysis in these patients. toms or both. Cough is the most frequently reported
A tuberculous cavity larger than 25 mm in diameter presenting feature. Initially, it is dry but later becomes
may persist long after successful antibiotic treatment. productive. The sputum may be mucoid, muco-puru-
Colonization with fungi forming a ball of mycetoma lent, blood-stained, or with massive hemoptysis. Chest
may rarely occur (Fig. 23.26b, 23.37a,b), and hemopty- pain due to associated pleurisy or pneumothorax may
sis may rarely be a presenting complaint. be a presenting symptom. Dyspnea due to tuberculous
Tuberculous granulomatous tissue nodules may pneumonia or bilateral fibrocavitary disease may be
be encapsulated with connective tissue during vari- the presenting symptom. Fever with sweating and
able stages of disease healing and activity, leading chills are common, particularly at night. Other consti-
to tuberculomas found in 3%-6% of patients. It may tutional symptoms including weakness, anorexia, and
be single or multiple with a central area of necrosis weight loss, which are nonspecific, may also be present
or calcification (Figs. 23.32a-d). Tuberculomas may (Dunlap et al. 2000; Johnson and Ellner 2000).
occur in both primary and post-primary pulmonary Symptoms related to extrapulmonary tuberculo-
tuberculosis. sis such as tuberculous monoarthritis, Pott's disease,
In bovine tuberculosis, the M. bovis bacilli are genitourinary symptoms, or other organ involve-
bacteriologically distinctive from M. tuberculosis but ment might accompany the respiratory symptoms at
otherwise cause identical pathogenesis, lung lesions, the time of presentation. We retrospectively reviewed
and clinical disease. In humans, M. bovis is a zoonotic the records of 176 adult patients with post-primary
disease that has been virtually eliminated in developed pulmonary tuberculosis who had a positive sputum
countries and other developing countries that practice culture for M. tuberculosis. They attended our hospi-
pasteurization of milk and immunization of dairy tal between 1998 and 2000. The frequency of history/
herds (Grange et al. 1994; Dannenberg 2001). The symptoms, physical findings, and imaging features of
organism is excreted in the milk from cows, goats, and these patients are presented in Table 20.1.
316 M. M. Madkour et al.

Table 20.1. Clinical features of 176 patients with culture-posi- be found (Fig. 23.21). Chest wall retraction due to
tive post-primary pubnonary tuberculosis fibrosis or prominence due to associated pleural
Clinical features No. of patients Percentage effusion may be found. Features of consolidations
with crepitation and bronchial breathing may be
Symptoms:
Cough 166 94.3
detected. Localized wheezes may be present, indicat-
Expectoration 128 72.7 ing endobronchial disease. Other systemic features
Hemoptysis 40 23.0 including pallor, rarely clubbing of the fingers and
Fever 135 76.7 toes in chronic cases, weight loss, dyspnea, and other
Sweating 80 45.4 extrapulmonary involvement may be found on physi-
Weight loss 98 55.7
Dyspnea 76 43.2
cal examination.
Chest pain 58 32.9 Patients with post-primary pulmonary tubercu-
Signs: losis may present with clinical features similar to
Crackles 105 59.6 those of community-acquired pneumonia. Al-Zeer
Localized wheezes 23 13.0 and colleagues (1998) reported a series of 64 patients
Bronchial sound 55 31.2
Diminished sound 66 37.5
admitted to hospital during the pilgrimage season
Clubbing 20 11.4 to Mecca in 1994 with an initial diagnosis of com-
Extrapulmonary 16 9.0 munity-acquired pneumonia. All patients came from
Bronchoscopy 28 16.0 developing countries. Microbiologically proven M.
tuberculosis was found in 13 patients (20.3%) and
was the most common cause of pneumonia among
The frequency of these symptoms may vary, being this group.
more intense among those co-infected with HIV
(Corbett et al. 2000).
The frequency of symptoms was also reported 20.2.1
from a large tuberculosis center in Riyadh for 1566 Hemoptysis in Pulmonary Tuberculosis
hospitalized patients with pulmonary tuberculosis
during 1983-1987. Fever and constitutional symp- Hemoptysis is often an alarming presenting symp-
toms were reported in 77.7%, cough with or without tom in patients with tuberculosis. It may vary in
expectoration in 94.3%, and hemoptysis in 40.7% (Al- severity from a slight tinge of blood mixed with the
Hajjaj et al. 1991). In another series from the south of sputum, to mild, moderate, or severe, massive, life-
Saudi Arabia, Al Wabel et al. (1995) reported on 190 threatening hemoptysis.
patients with post-primary pulmonary tuberculosis In our series of 176 patients with microbiologically
who were hospitalized over a 2.5-year period. Cough proved pulmonary tuberculosis, 28 patients (16%)
was noted in 84%, expectoration in 65%, hemoptysis had slight to mild hemoptysis that was managed with
in 23%, fever and constitutional symptoms in over chemotherapy and conservative treatment. Active
40% of patients. Pulmonary tuberculosis in elderly pulmonary tuberculosis with or without cavitation or
patients with underlying and concomitant other post-tuberculous bronchiectasis is the most common
chronic illnesses in 80 patients were reported from cause among these patients. Rarely, the development
Riyadh (Dahmash et al. 1995). These patients had of bronchial carcinoma in these patients as well or
diabetes mellitus or malignancies, and some were mycetoma or other unrelated illnesses could be the
on steroids, and these illnesses occurred in 86% of cause of hemoptysis in tuberculous patients (Steb-
them, but none had HIY. Cough was noted in 85%, bings and Lim 1999; Hirshberg et al. 1997).
expectoration in 60%, fever in 66%, hemoptysis in
17.5%, anorexia, weight loss, and other constitutional
symptoms in over 50%. Choyke and colleagues (1983) 20.2.2
reported on 103 patients with adult-onset pulmonary Massive Hemoptysis in Tuberculosis
tuberculosis: 85% were symptomatic, with fever in
40%, cough in 37%, weight loss in 23.6%, and hemop- Life-threatening massive hemoptysis due to active
tysis in 8%. tuberculosis or post-tuberculous bronchiectasis has
Physical examination of the chest may be normal been reported as 'common in comparison to other
despite the presence of pulmonary infiltrations causes listed in many publications', as reported by
depicted by chest radiography. Displacement of Hsiao et al. (2001) from the USA. Life-threatening,
the trachea due to lung fibrosis and collapse may massive hemoptysis is defined by most authors as
Post-primary Pulmonary Tuberculosis
expectoration of at least 200 ml of blood in 24 h,
significant drop of hemoglobin requiring blood
transfusion, or failure to respond to conservative
treatment such as oxygen supplement, morphine,
and antibiotics (Wong et al. 2002; Abal et al. 2001;
Lee et al. 2000; Hsiao et al. 2001; Conlan et al. 1983).
In South Africa, Conlan et al. (1983) reviewed 123
patients with massive hemoptysis due to different
causes. Tuberculosis was by far the most common
cause, either due to culture-proved active disease
or bronchiectasis as a sequela to tuberculosis. These
authors reported that active pulmonary tuberculosis
as a cause of massive hemoptysis was found in 47
patients (38%),24 men and 23 women aged between
19 and 60 years. Thirty-seven patients had bronchi-
ectasis, and 17 (45.9%) had bilateral upper lobe bron-
chiectasis secondary to former tuberculosis. The total
number of patients with massive hemoptysis with or
who had tuberculosis was 64 (52%). The remaining
causes were chronic narcotizing pneumonitis in II,
lung abscess in 6, lung cancer in 6, primary fungal
pneumonia in 4, bronchovascular fistula in 5, and
miscellaneous causes in 7 patients.
In France, Mal and colleagues (1999) reported the
intermediate and long-term outcome of bronchial
artery embolization (BAE) performed on 46 patients
with massive hemoptysis. Tuberculosis (active or
sequela) was found in 23 patients (50%) as a cause of
the massive hemoptysis. Idiopathic in 10, bronchiec-
tasis in 4, lung cancer in 2, and 1 patient each due to
other various causes.
The outcome of BAE was favorable with an imme-
diately successful result in stopping the bleeding, but
complications were also frequent. Revascularization
may occur with recurrence of hemoptysis, technical
failure in BAE, spinal cord injury related to invis-
ible anastomotic connections between the bron-
chial circulation and the anterior spinal artery. The
authors recommended that BAE should be avoided
in patients with minor hemoptysis.
In the USA, Hsiao et al. (2001) from Stanford
reported the assessment of modalities of investiga-
tions used for the localization of the site of bleeding
in patients with massive hemoptysis. They reviewed
the records of 28 patients seen between 1988 and
2000 who presented with massive hemoptysis. They
found 16 patients with tuberculosis (57%), 14 with
tuberculous bronchiectasis, and 2 with active disease,
bronchogenic carcinoma in 2, and other causes had
1 patient each. These authors noted, 'Contrary to the
statistics reported in many recent series, tuberculous
bronchiectasis is the most common underlying etiol-
ogy for massive hemoptysis in our patients'.
317
The authors also indicated that the overall mortal-
ity of patients with massive hemoptysis was 7%-80%
and operative mortality was 30%-40% as they found
from reviewing the literature on the subject. These
authors reported, 'The angiographic signs in hemop-
tysis include hyperplasia of the bronchial artery trunk
and branches, bronchopulmonary anastomoses, and
bronchial arterial aneurysms', which were seen on
angiography of these patients. They concluded that
fibroptic bronchoscopy before BAE is unnecessary in
patients with hemoptysis of known origin.
In Kuwait, Abal et al. (2001) prospectively studied
52 hospitalized patients with hemoptysis of variable
degrees of severity over a period of 1 year. Twenty had
blood-stained sputum, and 32 had frank hemoptysis
including 16 with massive blood expectoration. They
found that pulmonary tuberculosis (active or old) was
the most common cause, found in 17 patients (32.7%).
Other causes included carcinoma in 5, bronchitis in 3,
1 patient each due to other causes, and unknown cause
in 13 patients. These authors managed 80.8% of their
patients conservatively, and only 19% required BAE or
surgery. Recurrent hemoptysis occurred in 12% at the
I-year follow-up in this series.
In a retrospective study of BAE for massive hemop-
tysis, Wong et al. (2002) from South Africa reported
on 165 patients, and bilateral post-tuberculous
bronchiectasis was the most common cause (75%).
The short-term outcome of BAE was satisfactory in
controlling hemoptysis in all patients. Thoracic aor-
tography was done during the initial assessment, and
they noted pathologic enlargement of the bronchial
arteries and the presence of nonbronchial systemic
collaterals with arteriovenous shunting including
intercostal arteries. One patient had transient para-
paresis, and the authors noted that the presence of
spinal arteries was not considered a contraindication
to embolization.
20.2.3
Rasmussen Aneurysm
This is a rare phenomenon that can cause massive,
life-threatening hemoptysis. It involves invasion of a
peripheral pulmonary artery located within a tuber-
culous cavity often in the upper lobe. Invasion of the
arterial wall by tuberculous granulation tissue leads
to granulomatous vasculitis with replacement of the
adventitia and media with fibrin during the process
of healing. This will lead to weakening of the arterial
wall with mycotic pseudoaneurysmal formation in
one or more locations (see Chapter 43). Rupture of
318 M. M. Madkour et al.

the aneurysm will lead to massive, life-threatening Table 20.2. Imaging features of 176 patients with culture-posi-
hemoptysis as reported in 5% of postmortem cases tive post-primary pulmonary tuberculosis
(Winer-Muram and Rubin 1990; Kim et al. 2001). Imaging features No. of patients Percentage

Unilateral-parenchymal 119 67.6

Bilateral-parenchymal 57 32.3
20.2.4 Infiltrates and consolidation 144 81.8
Endobronchial Tuberculosis Cavities 61 34.6
Bronchiectasis 17 9.6
Endobronchial tuberculosis is a common complication Destroyed lungs 2 1.1
Calcifications 22 12.5
of post-primary pulmonary tuberculosis. It may occur (LN, parenchymal and pleural)
in up to 40% of patients, and the most common source Pneumothorax 8 4.5
of bronchial wall infection is a contiguous tuberculous Lymphadenopathy 19 10.8
cavity (Rohenberg and Shaw 1996). Rarely, the bron- (hilar and paratracheal)
chial wall may be affected by a hilar or paratracheal Lung infiltrates with pleural 46 26.1
effusion
tuberculous caseating lymphadenitis. Hematogenous Mycetoma 2 1.1
or lymphatic spread of infection to the bronchial wall Tuberculomas 3 1.7
may occur (Buckner and Walker 1990). The role of the Bronchopleural fistula 1 0.5
bronchial tree in the spread of tuberculosis to other
parenchymal segments or lobes on the same or oppo-
site side by transbronchial spread of caseous material
leading to bronchopneumonia is well recognized (1m et
al.1993). Hoarseness ofvoice due to laryngeal involve-
ment and localized wheezes due to bronchial disease apical and posterior segments of the upper lobes are
in the chest are common clinical features. The imag- the most common sites of tuberculous bronchiectasis
ing features of plain radiography and CT may show (Fig. 23.27c). Chest radiography may show ring shad-
multiple nodular opacities, bronchial wall thickening, ows with occasional fluid levels (Figs. 23.9 and 23.29).
post-stenotic dilatation, lobar hyperinflation, pulmo- Features on CT and HRCT will show the details of the
nary collapse, and atelectasis (Figs. 23.27 and 23.28). bronchial wall changes (Figs. 23.28c, 23.30, and 23.31)
Endobronchial tuberculosis, its definition, epide- (McAdams et al. 1995). Tuberculous bronchiectasis
miology, pathogenesis and pathology, classification, may also occur in the lower lobes of the lung. It is
clinical features, diagnosis, the role of bronchoscopy usually asymptomatic, but secondary bacterial infec-
in the diagnosis and management, and the treatment tion or hemoptysis may be a presenting feature.
are considered in a separate chapter (see Chap. 21).

20.2.5
Bronchiectasis as a Complication
of Post-primary PulmonaryTB
Bronchiectasis is a common sequela of pulmonary
tuberculosis (primary and post-primary). It may occur
primarily as a result of endobronchial tuberculosis
with irreversible bronchial wall dilatation (Lee et al.
1991). Secondary tuberculous bronchiectasis occurs as
a result of lung parenchymal destruction with fibrosis
(traction bronchiectasis). Based on high-resolution
CT, bronchiectasis was found in 27% of patients with
pulmonary tuberculosis, and the upper lobes were the
most common site (Cartier et al. 1999).
In our series of 176 patients with pulmonary
tuberculosis, 17 (9.6%, see Table 20.2) had imaging
features of tuberculous bronchiectasis (Figs. 23.9,
23.27c, 23.28c, 23.29, 23.30, and 23.31). Typically, the
20.2.6
Pneumothorax as Complication
of Post-primary PulmonaryTB
Pneumothorax may occur during active post-pri-
mary pulmonary tuberculosis. Rupture of a tuber-
culous cavity contiguous to the pleura may result in
pneumothorax. The incidence has been reported as
low, ranging from 0.6% to 1% in hospitalized tuber-
culous patients (Wilder et al.1962; Ihm et al.I972). In
our own series of 176 patients (see Table 20.2), pneu-
mothoraxoccurred in 8 (4.5%). In 1 patient,pneumo-
thorax failed to respond to antituberculous treatment
and chest tube insertion, and a bronchopleural fistula
was suspected. A fistulogram confirmed the presence
of a bronchopleural fistula and active tuberculosis in
the apical segment of the left upper lobe (Fig. 23.36).
Pneumothorax may also occur in treated and healed
pulmonary tuberculosis (Lambert 1956).
Post-primary Pulmonary Tuberculosis 319

20.2.7 20.2.8
ARDS and Tuberculosis Tuberculous Bronchopleural Fistula

Acute respiratory distress may occur in patients with Tuberculous bronchopleural fistula is rarely reported
bilateral chronic cavitary or bronchogenic pulmo- nowadays with the recent advances in chemotherapy.
nary tuberculosis with a high hospital mortality rate Most recent reports are on patients with late complica-
of up to 47%. Dyer and Potgieter (1984) described tions of collapse therapy for pulmonary tuberculosis
three adult patients from South Africa with adult that was done in the past (Johnson et al. 1973; Iseman
respiratory distress syndrome (ARDS) due to pul- and Madsen 1991; Uchida et al. 1999; Weissberg and
monary tuberculosis (nonmiliary). Weissberg 2001). It may develop after lung resection
A 31-year-old woman with cough, fever, and for pulmonary tuberculosis. The lung parenchymal
dyspnea lasting 2 months had bilateral pulmonary tuberculous cavity may rupture into the pleural space
tuberculosis, and because of the severe dyspnea and with pneumothorax, and the diagnosis can be made
the abnormal arterial blood gases, she was mechani- by fistulography (Fig. 23.36 radiology of pulmonary
cally ventilated. The diagnosis of tuberculosis was TB chapter). A tract may form between the bronchus
made by sputum-positive direct smear. Antituber- contiguous to the tuberculous cavity and the pleura,
culous treatment was started, but the patient died. producing a bronchopleural fistula. It is estimated that
Postmortem examination showed left upper and a bronchopleural fistula can occur in patients with
right lower lobe cavitary tuberculosis. The second active extensive pulmonary parenchymal tuberculosis
patient was a 22-year-old woman with a similar pre- in less than 1% (Miller 1981; Woodring 1986; Winer-
sentation who died in the ICU with postmortem evi- Muram and Rubin 1990). Patients are usually symp-
dence of bilateral tuberculous bronchopneumonia tomatic, and sputum production may increase. Plain
with a left bronchopleural fistula. The third patient chest radiography may show air in the pleural space,
had a similar presentation but did not require a changing air-fluid level, and contralateral spread of
ventilation and responded well to antituberculous tuberculous infiltration. CT may depict the sites of the
treatment. bronchopleural fistula (Kim et al. 2001).
Levy and colleagues (1987) from South Africa The management of these patients is usually
reported a retrospective study of 15 patients admit- difficult. Patients with a susceptible organism may
ted to the ICU between January 1982 and June 1985 respond to antituberculous treatment and intercos-
with a confirmed diagnosis of pulmonary tuberculo- tal tube drainage. Patients with multidrug-resistant
sis. These patients required ICU admission for respi- tuberculosis are treated with chemotherapy and
ratory failure. They comprised 1.5% of 933 patients lobectomy or pneumonectomy and pleural decorti-
with pulmonary tuberculosis hospitalized in Hillbow cation unless found to have respiratory insufficiency
Hospital. Eleven patients required ventilation, and 5 (Iseman and Madsen 1991). Endobronchial occlusion
died. Miliary tuberculosis was found in 6 patients, by coils has been found to be an effective method
tuberculous bronchopneumonia in 5 patients, lobar (Uchida et al. 1999).
pneumonia in 3 patients, bilateral lung parenchymal
destructive disease in 2.
Penner and colleagues (1995) from Canada 20.2.9
reviewed the records of 13 patients with confirmed Aspergilloma (Mycetoma)
pulmonary tuberculosis (7 women and 6 men) from
1984 to 1993, and all had respiratory failure requir- A chronic tuberculous cavity may be colonized by
ing mechanical ventilation. Seven patients had mili- the spores of Aspergillus fumigatus to form a fungal
ary or disseminated tuberculosis and 6, tuberculous ball (Aderaye and Jajaw 1996). It may occur in 11% of
pneumonia. Nine patients died, and only 4 survived. patients with a chronic tuberculous cavity, and approx-
M. tuberculosis was isolated in all patients from imately 25%-55% of patients with aspergilloma have
respiratory or nonrespiratory sites. These authors a history of the disease (Kim et al. 2001). Patients with
also quoted the work of Agarwal and colleagues aspergilloma are often asymptomatic but may present
(1977) regarding 16 patients with respiratory fail- with hemoptysis (Kaestel et al. 1999). Plain radiography
ure due to tuberculous pneumonia who required of the chest may depict a mobile, rounded mass with
mechanical ventilation. air-crescent ring (Fig.23.26b and 23.37a,b). Surgical
resection is essential if associated with hemoptysis
and systemic antifungal treatment is ineffective.
320
20.2.10
Tuberculomas
Solitary or multiple, round or oval, tuberculous
pulmonary mass lesions may be noted on the chest
radiographs in approximately 5% of patients (Kim et
al. 2001; Winer-Muram and Rubin 1990). They may
be the only radiographic manifestation of primary or
post-primary pulmonary tuberculosis. Their diam-
eter may range between 0.5 and 4.0 cm or larger with
a smooth or sharply defined margin. Central case-
ation or calcifications may be found in 20%-30% of
tuberculomas. Satellite lesions may be seen in up to
80% of these lesions.
20.2.11
Tuberculous Pulmonary Gangrene
An extremely rare but fatal complication of tuber-
culous cavitary lesion is the involvement of adjacent
vessels and the development of arterial and venous
vasculitis with thrombosis, leading to pulmonary gan-
grene (Reich 1993). Khan and colleagues (1980) from
New York reported on 4 patients with pulmonary
gangrene due to pulmonary tuberculosis, and only
1 survived. The first patient was a 55-year-old man
who presented with fever, cough, and hemoptysis
lasting 3 months. Chest radiography showed bilateral
upper lobe infiltrate with cavitation and intracavitary
mass. The sputum smear was positive for the bacilli.
Chemotherapy was started, but the patient developed
pneumothorax and died. Autopsy showed tuberculous
pneumonia with extensive tuberculous arteritis and
occlusion of the lumen by thrombosis.
The second patient died before establishing the
diagnosis, and subsequent sputum culture taken on
admission grew M. tuberculosis. Autopsy showed
pulmonary vasculitis in arteries and veins contigu-
ous with the tuberculous cavity. The third patient was
known to be tuberculous before and had undergone
lobectomy, and now presented with fever and a large
pulmonary cavity with an intracavitary mass in the
right upper lobe. The sputum smear was positive, and
culture grew M. tuberculosis. She responded well to
chemotherapy. The fourth patient had a similar chest
radiographic cavity with intracavitary mass and died.
The bacilli were found at autopsy as well as lung paren-
chymal cavities, tuberculous granulomatous vasculitis
with thrombosis of the pulmonary arteries and veins.
We reviewed the world literature and found 18
previously reported patients with pulmonary gan-
grene mostly due to Klebsiella pneumonia and other
M. M. Madkour et aI.
organisms but not due to M. tuberculosis. Lopez-Con-
treras et al. (1994) reported a 61-year-old alcoholic
man with a 4-month history of cough, fever, and
weight loss. Sputum smear and culture were positive
for M. tuberculosis. Chest radiography showed a large
cavity with air-fluid level and a free-floating mass.
The patient died, and autopsy was denied.
20.2.12
Tuberculosis, Lung Cancer and Other Neoplasia
The relationship between pulmonary tuberculous
and lung cancer has been frequently debated. Such a
relationship was raised by Greenberg and colleagues
(1964), indicating the co-existence of carcinoma and
tuberculosis of the lung. The co-existence between
bronchogenic carcinoma and tuberculosis creates
a difficult diagnostic problem for radiologists as
the radiographic changes may be misinterpreted as
progression of tuberculosis (Kim et al. 2001; Winer-
Muram and Rubin 1990).
Brown and Almenoff (1992) reviewed the litera-
ture of various retrospective studies on this co-exis-
tence. They reviewed other malignancies including
leukemia, lymphoma, myelofibrosis, head and neck
malignancies, as well as the use of immunosuppres-
sive chemotherapy, and their relationship with the
development of tuberculosis. They noted that tuber-
culosis was 6-9 times more common among patients
with Hodgkin's disease, lung cancer, and non-Hodg-
kin's lymphoma than those with other malignancies.
These authors noted, 'Tuberculosis was more likely
to be diagnosed at the time of tumor diagnosis in
patients with lung, head and neck malignancies,
and disease in these patients was predominantly
confined to the lungs'. Profound suppression of the
cell-mediated immune response caused by malignant
diseases or as a result of severe immunosuppressive
chemotherapy is the most likely cause of the high
risk of developing tuberculosis among these patients
(Brown and Almenoff 1992).
20.3
Diagnosis of Post-primary Pulmonary T8
20.3.1
Imaging Features of Post-primary PulmonaryTB
Various imaging modalities are used for the
depiction of features of post-primary pulmonary
Post-primary Pulmonary Tuberculosis
tuberculosis. The imaging features of post-primary
pulmonary tuberculosis can be broadly classified as
lung parenchymal disease with cavitation, endobron-
chial tuberculosis, pleural extension of the disease,
and other complications such as tuberculoma and
mycetoma.
Conventional plain chest radiography is the
mainstay imaging modality in depicting pulmonary
features of the disease. However, a normal chest
radiograph does not exclude pulmonary tuberculosis
and has been reported in approximately 10%-20%
of immunocompetent and immunocompromised
patients, respectively (Fitzgerald et a1.1991; Miller and
Miller 1993; Greenberg et al.1994; Lee and 1m 1995).
Computed tomography (CT) is useful in depicting
cavitation and in patients with pleural effusion that
may be masking lung parenchymal involvement on
the same side (Kuhlman et aI.1990). CT is more sensi-
tive than chest radiography (Hulnick et al. 1983; Lee
et a1.1996) in depicting mediastinal and paratracheal
lymphadenopathy, endobronchial tuberculosis, and
dissemination to the lung parenchyma and other
rare complications (Hatipoglu et al. 1996).
Bronchography, now replaced by CT, is still used as
the investigation of choice for the detection of bron-
chiectasis in poor-recourse countries of endemic
areas (Fig. 23.29). Arteriography for diagnostic or
therapeutic methods of bronchial artery emboliza-
tion (BAE) is used for these rare but life-threatening
massive hemoptysis cases.
Imaging features of post-primary pulmonary
tuberculosis, although suggestive, are not character-
istic as they can simulate other diseases (Lee and 1m
1995). Upper lobe infiltration or consolidation should
always be suspected as tuberculous. Cavitation may
occur in approximately 40%, and the diagnosis of
tuberculosis is usually not difficult if it is present
in the upper lobe and associated with bronchogenic
spread to other parts of the lung (Miller and Miller
1993). In some instances, upper lobe consolidation
should always be considered to be tuberculosis until
proven otherwise. The presence of lung parenchymal
scarring, fibronodular or calcific changes should not
be assumed as inactive, and follow-up is essential.
We retrospectively reviewed the radiological records
of 176 adult patients with post-primary pulmonary
tuberculosis who had positive sputum culture for
M. tuberculosis. They attended our hospital between
1998 and 2000, and their imaging features and fre-
quencies are presented in Table 20.2.
The imaging features of post-primary pulmonary
tuberculosis have been reported by us in a separate
chapter (Chap. 23).
321
20.3.2
Microbiology of the Sputum
and Bronchial Aspirate
Sputum microscopy for the detection of acid-fast
bacilli by the Ziehl-Nielsen stain remains the corner-
stone of rapid diagnosis of pulmonary tuberculosis
(Maartens 2002). In HIV patient's sputum, micros-
copy is positive less often. At least 3 single specimens
of sputum should be initially collected from patients
with productive cough. In those who have difficulty
in providing sputum, an aerosol inhalation of sterile
hypertonic saline can be used to stimulate sputum
production. Morning gastric aspiration, bronchoal-
veolar lavage, or transbronchoscopic brush or biop-
sies may be required.
At least 5,000 to 10,000 bacilli per milliliter of
sputum must be present to enable the detection of
the organism by stained smear (Hobby et al. 1973).
Sputum culture will require 10 to 100 bacilli to yield
a positive result (Yeager et al. 1967).
The American Thoracic Society (2000) reported
on the diagnostic standards and classifications of
tuberculosis in adult and children, and indicated that
50%-80% of patients with pulmonary tuberculosis
will have positive sputum smears. Traditional culture
media required 4-8 weeks to yield the bacilli, while
radiometric culture methods (BACTEC) combined
with a DNA probe allow identification of M. tubercu-
losis in 1-3 weeks.
Currently, PCR-based assays for the diagnosis
of tuberculosis and identification of drug-resistant
strains are configured to yield results in a few hours
to days (see Chapter PCR and Diagnosis of Tubercu-
losis by Dr. Diana L. Williams).
20.3.3
Bronchoscopic Diagnosis
The use of bronchoscopy to obtain diagnostic speci-
mens for patients with a chest radiograph suggesting
tuberculosis but with negative sputum specimen has
been reported (Willcox et al. 1982). Willcox and col-
leagues reported on 275 patients seen from 1976 to
1980, with imaging features of suspected tuberculosis
and negative sputum smear. Specimens collected by
bronchoscopy included bronchial brushings in 83,
transbronchiallung biopsies in 18,and post-bronchos-
copy sputum. Positive brushing yield was 67.5%, and
transbronchial biopsies were positive in 50%. They
also reported the co-existence of tuberculosis and
bronchial carcinoma in 4 of their patients (4%).
322
In our series of 176 patients, bronchoscopy was
done on 27 patients with a chest radiograph suggest-
ing pulmonary tuberculosis but with negative direct
smear. Tuberculosis was confirmed by positive cul-
ture of bronchoscopic specimens, and no bronchial
carcinoma was found.
Tuberculous infection after bronchoscopy may
be transmitted from patient to patient. Molecular
epidemiological studies by molecular typing of
DNA supported the transmission of M. tuberculosis
isolates to other patients via bronchoscopic contami-
nation (Michele et al. 1997; Argeton et al. 1997). It is
recommended that the instrument be cleaned prior
to its immersion in 2% aqueous solutions of glutar-
aldehyde for 45-min exposure times (Food and Drug
Administration 1992).
20.3.4
Tuberculin Skin Test
The tuberculin skin test is still widely used to
identify infection with M. tuberculosis. Antigenic
extracts of culture 'PPD' produces a delayed-type
hypersensitivity reaction. Intradermal injection
of 0.1 mg/0.1 ml of the standard 5-tuberculin unit
(TU) dose (Mantoux method) is done into the volar
or dorsal surface of the forearm (American Tho-
racic Society 2000). The test should be read between
48 and 72 h after injection. Three cut-off points
have been recommended for defining a positive
PPD test. A cut-off point of 2:5 mm of induration
using the ballpoint pen method of Sokal is consid-
ered positive in a person with recent contact or in
the presence of abnormal chest radiographs consis-
tent with tuberculosis. A cut-off point of 2:10 mm
is suggested for individuals who have normal or
mildly impaired immunity and a high likelihood of
being infected with the disease but without other
risk factors. A cut-off point of2:15 mm for individu-
als with no risk factors for tuberculosis is consid-
ered positive (American Thoracic Society 2000).
Tuberculin tests have several limitations including
difficult administration, anergy, poor specificity,
and the need for repeated testing to detect boosting
(Maartens 2002).A comparative study between skin
tests with PPD and measurement of the response
by in-vitro culture assays measuring IFN-y produc-
tion in response to tuberculin antigen stimulation
is described in detail in a separate chapter. Dr. Rohit
Katial has demonstrated the superiority of these
assays and compared it with the PPD skin test (see
Chap. 15).
M. M. Madkour et al.
20.3.5
Other Diagnostic Investigations
Ultrasound or CT-guided, transthoracic, percuta-
neous, fine-needle aspiration (FNA) cytology may
be useful in diagnosing pulmonary tuberculosis in
patients initially suspected of having malignancies.
Das et al. (1995) reported the use of FNA to diag-
nose tuberculosis in 29 patients in their series of 190
patients with malignancies and other causes.
Serological tests for antimycobacterial antibodies
in the serum using an ELISA immunoassay may be
positive in up to 88% (Barnes et al. 1993). DNA fin-
gerprinting PCR-based assays provide a noninvasive
method of diagnosing M. tuberculosis as well as iden-
tifying drug-resistant strains, with results ready in a
few hours or days (see Chap. 13).
Pulmonary function tests may be required in
tuberculous patients, particularly when lung sur-
gery is required. Radionuclide studies in pulmonary
tuberculosis are useful imaging tools in assessment
of the disease (see Chap. 24 and 26).
20.4
Treatment of Post-primary Pulmonary T8
The history of the management of tuberculosis has
been called 'the story of medical failure' (Holme
1998). 'The patients have been blamed for non-com-
pliance with the therapeutic regimen, but sociologic
studies have shown that, in most cases, the providers
of health care are at fault' (Grange and Zumla 2000).
The infectivity of tuberculous patients for close con-
tacts after starting chemotherapy has been studied
by several authors. Riley et al. (1962) reported that
the effluent air from the rooms of patients receiving
chemotherapy became noninfectious for guinea-pigs
within 2 weeks. Other authors found the rapid reduc-
tion in the number of viable bacilli by 1-2 logarith-
mic counts within 2 weeks (Yeager et a1.1967; Hobby
et al. 1973; Jindani et al. 1980). An editorial (1980)
reviewed the issue of isolation of infectious patients
with pulmonary tuberculosis, indicating that admis-
sion to a sanatorium was justified for supervised
treatment, considering its toxicity and to safeguard
their close contacts. It also referred to the work of
Jenkinson et al. (1979) who found viable bacilli in
the sputum of 15 patients who had received 6 weeks
of treatment, and active disease was produced when
injected into guinea-pigs. Studies on the duration of
antituberculous treatment for patients with a positive
Post-primary Pulmonary Tuberculosis
sputum smear to achieve three consecutive negative
results has been carefully studied prospectively by
Telzak and colleagues (1997) from the South Bronx
in New York. The study started from April 1993 to
March 1995 of all patients with culture-confirmed
tuberculosis. Data included the results of smears,
cultures, and drug susceptibility, HIV status, CD4
cell count (for HIV-positive patients). The main
objective of the study was to identify the time dura-
tion between initiating antituberculous treatment to
the first of three consecutive negative sputum smears
and the first of three consecutive negative sputum
cultures. During the period of the study, 199 patients
with culture-positive tuberculosis were diagnosed,
75% had lung parenchymal disease alone, 2% had
lung and pleural abscess, 14% had pulmonary and
extrapulmonary TB, 3% had only pleural disease,
and 7% had extrapulmonary disease alone. They had
complete information on 100 of the sputum smear-
positive patients (85%). They found that the mean
number of days before the first of three consecutive
negative sputum smears was 33, and the median was
23 days. The mean number of days until the first of
three consecutive negative sputum cultures was 32,
and the median was 26 days. These authors noted
that the following factors were associated with an
increased number of days to achieve their objective:
the high number of AFB on initial smear, the pres-
ence of cavitary disease, and no previous history of
tuberculosis. They noted that HIV had no effect on
the duration of treatment before achieving the study
objectives. HIV-positive patients are less likely to
have cavitary disease, and therefore more likely to
produce few AFB. In conclusion, patients with smear-
positive sputum require hospitalization and must be
isolated for a mean duration of 32 days after initiat-
ing appropriate treatment or longer if the initial AFB
count was high in the presence ofcavitary disease and
if the patient had no prior history of tuberculosis.
The aim of the appropriate chemotherapy is to
use drugs to which the bacilli are susceptible, with
bactericidal activity to cure the patient, reduce infec-
tivity to contacts, and prevent the emergence of drug
resistance (Centers for Disease Control 1994). Clini-
cians are often required to take the decision of either
to initiate the treatment of smear-negative patients
with a presumptive diagnosis of pulmonary tubercu-
losis based on the presence of clinical and imaging
features, or to wait for the sputum culture results that
may require 4-8 weeks to obtain and may not yield
the bacilli. In a prospective study of 139 patients with
a presumptive diagnosis of pulmonary tuberculosis,
Gordin et al. (1989) reported their findings. The
323
study was carried out in San Francisco between mid-
1981 to the end of 1982. The presumptive diagnosis
in this series was based on the presence of clinical
and radiological features suggestive of pulmonary
tuberculosis with negative sputum smear. Treatment
with isoniazid, rifampicin, and ethambutol was given
to all patients. Positive culture was reported later in
16 patients. Among the culture-negative patients, 43
showed radiographic improvement after 3 months
of initiating the treatment, clinical improvement in
7, and bronchoscopic confirmation of the diagnosis
in 1 patient.
There were 72 individuals with radiological stabil-
ity who were considered as having inactive previous
tuberculosis. The authors concluded that the treat-
ment was appropriate in 66 of 139 (48%) of patients.
The British Medical Research Council (BMRC) rec-
ommended the initiation of antituberculous treat-
ment for such patients after other causes for abnor-
mal chest radiography findings have been excluded
(Dutt and Stead 1994; Fox et al. 1999).
The modern short course of antituberculous
chemotherapy, Directly Observed Therapy (DOT), is
the best strategy for the treatment of post-primary
pulmonary tuberculosis to achieve the three goals,
that is, cure the patient, reduce infectivity to contacts,
and prevent the emergence of drug resistance if
appropriately adhered to. DOT is usually given as an
outpatient treatment and supervised by nurses. Such
visits to the outpatient facility may be difficult as it
involves traveling and may be costly for patients in
developing countries. Training of a family member
or a community lay person to supervise the adminis-
tration of DOT may be more practical in developing
countries (Wilkinson 1994). WHO indicated that for
the DOT strategy to be effective, it requires govern-
ment commitment, the availability of microscopic
and other diagnostic facilities, a continuous supply of
high-quality medication, direct observation of treat-
ment administration, and recording the response to
treatment (Netto et al. 1999).
DOT consists of an initial2-month phase of inten-
sive treatment using four drugs (rifampicin, isonia-
zid, pyrazinamide, and either ethambutol or strep-
tomycin) followed by a 4-month continuation phase
of rifampicin and isoniazid. These drugs are usually
given as daily treatment but can be given thrice
weekly, either throughout or during the continuation
phase, to facilitate the supervision of treatment. The
isoniazid daily dose is 5 mg/kg orally or intramus-
cularly (maximum 300 mg), rifampicin daily dose is
10 mg/kg orally (maximum 600 mg), pyrazinamide
daily dose is 15-30 mg/kg orally (maximum 2 g),
324
streptomycin daily dose is 15 mg/kg intramuscularly
(maximum 1 g) for persons below the age of 60 years
and 10 mg/kg intramuscularly (maximum 750 mg)
for persons above 60 years, ethambutol daily dose
is 15-25 mg/kg orally (maximum 2-5 g). The cure
rate of drug-susceptible disease when the patient
completes the course of treatment is up to 98%.
HIV co-infected patients have a similar outcome
of tuberculosis treatment success rate (Grange and
Zumla 2000). Although DOT has been successful in
some countries, endemic areas with HIV and MDR
tuberculosis have not achieved the WHO targets for
disease control (Netto et al. 1999).
Multidrug-resistant tuberculosis is a much more
serious problem to treat. Drugs used for treatment are
more toxic, more expensive, and require a prolonged
period of up to 24 months, and surgery may be con-
sidered to decrease the load of the organisms during
chemotherapy (Iseman 1993; Iseman et al.1990).
20.4.1
Response to Treatment
Appropriate chemotherapy is usually associated
with clinical and radiological improvements. Clini-
cal improvement with resolution of fever occurred
in 92% of patients 4 weeks after starting treatment
(Teklu and AI-Wabel 1994; Vanham et al. 1997).
Radiological improvement is noticed after 3 months
(Gordin et al. 1989). Paradoxical worsening of the
clinical and imaging responses to treatment may
occur particularly in the first few weeks or months.
HIV co-infected patients are more prone to paradoxi-
cal worsening after starting antituberculous treat-
ment, and this is reported to occur in 36% (Wendel
et al. 2001). In tuberculous patients without HIV,
paradoxical worsening may occur in 16% (AI-Majed
1996). Paradoxical worsening after antituberculous
treatment is defined as worsening of fever, cough,
shortness of breath or even development of ARDS,
enlargement of lymph nodes or other extrapulmo-
nary site of the disease during appropriate treatment
(Smith 1987; AI-Majed 1996; Rodriguez-Bano et al.
1997; Hung and Chang 1999; Wendel et al. 2001).
Radiological worsening of the pre-existing pul-
monary infiltrates, development of new lesions on
the same or apposite lung, enlargement of hilar or
paratracheal lymph nodes were noted in our series
(Figs.23.16a-c and 23.17a-c). The pathogenesis of
paradoxical worsening is not yet clear. It is believed that
it may be due to enhanced immune responses (immu-
nological rebound) with increased proliferation of
M. M. Madkour et al.
mononuclear cells and IFN-y production in response
to M. tuberculosis antigens (Wendel et al. 2001). Corti-
costeroid therapy has been used to modify the inten-
sity of symptoms (Rodriguez-Bano et al. 1997). There
have been no controlled trials on the management of
these paradoxical worsening responses.
References
Abal AT, Nair PC, Cherian J (2001) Haemoptysis: aetiology,
evaluation and outcome - a prospective study in a third-
world country. Respir Med 95:548-552
Aderaye G, Jajaw A (1996) Bilateral pulmonary aspergilloma:
case report. East Afr Med J 73:487-488
Agarwal MK et al. (1977) Respiratory Failure in Pulmonary
tuberculosis. Chest 72:605-609
Al-Hajjaj MS, Pandya L, Marie AA et al (1991) Pulmonary
tuberculosis in Saudi Arabia: a retrospective study of 1566
patients. Ann Saudi Med 11:443-447
Alland D, Kalkut GE, Moss AR (1994) Transmission of tubercu-
losis in New York City. An analysis by DNA fingerprinting
and conventional epidemiologic methods. N Engl J Med
33:1710-1716
Al-Majed SA (1996) Study of paradoxical response to che-
motherapy in tuberculous pleural effusion. Respir Med
90:211-214
Al-Wabel-A, Teklu B, Buluk 0 et al (1995) The spectrum of
pulmonary tuberculosis in the Asir region of Saudi Arabia.
SMJ 16:105-107
Alzeer A, Mashlah A, Fakim N et al (1998) Tuberculosis is the
commonest cause of pneumonia requiring hospitalization
during Hajj (Pilgrimage to Makkah). J Infect 36:303-306
American Thoracic Society (2000) Diagnostic standards and
classification of tuberculosis in adults and children. Am J
Respir Crit Care 161:1376-1395
Argeton T, VaIway S, Gore B et aI (1997) Transmission of a highly
drug-resistant strain (strain WI) of Mycobacterium tubercu-
losis: community outbreak and nosocomial transmission via
a contaminated bronchoscope. JAMA 278:1073-1077
Balasubramanian V, Wiegeshaus EH, Taylor BT et al (1994)
Pathogenesis of tuberculosis: pathway to apical localiza-
tion. Tuberc Lung Dis 75:168-178
Bandera A, Gori A, Catozzi L et al (2001) Molecular epidemiol-
ogy study of exogenous reinfection in an area with a low
incidence of tuberculosis. J Clin MicrobioI39:2213-2218
Barnes PF, Barrows SA (1993) Tuberculosis in the 1990 s. Ann
Intern Med 119:400-410
Barnes PF, Bloch AB, Davidson PT et al (1991) Tuberculosis in
patients with human immunodeficiency virus infection. N
Engl J Med 324:1644-1650
Bates JH (2001) Reinfection tuberculosis. How important is it?
Am J Respir Crit Care Med 163:600-601
Brande PVD, Dockx S, Demedts VM (1998) Pulmonary
tuberculosis in the adult in a low prevalence area: is the
radiological presentation changing? Int J Tuberc Lung Dis
2:904-908
Post-primary Pulmonary Tuberculosis
Brown ST, Almenoff PL (1992) Pulmonary mycobacterial
infections associated with neoplasia. Semin Respir Infect
7:104-113
Buckner CB, Walker CW (1990) Radiologic manifestation of
adult tuberculosis. J Thorac Image 5:28-37
Caminero JA, Pena MJ, Campos-Herrero Ml et al (2001) Exog-
enous reinfection with tuberculosis on a European island
with a moderate incidence of disease. Am J Respir Crit
Care 163:717-720
Cartier Y, Kavanagh PV, Johkoh T et al (1999) Bronchiectasis:
accuracy of high-resolution CT in the differentiation of
specific diseases. AJR Am J RoentgenoI173:47-52
Centers for Disease Control (1994) Guidelines for prevent-
ing the transmission of Mycobacterium tuberculosis in
health-care facilities. MMWR Morb Mortal Wkly Rep
43:1-132
Choyke PL, Sostman HD, Curtis AM et al (1983) Adult-onset
pulmonary tuberculosis. Radiology 148:357-362
Conlan AA, Hurwitz SS et al (1983) Massive hemoptysis. J
Thorac Cardiovasc Surg 85:120-124
Corbett EL, Churchyard GJ, Clayton TC et al (2000) HIV infec-
tion and silicosis: the impact of two potent risk factors on
the incidence of mycobacterial disease in South African
miners. AIDS 14:2759-2768
Dahmash NS, Fayed DF, Chowdhury MNH, Arora SC (1995)
Diagnostic challenge of tuberculosis of the elderly in hos-
pital: experience at a university hospital in Saudi Arabia. J
Infect 31:93-97
Dankner WM, Waecker NJ, Essey MA et al (1993) Mycobacte-
rium bovis infections in San Diego: a clinico-epidemiologi-
cal study of 73 patients and a historical review of a forgot-
ten pathogen. Medicine 72:11-37
Dannenberg AM Jr (2001) Pathogenesis of pulmonary Myco-
bacterium bovis infection: basic principles established by
the rabbit model. Tuberculosis 81:87-96
Das DK, Pant CS, Pant IN et al (1995) Transthoracic (percuta-
neous) fine needle aspiration cytology diagnosis of pulmo-
nary tuberculosis. Tuberc Lung Dis 76:84-89
Dunlap NE, Bass J, Fujiwara P et al (2000) Diagnostic stan-
dards and classification of tuberculosis in adults and chil-
dren. Am J Respir Crit Care Med 161:1376-1395
Dutt AK, Stead WW (1994) Smear-negative pulmonary tuber-
culosis. Semin Respir Infect 9:113-119
Dyer RA, Potgieter PD (1984) The adult respiratory distress
syndrome and bronchogenic pulmonary tuberculosis.
Thorax 39:383-387
Editorial (1980) Isolation of patients with pulmonary tuber-
culosis. Br Med J (5 April):962-963
Fitzgerald JM, Gizybouski S, Allen EA (1991) The impact of
human immunodeficiency virus infection on tuberculosis
and its control. Chest 100:191-200
Food and Drug Adminsitration, Infection Control Devices
Branch, Division of General and Restorative Devices, Office
of Device Evaluation (1992) Guidance on the content and
format of premarket notification (5IO[k]) Submissions
for Liquid Chemical Germicides. Rockville, Md: Food and
Drug Administration, 31 Jan
Fox W, Ellard GA, Mitchison DA (1999) Studies on the treat-
ment of tuberculosis undertaken by the British Medical
Research Council Tuberculosis Units, 1946-1986, with
relevant subsequent publications. Int J Tuberc Lung Dis 3
[SuppI2):S231-S279
Garcia M, Vargas JA, Castej6n R et al (2002) Flow-cytometric
325
assessment of lymphocyte cytokine production in tuber-
culosis. Tuberculosis 82:37-41
Geppert EF, Leff A (1979) The pathogenesis of pulmonary and
miliary tuberculosis. Arch Intern Med 139:1381-1383
Glynn JR, Shearer S, Kambashi B, Godfrey-Fausett P (2001)
HIV-I and recurrence, relapse, and reinfection of tubercu-
losis after cure: a cohort in South African mineworkers.
Lancet 358:1687-1693
Gordin FM, Slutkin G, Schecter G et al (1989) Presumptive diag-
nosis and treatment of pulmonary tuberculosis based on
radiographic findings. Am Rev Respir Dis 139:1090-1093
Grange JM, Zurnla A (2000) Advances in the management of
tuberculosis: clinical trials and beyond. Curr Opin Pul Med
6:193-197
Grange JM, Daborn C, Cosivi 0 (1994) HIV-related tuberculo-
sis due to Mycobacterium bovis. Eur Respir J 7:1564-1566
Greenberg SD, Jenkins DE, Bahar D (1964) Coexistence of
carcinoma and tuberculosis of the lung. Am Rev Respir
Dis 90:67-76
Greenberg SD, Frager D, Suster B et al (1994) Active pul-
monary tuberculosis in patients with AIDS: spectrum of
radiographic findings (including a normal appearance).
Radiology 193:115-119
Hatipoglu ON, Osma E, Manisali M et al (1996) High resolu-
tion computed tomographic findings in pulmonary tuber-
culosis. Thorax 1996:397-402
Heyderman RS, Goyal M, Roberts P et al (1998) Pulmonary
tuberculosis in Harare, Zimbabwe: analysis by spoligotyp-
ing. Thorax 53:346-350
Hirshberg B, Biran I, Glazer M et al (1997) Hemoptysis: etiol-
ogy, evaluation, and outcome in a tertiary referral hospital.
Chest 112:440-444
Hobby G, Holman A, Iseman M et al (1973) Enumeration
of tubercle bacilli in sputum of patients with pulmonary
tuberculosis. Antimicrob Agents Chemother 4:94-104
Holme C (1998) Tuberculosis: story of medical failure. Br Med
J 317:1260
Hsiao EI, Kirsch CM, Kagawa FT et al (2001) Utility of fiberop-
tic bronchoscopy before bronchial artery embolization for
massive hemoptysis. AJR Am J Roentgenoll77:861-867
Hulnick DH, Naidich DP, McCauley DL (1983) Pleural tuber-
culosis evaluated by computed tomography. Radiology 149:
759-765
Hung SC, Chang SC (1999) New pulmonary lesions during ther-
apy for extrapulmonary tuberculosis. Chest 116:1794-1797
Ihm HJ, Hankins JR et al (1972) Pneumothorax associated
with pulmonary tuberculosis. J Thorac Cardiovasc Surg
64:211-219
1m JG, Itoh H, Shim YS (1993) Pulmonary tuberculosis: CT
findings-early active disease and sequential change with
antituberculous therapy. Radiology 186:653-660
Iseman MD (1993) Treatment of multidrug-resistant tubercu-
losis. N Engl J Med 329:784-791
Iseman MD, Madsen LA (1991) Chronic tuberculous empyema
with bronchopleural fistula resulting in treatment failure
and progressive drug resistance. Chest 100:124-127
Iseman MD, Madsen L, Goble M et al (1990) Surgical inter-
vention in the treatment of pulmonary disease caused by
drug-resistant Mycobacterium tuberculosis. Am Rev Respir
Dis 141:623-625
Jindani A, Aber V, Edwards E et al (1980) The early bacterial
activity of drugs in patients with pulmonary tuberculosis.
Am Rev Respir Dis 121:939-949
326
Johnson JL, Ellner 11 (2000) Adult tuberculosis overview:
African versus Western perspectives. Curr Opin Pulm Med
6:180-186
Johnson TM, McCann et al (l973) Tuberculous bronchopleural
fistula. Am Rev Respir Dis 107:30-41
Kaestel M, Meyer W, Mittelmeier HO et al (1999) Pulmonary
aspergilloma - clinical findings and surgical treatment.
Thorac Cardiovasc Surg 47:340-345
Khan FA, Rehman M et al (l980) Pulmonary gangrene occur-
ring as a complication of pulmonary tuberculosis. Chest
77:76-80
Kim HY,Song KS et al (2001) Thoracic sequelae and complica-
tions of tuberculosis. Radiographics 21:839-860
Kuhlman JE et al (1990) CT features of thoracic mycobacterial
disease. Radiographics 10:413-431
Lambert HP (1956) Spontaneous pneumothorax and pulmo-
nary tuberculosis. Tubercle 37:207-209
Lee KS,Im JG (l995) CT in adult with tuberculosis of the chest:
Characteristic findings and role of management. AJR 164:
1361-1367
Lee KS, Kim YH, Kim WS et al (l99l) Endobronchial tubercu-
losis: CT features. J Comput Assist Tomogr 15:424-428
Lee KS, Hwang JW, Chung MP et al (l996) Utility of CT in the
evaluation of pulmonary tuberculosis in patients without
AIDS. Chest 110:977-984
Lee TW, Wan S, Choy DK et al (2000) Management of massive
hemoptysis: a single institution experience. Ann Thorac
Cardiovasc Surg 6:232-235
Levy H, Kallenbach JM, Feldman C et al (l987) Acute respi-
ratory failure in active tuberculosis. Crit Care Med 15:
221-225
Liesegang TJ, Cameron Douglas (l980) Mycobacterium bovis
infection of the conjunctiva. Arch Ophtalmol 98:1764-1766
Lopez-Contreras JL, Ris J et al (1994) Tuberculous pulmonary
gangrene: report of a case and review. Clin Infect Dis 18:
243-245
Maartens G (2002) Advances in adult pulmonary tuberculosis.
Curr Opin Pulm Med 8:172-177
Maartens G, Beyers N (2002) Tuberculosis in the tropics. Clin
Chest Med 23:341-350
Mal H, Rullon I, Mellot F (1999) Immediate and long-term
results of bronchial artery embolization for life-threaten-
ing hemoptysis. Chest 115:996-1001
McAdams HP, Erasmus J et al (l995) Radiologic manifesta-
tions of pulmonary tuberculosis. Radiol Clin North Am
33:655-678
McDonough KA, Florczyk MA, Kress Y (2000) Intracellular
passage within macrophages affects the trafficking of
virulent tubercle bacilli upon reinfection of other mac-
rophages in a serum-dependent manner. Tuberc Lung Dis
80:259-271
Michele TM, Cronin WA, Graham NMH et al (l997) Transmis-
sion of Mycobacterium tuberculosis by a fiberoptic bron-
choscope: identification by DNA fingerprinting. JAMA 278:
1093-1095
Millar JW, Horne NW (l979) Tuberculosis in immunosup-
pressed patients. Lancet 1:1176-1178
Miller WT (l981) Tuberculosis in the adult. Postgrad Radiol
1:147-167
Miller WT, Miller WT (l993) Tuberculosis in the normal host:
radiological findings. Semin Roentgenol 28:109-118
Moda G, Daborn q, Grange JM et al (l996) The zoonotic impor-
tance of Mycobacterium bovis. Tuberc Lung Dis 77:103-108
M. M. Madkour et al.
Netto EM, Dye C, Raviglione MC (l999) Progress in global
tuberculosis control 1995-1996, with emphasis on 22-high-
incidence countries. Global Monitoring and Surveillance
Project. Int J Tuberc Lung Dis 3:310-320
Penner C, Roberts D, Kunimoto D et al (l995) Tuberculosis as
a primary cause of respiratory failure requiring mechanical
ventilation. Am J Respir Crit Care Med 151:867-872
Reich JM (l993) Pulmonary gangrene and the air crescent
sign. Thorax 48:70-74
Riley R, Mills C, O'Grady F et al (l962) Infectiousness of
air from a tuberculosis ward. Ultraviolet irradiation
of infected air: comparative infectiousness of different
patients. Am Rev Respir Dis 85:511-525
Rodriguez-BaflO JR et al (l997) Systemic paradoxical response
to antituberculous drugs: resolution with corticosteroid
therapy. Clin Infect Dis 24:517-519
Rohenberg GT, Shaw P (l996) Radiology of pulmonary tuber-
culosis. Br J Hosp Med 56:195-199
Rook GAW,Zumla A (2001) Advances in the immunopathogenesis
of pulmonary tuberculosis. Curr Opin Pulm Med 7:116-123
Sauret J,Jolis R,Ausina Vet al (l992) Human tuberculosis due
to Mycobacterium bovis: report of 10 cases. Tuberc Lung
Dis 73:388-391
Small PM, Hopewell PC, Singh SP et al (l994) The epidemiol-
ogy of tuberculosis in San Francisco. A population-based
study using conventional and molecular methods. N Engl
J Med 330:1703-1709
Smith H (1987) Paradoxical responses during the chemo-
therapy of tuberculosis. J Infect 15:1-3
Sokal JE (1975) Measurement of delayed skin-test responses.
N Engl J Med 293:501-502
Stead WW, Bates JH (2000) Recurrent tuberculosis due to
exogenous reinfection. N Engl J Med 341:1050
Stebbings AE, Lim TK (l999) Cause, treatment and outcome
of patients with life-threatening haemoptysis. Singapore
Med J 40:67-69
Teklu B, Al-Wabel AH (l994) Resolution of fever in patients
on chemotherapy for pulmonary tuberculosis. Ann Saudi
Med 14:392-395
Telzak EE, Fazal BA, Pollard CL et al (1997) Factors influenc-
ing time to sputum conversion among patients with smear-
positive tuberculosis. Clin Infect Dis 25:666-670
Uchida T, Wada M, Sakamoto J et al (l999) Treatment for
empyema with bronchopleural fistulas using endobronchial
occlusion coils: report of a case. Jpn J Surg 29:186-189
Van Rie A, Warren R, Richardson M et al (1999) Exogenous
reinfection as a cause of recurrent tuberculosis after cura-
tive treatment. N Engl J Med 341: 1174-1179
Vanham G, Toossi Z, Hirsch CS et al (l997) Examining a para-
dox in the pathogenesis of human pulmonary tuberculosis:
immune activation and suppression/anergy. Tuberc Lung
Dis 78: 145-158
Wallis RS, Johnson JL (2001) Adult tuberculosis in the 21st
century: pathogenesis, clinical features, and management.
Curr Opin Pulm Med 7:124-133
Weissberg D, Weissberg D (2001) Late complications of collapse
therapy for pulmonary tuberculosis. Chest 120:847-851
Wendel KA, Alwood KS, Gachuhi R et al (2001) Paradoxical
worsening of tuberculosis in HIV-infected persons. Chest
120:193-197
Wilder RJ, Beacham EG et al (1962) Spontaneous pneumotho-
rax complicating cabitary tuberculosis. J Thorac Cardio-
vase Surg 43:561-573
Post-primary Pulmonary Tuberculosis
Wilkinson D (1994) High-compliance tuberculosis treatment
program in a rural community. Lancet 343:647-648
Willcox PA, Benatar SR, Potgieter (1982) Use of the flexible
fibreoptic bronchoscope in diagnosis of sputum-negative
pulmonary tuberculosis. Thorax 37:598-601
Winer-Muram HT, Rubin SA (1990) Thoracic complications of
tuberculosis. J Thorac Imag 5:46-63
327
Wong ML, Szkup P, Hopley MJ (2002) Percutaneous embolo-
therapy for life-threatening hemoptysis. Chest 121:95-102
Woodring JH (1986) Update: the radiographic features of
pulmonary tuberculosis. AJR 146:497-506
Yeager HJ, Lacy J, Smith L et al (1967) Quantitative studies on
mycobacterial populations in sputum and saliva. Am Rev
Respir Dis 95:998-1004
21 Endobronchial Tuberculosis
HEE SOON CHUNG

CONTENTS 21.15.1.1 Interventional Management 343

21.15.1.2 Tracheobronchial Stent 344
21.1 Introduction 329 21.15.1.3 Endobronchial Electrosurgery
21.2 Definition 330 vs Laser Photoresection 346
21.3 Epidemiology 330 21.15.1.4 Surgical Management 346
21.4 Pathogenesis 330 21.15.2 Bronchiectasis 346
21.5 Pathology 331 21.15.3 Broncholith 346
21.6 Classification 332 References 347
21.7 Characteristics of the Each Subtype
of Endobronchial Tuberculosis 333
21.7.1 Actively Caseating Type 333
21.7.2 Edematous-Hyperemic Type 333
21.7.3 Fibrostenotic Type 334 21.1
21.7.4 Tumorous Type 334 Introduction
21.7.5 Granular Type 335
21.7.6 Ulcerative Type 335
21.7.7 Nonspecific Bronchitic Type 336 Endobronchial tuberculosis is a specific form or a
21.8 Location of Bronchial Involvement 336 significant complication of pulmonary tuberculosis. It
21.9 Natural Course of Endobronchial Tuberculosis 337 is important to remember that endobronchial tuber-
21.10 Endobronchial Tuberculosis culosis has the following pitfalls: (1) its diagnosis is
in HIV-Infected Patients 337
frequently delayed, since the decreased incidence itself
21.11 Clinical Features 338
21.12 Diagnosis 338 diminishes the suspicion of tuberculosis (Greenbaum et
21.12.1 Clinical History 338 al. 1980; Chung et al. 1991a); (2) bronchostenosis may
21.12.2 Physical Examination 339 develop as a serious complication despite efficacious
21.12.3 Laboratory Testing 339 antituberculosis chemotherapy (Albert and Petty 1976;
21.12.4 Bacteriology 339
Hoheisel et al. 1994; Park et al. 1997); and (3) it is often
21.12.5 Tuberculin Testing 340
21.12.6 Pulmonary Function Test 340 misdiagnosed as bronchial asthma (Watson and Ayres
21.12.7 Radiology 340 1988; Williams et al.1988; Park et al.1995) or lung cancer
21.12.7.1 Simple Roentgenograms 340 (Matthews et al.1984; Smith et al. 1987).
21.12.7.2 Computerized Tomography 340 Recently there has been an unprecedented resur-
21.12.8 Bronchoscopy 341
gence of tuberculosis, which is related to the human
21.13 Differential Diagnosis 341
21.13.1 Pneumonia 341 immunodeficiency virus (HIV) epidemic, multidrug-
21.13.2 Fungal Infection and Actinomycosis 342 resistant strains, poverty and homelessness, immigra-
21.13.3 Lung Cancer 342 tion, and failures in the treatment system (Millard et al.
21.13.4 Bronchial Asthma 342 1994; Glynn 1998; Lerner 1999). The HIV epidemic may
21.14 Treatment 342
be associated with a higher incidence of endobronchial
21.14.1 Anti-Tuberculosis Chemotherapy 342
21.14.2 Corticosteroid Treatment 343 tuberculosis (Judson and Sahn 1994; Calpe et al.1995).
21.15 Complications and Their Management 343 Therefore, endobronchial tuberculosis continues to
21.15.1 Bronchostenosis 343 be a health problem, though the incidence of tubercu-
losis affecting respiratory organs including the trachea
and bronchi has been greatly reduced (Shulutko et al.
1979). Endobronchial tuberculosis should be borne
H. S. CHUNG, MD, FCCP in mind when patients are young female adults or
Associate Professor of Medicine, Seoul National University
College of Medicine and Seoul Municipal Boramae Hospital
adolescents who present with symptoms suggestive
affiliated to Seoul National University Hospital, # 395 Shindae- of asthma and/or with unusual roentgenographic pat-
bang-2-Dong, Dongjak-Gu, Seoul, 156-707, Korea terns, or in patients with HlV infection.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
330
21.2
Definition
Endobronchial tuberculosis was first described by
Richard Morton in 1694 (Jenks 1940), and is defined
as a specific inflammation of the trachea or major
bronchi caused by tubercle bacilli. Active disease can
be diagnosed when certain endobronchial lesions,
such as whitish gelatinous material, ulcer, tumor,
stenosis or inflammation, exist on bronchoscopy
and tuberculosis is proven by bronchoscopic biopsy
of these lesions.
Bronchial anthracofibrosis is defined as a luminal
narrowing associated with anthracotic pigmenta-
tion on bronchoscopy, without a relevant history
of pneumoconiosis or smoking, and may be a sign
of active endobronchial tuberculosis (Chung et al.
1998; Kim et al. 2000), but histologic confirmation is
necessary for definitive diagnosis (Garimella 2001).
Bronchial fibrostenosis is sometimes a surprise find-
ing in patients with a previous history of tuberculo-
sis, which must be the presumed cause of stenosis.
Since the sequela of endobronchial tuberculosis and
reactivated pulmonary tuberculosis can be present
coincidently, fibrotic stenosis of bronchi may be
inactive lesions resulting from prior endobronchial
tuberculosis even though sputum examination for
acid-fast bacilli is positive. Therefore, it is necessary
to obtain histologic proof of tuberculosis for a defi-
nite diagnosis of endobronchial tuberculosis.
Localized tuberculous bronchitis in segmental
bronchi communicating with diseased portions of
lung is common in pulmonary tuberculosis (Daniel
1994). Resected lung specimens frequently show either
ulceration or stenosis of the draining bronchioles or
bronchi, and the same endobronchial processes may
result in bronchiectasis due to destruction of the bron-
chial wall (Rossman and Oner-Eyuboglu 1998). Since
these endobronchial lesions distal to lobar bronchi do
not have clinical significance, they should be included
in the disease entity of pulmonary tuberculosis rather
than that of endobronchial tuberculosis.
21.3
Epidemiology
Tuberculosis is still a common disease despite its
great decline in recent years. In high-prevalence
areas, tuberculosis afflicts chiefly young adults. In
countries where HIV infection is endemic, tubercu-
losis is one of the most important causes of morbid-
H. S. Chung
ity and mortality in acquired immune deficiency
syndrome (AIDS) patients (Daniel 1994).
The incidence of endobronchial tuberculosis
among tuberculous postmortem specimens was as
high as 40% prior to the chemotherapy era (Wolinsky
1989). With modern treatment, endobronchial tuber-
culosis has been reported in about 10% of cases of
pulmonary tuberculosis, when fiberoptic bronchos-
copy is routinely performed (Jokinen et al. 1977).
In my previous study, the incidence of endobron-
chial tuberculosis in pulmonary tuberculosis was
5.88 percent. However, it should be borne in mind
that the actual incidence is somewhat higher than
this, because bronchoscopy was not performed rou-
tinely in all tuberculosis patients but only when an
endobronchial lesion was highly suspected. Endo-
bronchial tuberculosis shows a marked preference
for female patients, and the male-to-female ratio is
approximately 1:5. The disease is usually detected
in the third decade. The reason why endobronchial
tuberculosis is more common in young females is
not clear. One of the possible reasons is that organ-
ism implantation from infected sputum may occur
more easily in females, especially in teenagers and
those in their twenties, because they do not expecto-
rate sputum, for sociocultural and cosmetic reasons
(Chung and Lee 2000). In addition, it is well recog-
nized that endobronchial tuberculosis frequently
occurs as a part of primary pulmonary tuberculosis
in young adults (Smith et al. 1987).
21.4
Pathogenesis
First infection with the tubercle bacillus is known
as primary tuberculosis, and usually includes the
involvement of the draining lymph nodes in addition
to the initial lesion. The combination of the primary
or Ghon focus and draining lymph nodes is termed
the primary complex. Although primary tuberculo-
sis was formerly relatively common in intestines or
tonsils, due to infection from milk, and may occur
in various other unusual sites, in the vast majority
of cases the route of infection is by inhalation and,
consequently, the primary lesion is pulmonary. All
other tuberculous lesions are regarded as post-pri-
mary. Multiple terms have been used to describe this
stage of tuberculosis - for example, chronic tubercu-
lous postprimary disease, reinfection or recrudescent
tuberculosis, adult-type progressive tuberculosis,
endogenous reinfection and reactivation tubercu-
Endobronchial Tuberculosis
losis (Garay 1996). Regardless of the nomenclature,
these tuberculous lesions are not accompanied by
major involvement of the draining lymph nodes.
In primary pulmonary tuberculosis, the first
focus of exudative inflammation may occur in any
bronchopulmonary segment, more often near the
periphery in the middle or lower lung, i.e. that por-
tion of the lung receiving the greatest ventilation, and
this is where bacilli within droplet nuclei tend to be
implanted. Tubercle bacilli multiply in the exudates
and may be carried into some of the very abundant
pulmonary lymphatics, and then to the regional
lymph nodes. Hematogenous seeding to the distant
organs then probably occurs with considerable fre-
quency, either by way of lymphatic channels or more
directly from smaller pulmonary veins within the
exudate. However, lesions of clinical magnitude do
not develop by hematogenous seeding unless native
resistance is low or the number of bacilli is great.
Regional lymph node involvement is a fairly
consistent characteristic of primary pulmonary
tuberculosis. The hilar lymph node is most com-
monly involved, but the paratracheal node is also
frequently enlarged, and a substantial minority of
cases show enlargement of both the hilar and para-
tracheal nodes. These nodes develop an intense
cellular inflammatory response and may become so
large that they compress the major bronchi, with
obstructive atelectasis or obstructive pneumonitis.
This phenomenon was formerly called epituberculo-
sis. Caseation often follows in the nodes, but healing
by partial resorption of the caseum, and the calcifica-
tion of the lesion in the lung and lymph nodes, is the
rule. However, such nodes occasionally rupture or
protrude into bronchi. In children especially, rupture
of the caseous glands into the trachea or major bron-
chi causes collapse of the lung or even sudden death
by suffocation in young children. In adults, the lung
component of the primary complex is usually more
obvious and the nodal component may not be seen,
whereas in children often only an enlarged hilar or
paratracheal node is apparent (Seaton et al.1989).
In post-primary tuberculosis, lymph node
involvement is seldom extensive and pulmonary
lesions on chest roentgenograms are usually located
in the apical or subapical areas of the upper lobe. The
predilection for the upper lung probably relates to the
high oxygen tensions at the lung apex (Riley 1960).
Even though the pathogenesis of endobronchial
tuberculosis is not yet fully established, sources
of endobronchial tuberculosis may include direct
implantation of tubercle bacilli into the bronchus
from an adjacent pulmonary parenchymal lesion,
331
direct airway infiltration from an adjacent tubercu-
lous mediastinal lymph node, erosion and protru-
sion of an intrathoracic tuberculous lymph node into
the bronchus, hematogenous spread and extension
to the peribronchial region by lymphatic drainage
(Myerson 1944; Smart 1951; Matthews et al. 1984;
Smith et al.1987; Lee et al. 1992; Kim et al.1993).
Endobronchial tuberculosis, before the era of
chemotherapy, was considered a complication of
advanced post-primary disease. Effective antituber-
culosis chemotherapy and preventive measures have
reduced childhood tuberculosis exposure, which has
resulted in an increase in adult primary tuberculosis
with unusual clinical and roentgenographic presenta-
tions. In the 1970s and early 1980s, a number of reports
concerning the so-called unusual radiographic mani-
festations of adult tuberculosis were published. Up to
one-third of adult cases were reported to have atypical
findings, such as mediastinal adenopathy, lower lung
consolidation and miliary disease. However, these
"unusual manifestations" are the usual manifestations
of primary tuberculosis. The only unusual aspect is
the increasing incidence of primary tuberculosis in
adults (McAdams et al.1995). In the modern drug era,
endobronchial tuberculosis is more likely to be discov-
ered in adults with primary tuberculosis (Smith et al.
1987), whereas post-primary disease is by far the most
common type of pulmonary tuberculosis.
21.5
Pathology
The pathologic changes that occur in the involved
bronchi differ in primary tuberculosis from those of
post-primary tuberculosis. The involvement of the
trachea and bronchi in primary tuberculosis usually
results from the pressure created by the enlarged
lymph nodes, which may cause partial or complete
bronchial obstruction, necrosis and ulceration of the
bronchial wall, often in association with the rupture of
caseous nodes into the bronchi. In post-primary tuber-
culosis, the pathogenesis differs, and bronchial lesions
are usually secondary to repeated implantation of
tubercle bacilli from a sloughing parenchymal source
distal to the lesion (Medlar 1955). The earliest lesions
often are lymphocytic infiltrations of the mucosa
with or without congestion or edema. These usually
clear when the parenchymal focus ceases to slough.
However, bronchial lesions may progress to tubercle
formation in the mucosa and submucosa. Rarely,
necrosis with ulceration and sloughing may involve
332
portions of the bronchial wall, and it is this type of
lesion that causes bronchial narrowing and stenosis,
distal to which further bronchial inflammation may
lead to obstructive pneumonitis or bronchiectasis. The
healing of more extensive lesions is frequently accom-
panied by cicatricial stenosis (Wolinsky 1989).
21.6
Classification
The clinical course of endobronchial tuberculosis
is variable, not only because there are several pos-
sible pathogenetic mechanisms, but also because the
interactions between the effects of mycobacteria,
host immunity and antituberculous drugs is com-
plex, and any variation in these three factors may
result in an altered course (Chan and Pang 1989).
Therefore, it is an oversimplification to view cases
of endobronchial tuberculosis as a homogeneous
group, and endobronchial tuberculosis is probably
better divided into subtypes.
Over the past five decades, many classifications
of endobronchial tuberculosis have been published.
Judd (l947) classified endobronchial tuberculosis into
intrabronchial and extrabronchial types on the basis
of their pathogenetic mechanisms. In 1957, the Sixth
All-Union Congress of Phthisiologists adopted a clas-
sification which categorized endobronchial tuberculo-
sis into four forms: infiltrative, ulcerative, cicatricial
and fistulous (or glandular). After the introduction of
the flexible bronchoscope, Oho and Amemiya (l984)
suggested that endobronchial tuberculosis be classi-
fied as edematous-hyperemic, infiltrative-prolifera-
tive, ulcerative-granulative and fibrostenotic.
In 1991, my colleagues and I suggested a clas-
sification of endobronchial tuberculosis which
considered the above-mentioned classifications and
bronchoscopic features. This classification categorized
endobronchial tuberculosis into seven forms: actively
caseating, stenotic without fibrosis, stenotic with
fibrosis, tumorous, granular, ulcerative and nonspe-
cific bronchitic (Chung et al. 1991a), and was partially
reported at the Sixth World Congress for Bronchology
(Han et al. 1989). However, the terms "stenotic with-
out fibrosis" and "stenotic with fibrosis" were later
renamed "edematous-hyperemic" and "fibrostenotic:'
respectively, due to the fact that gross bronchoscopic
findings cannot reflect the presence of fibrosis.
As a result, the forms of endobronchial tuberculosis
were classified into seven subtypes by bronchoscopic
findings: actively caseating, edematous-hyperemic,
H. S. Chung
fibrostenotic, tumorous, granular, ulcerative and non-
specific bronchitic (Chung and Lee 2000). The actively
caseating, edematous-hyperemic, fibrostenotic and
tumorous forms of endobronchial tuberculosis have
varying degrees of bronchostenosis proximal to the
segmental bronchi, whereas the granular, ulcerative
and nonspecific bronchitic forms do not have luminal
narrowing of the bronchi.
This new classification is valuable for predicting the
therapeutic outcome of endobronchial tuberculosis,
because it is closely related to the extent of disease pro-
gression (Chung and Lee 2000). Pathologically, the initial
lesion, which presents as simple erythema and edema of
the mucosa with lymphocytic infiltration of the submu-
cosa (Medlar 1955; Shulutko et al.1979),corresponds to
nonspecific bronchitic endobronchial tuberculosis. This
lesion is followed by submucosal tubercle formation,
which produces the erythema and granularity seen at
bronchoscopy (Myerson 1944), and partial bronchial
stenosis, which is caused by considerable congestion
and edema of the mucosa (Medlar 1955). These are the
granular and edematous-hyperemic types, respectively.
At this point, the development of caseous necrosis, with
the formation of tuberculous granuloma, can be found
at the mucosal surface (Medlar 1955), which constitutes
actively caseating endobronchial tuberculosis. However,
when the inflammation erupts through the mucosa, an
ulcer is seen, which may be covered by caseous material
(Myerson 1944), and the disease is considered to be of
the ulcerative type. Finally, the bronchial mucosal ulcer
evolves into hyperplastic inflammatory polyps, and
the endobronchial tuberculous lesion heals by fibro-
stenosis (Salkin et al. 1943; Smith et al. 1987). During
this process, the lesion has moved through either the
tumorous or fibrostenotic type. In addition, tumorous
endobronchial tuberculosis can develop via a pathway
involving erosion and later protrusion of an intra-
thoracic tuberculous lymph node into the bronchus
(Shulutko et al.1979; Judson and Sahn 1994). This may
be the principal mechanism oftumorous endobronchial
tuberculosis, given that computed tomography usually
reveals an endobronchial mass, as well as enlarged
lymph nodes adjacent to the bronchus, and anthracotic
pigment is frequently seen in biopsy specimens (Yee et
al. 1985; Smith et al.1987). Interestingly, endobronchial
tuberculosis in patients with HIV infection shows such
bronchoscopic and computed tomography findings
(Judson and Sahn 1994; Alame et al. 1995; Calpe et al.
1995; Saadoun et al.1998).
In my previous study (Chung and Lee 2000), the
actively caseating type was the most common form,
the edematous-hyperemic, fibrostenotic, tumorous
and granular types were relatively common, the non-
Endobronchial Tuberculosis 333

specific bronchitic type occurred less frequently and
the ulcerative type was the rarest form (Table 21.1).
21.7
Characteristics of the Each Subtype
of Endobronchial Tuberculosis
In my previous study, about a third of the cases
of the edematous-hyperemic type became of the
nonspecific bronchitic type and healed within three
months of treatment. The remaining two-thirds
became fibrostenotic within three months of drug
treatment, and complete obstruction of the bronchial
lumen often occurred. The prognosis of edematous-
hyperemic endobronchial tuberculosis is poor.

21.7.1
Actively Caseating Type

Actively caseating endobronchial tuberculosis is diag-

nosed when the bronchial mucosa is swollen, hyper-
emic and diffusely covered with a whitish cheese-like
material. This form is usually accompanied by luminal
narrowing at diagnosis, whether granulation tissue is
present or not (Fig. 21.1).
In my previous study, approximately one-third
of the cases of actively caseating endobronchial
tuberculosis healed via the edematous-hyperemic,
granular or nonspecific bronchitic types without
complication. The remaining two-thirds transformed
into the fibrostenotic type within three months of
antituberculosis and corticosteroid treatment. The
formation of granulation tissue is a poor prognos-
tic factor, because all cases that showed granulation
tissue on follow-up bronchoscopy changed into the
Fig.21.1. Bronchoscopic finding of actively caseating endo-
fibrostenotic type. The prognosis of actively caseat- bronchial tuberculosis. From Chung and Lee (2000). Repro-
ing endobronchial tuberculosis is poor. duced, with permission, from Chest 117:385

21.7.2
Edematous-Hyperemic Type

In cases of edematous-hyperemic endobronchial

tuberculosis, the bronchial lumen is narrowed due
to severe mucosal swelling with surrounding hyper-
emia. However, neither caseous material nor fibrous
contracture is found at diagnosis (Fig. 21.2).

Table 21.1. Classification of 114 endobronchial tuberculoses

by Bronchoscopic Features

Type Cases %

Actively caseating type 49 43.0

Edematous-Hyperemic type 16 14.0
Fibrostenotic type 12 10.5
Tumorous type 12 10.5
Granular type 13 11.4
Nonspecific bronchitic type 9 7.9
Ulcerative type 3 2.7
Fig. 21.2. Bronchoscopic finding of edematous-hyperemic
From Chung and Lee (2000). Reproduced, with permission, endobronchial tuberculosis. From Chung and Lee (2000).
from Chest 117:385 Reproduced, with permission, from Chest 117:385
334
21.7.3
Fibrostenotic Type
Fibrostenotic endobronchial tuberculosis presents
as a marked narrowing of the bronchial lumen with
fibrosis. Usually) endobronchial tuberculous lesions
do not encircle the bronchial mucosa and normal
mucosa is partly spared, and the stenotic bron-
chial lumen becomes a crushed waterdrop-shape,
as shown in Fig. 21.3. In some cases, the bronchial
lumen is completely occluded. The edges of this
stricture are avascular and pale because of dense
fibrosis, and it is impossible to biopsy due to its very
hard consistency. However, active tuberculosis can be
diagnosed by bronchoscopic biopsy of the chronic
inflamed mucosa at the periphery of the lesion, which
is usually swollen and reddened.
In my previous study, all cases remained in a fibro-
stenotic state inspite of drug therapy. In addition,
roughly half of the cases showed progressive fibro-
stenosis and this resulted in complete obstruction of
the bronchial lumen two or three months after treat-
ment. The prognosis of fibrostenotic endobronchial
tuberculosis is very poor.
21.7.4
Tumorous Type
Tumorous endobronchial tuberculosis is characterized
by an endobronchial mass that has a surface often cov-
Fig.21.3. Bronchoscopic finding of fibrostenotic endobron-
chial tuberculosis. From Chung and Lee (2000). Reproduced,
with permission, from Chest 117:385
H. S. Chung
ered with caseous material, and which nearly totally
occludes the bronchial lumen (Fig. 21.4). This form of
endobronchial tuberculosis is frequently mistaken for
lung cancer because of its bronchoscopic appearance
and the fact that the computerized tomography find-
ings mimic those of lung cancer.
Anthracotic pigmentation or anthracofibrosis can
appear after the endobronchial lesion has been suc-
cessfully treated. If the lymph nodes caseate, they may
point towards and discharge intrabronchially. Fol-
lowing discharge, the resultant fistulae usually heal
slowly, leaving small, sometimes pigmented, pitted
scars in the bronchial wall. These can be visualized
as hard, depressed, black plaques by bronchoscopy,
and have been observed to follow the intrabronchial
perforations of tuberculous lymph nodes. The black
pigment derives from the anthracotic material pres-
ent in the tuberculous nodes, which is subsequently
incorporated into scarred areas.
In my previous study, the evolution of tumorous
endobronchial tuberculosis was very complicated and
unpredictable. Approximately 70% of cases eventually
changed into the fibrostenotic type, which included
the complete obstruction of the bronchial lumen
with fibrosis. Six months after treatment, another
20% were found to exhibit impending obstruction
of the left main bronchus due to a re-growing mass,
although this were successfully corrected by broncho-
scopic electrocautery. In 20% of cases, new tumorous
lesions appeared five or six months after treatment,
and anthracofibrosis or anthracotic pigmentation was
Fig.21.4. Bronchoscopic finding of tumorous endobronchial
tuberculosis. From Chung and Lee (2000). Reproduced, with
permission, from Chest 117:385
Endobronchial Tuberculosis
observed in 30% of cases. The prognosis of tumorous
endobronchial tuberculosis is grave.
It is very difficult to analyze the evolution of
tumorous endobronchial tuberculosis; this is due
to the fact that the disease exhibits diverse progress
and unexpected changes. However, the observations
of Shulutko and coworkers (l979) are extremely
valuable. In bronchoglandular tuberculosis, which is
equivalent to tumorous endobronchial tuberculosis,
bronchoglandular fistulas occur when the necrotic
foci of tuberculous lymph nodes rupture into the
bronchial lumen, extruding the lymph node con-
tents. During this stage, bronchial stenosis is often
temporary, and the bronchoglandular fistula usually
heals, leaving a thin tender scar that neither deforms
the wall nor narrows the lumen of the bronchus,
provided endobronchial treatment (removal of case-
ous masses and granulations, cauterization of the
fistulous opening or peribronchial blockades) is per-
formed. However, if scarring of the bronchoglandular
fistula continues, the bronchial lumen may be perma-
nently narrowed by persistent cicatricial stenosis, or
it may even be completely obliterated. Most cicatri-
cial stenoses of endobronchial tuberculosis appear
to be the sequelae of tumorous bronchadenitis, and
occasionally more lymphoglandular fistulas arise
near the first fistula. When the insights of Shulutko
et al. were applied to my previous series, unexpected
changes were no longer unexpected, but probable. In
order to maintain bronchial patency, early diagnosis
and efficacious treatment, including interventional
Fig.21.5. Bronchoscopic finding of granular endobronchial
tuberculosis. From Chung and Lee (2000). Reproduced, with
permission, from Chest 117:385
335
modalities (Smith et al.1987; Watson and Ayres 1988;
Chung et al. 1991a), are very important in tumorous
endobronchial tuberculosis.
21.7.5
Granular Type
Granular endobronchial tuberculosis appears macro-
scopically like scattered grains of boiled rice, and the
underlying bronchial mucosa shows severe inflam-
matory change (Fig. 21.5).
In my previous study, roughly 20% of the cases
showed fibrostenosis of the bronchial lumen two
months after treatment, and the other 80% of cases
healed without endobronchial sequelae three or four
months after treatment. The prognosis of granular
endobronchial tuberculosis is good.
21.7.6
Ulcerative Type
The appearance of the bronchial ulcer in ulcerative
endobronchial tuberculosis is very similar to that of
a peptic ulcer (Fig. 21.6), and it can result from the
submucosal lymphatic spread of organisms from
adjacent parenchymal disease or implantation.
In my previous study, the prognosis of ulcerative
endobronchial tuberculosis was excellent, as all cases
were completely resolved within three months of
Fig.21.6. Bronchoscopic finding of ulcerative endobronchial
tuberculosis. From Chung and Lee (2000). Reproduced, with
permission, from Chest 117:385
336 H. S. Chung

treatment commencement. However, the number of

cases involved was too small to serve as a basis for
evaluating prognosis.

21.7.7
Nonspecific Bronchitic Type

In nonspecific bronchitic endobronchial tuberculo-

sis, only mild mucosal swelling and/or hyperemia are
evident by bronchoscopy (Fig. 21.7). Tuberculosis is
proven by bronchoscopic biopsy of the lesions.
As my previous investigation indicated, all cases
of nonspecific bronchitic endobronchial tuberculosis
that were studied healed within two months of treat-
ment, suggesting that the prognosis for this disease
is excellent.

Fig. 21.7. Bronchoscopic finding of nonspecific bronchitic

21.8 endobronchial tuberculosis. From Chung and Lee (2000).
Reproduced, with permission, from Chest 117:385
Location of Bronchial Involvement

As pulmonary tuberculosis shows a right-sided most common site of endobronchial tuberculous

predominance in most series (McAdams et al. 1995),
so endobronchial tuberculosis may occur more fre-
quently in the right side (Garay 1996).
In my previous study (Chung et al. 1991a), endo-
bronchial tuberculosis showed a mild right-sided pre-
dominance, whereas luminal narrowing of the bronchi
was more frequently observed in the left. Moreover,
tracheal involvement was not rare. Interestingly, the
lesions was found to be the left mainstem bronchus.
Lowet al. (2001) confirmed this finding. Endobronchial
tuberculosis of both lower lobe bronchi and the right
middle lobe bronchus was found to be mainly of the
nonspecific bronchitic type, and tumorous endobron-
chial tuberculosis was found to occur mainly in the
left mainstern bronchus. The distributions of involved
sites and their types are illustrated in Fig. 21.8.

Trachea: 17.3
Rt main B: 6.1
Cas' 48.6
Cas: 56.2 E-H . 37.1
E-H :18.7 Fs : 8.6
Fs : 12.5 NsB. 5.7
Tum: 6.3
Gra . 6.3 Lt Main B: 18.4
Lt upper B: 4.8
E-H : 35.6
Cas: 26.7 Fig. 21.8. Distributions of
Rt upper B: 17.9 Fs : 24.4
Fs : 37.5
Cas: 37.5 involved sites and their
Cas : 40.0 Tum: 8.9 E-H : 25.0
Gra : 4.4 types versus their relative
E-H : 29.3
Fs . 17.3 frequencies (%) in endo-
NsB : 6.7 Upper div: 5.8 bronchial tuberculosis.
Ulcer: 6.7 *Rt right, Lt left, Interm
Cas: 30.8
E-H : 30.8 intermediate, B bronchus,
Fs : 23.0
Rt lower B : 15.1 E-H : 50.0 Tum: 15.4 div division. Cas actively
Cas: 25.0 caseating type, E-H
NsB : 55.6 Tum: 25.0 Lingular div: 1.2
edematous-hyperemic
E-H : 18.5
Fs : 18.5 Rt middle B : 3.7 Fs: 100.0 type, Fs fibrostenotic
Cas. 7.4 type, Tum tumorous type,
NsB 42.8 Lt lower B: 7.6
Fs 28.6 era granular type, Ulcer
E-H 14.3 NsB : 84.6 ulcerative type, NsB non-
Tum 14.3 Tum: 15.4
specific bronchitic type
Endobronchial Tuberculosis 337

21.9 21.10
Natural Course of Endobronchial Endobronchial Tuberculosis
Tuberculosis in HIV-Infected Patients

The presumptive natural course of endobronchial
tuberculous lesions (Fig. 21.9, dashed arrow) can
be deduced by observing the healing process
(Fig. 21.9, solid arrow). The desired end result of
endobronchial tuberculosis is healing without sig-
nificant sequelae; the other possibility is fibrosteno-
sis. All subtypes of endobronchial tuberculosis are
situated between these two conclusions, and they
can transform into other subtypes during treat-
ment. However, there is a critical point between
these two extremes, the position of which is mainly
determined by the extent of disease progression
(Kim et al. 1993) and the formation of granula-
tion tissue (Shulutko et al. 1979; Smith et al. 1987).
Bronchial stenosis is inevitable (Albert and Petty
1976; Caligiuri et al. 1984) if the disease progresses
beyond this critical point. Therefore, prompt diag-
nosis and efficacious treatment are of paramount
importance in cases of endobronchial tuberculosis
in order to minimize the resultant bronchial steno-
sis (Park et al. 1997). To alleviate bronchostenosis
that has already developed, aggressive therapeutic
modalities should be considered before dense
fibrosis progresses or complete obstruction of the
bronchus occurs.
In a normal host the immunologic response to infec-
tion by the tubercle bacillus provides a degree of pro-
tection against additional tubercle bacilli that may be
subsequently inhaled in droplet nuclei. The likelihood
of reinfection is a function of the risk of re-exposure,
the intensity of such exposure, and the integrity of the
host's immune system. In developed countries the risk
of re-exposure to an infectious case is low. Further-
more, in the otherwise healthy, but previously infected
person, any organisms deposited in the alveoli are
likely to be killed by cell-mediated immune response.
Mycobacterial infections have been commonly
observed in patients infected with HIV. These indi-
viduals are 200 times more likely to contract tubercu-
losis than HIV-negative individuals (Broughton and
Bass 1999). In contrast to tuberculosis infections in
normal adults, tuberculosis in AIDS patients cannot
be readily identified by clinical or radiological cri-
teria. It has become apparent that infection with
HIV, because of its profound suppression of normal
immune response, predisposes the individual to
much more severe forms of tuberculosis. In HIV-
infected persons with tuberculosis, dissemination of
the tubercle bacilli and a variety of extrapulmonary
manifestations are common. Thus, unusual clinical

Observed Healing Protess

HEALING WITHOUT SEQUELAE .....- - - - - Critical Point - - - . . BRONCHIAL STENOSIS

•I
,I
,,
ULCERATIVE TYPE .--------.- TUMOROUS TYPE
,,
,,
I
,,
:,
I
,,
I
I
I
/:/'/ \\\\ ,,
,,
,,

,,'
I
,,
,,,
I
,
,, l \\
\
NONSPECIFIC • - - - - - - - - - - .- EDEMATOUS· - - - - - - - - - . Formation of - - - - . FIBROSTENOTIC
BRONCHITIC TYPE .. HYPEREMIC TYPE .. Granulation Tissue - TYPE
,

t \\\ /// \\\

I
I
I Fig. 21.9. A scheme sum-
I
I
marizing the observed
,,,
I

healing process (solid

Bro"chiolln\lO!vement " , ' \ , ,
arrow) and the pre-
ofTuberculosis \ , \ , I
,
I
sumptive natural course
I
• GRANULAR TYPE -:---------.- ACTIVELY CASEATING TYPE I
(dashed arrow) of endo-
I

NORMAL BRONCHUS
Presumptive Natural Course
• - - - - - -- -- -- -- - -.-
+ bronchial tuberculous
lesions. From Chung and
Critical Point I ----------. BRONCHIAL StENOSIS
Lee (2000). Reproduced,
with permission, from
Chest 117:385
338 H. S. Chung

presentations of tuberculosis in HIV-infected per-
sons present a special diagnostic challenge (Ameri-
can Thoracic Society 1990).
Tuberculosis in HIV-infected patients may have
the radiographic characteristics of primary disease.
Infiltrates may appear in any lung zone, cavitation is
uncommon, and mediastinal or hilar lymphadenopa-
thy is often present (Mangura and Reichman 1989).
The tumorous type of endobronchial tuberculosis
predominates in patients with HIV infection, and the
condition may be the result of the erosion and protru-
sion of a tuberculous lymph node into the bronchus, a
pathogenetic mechanism of primary tuberculosis.
The diagnosis of endobronchial tuberculosis in
HIV patients should always be considered, since
it may be more frequent than suspected (Calpe et
al. 1995). Consequently, a more liberal indication
of bronchoscopy helps in the early detection of
endobronchial tuberculosis in patients with AIDS
(Alame et al. 1995). Even if no endobronchial lesion
is apparent by bronchoscopy, bronchoalveolar lavage
and transbronchial biopsy are particularly useful
for diagnosing HIV-related pulmonary tuberculo-
sis, since only two-thirds to three-quarters of the
patients have a positive acid-fast smear, and the
tuberculin skin test is less sensitive in these patients
(Rossman and Oner-Eyuboglu 1998).
A cough is the most common symptom, which has
an almost imperceptible onset. Coughs slowly progress
over weeks or months and become more frequent; they
may be nonproductive at first, but occasionally may be
associated from the onset with mucopurulent sputum
or blood, and later, the expectoration of sputum usu-
ally appears. Shortness of breath frequently occurs,
and may be associated with localized wheezes, stri-
dors or the absence of a breathing sound. Fever is
frequently associated with the disease, and though
low-grade at the onset it may become marked as the
disease progresses. Characteristically, the fever devel-
ops in the late afternoon and may not be accompanied
by pronounced symptoms. With defervescence, usu-
ally during sleep, sweating occurs - the classic «night
sweats." In more extensive disease, hemoptysis may
occur but is seldom massive.
The clinical features of endobronchial tuberculo-
sis as detailed in several reports are summarized in
Table 21.2. Symptoms are usually nonspecific and are
caused mainly by the coexisting pulmonary tubercu-
losis. In contrast to uncomplicated pulmonary tuber-
culosis, hemoptysis was less frequent and dyspnea
was much more common in my previous study.
21.12
Diagnosis

21.11 21.12.1
Clinical Features Clinical History

The clinical features of endobronchial tuberculosis When endobronchial tuberculosis is suspected, inqui-
vary widely, depending on the site(s) and the extent ries concerning patient exposure to a person with an
of involvement, and endobronchial tuberculosis may open case of pulmonary tuberculosis may be helpful,
occur in the absence of recognized symptoms. especially if this contact is long-standing and close. A

Table 21.2. Clinical features of endobronchial tuberculosis as detailed in several reports

Symptom and sign Song Chung Ip Hoheisel Low

(%) et al. (1985) et al. (1991) et al. (1986) et al. (1994) et al. (2001)

Cough 87.5 75.3 100 87 86

Sputum 75 68.7 95 89
Dyspnea 24 33.1 35 57
Hemoptysis 17.5 1.2 25 8
Chest pain 45 15 15
Fever 35 26.5 50 87 24
Absence of a BS 27.5 28.3 19
Wheezing/Stridor 27.5 27.7 15 9 19
Rhonchi 14.5 12
Weight loss 29
None 35

BS breathing sound
Endobronchial Tuberculosis
past diagnosis of pneumonia that has recurred from
time to time should always arouse suspicion regarding
endobronchial tuberculosis. All previous chest films
should be obtained. Medical conditions that increase
the risk of tuberculosis (i.e. diabetes mellitus, gastrec-
tomy, drug abuse, etc) should also be reported.
21.12.2
Physical Examination
Acomplete physical examination should be performed,
although an examination of the chest will usually fur-
nish the main clues.
Wheezing is frequently heard on auscultation.
Persistent unilateral wheezing is more indicative of
endobronchial tuberculosis, while transient wheezing
due to bronchial secretions usually clears with cough.
Stridor may occur with an ulceration and cicatrix of
the trachea or larynx. It is nearly impossible to differ-
entiate wheezing and stridor heard in endobronchial
tuberculosis from that heard in bronchial asthma.
Decreased breathing or the absence of a breathing
sound is another frequent finding. When tuberculo-
sis involves the larynx, hoarseness is usually present
and is often accompanied by severe pain.
21.12.3
Laboratory Testing
Routine laboratory examinations are rarely helpful in
establishing or suggesting a diagnosis. A broad range
of hematologic manifestations has been reported in
endobronchial tuberculosis. The most common of
these are increases in peripheral blood leukocyte
counts and modest anemia, each of which occurs in
approximately 10% of patients. WBC counts of over
20,000/f.lL suggest another infectious process, and the
erythrocyte sedimentation rate is usually elevated.
With endobronchial tuberculosis, HIV testing is rec-
ommended in patients who have known or suspected
risk factors for the acquisition of HIV infection.
21.12.4
Bacteriology
Endobronchial tuberculosis, especially the extensive
form, is usually higWy infectious. In my previous
study, the sputum smear for acid-fast bacilli was
positive in about half of the patients, and the sputum
culture for tubercle bacilli was positive in 70% or more.
339
Conversely, some recent studies have found a low yield
on sputum acid-fast bacilli smears. The explanation
given was that the expectoration of sputum is difficult
because of mucus entrapment by proximal bronchial
granulation tissue (Ip et al. 1986; Lee et al. 1992; van
den Brande et al. 1990). It should not be surprising
that large differences are observed in the positive
rates of acid-fast bacilli, because culture yield, like
microscopic examination, is affected by the clinical
status of patients.
Sputum examination for acid-fast bacilli is much
more important, since patients with endobronchial
tuberculosis occasionally present with apparent
acute pneumonia. In fact, the diagnosis of tubercu-
losis is made only on routine sputum examination or
because of a failure of clinical or radiological resolu-
tion using broad-spectrum antibiotics. However, it
should be noted that a positive acid-fast smear is not
specific for Mycobacterium tuberculosis. Other myco-
bacteria, both saprophytes and potential pathogens,
can be acid-fast. Thus, the only absolute way of con-
firming a diagnosis of M tuberculosis is by culture.
FresWy expectorated sputum is the best sample
to stain and culture for M tuberculosis. Twenty-four
hour old sputum samples are frequently overgrown
with mouth flora and are much less useful. If a patient
is not producing sputum spontaneously, induced
sputum is the next best specimen for study. When a
patient cannot provide a spontaneous sputum sample,
a gastric aspirate to obtain swallowed sputum may be
useful. This sample must be obtained in the morning
before the patient arises or eats.
Bronchoscopy is preferable to gastric aspiration,
since bronchial secretions are readily available for
acid-fast smear and culture, as well as for cytologic
study, and a careful evaluation of the tracheobron-
chial tree can be carried out at the same time. Speci-
mens for culture should be obtained using a minimal
amount of anesthesia, because the local anesthetics
used for fiberoptic bronchoscopy may be lethal to M
tuberculosis. The portion of the biopsy specimen to
be used for culture should not be placed in formalin.
Post-bronchoscopy sputum can be another valu-
able source of diagnostic material, and bronchoscopy
may cause the patient to continue producing sputum
for several days. These later specimens should also be
collected and examined, as they may reveal tubercle
bacilli absent from the sample taken on the day of the
bronchoscopy.
Newer technologies, such as radiometric technol-
ogy (the BACTEC system), polymerase chain reaction
and genetic probes, immunoassay of mycobacterial
antigens, and the detection of biologic compounds,
340 H. s. Chung

can allow an early diagnosis or improve our ability to
isolate mycobacteria from clinical specimens (Glassroth
1993).
21.12.5
Tuberculin Testing
The tuberculin skin test is used as an indicator of
M. tuberculosis infection, and relies on a cell-medi-
ated immune response. However, a positive delayed
hypersensitivity reaction to tuberculin indicates only
the occurrence of a prior primary infection and not
the presence of clinically active disease.
Very large reactions (greater than 25 mm of indu-
ration) are more frequently associated with active
tuberculosis. However, a negative reaction to tubercu-
lin does not rule out the diagnosis, because the patient
may be anergic or have a specific anergy to tuberculin
(Rossman and Oner-Eyuboglu 1998). Nevertheless, the
absence of skin reactivity to tuberculin makes the pos-
sibility of tuberculosis unlikely in most situations.
21.12.6
Pulmonary Function Test
Bronchial disease and pulmonary parenchymal lesions
coexist in endobronchial tuberculosis. Although this
may have an influence on pulmonary function, no
pathognomonic pattern of physiologic disturbance
exists in endobronchial tuberculosis. Nevertheless,
the usual pattern is that of a predominantly restric-
tive ventilatory defect. The reason why endobronchial
tuberculosis shows this restrictive pattern may be
due to organic obstruction of the bronchial tree and
chronic inflammatory or bronchiectatic changes of the
lung parenchyme. The pulmonary function test may
be helpful in the differential diagnosis of endobron-
chial tuberculosis and bronchial asthma. The results of
pulmonary function testing in my previous study are
illustrated in Fig. 21.10.
Pulmonary function testing is useful for the follow-
up of endobronchial tuberculosis. Changes in pul-
monary function during treatment in endobronchial
tuberculosis are significantly correlated with changed
bronchoscopic findings, although the indices of pul-
monary function testing at diagnosis do not correlate
significantly with the gross bronchoscopic findings of
endobronchial tuberculosis (Chung and Lee 1996).
21.12.7
Radiology
21.12.7.1
Simple Roentgenograms
Despite the fact that roentgenographic examination
itself is not diagnostic of endobronchial tuberculosis,
it is a necessary procedure. A lateral chest film should
be a part of every roentgenographic inquiry, since
approximately 25 percent of the lung fields cannot be
visualized on the conventional posteroanterior film.
Endobronchial tuberculosis, as well as pulmo-
nary tuberculosis, may produce almost any form
of pulmonary radiographic abnormality. However,
endobronchial tuberculosis frequently presents with
unusual roentgenographic findings, such as pneu-
monic consolidation, lobar or segmental collapse, a
mass-like lesion or hilar or paratracheal adenopa-
thy. A normal chest radiograph cannot completely
exclude endobronchial tuberculosis, because roughly
10% of patients show normal chest roentgenograms.
Radiologic findings of endobronchial tuberculosis in
my previous study are shown in Table 21.3.
21.12.7.2
Computerized Tomography
Special imaging techniques, such as computerized
tomography and magnetic resonance imaging, may be
of particular value in defining nodules, cavities, cysts,
Table 21.3. Radiologic findings of 166 endobronchial tuber-
culoses
Characteristics Cases %

Pneumonic patchy infiltration 42 25.3

Atelectasis or Collapse 33 19.9
Fibrostreaky densities 27 16.3
Cavitary lesion 23 13.9
Mass-like lesion 12 7.2
Bronchiectatic change 7 4.2
Mediastinal widening 18 10.8
Fig.21.10. The results of pulmonary function testing in 85 No active lesion 21 12.7
cases of endobronchial tuberculosis
Endobronchial Tuberculosis 341

calcifications, the contours oflarge bronchi and vascular

essential that aggressive therapy be performed before
details in lung parenchyma. Moreover, since the lengththe disease progresses too far and bronchostenosis
of bronchial involvement, the thickness of bronchial becomes inevitable. Therefore, the bronchoscopic
wall and mediastinal lymph node enlargement can approach is mandatory not only for the prompt diag-
be evaluated noninvasively, computerized tomography nosis of endobronchial tuberculosis, but also for the
prevention of further bronchostenosis.
is a useful adjunct to direct endoscopic visualization,
particularly when performed at a 5-mm intervals with The suspicion of endobronchial tuberculosis pro-
5-mm slice collimation through the hila. The techniquevides an important indication for bronchoscopy. The
accurately depicts bronchial abnormality in 93 to 100%presence of a wheeze or of a persistent and uncontrol-
of all cases (McAdams et al. 1995), and it can also recon-
lable cough, or the presence of tubercle bacilli in the
struct three-dimensional images of the trachea and of sputum without an obvious source in the lung paren-
the major bronchi (Choi et al. 2002). chyma, provides sufficient reason for the procedure.
Computerized tomography findings of endobron- Roentgenographic findings of atelectasis or selective
chial tuberculosis include: isolated long segment segmental or lobar collapse, unexplained shadows
bronchial narrowing with concentric wall thicken- near the hilum, the sudden or gradual appearance of
ing, complete endobronchial obstruction, extrinsic localized obstructive over-distention or ballooning
compression by adjacent adenopathy and even direct cavities should also be investigated by bronchoscopy.
bronchial invasion of a caseous node. The erosion of A ballooning cavity is often referred to as the tension
calcified lymph nodes into adjacent bronchi, known ascavity, regardless of whether or not the intracavitary
broncholithiasis, can also be observed, along with the
pressure is known. Ballooning cavities are frequently
secondary to the involvement of bronchi in tubercu-
resultant segmental collapse or overinflation. Comput-
erized tomographic scans of the central airways usually
losis (Wolinsky 1989).
show active endobronchial tuberculosis as an irregular Tuberculosis produces two main bronchoscopi-
narrowing of the airways with marked wall thickening;cally visible changes: endobronchial inflammation
the scans also frequently show mediastinal lymph nodeand distortion due to extrabronchial lymph-node
enlargement, whereas fibrotic tuberculosis shows as aenlargement. One may see inflamed, swollen mucosa
rather smooth narrowing of the airways with minimal and purulent secretions or blood; alternatively, one
wall thickening (Kim et al.1997; Moon et al.1997). may sometimes see fibrous masses ofwhite-pink color
resembling cauliflower and simulating a cancerous
tumor, or even ulceration, in the bronchial draining
21.12.8 lobes or segments afflicted with active tuberculosis.
Bronchoscopy Acute inflammatory changes can respond rapidly to
chemotherapeutic treatment, leaving normal bron-
Bronchoscopy is the single most useful modality chi, but healing may lead to bronchial scarring and
in the diagnosis of endobronchial tuberculosis, sometimes to marked contractive stenosis.
whereas the chest radiograph is the most important In summary, the diagnostic use of bronchoscopy
for suggesting a diagnosis of pulmonary tuberculo- in obtaining a biopsy from an area of bronchial
sis. Bronchoscopy is necessary not only to make the ulceration or obstruction can clinch the diagnosis of
diagnosis of endobronchial tuberculosis, but also endobronchial tuberculosis.
to exclude bronchogenic carcinoma. Bronchoscopy
has a unique value in the management of endo-
bronchial tuberculosis. When the different forms of
endobronchial tuberculosis are classified into seven 21.13
subtypes (actively caseating, edematous-hyperemic, Differential Diagnosis
fibrostenotic, tumorous, granular, ulcerative and
nonspecific bronchitic), the therapeutic outcome of 21.13.1
each subtype, except the tumorous type, can be pre- Pneumonia
dicted by follow-up bronchoscopy during the initial
two to three months of treatment. However, in the In contrast to endobronchial tuberculosis, in which
case of tumorous endobronchial tuberculosis, close symptoms are nonspecific, the acute pneumonic
and long-term follow-up is advisable, given that the patient usually has typical symptoms of fairly recent
evolution of lesions during treatment is very compli- onset. If the patient is symptomatic, an oral antibi-
cated and bronchial stenosis may develop later. It is otic should be prescribed, and if the radiographic
342
opacities do not clear or improve in 2 to 3 weeks,
then tuberculosis is a possibility. If the pneumonia
is refractory to standard treatment, especially if the
white count is normal, then repeated sputum speci-
mens should be examined for acid-fast bacilli.
21.13.2
Fungal Infection and Actinomycosis
Polesky et al. (1999) presented data on 38 cases of
coccidioidomycosis of the airways, which included 6
cases detailed from their own experience and 32 from
the literature. In histoplasmosis, regional lymph node
involvement is invariable during the initial infection.
The extrinsic pressure of enlarged nodes on the air-
ways may cause obstruction and distal infection
or atelectasis. Complications of histoplasmosis are
similar to those of primary pulmonary tuberculosis
(Fraser et al. 1994).
In 1999, Lee et al. reported a case of biopsy-proven
endobronchial actinomycosis. The bronchoscopic
findings of this disease are very similar to those of
endobronchial tuberculosis.
21.13.3
Lung Cancer
This differential diagnosis occurs primarily in the
middle-aged and elderly age groups. Consolidation
distal to a proximal carcinoma may be cavitated
and closely mimic tuberculosis. Moreover, carci-
noma of the lung and tuberculosis may be present
simultaneously; in fact, the frequency of coexis-
tent lung cancer and tuberculosis is as high as 5%
(McAdams et al. 1995). In cases with a simultane-
ous presentation of carcinoma and tuberculosis,
the diagnosis of tuberculosis is frequently made
first, and the carcinoma diagnosis is often delayed
for several months. Thus, if the radiographic and
clinical findings suggest carcinoma, but the sputum
has acid-fast bacilli, further procedures to diagnose
carcinoma may still be indicated (Rossman and
Mayock 1988).
On bronchoscopy, tuberculous granulation tissue
may be observed to erupt through the bronchial
mucosa to form a tumor-like mass. The caseous tra-
cheobronchial lymph nodes can produce ominous,
irregular swellings into the main bronchi or trachea
that suggest the malignant invasion of lymph nodes
(Stradling 1981). Bronchoscopic biopsy will reveal
the true situation.
H. S. Chung
21.13.4
Bronchial Asthma
Some patients with endobronchial tuberculosis show
wheezing on physical examination but have normal
chest films, which mimics bronchial asthma. One should
always consider the possibility of endobronchial tuber-
culosis during the differential diagnosis of bronchial
asthma if patients with wheezing show normal airway
responsiveness (Park et al. 1995) or a poor response to
bronchodilation (Williams et al. 1988).
21.14
Treatment
21.14.1
Anti-Tuberculosis Chemotherapy
In principle, the short-course regimens are the regimens
of choice for patients with endobronchial tuberculosis.
Many short-course regimens have been investigated
and proposed for pulmonary tuberculosis, and these
can also be applied in endobronchial tuberculosis.
A 6-month regimen consisting of isoniazid, rifampin
and pyrazinamide for 2 months, followed by isoniazid
and rifampin for 4 months, is the preferred treatment
for patients with fully susceptible organisms and who
can adhere to treatment. Ethambutol (or streptomycin
in children too young to be monitored for visual acuity)
should be included in the initial regimen until the
results ofdrug susceptibility studies are available, unless
there is only a small possibility of drug resistance (i.e.
if there is less than 4% primary resistance to isoniazid
in the community, and the patient has had no previous
treatment with antituberculosis medications, is not
from a country with a high prevalence of drug resis-
tance and has no known exposure to a drug-resistant
case). The four-drug,6-month regimen is effective even
when the infecting organism is resistant to isoniazid.
This recommendation applies for both HIV-infected
and uninfected persons. However, in the presence of
HIV infection, it is critically important to comprehen-
sively assess the clinical and bacteriologic response.
Any problem with the response to treatment indicates
that the usual evaluation should be undertaken and
the therapy possibly prolonged. Consideration should
be given to treating all patients with directly observed
therapy, and the duration of treatment should total at
least 6 months and 3 months beyond culture conver-
sion. Children should be managed in essentially the
same ways as adults by using appropriately adjusted
Endobronchial Tuberculosis
doses of the drugs. In multidrug-resistant tuberculosis
(i.e. resistance to at least isoniazid and rifampin), treat-
ment must be based on susceptibility studies (American
Thoracic Society 1994).
With endobronchial tuberculosis, therapy needs
to be prolonged as decided on a case-by-case basis
and should be individualized after considering the
clinical, bacteriologic and bronchoscopic response,
especially if the dominant form is of the tumorous
type or if interventional management is performed.
21.14.2
Corticosteroid Treatment
The use of corticosteroids has been and remains con-
troversial. Nevertheless, under certain conditions corti-
costeroids are of benefit in patients with endobronchial
tuberculosis. It appears that steroids are effective in rap-
idly reducing the mass effects of mediastinallymphade-
nopathy in patients with primary tuberculosis. Steroids
may, therefore, decrease the severity of local obstruc-
tive complications (Nemir et al. 1963). Corticosteroids
are more likely to be beneficial in the earlier stages of
endobronchial tuberculosis, when hypersensitivity is
the predominant mechanism (Dooleyet al. 1997), and
they are unlikely to be helpful in more advanced cases
when extensive fibrosis is present. Therefore, corticoste-
roids can be cautiously prescribed for actively caseating
and edematous-hyperemic endobronchial tuberculosis
(which occur in the earlier stages of the disease) and
for tumorous endobronchial tuberculosis, which may
be derived from primary tuberculosi~. The usual dose
required is 40-60 mg of prednisone (or approximately
1 mg/kg ofbody weight) orally daily for 4-6 weeks, with
gradual tapering over the next few weeks.
It should be emphasized that before corticosteroids
are prescribed, one must be confident that adequate
antituberculosis chemotherapy is being given. Prompt
treatment upon early diagnosis, before the formation
of dense fibrosis, is a prerequisite for the prevention
and amelioration of bronchostenosis (Park et al.1997).
21.15
Complications and Their Management
21.15.1
Bronchostenosis
Traditional strategies for repairing benign tracheo-
bronchial stenosis include repeated endoscopic bal-
343
loon dilatation, which tends to be only a temporary
measure, Nd-YAG laser resection and surgical recon-
structive procedures, which are generally considered
to be the gold standard, although some patients will
have contraindications to surgery. Stent placement,
in contrast, is a new therapeutic modality. For select
patients, a multidisciplinary approach, whereby pul-
monologists, otolaryngologists and thoracic surgeons
decide on the most appropriate use of laser resection,
stent placement and surgical techniques has been
advocated (Dineen et al. 2002).
21.15.1.1
Interventional Management
Since bronchostenosis develops despite adequate anti-
tuberculosis therapy and/or corticosteroid treatment,
more aggressive interventional management may be
indicated to restore the patency of the involved bron-
chus in selected cases of endobronchial tuberculosis.
In tumorous endobronchial tuberculosis, because
the prognosis is grave if the condition is not treated
aggressively, the endobronchial tumorous lesion
itself should be removed by laser resection or elec-
trosurgery to prevent further bronchostenosis.
The fibrous membrane can simply be resected by
laser photoresection or electrocautery (in the fibro-
stenotic type of endobronchial tuberculosis), if the
membrane has a concentric web-like stricture (Becker
et al. 1991). In cases where the fibrostenosis is relatively
long, an endobronchial stent can be placed after dilata-
tion with a high-pressure balloon catheter. Some, but
not all, patients with tuberculous bronchostenosis may
be provided with long-term control of their disease
process by stent placement. Restenosis by granulation
tissue formation can develop, but it may be success-
fully treated by laser ablation or electrosurgery.
The therapeutic results of stent placement have
been very poor in actively caseating, edematous-
hyperemic and tumorous endobronchial tubercu-
losis, which have active severe inflammation of the
involved bronchus. Thus, appropriate antitubercu-
losis chemotherapy and/or corticosteroid treatment
should be administered prior to stent placement,
until active inflammation disappears. This may take
up to 5 months, after which, the endobronchial tuber-
culosis is usually of the fibrostenotic type, which is
the main indication of stent placement. During drug
therapy, repeated balloon dilatation may be needed
to prevent complete obstruction and to maintain the
patency of the narrowed bronchus (an example is
shown in Figs. 21.11 and 21.12). Once active inflam-
mation has been ameliorated by drug treatment, a
344 H. S. Chung

tracheobronchial stent can be placed, if necessary
(Figs. 21.13-15}. Further chemotherapy should be
given for a minimum of 3 months after the procedure
to prevent recurrence. The stent may be successfully
removed one year after placement, and the success
rate of stenting is roughly 50% (Kim, personal com-
munication). Granular, ulcerative and nonspecific
bronchitic endobronchial tuberculosis do not indi-
cate stent placement, because they do not involve
significant bronchostenosis.
27.75.7.2
TracheobronchialStent
The ideal tracheobronchial stent should possess
several vital characteristics. It should be capable of
establishing and maintaining airway patency; it should
be easy and safe to place and (if necessary) remove;
it should be biocompatible; and, it should be flexible
enough to fit appropriately in irregular anatomy.
In addition, it should not cause mucosal injury or
granulation tissue formation, hamper the ability to
clear secretions, migrate from its desired position or
obstruct otherwise patent lumens (Jantz and Silvestri
2000). As of yet, no such stent has been developed.
Essentially, five types of stent have been developed
and used for tracheobronchial obstructions. The earli-
est types were of silicone, most notably the Dumon or
Endoxane stent. The next stents used were of uncov-
ered stainless steel, such as the Gianturco Z and the
Palmaz stents, which were the first stents capable of
being deployed by flexible bronchoscopy. The third
type were the second generation of metal stents as
represented by the Wallstent and the Ultraflex. These
stents are alloy-based mesh or interwoven loop stents
and are also available with a polyurethane covering.

Fig. 21.11a-d. Repeated balloon dilatation of the left mainstem bronchus in a case with endobronchial tuberculosis. a At diag-
nosis. b At 1 month of treatment. c At 2 months of treatment; d At 3 months of treatment

Fig. 21.12a, b. Three-dimensional images of central airways in the case illustrated in Fig. 21.11. a At diagnosis. b At 2 months
of treatment
Endobronchial Tuberculosis 345

Fig. 21.13a-d. Stent placement for endotracheal tuberculosis. a Before stenting. b After stenting. c Just after stent-removal. d At
4 months after stent-removal

Fig. 21.14a, b. Three-dimensional images of the trachea in the case illustrated in Fig. 21.13. a Before stenting. b After stenting

Fig.21.15a, b. Stent placement for mainstem endobronchial tuberculosis. a Before stenting. b Just after stenting
346
The polymer coating prevents the ingrowth of tumor
or granulation tissue through the stent mesh frame-
work. The fourth type of stent, the bifurcated Y-stent,
was designed specifically to deal with the anatomic
complexities of the central airways and the carina, and
the fifth type of stent was constructed from a variety
of different materials, such as polyester/silicone and
nitinollsilicone. Generally, the most common com-
plications associated with stents are stent migration,
secretion retention and granulation tissue formation.
Refinements in stent designs are required to improve
biocompatibility and reduce migration and granula-
tion tissue formation (Dineen et al. 2002).
Overall, silicone stents have been very successful
in treating benign airway obstructing lesions and
may well be the gold standard for years to come.
Recently, animal model trials have been reported
with a bioabsorbable stent (Korpela et al. 1999). In
the future, bioabsorbable stents may have a role in the
management of benign airway stenoses.
21.15.1.3
Endobronchial Electrosurgery vs Laser Photoresection
Nd-YAG laser photoresection is the most effective
therapy for treating tracheobronchial obstructions
from benign or malignant lesions. However, the wide-
spread use of this technique is limited by the perceived
need for rigid bronchoscopy, expensive equipment
and special training and fear of major complications.
Advances in flexible bronchoscopy have allowed devel-
opments in other techniques directed at alleviating
airway obstruction, including cryotherapy, brachy-
therapy and photodynamic therapy. Although cost-
effective, their effects are delayed and they may require
repeated treatments (Coulter and Mehta 2000).
Labeled "the poor man's laser:' electrosurgery
offers equivalent laser-like tissue effects at a fraction
of the cost. The ability to perform endobronchial
electrosurgery in the outpatient setting is more
accommodating for patients and time-saving for
physicians, and it offers Significant cost savings. The
lesions found to be most amenable to endobronchial
electrosurgery are polypoid in morphology and
attached to the airway by a stalk. Flat or sessile lesions
can be treated by fulguration with the coagulation
probe (Coulter and Mehta 2000).
Endobronchial lesions in the tumorous type of
endobronchial tuberculosis are good candidates for
electrosurgery, because they usually present as pol-
ypoid masses or sessile lesions. Although the number
of cases was small, my previous study found that
electrosurgery produced successful results in this
H. S. Chung
type of endobronchial tuberculosis (Chung and Lee
2000). Granulation tissue caused by metallic stents
can also be removed by endobronchial electrosur-
gery. Special care must be taken not to touch the
wires when ablating these lesions, since electrical
conduction may occur through the wire mesh. The
circumferential treatment of endobronchial lesions
should be avoided in endobronchial electrosurgery,
as this may lead to cartilaginous damage, fibrosis and
subsequent stenosis.
21.15.1.4
Surgical Management
Severe bronchostenosis with poor response to medi-
cal treatment and intervention usually requires later
resection. Bronchoscopy and computed tomography
are the treatments of choice in the accurate diagnosis
of bronchial involvement and the assessment of sur-
gical indications (Watanabe et al. 1997). To preserve
lung function, bronchoplastic surgery is essential,
especially for bronchial stricture of the trachea or
larger bronchi, and appropriate antituberculosis
chemotherapy should be given for 9 to 12 months
perioperatively to prevent recurrence and restenosis
(Hsu et al. 1997).
21.15.2
Bronchiectasis
Distention by mucus, caseous tissue or secondary infec-
tion beyond a bronchial stenosis may result in bronchi-
ectasis, especially following lobar or segmental lesions.
(Seaton et al. 1989). The incidence of bronchiectasis is
reduced by prompt antituberculosis chemotherapy.
21.15.3
Broncholith
In primary tuberculosis, gross tracheobronchial
lymph-node enlargement may take place. In many
cases, these caseous nodes do not rupture but instead
inspissate, contract and calcify, and occasionally such
calcifications may later begin to ulcerate through the
bronchial wall to produce repeated hemoptysis and
eventually be extruded into a bronchus as a "broncho-
lith:' A partially protruded broncholith can sometimes
be removed bronchoscopically if not expectorated by
the patient. However, attempts to remove broncholiths
bronchoscopically can lead to profuse hemorrhage
and, therefore, are best avoided (Stradling 1981).
Endobronchial Tuberculosis
Acknowledgements. I am very grateful to Dr. Ho
Joong Kim, and Miss Jin Ha Park for their generous
contribution.
References
Alame T et al (1995) Lymph node perforation into the airway
in AIDS-associated tuberculosis. Eur Respir J 8:658-660
Albert RK, Petty TL (1976) Endobronchial tuberculosis pro-
gressing to bronchial stenosis. Chest 70:537-539
American Thoracic Society, Medical Section of the American
Lung Association (1990) Diagnostic standards and classifi-
cation of tuberculosis. Am Rev Respir Dis 142:725-735
American Thoracic Society, Medical Section of the American
Lung Association (1994) Treatment of tuberculosis and
tuberculosis infection in adults and children. Am J Respir
Crit Care Med 149:1359-1374
Becker HD et al (199l) Diagnostic studies, chap 3. In: Becker
HD, Kayser K, Schulz V, Tuengrthal S, Vollhaber H-H (eds)
Atlas of bronchoscopy: technique, diagnosis, differential
diagnosis, therapy. Decker, Philadelphia, pp 15-133
Broughton WA, Bass JB Jr (1999) Tuberculosis and diseases
caused by atypical mycobacteria. In: Albert RK, Spiro SG,
Jett JR (eds) Comprehensive respiratory medicine. Mosby,
London, pp 1-16
Caligiuri PA, Banner AS, Jenski RJ (1984) Tuberculous main-
stem bronchial stenosis treated with sleeve resection. Arch
Intern Med 144:1302-1303
Calpe JL, Chiner E, Larramendi CH (1995) Endobronchial
tuberculosis in HIV-infected patients. AIDS 9:59-64
Chan HS, Pang JA (1989) Effect of corticosteroids on dete-
rioration of endobronchial tuberculosis during chemo-
therapy. Chest 96, 1195-96
Choi YW et al (2002) Low-dose spiral CT: application to
surface-rendered three-dimensional imaging of central
airways. J Comput Assist Tomogr 26:335-341
Chung HS, Lee JH (1996) Study about the clinical features and
pulmonary function test of endobronchial tuberculosis (in
Korean). Tuberc Respir Dis 43:147-158
Chung HS, Lee JH (2000) Bronchoscopic assessment of the evo-
lution of endobronchial tuberculosis. Chest 117:385-392
Chung HS et al (1991a) Classification of endobronchial tuber-
culosis by the bronchoscopic features (in Korean). Tuberc
Respir Dis 38:108-115
Chung HS, Hyun IG, Han SK (1991b) Bronchoscopic elec-
trocautery for airway obstruction in the tumorous type
of endobronchial tuberculosis (in Korean). Tuberc Respir
Dis 38:347-356
Chung MP et al (1998) Bronchial stenosis due to anthracofi-
brosis. Chest 113:344-350
Coulter TD, Mehta AC (2000) The heat is on: impact of endo-
bronchial electrosurgery on the need Nd-YAG laser pho-
toresection. Chest 118:516-521
Daniel TM (1994) Tuberculosis, chap 130. In: Isselbacher KJ,
Braunwald E, Wilson JD, Martin JB, Fauci AS, Kasper DL
(eds) Harrison's principles of internal medicine, 13th edn.
McGraw-Hill, New York, pp 710-718
Dineen KM, Jantz MA, Silvestri GA (2002) Tracheobronchial
stents. J BronchoI9:127-137
Dooley DP, Carpenter JL, Rademacher S (1997) Adjunctive cor-
347
ticosteroid therapy for tuberculosis: a critical reappraisal
of the literature. Clin Infect Dis 25:872-887
Fraser RS et al (1994) Infectious disease of the lungs, chap 7.
In: Fraser RS, Pare JAP, Fraser RG, Pare PD (eds) Synopsis
of diseases of the chest, 2nd edn. Saunders, Philadelphia,
pp 287-391
Garay SM (1996) Pulmonary tuberculosis, chap 32. In: Rom
WN, Garay SM (eds) Tuberculosis. Little Brown, Boston,
pp 373-412
Garimella PS (2001) Anthracofibrosis. Chest 120 [Suppl):
327s
Glassroth J (1993) Diagnosis of tuberculosis. In: Reichman LB,
Hershfield ES (eds) Tuberculosis: a comprehensive interna-
tional approach. Dekker, New York, pp 149-166
Glynn JR (1998) Resurgence of tuberculosis and the impact of
HIV infection. Br Med Bull 54:579-593
Greenbaun M, Beyt BE, Murray PR (1980) The accuracy of
diagnosing pulmonary tuberculosis at a teaching hospital.
Am Rev Respir Dis 21:447-481
Han SK et al (1989) Balloon dilatation of bronchial stenosis
in endobronchial tuberculosis. Program and book of
abstracts. The 6th world congress for bronchology, Tokyo/
Kyoto,p 130
Hoheisel G et al (1994) Endobronchial tuberculosis: diagnostic
features and therapeutic outcome. Respir Med 88:593-597
Hsu HS et al (1997) Surgical treatment of endobronchial
tuberculosis. Scand Cardiovasc J 31:79-82
Ip MSM et al (1986) Endobronchial tuberculosis revisited.
Chest 89:727-730
Jantz M, Silvestri GA (2000) Silicone stents versus metal stents
for management of benign tracheobronchial disease-pro:
metal stents. J Bronchol 7:177-183
Jenks RS (1940) Tuberculous tracheobronchitis: a review. Am
Rev Tuberc 37:315-326
Jokinen K, Palva T, Nuutinen J (1977) Bronchial findings in
pulmonary tuberculosis. Clin Otolaryngol 2: 139-148
Judd AR (1947) Tuberculos tracheobronchitis. J Thoracic Surg
16:512-521
Judson MA, Sahn SA (1994) Endobronchial lesion in HIV-
infected individuals. Chest 105:1314-1323
Kim HY et al (2000) Bronchial anthracofibrosis (inflamma-
tory bronchial stenosis with anthracotic pigmentation): CT
findings. Am J Roentgenol 174:523-527
Kim Y et al (1997) Tuberculosis of the trachea and main
bronchi: CT findings in 17 patients. Am J Roentgenol168:
1051-1056
Kim YH et al (1993) Serial fiberoptic bronchoscopic observa-
tions of endobronchial tuberculosis before and early after
antituberculosis chemotherapy. Chest 103,673-77
Korpela A et al (1999) Bioabsorbable self-reinforced poly-L-
lactide, metallic, and silicone stents in the management of
experimental tracheal stenosis. Chest 115:490-495
Lee JH et al (1992) Endobronchial tuberculosis: clinical and
bronchoscopic features in 121 cases. Chest 102:990-994
Lee SH et al (1999) Endobronchial actinomycosis simulating
endobronchial tuberculosis: a case report. J Korean Med
Sci 14:315-318
Lerner BH (1999) Catching patients: tuberculosis and deten-
tion in the 1990 s. Chest 115:236-241
Low SY, Eng P, Hsu A (200l) Interventional bronchoscopy in
tuberculous tracheobronchial stenosis. Chest 120 [Suppl):
160s
Mangura BT, Reichman LB (1989) Pulmonary tuberculosis. In:
348
Pennington JE (ed) Respiratory infections: diagnosis and
management, 2nd edn. Raven, New York, pp 528-543
Matthews JI, Matarese SL, Carpenter JL (1984) Endobronchial
tuberculosis simulating lung cancer. Chest 86:642-644
McAdams HP, Erasmus J, Winter JA (1995) Radiologic mani-
festation of pulmonary tuberculosis. Radiol Clin North Am
33:655-678
Medlar EM (1955) The behavior of pulmonary tuberculous
lesion: a pathological study. Am Rev Tuberc 71:132-145
Millard PS et al (1994) Rurality and tuberculosis incidence
trends in North and South Carolina, 1982 to 1992. J Rural
Health 10:226-236
Moon WK et al (1997) Tuberculosis of the central airways: CT
findings of active and fibrotic disease. Am J Roentgenol
169:649-653
Myerson MC (1994) Tuberculosis of the trachea and bronchus.
Thomas, Springfield, pp 250-275
Nemir RL et al (1963) Prednisone therapy as an adjunct in the
treatment of lymph node-bronchial tuberculosis in child-
hood: a double-blind study. Am Rev Respir Dis 88:189-198
Oho K, Amemiya R (1984) Pathological bronchoscopic
findings, chap 4. In: Oho K, Amemiya R (eds) Practical
fiberoptic bronchoscopy, 2nd edn. Igaku-Shoin, Tokyo,
pp 94-150
Park CS et al (1995) Bronchial hyperreactivity in patients with
endobronchial tuberculosis. Respir Med 89:419-422
Park IW, Choi BW, Hue SH (1997) Prospective study of corti-
costeroid as an adjunct in the treatment of endobronchial
tuberculosis in adults. Respirology 2:275-281
Polesky A et al (1999) Airway coccidioidomycosis: report of
cases and review. Clin Infect Dis 28:1273-1280
Riley RL (1960) Apical localization of pulmonary tuberculosis.
Bull Johns Hopkins Hosp 106:232-239
Rossman MD, Mayock RL (1988) Pulmonary tuberculosis,
chap 7. In: Schlossberg D (ed) Tuberculosis, 2nd edn.
Springer, Berlin Heidelberg New York, pp 71-76
Rossman MD, Oner-Eyuboglu AF (1998) Clinical presentation
and treatment of tuberculosis, chap 162. In: Fishman AP,
Elias JA, Fishman JA, Grippi MA, Kaiser LR, Senior RM
H. S. Chung
(eds) Fishman's pulmonary diseases and disorders, 3rd
edn. McGraw-Hill, New York, pp 2483-2503
Saadoun R et al (1998) Endobronchial tuberculosis present-
ing as an obstructive tumor in an HIV-l-positive patient:
apropos of a case and review of the literature. Rev Med
Interne 19:344-347
Salkin D, Cadden V, Edson RC (1943) The natural history of
tuberculous tracheobronchitis. Am Rev Tuberc 47:351-359
Seaton A, Seaton D, Leitch AG (1989) Clinical features of
tuberculosis, chap 14. In: Seaton A, Seaton D, Leitch AG
(eds) Crofton and Douglas's respiratory diseases, 4th edn.
Blackwell, Oxford, pp 395-422
Shulutko et al (1979) Tuberculosis. In: Lukomsky GI (ed) Bron-
chology. Mosby, St Louis, pp 287-305
Smart J (1951) Endobronchial tuberculosis. Br J Dis Chest 45:
61-68
Smith LS, Schillaci RF, Sarlin RF (1987) Endobronchial tuber-
culosis: serial fiberoptic bronchoscopy and natural history.
Chest 91:644-647
Song JH, Han SK, Heo 1M (1985) Clinical observation of
endobronchial tuberculosis(in Korean). Tuberc Respir Dis
4:276-284
Stradling P (1981) Inflammatory and associated changes, chap
5. In: Stradling P (ed) Diagnostic bronchoscopy, 4th edn.
Churchill Livingstone, New York, pp 61-81
Van den Brande PM et al (1990) Clinical spectrum of endo-
bronchial tuberculosis in elderly patients. Arch Intern Med
150:2105-2108
Watanabe Y et al (1997) Treatment of bronchial stricture due
to endobronchial tuberculosis. World J Surg 21:480-487
Watson JM, Ayres JG (1988) Tuberculous stenosis of the tra-
chea. Tubercle 69:223-226
Williams DJ et al (1988) Endobronchial tuberculosis present-
ing as asthma. Chest 93:836-838
Wolinsky E (1989) Tuberculosis, chap 26. In: Baum GL, Wolin-
sky E (eds) Textbook of pulmonary diseases, 4th edn. Little
Brown, Boston, pp 465-520
Yee A et al (1985) Pigmented polypoid obstructive endobron-
chial tuberculosis. Chest 87:702-703
22 Pleural Tuberculosis
M. MONIR MADKOUR, MAJDY IDREES, MONA
CONTENTS
22.1 Introduction 349
22.2 Epidemiology 349
22.3 Pathogenesis of Tuberculous Pleural Effusion 350
22.4 Clinical Features 351
22.5 Diagnosis of Tuberculous Pleural Effusion 352
22.5.1 Diagnosis 352
22.5.2 Pleural Fluid Examination 352
22.5.3 Pleural Biopsy 353
22.5.4 Thoracoscopy in Tuberculous Patients 354
22.5.5 Tuberculin Skin Test (PPD) 354
22.5.6 Imaging Features of Pleural Tuberculosis 355
22.5.7 Treatment 355
References 356
22.1
Introduction
Tuberculous pleurisy is a common disease in develop-
ing countries and its incidence is increasing in devel-
oped countries. It is considered an extrapulmonary
manifestation of tuberculosis despite its intimate
anatomical relationship with the lung parenchyma.
In the United States, because of the recent increase
in the incidence of tuberculosis due to the HIV epi-
demic and immigrants from developing endemic
countries, the incidence of pleural tuberculosis has
increased in parallel with, and has been found in 20%
of, AIDS patients (Ankobiah et al. 1990).
Although tuberculous pleural effusion is associ-
ated with parenchymal infiltrates and cavitation in
approximately 50% of patients, isolated pleural effu-
sion without lung parenchymal disease can present
M. M. MADKOUR, MD, DM, FRCP
Consultant, Department of Medicine, Riyadh Armed Forces
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
M. IDREES, MD, FRCP (C), FCCP
Head, Pulmonary Function Laboratory, Division of Pulmonary
Medicine, Department of Medicine, Riyadh Armed Forces
Hospital, ClIO, P.O. Box 7897, Riyadh 11159, Saudi Arabia
M. AL SHAHED, MBBS, FRCR
Consultant Radiologist, Department of Radiology, Riyadh Armed
Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi Arabia
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
AL SHAH ED
a diagnostic challenge; malignancy is of particular
concern (Hunlnick et al. 1983; Hirsch et al. 1979). It
is estimated that between 15 to 20% of patients with
exudative pleural effusion in general are difficult to
diagnose. In developing countries with poor resources,
conventional methods of diagnosis by pleural fluid
culture may only be positive in 25-50% of patients
(Idell 1994). Presumptive diagnoses of tuberculous
origin in exudative pleural effusion, based on clinical
features, radiography and response to antituberculous
chemotherapy in endemic countries are still adopted
by many clinicians. Patients who are not treated for
tuberculous pleural effusion often develop reactiva-
tion pulmonary disease (Palmer 1979).
In developed countries, due to recent advances
in investigative facilities, particularly in the field of
molecular biology assays, the diagnosis can be made
in few hours or days if other investigations are not
helpful (Takagi et al.1998; Querol et al.1995; Lassence
et al. 1992).
22.2
Epidemiology
Pleural tuberculosis is the second commonest form of
extra pulmonary parenchymal disease, after the tuber-
culous lymphadenitis. Pleural tuberculosis is also more
common in patients that are co-infected with AIDS
(11%) than among those without AIDS (6%).
Other reports indicating a higher incidence of
pleural tuberculosis (20%) in AIDS patients have
found it to be the most common cause of pleural
effusion in this patient population (Modilevsky et al.
1989; Shivaram et al.1989;Ankobiah et aI.1990).
In developed countries, pleural tuberculosis is
more commonly caused by post-primary disease than
by primary infection. Reactivation of the disease was
the cause of tuberculous pleural effusion in 64% of
patients seen in Scotland and reported by Moudgil et
al. (1994). Primary pleural tuberculosis was noted in
7% of patients with active pulmonary disease (Aktogu
350
et al. 1996). In young adults with HIV who are co-
infected with tuberculosis, the primary disease as a
cause of pleural tuberculosis was more common than
reactivation (post-primary) (Sudre et al.1992; Batung-
wanayo et al.1993; Richter et al.1994).
Seibert et al. (1991) from Alabama, USA, reviewed
the medical records of 1,738 patients with tubercu-
losis seen between 1968 and 1988, to determine the
age distribution, pleural fluid culture and frequency
of pleural effusion among those with parenchymal
disease, as well as the prevalence of positive tubercu-
lin test. Pleural tuberculosis was found in 70 patients
(4.0%) of all forms of the disease. The mean age was
47±17.7 years, with male predominance and pul-
monary parenchymal involvement noted in 50% of
patients. In Spain, two large studies of patients with
pleural effusion were undertaken to determine the
causes. The first series was a retrospective study of
414 patients with pleural effusion of unknown cause,
who were seen between 1979 and 1986. Tuberculous
cause was identified in 107 patients (28.5%) (Bueno
et al. 1990). Male predominance was noted and the
average age was 55 years.
In the second series, 642 patients with pleural effu-
sion were seen between 1989 and 1993. Tuberculous
pleural effusion was found in 25% of these patients,
similar to the earlier series, and the average age was
also similar (57.1±21.1 years) with male predomi-
nance as well (Valdes et al.I996b).
In Australia, Christopher et al. (1998) reported a
prospective study of 27 patients with plural effusion
that were hospitalized for the purpose of etiological
diagnosis. Tuberculosis was found in 16 patients
(59%) and was considered as the single most common
cause of exudative pleural effusion in that series.
In the United States, the incidence of tuberculosis as
a cause of pleural effusion in AIDS patients was found
to be increased. In New York, Relkin and colleagues
(1994) retrospectively studied pleural tuberculosis
in HIV-positive patients and found it to occur at a
younger age. They found 70 patients with tuberculo-
sis pleural effusion, including 43 HIV-positive (mean
age 38±1 years) and 27 HIV-negative (mean age 52±3
years). They also confirmed previous observations
that pleural tuberculosis in HIV-negative patients
occurs more frequently in the older age group.
In developing countries, such as some sub-Saha-
ran African countries, the prevalence of tuberculosis
has increased as a result of the HIV epidemic and the
incidence of pleural tuberculosis has also increased
in parallel. Richter and colleagues from Tanza-
nia (Richter et al. 1991) prospectively studied 127
patients with pleural effusion. Tuberculosis was the
M. M. Madkour et al.
most common cause (70%) in this series. The authors
also found that among the 89 patients with tubercu-
lous pleurisy 57 (64%) were HIV-positive.
In Zimbabwe where HIV is endemic, tuberculosis
has become the single most likely cause of opportu-
nistic infection (Heyderman et al. 1998). The preva-
lence of HIV among male factory workers was 19.4%
and in females was over 30% (Mbizvo et al. 1996).
In 1995, Harare provided an annual report of the
incidence of tuberculosis, which was 195 cases per
100,000 population, with over 40% of the TB cases
co-infected with HIV.
In our series of 176 patients with culture positive
sputum for M. tuberculosis, 46 patients had pleural
effusion (26.1 %) - see Table 2 in the chapter on post-
primary tuberculosis. The average age group was 40
years, with male predominance (75%). The major-
ity of the patients were from the central (57%) and
southern (29%) provinces of Saudi Arabia. Unlike
the situation in the sub-Saharan African countries,
HIV was not found to be a contributing factor for the
development of tuberculous pleural effusion.
22.3
Pathogenesis of Tuberculous Pleural
Effusion
Tuberculous pleurisy has been described by an
immunologist as an excellent model for studying the
immune response in vivo. The mode of transmission
of M. tuberculosis to the pleura may be hematogenous
spread, as in primary tuberculosis, or it may be second-
ary to the direct invasion from adjacent tuberculous
lymphadenitis or the rupture of subpleural tubercu-
lous lung parenchymal cavity; it may also result from
extrapulmonary disease such as spinal tuberculosis
(Sahn 1988; Stevenson 1955; Mohammed et al. 1998;
Maeda et al. 1993; Wallis 1996; Zhang et al. 1994;
Morehead 1998). The pleura respond to the myco-
bacterial antigens with an extensive infiltration of
mononuclear phagocytic cells and increased produc-
tion of cytokines; however, the mechanisms by which
these cells are recruited to the pleural space remains
unclear. Several recent immunological studies have
suggested that activated macrophages and CD4 lym-
phocytes play the most important role in cell-medi-
ated immunity with granuloma formation against
M. tuberculosis. Antigens from the wall of the bacilli
have a potent stimulant effect and lead to the exces-
sive production of cytokines (TNF-a and IFN-y) from
these inflammatory cells, which in turn contribute to
Pleural Tuberculosis
the formation of granuloma and to the defense mecha-
nism against infection (Epstein et al. 1987; Shimokata
et al. 1986; Barnes et al. 1989, 1990; Lorgat et al. 1992;
Kunkel et al. 1989; Willis et al. 1990; Zhang et al. 1994;
Wallis 1996). Several observations have suggested that
a vigorous in-situ or "compartmentalized" intrapleu-
ral immune response with activation of cell-mediated
immunity (CMI) and delayed type hypersensitivity
(DTH) with cytokine production in response to the
bacilli cell-wall antigens take place (Barnes et al. 1989).
Mycobacterium tuberculosis bacilli contain multiple
antigens, including polysaccharides (arabinogalactan
and arabinomannan) and proteins (mainly glycolip-
ids and lipoproteins). Polysaccharides from the bacilli
generally do not elicit delayed type hypersensitivity
(DTH) (Chaparas et al. 1971; Yamamura et al. 1968;
Barnes et al. 1993). Lipoprotein and glycolipids anti-
gens elicit DTH and stimulate T lymphocyte and
cytokine production, which can be detected in high
concentration in the pleural fluid. Pleural mesothelial
cells are the first cells to respond to bacilli invading the
pleural space. These cells are metabolically active and
likely initiate and propagate an inflammatory reaction
(Mohammed et al. 1998). Mesothelial cells have been
found to initiate inflammation in response to various
agonists by the production and release of chemokines
(Boylan et al. 1992; Goodman et al. 1992; Jonjic et al.
1992).lnterleukin 12 (IL-12) cytokine has been found
in high concentration in the pleural fluid of patients
with tuberculous pleurisy (Zhang et al. 1994).
Zhang and colleagues from California found that
IL-12 play an important role in the immune response
by enhancing the production of interferon-y, facili-
tating development of Thl cells and augmenting
cytotoxicity of antigen-specific T-cells and natural
killer cells. IL-12 enhances natural killer cell-medi-
ated cytotoxicity and augments antigen-dependant
proliferation by CD8+ cytotoxic T lymphocytes
(Zhang et al. 1994). It has also been suggested that IL-
12 enhances cytotoxicity by activating CD4 + T cells
against the bacilli loaded phagocytic macrophages,
leading to its apoptosis and the destruction of the
bacilli within it. Evidence for the excessive produc-
tion ofantigen-reactive T lymphocytes,IFN-y, TNF-u
and IL-12 is provided by the elevated levels of these
cytokines in the pleural fluid of tuberculous patients.
There is evidence that IL-12 is produced locally in
the pleura in response to M. tuberculosis antigens,
and that when IL-12 is experimentally suppressed,
the bacilli proliferate again. These findings suggest
that IL-12 contributes to lymphocyte recognition of
M. tuberculosis antigens, probably by enhancing pro-
liferation of activated T cells. The high concentration
351
of IFN-y and TNF-u (both have anti-mycobacterial
activity) in the pleural fluid are a likely indication
that they playa role in immune resistance. TNF-u
augments the macrophages' capacity to phagocytose
and kill the bacilli (Zhang et al. 1994). The authors
conclude that IL-12 contribute to the immune
response against M. tuberculosis by enhancing pro-
duction of IFN-y, the T cells and orchestrating anti-
gen-specific cytolytic mechanisms.
CD4 T cells also playa critical role in the immune
response to M. tuberculosis antigens. In 1996, Wallis
(from Ohio), studied the response of T-cells to alpha
antigen, a major protein from the wall of the bacilli,
and reported the increased IFN-yproduction.
As a result of the immune response and the out-
come of the infection, inflammatory cellular infil-
trates and granulomata formations may occur. As
the infection progresses, inflammatory responses
and cytokines release will lead to increased vascular
permeability of the local capillaries. Subsequently,
plasma protein exude fluid into the pleural space
forming pleural effusion (Ellner et al. 1988). In post-
primary pleural effusion, the disease may be com-
plicated by empyema, lung parenchymal spread or
bronchopleural fistula. Pleural fibrosis and calcifica-
tion may occur in chronic tuberculous empyema.
22.4
Clinical Features
Pleural effusion due to tuberculosis most commonly
occurs 3 to 12 months, or even more, after the pri-
mary infection in adolescents and young adults, but
it may occur at any time during the course of the
disease (Gedde-Dhal 1952; Berger and Mejia 1973;
Sahn 1988; Ansari and Idell 1998). Pleural tuber-
culosis, due to reactivation of the disease, may be
noted in up to 50% of patients with post-primary
lung parenchymal involvement, particularly with
cavitations (Seibert et al. 1991).
The onset of symptoms is acute in about 70% of
patients with cough, fever and chest pain. Gradual
onset may be noted in about 30% with dyspnea,
weight-loss and weakness. The frequency of these
symptoms may vary from one series to another.
The presence or absence of co-infection with
HIV does not change the frequency of symptoms,
but non-HIV patients are more symptomatic (See
Table 22.1).
Non-symptomatic pleural effusion may occur in
HIV co-infected patients and was reported in 4 of 21
352
Table 22.1.
Symptoms Our Own Hong Kong USA
series (Chang et al. 1991) (Ankobiah et al. 1990) (Heyderman et al. 1998)
No HIV No HIV ±HIV
Cough 0/0 63 71 60
Dyspnea 0/0 77 48 38
Fever 0/0 86.5 71 72
Weight loss 0/0 90 55 59
Chest pain 0/0 44 53 44
HIV patients (20%) by Ankobiah et al. (1990) from
New York. Clinically, sharp pain with pleural rub
during deep inspiration on the affected side may be
noted. The clinical features of pleural effusion may
depend on its size, and are manifested as decreased
chest movement, dullness of percussion, decreased
breath sound or retraction of chest wall due to exten-
sive pleural fibrosis. These features are not character-
istic of tuberculous pleurisy. Pleural tuberculosis is
most commonly unilateral but may be bilateral in 5%
of patients (Chan et al.1991; Heyderman et al.1998).
In our experience with 46 patients with proven
pleural tuberculosis, 44 (96%) had unilateral pleural
effusion. One patient (2%) had transudative effusion
based on Light's criteria. Thirty patients (65%) had
moderate effusions and 11 (24%) had massive effu-
sions. Loculated effusion was reported in 9 patients
(20%) and pleural calcification in 7 (15%).
22.5
Diagnosis of Tuberculous Pleural Effusion
22.5.1
Diagnosis
The diagnosis of pleural tuberculosis begins with a
high index of suspicion, clinical features, PPD skin
test, imaging features and a culture of all potentially
diagnostic specimens (body fluids or tissue biopsies).
Screening methods used to differentiate tuberculous
from non-tuberculous pleural effusion have been
evaluated to identify the discriminate power of each
investigative parameter. In a recent report from
Spain, Carrion-Valero and Perpina-Tordera (2001)
screened tuberculous pleural effusion by discrimi-
nant analysis of 189 patients. In this retrospective
study, the discriminating power of routine imagin-
ing features and laboratory parameters was reviewed.
Using the backward elimination method, the authors
found that age, the tuberculin skin test, pleural white
M. M. Madkour et al.
Zimbabwe
±HIV
97
82
73
51
73
cell count and bloodstained exudative effusion were
the best discriminating functions in the screening
for pleural tuberculosis. The diagnosis, however, can
only be confirmed by the presence of M. tuberculo-
sis determined by direct staining or by culture. The
diagnosis of pleural tuberculosis has been defined
for the purpose of clinical studies by several authors
when one or more of the following were confirmed:
(1) positive microbiological findings (smear or cul-
ture) in pleural fluid or biopsy; (2) histopathological
evidence of granulomatous pleuritis with clinical
andlor radiological response to anti-tuberculous
treatment; (3) positive skin testing, recent conver-
sion in young patient with a lymphocytic, exudative
effusion (Ankobiah et al. 1990; Seibert et al. 1991;
Morehead 1998).
22.5.2
Pleural Fluid Examination
Pleural fluid examination remains an important
tool in the of investigation of tuberculous pleu-
risy. Tuberculous pleural fluid is an exudate which
may be cloudy (turbid), yellow, serosanguineous or
hemorrhagic in appearance. The PH usually ranges
between 7.30 and 7.40 or lower, particularly among
those co-infected with HIV (George et al. 1985;
Pablo et al. 1997). Total WBC may range from 5,000
to 10,000/mm3, with more than 50% lymphocytes;
however, polymorphs may be seen early after the
onset of symptoms. Red cell count may be raised to
a variable degree in the pleural fluid.
Mesothelial cell count is usually less than 5% of the
pleural fluid white cell count. Raised mesothelial count
above 5% is considered by many authors to argue
against tuberculous pleurisy (Spriggs and Boddington
1960). However, some authors have reported mesothe-
lial cell counts above 5% in patients with tuberculous
pleurisy (Hirsch et al. 1979; Lau 1989; Santos-Santre et
al. 1990). Pleural fluid protein is usually greater than
3.0 Gldl, and glucose may be decreased. Pleural fluid
Pleural Tuberculosis
lactic dehydrogenase (LDH) is usually above 200 LU.I
L. Pleural fluid adenosine deaminase (ADA) measure-
ment has been used as a biochemical parameter in
an attempt to differentiate between tuberculous and
other exudative non-tuberculous pleural effusions
(Valdes et al. 1996).
In a prospective study from India, Sharma et al.
(2001) reported the diagnostic value of ADA as a
marker for tuberculous pleurisy. Seventy-five patients
with exudative pleural effusion were assessed. Pleural
fluid ADA, as well as serum ADA, levels were signifi-
cantly higher in tuberculous pleurisy compared with
non-tuberculous effusions. The authors reported
the sensitivity and specificity of pleural ADA at two
cut-off points, 35 lUlL and 100 lUlL. At 100 lUlL, the
sensitivity of ADA was 40% and the specificity was
100%. The authors suggested that using 100 lUlL
pleural ADA level may spare as many as 40% of
patients from having a pleural biopsy. However, ADA
level estimation is complicated by false-positive and
false-negative results (Yamada et al. 2001). Maartens
and Bateman (1990) from South Africa found in their
prospective study of 111 patients with pleural effu-
sion that ADA did not provide a valuable diagnostic
test of pleural tuberculosis as has been suggested.
In recent years, the use of pleural fluid cytokines as
parameters for the diagnosis of tuberculous pleurisy
has been reported (Yamada et al. 2001; Pablo et al.
1997; Xirouchaki et al. 2002). In a comparative study
between raised pleural levels of ADA and cytokines,
interleukin-8 (IL-8), tumor necrosis factor alpha
(TNF-a) and interferon gamma (IFN-y), Yamada et al.
reported their findings. The study involved samples of
pleural fluid obtained from 21 tuberculous, 21 inflam-
matoryand 18 malignancy cases. The authors indicated
that IFN-y was a "very reliable marker" of tuberculous
pleurisy. Xirouchaki and colleagues (2002) indicated
that the measurement of pleural fluid cytokines might
be helpful in differentiating malignancy from tubercu-
10sis in exudative pleural effusion.
Methods for measuring antigens and antibodies to
M. tuberculosis in the pleural fluid have been reported
by many authors. It has further been reported that the
sensitivity of these tests are only about 50% (Murante
et al.1990; Hara et al. 1992; Caminero et al.1993). See
chapter on Immunological Tests for Tuberculosis.
Molecular biology techniques utilizing various PCR
protocols have been developed for detecting the M.
tuberculosis genome in pleural and other body fluid
samples, as well as in tissue biopsies. The reported
sensitivity and specificity has varied from one tech-
nique to another. Querol and colleagues (1995) from
Spain reported 107 patients with pleural effusion
353
between 1991 and 1993. Twenty-one patients had
proven tuberculosis and 86 non-tuberculous (cancer,
parapneumonia, cirrhosis, heart failure, emphysema,
lymphoma, SLE and non-specific pleuritis). The PCR
assay was based on the detection of a 123-bp DNA
segment belonging to the insertion sequence IS6100,
specific for M. tuberculosis. The diagnoses of twenty-
one tuberculous patients were confirmed by clinical,
PPD skin test, pleural ADA, cytology, microbiology
and histopathology of pleural biopsies. Positive pleu-
ral biopsy with granuloma formation was found in
72%, positive culture in 67% and raised ADA activity
in 86%. PCR sensitivity and specificity were found to
be 81% and 100%, respectively. Other authors applied
PCR to detect DNA (IS 6110) in pleural biopsy speci-
mens (Takagi et al.1998). These authors reported PCR
sensitivity and specificity of 89% and 100%, respec-
tively, of patients, which was similar to the results of
Querol and colleagues. These two studies demonstrate
the excellent sensitivity and specificity of PCR using
either pleural fluid or pleural tissue for rapidly diag-
nosing pleural tuberculosis.
Pleural fluid microbiological examination (smear
and culture) for M. tuberculosis may detect only
25% to 50% of patients (Idell 1994). In patients co-
infected with HIV, pleural fluid cultures are more
often positive than in non-HIV patients. Parenchy-
mal lung infiltrates are more severe and depicted by
various imaging modalities in HIV positive patients,
suggesting more mycobacterial extension from the
lung into the pleural space, in contrast to HIV nega-
tive individuals (Luzze et al. 2001). In a prospective
study of 111 patients with pleural effusion, and from
an area with a high prevalence of tuberculosis in
South Africa, radiometric cultures using BACTEC
and conventional cultures were compared (Maartens
and Bateman 1990). Tuberculosis was confirmed by
histopathology, microbiology or both in 62 patients
(56%). Positive pleural fluid culture was found in
47%, while histology and tissue culture were positive
in 84% and 71%, respectively. BACTEC was faster
than conventional mycobacterial culturing, and the
yield appeared after 18 versus 33 days. The incidence
of a positive pleural fluid culture as reported in some
series is shown in Table 22.2.
22.5.3
Pleural Biopsy
Closed needle biopsy of the pleura should be con-
sidered for further evaluation if clinical features,
PPD skin test, imaging findings and pleural fluid
354
Table 22.2. Pleural tuberculosis in some recent reports: diagnostic procedures, % yield
Series Year Fluid culture Biopsy culture
% positive % positive
Maartens and Bateman (l990) 47 71
Ankobiah et al. (l990) 57 45.6
Seibert et al. (1991) 58 66.7
Chang et al. (l991) 23 40
Relkin et al. (1994) 83.3 65.5
Kitinya et al. (1994) 28 45.6
Heyderman et al. (1998) 14b (smear) 40
Our Own Series 23 65
'Had problem with PPD reading compliance
bCulture was not reported
biochemical parameters and culture do not provide
a definitive diagnosis. It is an invasive procedure with
significant discomfort and potential complications
(Ansari and IdellI998). Good quality single pleural
biopsy sampling is sufficient to diagnose pleural
tuberculosis, in contrast to the increased diagnostic
yield with multiple biopsies for malignancy (Jimenez
et al. 2002). Positive pleural biopsy culture was found
to be higher than pleural fluid (39% versus 13%) by
Bueno and colleagues (1990).
Histopathological findings of granuloma in
pleural biopsy may be found in up to 80% of cases.
PCR of pleural tissue biopsy for M. tuberculosis was
found to be rapid, sensitive and specific in 89-100%
of cases for diagnosing pleural tuberculosis (Takagi
et al. 1998).
An audit to value the routine practice of pleural
fluid aspiration and biopsy for the investigation of
pleural effusion has been reported by Walshe and
colleagues (1992). The authors reviewed the investi-
gative results of pleural fluid and biopsy procedures
of 112 patients. They reported that the protein con-
tent of pleural fluid was of little value as a diagnostic
indicator, as it overlapped substantially between vari-
ous other diagnostic groups. A low positive yield for
microbiological findings was of significance. Pleural
biopsy was performed only in 30% in this series. The
authors, however, indicated that the practical diffi-
culties regarding pleural biopsy included the lack of
experience among junior staff, lack of pleural biopsy
needles and inadequate or poor quality samples. The
authors reported the rate of complications (pneumo-
thorax and emphysema) as 2% for aspiration alone
and 4% for aspiration combined with biopsy. The
incidence of pleural tissue biopsies, positive cultures
and the presence of granulomata on histopatho-
logical examination in different series are shown in
Table 22.2.
M. M. Madkour et al.
Biopsy histology PPD %
% positive positive
84 78
80.7 60
84.6 93
97 77
80 56
100 74
60 30'
80 76
22.5.4
Thoracoscopy in Tuberculous Patients
Thoracoscopic examination was first described by
Professor Hans Christian Jabobaeus from Sweden in
1910. He inserted a cystoscope into the pleural space
of two patients with exudative pleuritis (Thomas
1994). He reported his observations of patients with
pulmonary tuberculosis that had been treated with
pneumothorax in which there was no pleuritis. He
also explored the therapeutic application of tho-
racoscopy in breaking adhesions in patients with
pulmonary tuberculosis that were selected for lung
collapse therapy. He noted the complications of tho-
racoscopy, including haemorrhage, pleural effusion
and subcutaneous emphysema.
At present, the role of thoracoscopy in obtaining
pleural tissue biopsies in the diagnosis of pleural
tuberculosis is restricted, as a closed-needle pleural
biopsy is the most reasonable method of initially
obtaining a diagnostic tissue specimen.
22.5.5
Tuberculin Skin Test (PPD)
The use ofPPD skin testing is valuable in the diagnosis
of patients with pleural tuberculosis. In patients with
recent conversion, pleural tuberculosis will be classi-
fied as primary tuberculous pleurisy. Patients with a
past history of positive PPD or active tuberculosis will
be classified as secondary (Ankobiah et al. 1990).
The PPD skin test is considered positive when
the induration measurement is over 9 mm, and it is
found in 60 to 90% of patients with pleural tubercu-
losis as reported in several series (See Table 22.2). A
negative PPD skin test is more frequently reported in
patients with tuberculous pleurisy co-infected with
Pleural Tuberculosis 355

HIV than in those with negative HIV. Ankobiah and sion (Fig.22b in chapter 23). Bronchopleural fistula
colleagues (1990) from New York found positive PPD may be depicted by CT or fistulograrn (see Fig. 36 in
reaction in only 12% of AIDS patients and 80% in chapter 23). Fibrothorax with diffuse pleural thicken-
non-AIDS patients. Relkin et al. (1994) found a simi- ing but without effusion may be seen on CT.
lar low incidence of positive PPD in their patients Ultrasonography (US) may play a role in the
with tuberculous pleurisy co-infected with HIV. investigation of tuberculous pleural effusions. It may
compared with HIV negative ones (41 versus 76%). help in detecting pleural thickening, nodularity and
The high frequency of negative PPD skin tests among in obtaining US-guided closed-needle pleural biopsy
those co-infected with HIV reflects the alternation of (Chang et al. 1991; Akhan et al. 1992). The imaging
delayed hypersensitivity due to HIV. features of various modalities for pleural tuberculo-
sis are non-specific, but may help in distinguishing
tuberculous from non-tuberculous pleural disease.
22.5.6
Imaging Features of Pleural Tuberculosis
22.5.7
The imaging modalities most commonly used in Treatment
tuberculous pleural disease are plain chest radiogra-
phy, CT and ultrasonography. Rarely, a fistulogram Treatment of pleural tuberculosis is similar to treat-
for bronchopleural fistula is used when CT localiza- ment of pulmonary disease. Full assessment is essen-
tion of the tract is not clear. CT is more sensitive tial, particularly in regard to prior treatment for TB,
and useful than plain radiography in the evaluation drug susceptibility or resistance. The co-infection
of pleural disease. It may show unrecognized small with HIV has to be determined. Detection of com-
subpleural cavities, parenchymal lung involvement, plicated pleural effusion is essential, as it may lead
lymphadenitis or rib involvement (see Figs. 36a, bin to therapeutic failure if medical treatment is used
chapter on "Radiology of Pulmonary Tuberculosis") alone without surgical intervention. If left untreated,
(Moon et al. 1999; Yilmaz et al. 1998; Winer-Muram pleural tuberculosis will develop into active pulmo-
and Rubin 1990; Hunlnick et al. 1983). nary or extra-pulmonary disease within 5 years in at
Unilateral, moderate or large pleural effusion can least 65% of immunocompetent patients (Roper and
easily be depicted by plain radiography (see Figs. 12a, Waring 1955).
b, 34b in the chapter on "Radiology of Pulmonary Uncomplicated, HIV negative tuberculous pleu-
Tuberculosis"). Bilateral small or moderate pleural risy can be treated with a short course of 6-9 months
effusions may be depicted by plain radiography and of anti-tuberculous chemotherapy.
CT (see Fig. 12c in chapter 23). Imaging features of Six months of anti-tuberculous chemotherapy
contiguous subpleural tuberculous cavities in the lung using two drugs for pleural tuberculosis was reported
parenchyma as localized empyema may be seen on by Dutt et al. (1992). The authors used isoniazid 300 mg
plain radiography or CT (see Figs. 22b, 26 in chapter 23) and rifampicin 600 mg daily for one month, followed
(Seibert et al. 1991). Other imaging features of lung by twice weekly treatment with isoniazid 900 mg and
parenchymal disease with bronchiectasis may be rifampicin 600 mg for five more months. This regimen
noted in patients with tuberculous pleural effusion used by Dutt and colleagues on 161 patients with TB
(see Fig. 30b in chapter 23). Features of large empy- pleurisy and follow-up for about 4-6 months revealed
ema with increased thickness of the pleura may be no relapses. In areas with high risk for HIV infection
found on plain radiography and CT (see Fig.34b in or with high incidence of drug-resistant mycobacteria,
chapter 23). Tuberculous pleural empyema may be initial treatment should include at least four drugs
localized with thick, calcified wall better depicted on until drug susceptibility testing is available.
CT than plain chest radiography (see Figs. 13, 22b, 23, Immunosuppressed patients with pleural tuber-
34a, 35a in chapter 23). CT may depict focal pleural culosis should be treated for a much longer period,
thickness with calcifications in the absence of fluid at least 12 months. Patients with multi-drug resistant
collection (see Figs. 23, 32d in chapter 23). An asso- tuberculosis should be treated with 5-6 drugs.
ciated destruction of adjacent ribs with cold abscess Complicated tuberculous pleural effusion may
formation and empyema may be depicted better by CT be due to the development of loculated chronic
(35a, b in chapter 23). Other calcifications in the lung empyema, drug resistance or bronchopleural fistula.
parenchyma or mediastinal lymph nodes may also be Medical treatment alone will not be sufficient in such
depicted in association with tuberculous pleural effu- patients and surgical intervention will be required
356 M. M. Madkour et al.

(Brown and Pomerantz 1995; Iseman et al. 1991). Immunol107:149-153

Surgical treatment is discussed in detail in a separate Christopher DJ, Peter JV, Cherian AM (1998) Blind pleural
biop~y using a tru-cut needle in moderate to large pleural
chapter (see chapter XX).
effUSIOn - an experience. Singapore Med J 39:196-199
In a recent report from India, Lahiri et al. (1998) Dutt AK et al (1992) Tuberculous pleural effusion: 6-month
analyzed the surgical procedures for thoracic tuber- therapy with isoniazid and rifampicin. Am Rev Respir Dis
culosis in 1655 cases seen over a 20-year period. 145:1429-1432
These procedures were needed in only 2.2% of the Ellner JJ et al (1988) The immunology of tuberculous pleurisy.
Semin Respir Infect 3:335-342
patients, including tuberculous empyema with or
Epstein DM et al (1987) Tuberculosis pleural effusions. Chest
without broncho-pleural fistula. Intercostal drainage 91:106-109
with irrigation, thoracostoma and chest wall tube Gedde-Dahl T (1952) Tuberculosis infection in the light of
drainage were the most frequent minor procedures. tuberculin maticuolation. Am J Hyg 56:139-214
Major procedures were less frequently required and George RB et al (1985) Mycobacterial, fungal, actinomycotic and
nocardial infection of the pleura. Clin Chest Med 6:63-75
included decortication and thoracoplasty.
Goodman RB, Wood RG, Martin TR et al (1992) Cytokine-
stimulated human mesothelial cells produce chemotactic
activity for neutrophils including NAP-lIlL-8. J Immunol
148:457-465
References Hara N et al (1992) Pleural SC56-9 in differential diagnosis
of tuberculous, malignant and other effusions. Chest 102:
1060-1064
Akhan 0 et al (1992) Tuberculous pleural effusions: ultrasonic
Heyderman RS,Makunike R,Muza T et al (1998) Pleural tuber-
diagnosis. J Clin Ultrasound 20:461-465
culosis in Harare, Zimbabwe: the relationship between
Aktogu S et al (1996) Clinical spectrum of pulmonary and
human immunodeficiency virus, CD4 lymphocyte count,
pleural tuberculosis: a report of 4, 480 cases. Eur Respir
granuloma formation and disseminated disease. Trop Med
J 9:2031-2035
Int Health 3:14-20
Ankobiah WA, Finch P Powell S et al (1990) Pleural tuberculo-
Hirsch A et al (1979) Pleural effusion: laboratory test in 300
sis in patients with and without AIDS. J Assoc Acad Minor
cases. Thorax 34:106-112
Phy 1:20-23
Hunlnick D et al (1983) Pleural tuberculosis evaluated by
Ansari T, Idell S (1998) Management of undiagnosed persis-
computed tomography. RadioI149:759-765
tent pleural effusions. Clin Chest Med 19:407-417
Idell S (1994) Evaluation of perplexing pleural effusion. Con-
Barnes PF, Mistry SD, Cooper CL et al (1989) Compartmental-
temp Intern Med 6:31-39
ization of a CD4+ T lymphocyte subpopulation in tubercu-
Iseman MD et al (1991) Chronic tuberculous empyema with
lous pleurisy. J ImmunoI142:1114-1119
bronchopleural fistula resulting in treatment failure and
Barnes PF, Fong S-J, Brennan PJ et al (1990) Local production
progressive drug resistance. Thorax 100:124-127
of tumor necrosis factor and IFN-'Yin tuberculous pleurisy.
Jimenez D et al (2002) Determining the optimal number of
J ImmunoI145:149-154
specimens to obtain with needle biopsy of the pleura.
Barnes P et al (1993) Cytokine production at the site of disease
Respir Med 96:14-17
in human tuberculosis. Infect Immun 61:3482-3489
Jonjic N, Peri G, Bernasconi S et al (1992) Expression of adhe-
Batungwanayo J et al (1993) Pleural effusion, tuberculosis and
sion molecules and chemotactic cytokines in cultured
HIV-l infection in Kigali, Rwanda. AIDS 7:73-79
human mesothelial cells. J Exp Med 176:1165-1174
Berger H, Mejia E (1973) Tuberculous pleurisy. Chest 63:88-92
Kitinya IN, Richter C, Perenboom R et al (1994) Influence pf
Boylan AM, Ruegg C, Kim KJ et al (1992) Evidence of a role for
HIV status on pathological changes in tuberculous pleuri-
mesothelial cell derived interleukin-8 in the pathogenesis
tis. Tuberc Lung Dis 75:195-198
of asbestos-induced pleurisy in rabbits. J Clin Invest 89:
Kunkel SL, Chensue SW, Strieter RM et al (1989) Cellular and
1257-1267
molecular aspects of granulomatous inflammation. Am J
Brown J, Pomerantz M (1995) Extrapleural pneumonectomy
Respir Cell Mol Bioi 1:439-447
for tuberculosis. Chest Surg Clin North Am 5:289-296
Lassence AD et al (1992) Detection of mycobacterial DNA in
Bueno CE et al (1990) Cytologic and bacteriologic analysis
pleural fluid from patients with tuberculous pleurisy by
of fluid and pleural biopsy specimens with Cope's needle.
means of the polymerase chain reaction: comparison of
Arch Intern Med 150:1190-1994
two protocols. Thorax 47:265-269
Caminero JA et al (1993) Diagnosis of tuberculosis by detec-
Lau KY (1989) Numerous mesothelial cells in tuberculous
tion of specific IgG antigen 60 in serum and pleural fluid.
pleural effusion. Chest 96:438-439
Respiration 60:58-62
Lorgat F et al (1992) Evidence for in vitro generation of cyto-
Carrion-Valero F, Perpina-Tordera M (2001) Screening tuber-
toxic T-cells. PPD-stimulated lymphocytes from tuberculous
culous pleural effusion by discriminant analysis. Int J
effusion demonstrate enhanced cytotoxicity with acceler-
Tuberc Lung Dis 5:673-679
ated kinetics of induction. Am Rev Respir Dis 145:418-423
Chang DB et al (1991) Ultrasound-guided pleural biopsy with
Luzze H etal (2001) Evaluation ofsuspected tuberculous pleurisy:
tru-cut needle. Chest 100:132-133
clinical and diagnostic findings in HIV-positive and HIV-neg-
Chaparas SD, Thor DE, Hedrick SR (1971) Comparison oflym-
ative adults in Uganda. Int J Tuberc Lung Dis 5:745-755
phocyte transformation, inhibition of macrophage migra-
Maartens G, Bateman ED (1990) Tuberculous pleural effusion:
tion and skin tests using dialyzable and nondialyzable
increased culture yield with bedside inoculation of pleural
tuberculin fractions from Mycobacterium bovis (BCG). J
Pleural Tuberculosis
fluid and poor diagnostic value of adenosine deaminase.
Thorax 46:96-99
Maeda J, Ueki N, Ohkawa T et al (1993) Local production and
localization of transforming growth factor-beta in tuber-
culous pleurisy. Clin Exp Immunol 92:32-38
Mbizvo MT et al (1996) HIV seroincidence and correlates
of seroconversion in a cohort of male factory-workers in
Harare, Zimbabwe. AIDS 10:895-901
Modilevsky T et al (1989) Mycobacterial disease in patients
with human immunodeficiency virus infection. Arch
Intern Med 147:2201-2205
Mohammed KA, Nasreen N, Ward MJ et al (1998) Mycobacte-
rium-mediated chemokine expression in pleural mesothe-
lial cells: role of C-C chemokines in tuberculous pleurisy. J
Infect Dis 178:1450-1456
Moon WK et al (1999) Complicated pleural tuberculosis in
children: CT evaluation. Pediatr RadioI29:153-157
Morehead RS (1998) Tuberculosis of the pleura. South Med J
91:630-636
Moudgil H et al (1994) Reactivation disease: the commonest
form of tuberculous pleural effusion in Edinburgh. Respir
Med 88:301-304
Murante T et al (1990) Antipurified protein-derivative anti-
body in tuberculous pleural effusion. Chest 97:670-673
Pablo AD et al (1997) Are pleural fluid parameters related to
the development of residual pleural thickening in tubercu-
losis? Chest 112:1293-1297
Palmer PES (1979) Thoracic tuberculosis. Semin Roentegenol
14:204-243
Querol JM et al (1995) Rapid diagnosis of pleural tuberculosis
by polymerase chain reaction. Am J Respir Crit Care Med
152:1977-1981
Relkin F,Aranda CP,Garay SM et al (1994) Pleural tuberculosis
and HIV infection. Chest 105:1338-1341
Richter C et al (1991a) Extrapulmonary tuberculosis - a
simple diagnosis? A retrospective study in Dar es Salam.
Trop Geog Med 43:375-378
Richter C et al (1991b) Pleural effusion aetiology, diagnosis
and clinical picture in HIV-positive and negative patients.
A prospective study in Dar es Salam. Proceedings of the
VIth international congress on AIDS in Africa, Dakar, Sen-
egal, Dec 1991; WA.220
Richter C et al (1994) Clinical features ofHIV-seropositive and
HIV-seronegative patients with tuberculous pleural effu-
sion in Dar es Salam, Tanzania. Chest 106:1471-1476
Roper WH, Waring JJ (1955) Primary serofibrinous pleural
effusion in military personnel. Am Rev Tuberc 71:616-635
Sahn SA (1988) State of the art - the pleura. Am Rev Respir
Dis 138:184-234
Santos-Sastre et al (1990) Mesothelial cells and tuberculous
pleuritis. Chest 98:518
Seibert AF et al (1991) Tuberculous pleural effusion. Twenty-
five year experience. Chest 99:883-887
357
Sharma SK et al (2001) A prospective study of sensitivity and
specificity of adenosine deaminase estimation in the diag-
nosis of tuberculous pleural effusion. Indian J Chest Dis
Allied Sci 43:149-155
Shimokata K, Kishimoto H, Takagi E et al (1986) Determina-
tion of the T-cell subunit producing g-interferon in tuber-
culous pleural effusion. Microbiol Immunol 30:353-361
Shivaram I et al (1989) Pleural effusion in AIDS/ARC. Chest
96:2215
Spriggs AL, Boddington MM (1960) Absence of Mesothelial cells
from tuberculous pleural effusions. Thorax 15:169-171
Stevenson F (1955) The natural history of pleural effusion and
orthopaedic tuberculosis. J Bone Joint Surg Br 37:80-91
Sudre P et al (1992) Tuberculosis: a global overview of the
situation today. Bull WHO 70:149-159
Takagi N et al (1998) Polymerase chain reaction of pleural
biopsy specimens for rapid diagnosis of tuberculous pleu-
ritis. Int J Tuberc Lung Dis 2:338-341
Takagi N, Hasegawa Y,Ichiyama S et al (1998) Polymerase chain
reaction of pleural biopsy specimens for rapid diagnosis of
tuberculous pleuritis. Int J Tuberc Lung Dis 2:338-341
Thomas PA (1994) A thoracoscopic peak: what did Jacobaeus
see? Ann Thorac Surg 57:770-771
Valdes L et al (1996a) Adinosine deaminase (ADA) isoenzyme
analysis in pleural effusion: diagnostic role and relevance
to the origin of increased ADA in tuberculous pleurisy. Eur
Respir J 9:747-751
Valdes L et al (1996b) The etiology of pleural effusion in an area
with high incidence of tuberculosis. Chest 109:158-162
Wallis RS (1996) New approaches to identification of anti-
gens of Mycobacterium tuberculosis. Infect Agents Dis 5:
119-125
Walshe AD, Douglas JG, Kerr KM et al (1992) An audit of the
clinical investigation of pleural effusion. Thorax 47:734-737
Willis RS, Amiy-Tahmasseb M, Ellner JJ (1990) Induction of
interleukin-l and tumor necrosis factor by mycobacterial
proteins: the monocyte Western blot. Proc Nat! Acad Sci
USA 87:3348-3352
Winer-Muram HT, Rubin SA (1990) Thoracic complications of
tuberculosis. J Thorac Imag 5:46-63
Xirouchaki N et al (2002) Diagnostic value of interleukin-l-
alpha, interleukin-6 and tumor necrosis factor in pleural
effusion. Chest 121:815-820
Yamada Yet al (2001) Cytokines in pleural liquid for diagnosis
of tuberculous pleurisy. Respir Med 95:577-581
Yamamura Y, Onoue K, Azuma I (1968) Biology of the myco-
bacterioses. Chemical and immunological studies on pep-
tides and polysaccharides from tubercle bacilli. Ann NY
Acad Sci 154:88-97
Yilmaz MU et al (1998) Computed tomography findings of
tuberculous pleurisy. Int J Tuberc Lung Dis 2:164-167
Zhang M, Gately MK, Wang E et al (1994) Interleukin 12 at the
site of disease in tuberculosis. J Clin Invest 93: 1733-1739
23 Radiology of Pulmonary Tuberculosis
MONA AL SHAHED, MOHAMMED ABD EL BAGI, M. MONIR MADKOUR

CONTENTS 23.1
Introduction
23.1 Introduction 359
23.2 Imaging 359
23.2.1 Conventional Radiography 359 Tuberculosis (TB) is a disease caused by infection
23.2.2 Computed Tomography (CT) 360 with Mycobacterium tuberculosis and accounts
23.2.3 High Resolution Computed Tomography 360 for more than 95% of pulmonary mycobacterial
23.2.4 Bronchography 360 infection. Other non-tuberculous mycobacteria, e.g.
23.2.5 Arteriography 360
M. Kansasii, M. avium-intercellular complex and
23.3 Classification and Radiological Presentation 360
23.3.1 Primary Tuberculosis 360 others, account for the remainder. Tuberculosis (TB)
23.3.1.1 Parenchymal Disease 361 can affect virtually any organ system in the body. It is
23.3.1.2 Lymphadenopathy 362 a topic of universal concern due to the recent resur-
23.3.1.3 Pleural Disease in Primary Tuberculosis 364 gence of (TB) in both immunocompetent and immu-
23.3.1.4 Miliary TB 365
nocompromised individuals (Davis et al. 1993).
23.3.2 Paradoxical Transient Worsening
Phenomenon 366 Much of this increase has been ascribed to human
23.3.3 Post Primary Tuberculosis 368 immunocompromised virus infection (HIV), poverty
23.3.3.1 Parenchymal Disease 369 and homelessness. The radiology of pulmonary (TB)
23.3.3.2 Cavitation 370 reflects, and is influenced by, the immune status of
23.3.3.3 Endobronchial Disease 371
the host, the virulence of the organisms, the presence
23.3.3.4 Tuberculoma 374
23.3.3.5 Pleural Disease in Post-primary Tuberculosis 375 of delayed hypersensitivity due to mycobacterium
23.3.3.6 Mycetoma Formation 378 tuberculosis and the method of spread.
23.4 Assessment of Activity of Pulmonary TB 379 Pulmonary (TB) is classically divided into pri-
23.5 Non-tuberculous Mycobacteria 379 mary and post primary (reactivation) disease. There
23.6 Mycobaqerial Infection and Human
is considerable overlap in the radiological manifesta-
Immunocompromised Virus (HIV) 379
23.6.1 Background 379 tion of these two entities. Should primary (TB) pass
23.6.2 Mycobacterial Tuberculosis 380 into the post primary form without a break, the term
23.6.3 Non-Tuberculous Infections 380 progressive primary tuberculosis is used.
23.6.3.1 Mycobacterial Avium-intercellular Complex
(MAl) 380
23.6.3.2 Mycobacterial Kansasii 381
23.6.3.3 Other Non-Tuberculous Infections 381
23.6.3.4 Non-Tuberculous Bacteria in Immunodeficiency 23.2
States Other than HIV 382 Imaging
References 382
23.2.1
Conventional Radiography
M. AL SHAHED, MBBS, FRCR
Senior Consultant Radiologist, Department of Radiology,
Riyadh Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, Conventional chest radiography is the mainstay in
Saudi Arabia the detection and follow up examination of patients
M. ABD EL BAGI, MB BCh, DMRD, FSRRCSI with pulmonary TB. Good quality chest radiographs
Senior Consultant Radiologist, Department of Radiology, are essential and remain the first line of investigation.
Riyadh Armed Forces Hospital, P.O. Box 7897, Riyadh 11159,
Normal radiographs do not exclude tuberculosis and
Saudi Arabia
M. M. MADKOUR, MD, DM, FRCP have been reported in up to 10% of immunocompetent
Consultant, Department of Medicine, Riyadh Armed Forces patients and in up to 20% of immunocompromised
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia patients (Fitzgerald et al. 1991; Greenberg et al. 1994).

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
 360
23.2.2
Computed Tomography (CT)
Recently, conventional Computerized Tomography
(CT) has been used in selected cases. It is required
in many circumstances such as in the detection of
cavitation and in the evaluation of the route of spread
(Kuhlman et al. 1990). CT is most useful in assessing
pleural disease and in patients with extensive opaci-
fication on the conventional radiographs because of
the marked destruction of the lung by the disease
process. CT has proven to be more sensitive than
conventional radiography in detecting adenopathy,
parenchymal shadowing, particularly in miliary dis-
ease, and in detecting bronchogenic dissemination.
CT is used in the evaluation of complications that
might occur in the process of the disease, including
bronchopleural fistula formation and extension into
chest wall. CT is also used to guide percutaneous
intervention procedures and pre-surgical planning.
23.2.3
High Resolution Computed Tomography
High resolution CT (HRCT) is extremely useful
in understanding the pathological process of the
disease and route of spread; it is also useful in the
evaluation of disease activity (1m et al. 1993).
a b
Fig. 23.1. a Primary TB. Chest x-ray of 7 year old showing ill-defined air space consolidation (arrowheads). b CT the same patient
showing dense air space consolidation with air bronchogram (arrowhead)
M. Al Shahed et al.
23.2.4
Bronchography
Bronchography was once the tool of choice for the
detection of bronchiectasis, but it has now been
replaced by HRCT.
23.2.5
Arteriography
Bronchial artery angiography is currently performed
for therapeutic reasons in the treatment of life-
threatening haemoptysis, which is not a common
presentation of bronchiectasis due to tuberculosis.
23.3
Classification and Radiological Presentation
23.3.1
Primary Tuberculosis
Primary (TB) is a term used when the infected
individual has not been previously exposed to M.
tuberculous and lacks hypersensitivity to tuberculo-
protein. Initially, primary TB was described in infants
and children, in whom it remains the most common
Radiology of Pulmonary Tuberculosis
form (Miller and Miller 1993; Lamont et al. 1986).
Increasingly, however, it has been encountered in
adult population, and it is now estimated to account
for 23-34% of all adult cases (Buckner et al. 1990;
McAdams et al. 1995; Miller and Miller 1993; Miller
1994; Woodring et al. 1986).
Radiologically, primary pulmonary TB typically
manifests in four major ways, singly or in aggregate:
(l) parenchymal disease; (2) lymphadenopathy;
(3) pleural effusion; and (4) miliary disease (Agrons
et al. 1993; 1m et al. 1995; Leung et al. 1992; Palmer
1979; Stansberry 1990). Ten percent (lO%) of all cases
of primary TB progress directly into the chronic
form, which is indistinguishable from reactivation
(Gepport and Left 1979). Moreover, the result of
a chest radiograph may be normal in 15-50% of
cases (McAdams et al. 1995; Miller and Miller 1993;
Woodring et al.1986).
23.3.7.7
ParenchYn1alDisease
By radiological examination, parenchymal shad-
owing is commonly manifested as homogenously
dense air space consolidation with ill-defined mar-
gins, frequently segmental or lobar in distribution
(Fig. 23.1a). Expansion of a consolidated lobe may
occur (Fig. 23.2). The focal parenchymal lesion may
be mass-like, and in adults it may be confused with
neoplasia (Fig. 23.3). It is usually unilateral and
single, but multilobar consolidation can be seen in
up to 25% of cases (Leung et al. 1992; Woodring
et al. 1986). A tendency to favor the right lung has
Fig. 23.2. Chest x-ray demonstrating dense air space consoli-
dation with expansion of right upper lobe in 10 years old child
with primary tuberculosis.
361
emerged in several series (Lamont et al. 1986; Leung
et al. 1992; Weber et al. 1968). This was also noted
in our patients. No consensus has been reached as
to regional preference within the lung. Upper lobe
predominance (Weber et al. 1968; Nagakura 1960;
Joffe 1960; Choyke et al. 1983), mid and lower lobes
predominance (Woodring et al. 1986; Choyke et al.
1983) and no anatomic bias (Lamont et al.1986) have
each been reported. Cavitation may occur in pri-
mary TB (Fig. 23.4). The incidence varies from 10%
(Choyke et al. 1983) to 30% (Woodring et al. 1986).
It is more common in infants and children raised
in communities in which TB has been introduced
comparatively recently (Weizman et al. 1980). At
Computed Tomography (CT), primary TB typically
manifests as dense hemogenous air space consolida-
tion with well-defined margins (Harishigani et al.
2000) (Fig. 23.1b).
Parenchymal shadowing usually resolves with no
radiological sequelae over 6-24 months. Some may
be left with residual scarring or a calcified focus
(Ghon focus). This is seen in up to 20% of patient
(Fig. 23.5). Occasionally during the first three months
of treatment, worsening of parenchymal shadowing
occurs despite appropriate therapy, a phenomenon
referred to as paradoxical transient worsening (see
below). Single or multiple tuberculomas may develop
in primary TB, but they are seen much less frequently
than in post primary TB.
Fig.23.3. Chest x-ray of a 35-year old female with primary
tuberculosis presenting with a focal lung lesion (arrow), right
paratracheal and left hilar lymph adenopathy (arrowhead)
 362
a b
Fig. 23.4. a Chest x-ray showing cavitation within tuberculous consolidation in the right upper lobe. b The lesion has resolved
with minimal residual scaring after the completion of antituberculous treatment
Fig.23.5. Chest x-ray showing foci of the calcification in the
right perihilar region (arrowhead) a stigma or previous tuber-
culous infection
23.3.1.2
Lymphadenopathy
Lymphadenopathy is the hallmark of primary TB,
with or without concomitant parenchymal abnor-
mality. Age stratification qas been identified in many
series (Lamont et al. 1986; Leung et al. 1992; Weber
et al. 1968), with a higher prevalence of lymphade-
nopathy in children under 3 years of age than in
M. AI Shahed et al.
older children and adults with primary tuberculosis
(McAdams et al.1995; Woodring et al.1986). It can be
the only manifestation in young children (Campbell
and Dyson 1977). Any lymph node group may be
involved, but the patterns commonly seen are uni-
lateral hilar, unilateral hilar plus right paratracheal
or isolated right paratracheal adenopathy (Fig. 23.6).
Right paratracheal nodal involvement predominates
in our experience, similar to the experience of others
(Amorosa et al. 1978; Rottenberg and Shaw 1996;
Weber et al.1968). Bilateral adenopathy has also been
described in up to 32% of cases (Leung et al. 1992);
when present, it is almost always asymmetrical and
can be strikingly extensive, resembling lymphoma,
metastatic disease and sarcoidosis (Fig. 23.7). CT
demonstrates lymphadenopathymore accuratelythan
chest radiographs. Characteristically, on post-con-
trast studies the nodes larger than 2 cm in diameter
consistently display a low density center, 40-50 HU,
with enhancing peripheral rims (Harishigani et al.
2000) (Fig. 23.8), and hence citing dissimilarity to
lymphoma, sarcoidosis or histoplasmosis. Lymph-
adenopathy usually resolves at a slower rate than
parenchymal lesions. Nodal calcification occurs 6
months or more after infection and is more fre-
quently seen than parenchymal calcification; it is
encountered in up to 35% of all cases (Weber et al.
1968; Leung et al. 1992). There are significant differ-
ences in the distribution and pattern of calcification
of lymph nodes in TB and sarcoidosis, which can be
explained by the route of lymphatic drainage and the
Radiology of Pulmonary Tuberculosis
Fig.23.6. Primary tuberculous presenting with isolated right
paratracheallymphadenopathy in 12-year old male patient
Fig. 23.8. Post contrast CT of the chest demonstrating charac-
teristic appearance of tuberculous lymphadenopathy with low
density centre and enhancing peripheral rim (arrowheads)
caseating nature of tuberculous granulomas. When
hilar node calcification is present, it is more likely to
be unilateral in TB and bilateral in sarcoidosis. A focal
pattern of calcification is more common in sarcoid-
osis, while complete nodal calcification is common
in TB (Fig. 23.9). Nodal pressure and/or erosion
into adjacent structures may lead to the following
complications: (1) Obstruction of airways leading to
over-inflation and/or segmental/lobar collapse, usu-
ally in the anterior segment of the right middle lobe;
(2) Perforation into an airway, leading to Widespread
bronchogenic dissemination. The healing of these
363
Fig. 23.7. a Frontal and lateral. Chest x-ray of primary tubercu-
losis presenting with extensive lymphadenopathy resembling
lymphoma. Note large hilar (arrow) and bilateral paratracheal
(arrowhead) lymph node involvement. b Lateral x-ray of the
same patient demonstrating hilar adenopathy (arrows)
lesions, though occasionally without sequelae, often
results in various combinations of bronchostenosis,
bronchiectasis, parenchymal fibrosis and loss of lung
volume; (3) Haematogenous spread from infected
nodes, which can lead to "metastatic" lesions such as
soft tissue abscesses. These "metastatic" foci may lie
dormant for years before they become active, mani-
festing as bone, joint or renal TB; (4) Erosion of the
node into the pericardium, leading to tuberculous
pericarditis. Pericarditis can also be a complication of
miliary TB. Involvement of the pericardium can pres-
ent in the form of pericardial effusion (Fig. 23.10) or
364 M. Al Shahed et al.

Fig. 23.10. Chest x-ray demonstrating cardiac enlargement due

to pericardial effusion caused by tuberculous pericarditis

Fig.23.9. Chest x-ray demonstrating complete lymph node

calcification typical of tuberculosis (arrowhead). Note exten-
sive cystic bronchiectasis changes in the right middle and
lower lobes (arrows), some showing air fluid level (small
arrows). Also node calcined granuloma in the left mid zone
(short arrow)

in the form of constrictive pericarditis, with or with-

out calcification (Fig. 23.11). Other complications of
nodal involvement include erosion into oesophagus,
phrenic and recurrent laryngeal nerve paresis, supe-
rior vena cava obstruction and fistula formation.

23.3.1.3
Pleural Disease in Primary Tuberculosis

Pleural effusion is fairly common, occurring in up to

25% of cases. Unlike lymphadenopathy, it is seen more
in adults than in children (Weber et al. 1968; Stead et
al. 1968; Derham 1956). The effusions are generally
unilateral, the exception being when the disease is
a complication of miliary TB. Associated pulmonary
parenchymal lesions or adenopathy may be radio-
logically occult. Tuberculosis should be considered in
any young patient with moderate size or large unilat-
eral pleural effusion in the absence of demonstrable
pulmonary disease (Fig. 23.12a, b). Pleural effusion
in primary tuberculosis can be self-limiting, or it can
lead to serious disease years later. The effusions are
usually painless and can be very large at presenta-
tion. The majority of these effusions are reactive in
nature, rather than due to direct pleural involvement
Fig. 23.11. Lateral chest x-ray showing pericardial calcification
as sequelae to tuberculous pericarditis (arrowheads)
by tuberculosis. Therefore, mycobacterium bacilli are
rarely isolated from culture of the pleural fluid. The
diagnosis is best established by pleural biopsy. CT is
more sensitive than plain radiographs in depicting
associated pulmonary infiltrate and adenopathy in
addition to pleural fluid (Fig. 23.12c, d). CT will also
demonstrate evenly thickened parietal pleura, which
will enhance when contrast material is used. Com-
 Radiology of Pulmonary Tuberculosis 365

a b

c d

Fig. 23.12. a Chest x-ray demonstrating left sided pleural effusion with no visible parenchymal lesion or adenopathy in a young
patient presenting with fever, night sweat, loss of weight and high ESR. Pleural biopsy confirms the diagnosis. b Post enhanced
CT of the chest of the same patient demonstrating right sided simple effusion (arrow). c Thoracic post-enhanced CT of another
patient presented with fever and weight loss showing thin bilateral pleural effusion (arrowheads) and lymph nodes (arrows).
d CT scan of the same patient at lower level showing posterior mediastinal adenopathy. No parenchymal lesions were seen
Bronchial aspirate was positive for tuberculosis

plete resolution is the usual sequelae to antitubercu-
lous treatment. Incompletely resolved effusions will
commonly develop into secondary disease (Palmer
1979). Tuberculous empyema is a less common form
of pleural disease in primary tuberculosis, resulting
from discharge of mycobacterium bacilli into pleural
space (Fig. 23.13). It is usually seen in post primary
cases (see below).
23.3.7.4
Miliary 18
Miliary TB results from haematogenous dissemina-
tion of the organisms and is seen in both primary and
reactivation tuberculosis. It is increasingly encoun-
tered in adults, particularly the immunosuppressed,
in whom it has been reported in up to 13% of cases
(Lee and 1m 1995). It is associated with a very high
mortality if untreated, respiratory failure being the
major cause of death. The characteristic radiologiCal
appearance consists of Widespread nodular shadows
measuring 2-3 mm in diameter and randomly dis-
tributed throughout the parenchyma (Fig. 23.14). In
approximately 50% of cases the initial radiographs
appear normal. Ill-defined haze can be an early
manifestation before the characteristic miliary pat-
tern became discernable. High resolution CT (HRCT)
can detect diffuse lung changes where the initial
radiographic appearance is normal or undetermined
(Fraser et al. 1989). Miliary TB at HRCT manifests as
numerous fine, discrete nodular or reticulonodular
shadows. An associated nodular appearance of the
interlobular septa and vessels may also be seen
(McGuinness et al. 1992) (Fig. 23.15). Miliary tuber-
 366 M. Al Shahed et al.

culosis differs from bronchogenic spread in its even
distribution throughout the lung and its uniform
size. Diseases which have a similar appearances to
that of miliary TB are varicella pneumonia, sarcoid-
osis, histoplasmosis, metastasis, pneumoconiosis and
haemosiderosis. The resolution of these nodules is
usually slow and complete, with no residual calcifica-
tion after treatment. Radiological improvement can
be seen within 3 weeks of starting therapy.
23.3.2
Paradoxical Transient Worsening Phenomenon

Paradoxical transient worsening of the radiographic

Fig. 23.13. Tuberculous empyema as a complication of pleural
disease. Note the presence of irregular pleural calcification
(arrowheads)
findings, both parenchymal and extraparenchymal,
is common in the first three months following the
initiation of therapy (Akira et al. 2000; Amodio et
al. 1986; Bobrowitz 1980; Campbell and Dyson 1977;
Lamont et al. 1986; Matthay et al. 1974; Weber et al.
1968). Paradoxical response refers to progression
of the original lesions or development of new ones
during apparently adequate antituberculous treat-
ment (Bobrowitz 1980), usually regresses without a
change in the initial drug regimen. As we and others
have observed (Akira et al. 2000), progression often
manifests as enlargement of the original parenchy-
mal shadowing. Areas of ground glass opacities, new
areas of consolidations ipsilateral or contralateral
to the initial lesion and new macronodules in the
ipsilateral site were also encountered; none were
with cavitations (Figs. 23.16, 23.17). It is important
to recognize this transient benign phenomenon to
avoid unnecessary invasive procedures or changes of
appropriate therapy. The mechanisms of these clini-
cally and radiologically impressive changes remain
unclear. Several hypotheses have been considered,

Fig. 23.14. a Chest x-ray of a child presenting with miliary tubeculo-

sis. Note widespread nodular shadowing involving both lung fields.
b Miliary tuberculosis in another child presenting with massive
mediastinal adenopathy (arrowheads) and superadded consolida-
a tion in the right middle lobe (arrow)
 Radiology of Pulmonary Tuberculosis 367

Fig. 23.15. a CT scan of the chest demonstrating numerous fine discrete nodular shadow. Note the nodular appearance of the
interlobular septa (arrows). b CT scan of same patient more caudally showing the miliary pattern of tuberculosis. Note the
nodular irregular outline of the blood vessels (arrows)

Fig. 23.16. a Paradoxical transient worsening: A young female

patient presented with malaise, fever, weight loss and night
sweats. Chest x-ray showed right paratracheal adenopathy
(arrow) and ill-defined consolidation in the right perihilar
region (arrowhead). Tuberculosis was diagnosed based on
positive bronchial aspirates and was started on antitubercu-
lous treatment. b Chest x-ray 8 weeks after initiation of antitu-
berculous treatment showed progression of the parenchymal
lesion with development of new macronodular lesions (arrow-
heads). Note significant regression of the right paratracheal
adenopathy. c Chest x-ray demonstrates complete resolution
after completion of antituberculous treatment c
 368
Fig.23.17. a Paradoxical transient worsening: A middle aged
male patient with positive bronchial aspirate. Initial chest x-ray
showed patchy infiltrates in the right lower lobe. b Six weeks
after initiation of antituberculous treatment a new lesion with
thick wall developed in the right upper lobe (arrowhead). Note
resolution of the initial changes. c Chest x-ray 6 months after
initiation of treatment showing almost complete resolution
including enhanced focal immune responses (Iwai
et al. 1979; Marshall and Chambers 1988), local
hypersensitivity to sudden destruction of the bacilli
and tuberculoprotein (Onwubalili et al. 1986) and
local hypersensitivity to drug reaction (Akira et al.
2000). Interestingly, new pulmonary lesions during
therapy for extrapulmonary TB have been reported
by Sze-chunhung et al. (Hung and Chang 1999).
Transient paradoxical worsening should be differ-
entiated from true worsening of tuberculous lesions,
complication by bacterial and fungal infection and
drug reaction. Thin-cut CT in this aspect may playa
role. In a study by (Akira et al. 2000), the dominant
CT findings of transient progressions were ground
glass opacities and/or consolidation, whereas the
dominant CT findings of true worsening of TB were
macronodules and centrilobular nodules, often with
cavitations. Currently, the American Thoracic Society
M. AI Shahed et al.
and Centers for Disease Control recommend a radio-
graphic evaluation at 2-3 months after initiation of
therapy (Bass et al. 1986). Parenchymal abnormalities
can then be evaluated every 2-3 months until they
clear, and lymphadenopathy at yearly intervals until
radiographically stable (Abernathy 1989).
23.3.3
Post Primary Tuberculosis
The term post-primary TB is used to describe tuber-
culosis in patients who have acquired tubercula pro-
tein hypersensitivity from a previous infection or
BCG vaccination. Other terminology used include:
reactivation, secondary, reinfection or adult tubercu-
losis. Postprimary disease results from reactivation
of a previously dormant primary infection in 90% of
 Radiology of Pulmonary Tuberculosis 369

cases, and as extension of primary TB in 10% of cases that CT findings of fluid-bronchogram, in an area
(Woodring et al. 1986). It is a disease of adolescence of consolidation, bronchial dilatation and proxi-
and adulthood characterized by its chronicity, strong mal bronchial wall thickening is significantly more
site preference, cavitation and fibrosis. Radiologic prominent in tuberculous pneumonia than in non-
features of postprimary TB can be broadly classi- tuberculous, non obstructive pneumonia (Fig. 23.19,
fied as lung parenchymal disease with cavitation, 23.24b).
endobronchial airway disease, tuberculoma, pleural Hilar and mediastinal lymph node involvement is
extension and other complications. veryrare,being reported in only5% ofcases (Woodring
et al. 1986) (Fig. 23.20). The initial exudative infiltrates
23.3.3.1 will be replaced by a more sharply defined nodular
ParenchYlnalDisease pattern (acinar nodules), interspersed with reticular
opacities of fibroproductive lesions (Fig. 23.21). The
The earliest radiological manifestation is the devel-
opment of exudative lesions of patchy ill-defined
segmental or subsegmental infiltrations in typical
anatomic regions: apical and posterior segments of
an upper lobe or a superior segment of a lower lobe
in 95% of patients. This site of preference has been
attributed to the oxygen tension being relatively high,
and to lymphatic drainage being relatively ineffec-
tive in these areas. The lesions consist of peripheral
consolidation, often patchy and frequently associ-
ated with accentuated bronchovascular markings,
extending to the ipsilateral hilar region (Fig. 23.18).
Widespread bronchopneumonia is usually a result of
endobronchial seeding (Spencer et al. 1990). Despite
this strong site preference, no portion of the lung is
immune; isolated anterior and basal involvement
have been reported (Woodring et al. 1986). Pulmo-
Fig. 23.19. CT chest demonstrating dense consolidation with
nary TB presenting as segmental consolidation has fluid-bronchogram (arrowheads), and evidence of endobron-
similar radiological appearances as non-tuberculous chial spread (small arrows). The posterior mass (large arrows)
segmental pneumonia. Park et al. (1999) have shown is due to enlarged lymph nodes

a b

Fig. 23.18. a Chest x-ray of an adult female showing patchy consolidation with element of collapse in the right upper lobe. Note
bilateral hilar calcification as evidence of previous tuberculous infection (arrowheads). b Chest x-ray of an adult male showing
consolidation in the apical segment of both upper lobes. Note calcification of the hilar lymph nodes (arrows)
370
Fig.23.20. Chest x-ray of 40 year old male presented with
fever and malaise, chest x-ray showed upper lobe and lower
lobe consolidation with associated paratracheal adenopathy
(arrowheads)
Fig.23.21. Post primary TB. Chest x-ray demonstrating nodulo-
reticular opacities in the apical segment of the right upper lobe
with element of fibrosis. Note shifting of the trachea to the
ipsilateral site
majority of patients show both patterns. Marked tissue
destruction may occur at this stage, with cavitation
and necrosis. Healing may occur by fibrosis, result-
ing in cicatrization, considerable loss of lung volume,
traction bronchiectasis and atelectasis. Scattered
calcification may also be seen (Fig.23.22) (Lee et al.
1991). Secondarily, compensatory signs include eleva-
tion of ipsilateral hilum, over-infiltration of adjacent
lung tissue and bullate formation (Figs. 23.23, 23.24).
Extensive fibrosis and bullae formation with scattered
M. Al Shahed et al.
Fig. 23.22. a Post primary TB. Chest x-ray demonstrating fibrotic
changes with cicatrization, deviation of trachea to the right
indicating loss of lung volume in the right upper lobe (black
arrowheads). Bronchial distortion is also present (white arrow-
head). b CT of different patient with post primary TB showing
evidence of fibrosis and thickened septi in the left upper lobe
and scattered calcific densities (arrows). Patchy emphysematous
changes (white arrowhead) and thickened pleura (black arrow-
heads). Note complete calcification of lymph nodes (large black
arrowhead)
macronodular shadows are seen in end-stage pulmo-
nary tuberculosis (Fig. 23.25).
23.3.3.2
Cavitation
Cavitation is a distinct feature of postprimary TB
and is of considerable diagnostic significance, since
it indicates the likelihood of activity (1m et al. 1993).
It is seen in up to 40-80% of patients with postpri-
Radiology of Pulmonary Tuberculosis 371

Fig. 23.23. Post primary TB. Chest x-ray demonstrating
fibrotic changes with considerable loss of lung volume of the
right upper lobe. Note deviation of trachea to the ipsilateral
site, elevation of right hilum, (small arrow) and compensa-
tory hyperinflation of right lower lobe (large arrow). Pleural
thickening and calcification (arrowheads)
Fig. 23.24. Post primary TB chest x-ray demonstrate end stage
tuberculous changes involving the right lung with extensive
fibrosis, loss of lung volume, deviation of the mediastinum,
bullae formation (arrowheads) and hyperinflation of the left
lung

mary TB (Rottenberg and Shaw 1996). Cavities tend

to occur in the areas of consolidation. They have no
diagnostic features and may be single or multiple,
small or large, thin or thick-walled (McAdams et
al. 1995). Air fluid level may be present, but it is
uncommon and usually indicates superadded infec-
tion (Fig. 23.26). Spontaneous pneumothorax may be
seen (Woodring et al. 1986). A residual tuberculous
cavity may lead to endobronchial spread or become
a site for the development of mycetoma. Rasmussen
aneurysm is a rare but a life-threatening complica-
tion of cavitary lesion involving the wall of the pul-
monary artery, and may lead to massive haemoptysis
if ruptured.

23.3.3.3
Endobronchial Disease

Endobronchial disease is a common complication

seen in up to 40% of patients with active TB. The
source of bronchogenic spread is usually from an
adjacent tuberculous cavity or tuberculous lymph
node containing liquefied materials rich in tubercu-
lous bacilli (Smith and Schillaci 1987; McLoud and
Naidich 1992). The spread through peribronchial
lymphatic channels or directly via infected sputum
are less common routes (Collins et al. 1998; Rot-
tenberg and Shaw 1996). Rarely, the airways can be
involved through a haematogenous route via bron-
chial arteries (Auerbacho 1949; Buckner and Walker
Fig.23.25. Post primary TB. Chest x-ray showing end stage
tuberculous changes with extensive fibrosis, macro nodulo-
reticular shadows, hyperinflation and bullae formation
involving both lung fields. Note scattered calcifications
(arrowheads)
1990). Bronchogenic spread can be ipsilateral or con-
tralateral, and bilateral dissemination can also occur.
(see chapter 21, Endobronchial Tuberculosis)
The radiological manifestations of bronchogenic
spread include multiple small nodular opacities that
 372 M. Al Shahed et al.

c d

Fig. 23.26. a Post-primary TE. Chest x-ray demonstrating a large thick wall cavity in a consolidated left upper lobe with an air
fluid levels (arrow). Patchy infiltrate with macronodular shadows in the right lower lobe (small arrowheads) indicating bron-
chogenic spread. Note two nodular opacities in the left midzone representing tuberculous granuloma (open arrows). b CT chest
of a different patient demonstrating thick-walled subpleural tuberculous cavity with air fluid level. Note localized empyema as
a complication (arrowheads). c Tomography of the chest demonstrates a thin-wall tuberculous cavity (arrow). d Tuberculous
abscess. Chest x-ray of a female patient with bilateral thin wall cavities demonstrating air fluid level (arrowheads)

may later become confluent (Fig. 23.27). Bronchial
wall thickening, post-stenotic dilatation, lobar hyper-
inflation, persistent pulmonary collapse, obstructive
pneumonia and mucoid impaction are other radio-
logical features of bronchogenic disease. HRCT is
more sensitive than conventional radiographs in
demonstrating early endobronchial spread. The most
common findings are the formation of 2-4 mm cen-
trilobular nodules and branching linear structures
((tree-on bud" appearance), which represent case-
ation necrosis within and around the bronchioles.
The centrilobular nodules are well defined and of
high attenuation, a finding which is unusual in other
causes of bronchopneumonias (1m et al. 1993, 1995)
(Fig. 23.28). Other findings include ill defined 8 mm
centrilobular nodules, lobular consolidation, thick-
ened interlobular septa and long segment narrowing
with acentric wall thickening of the bronchi (1m et al.
1993; Lee et al.I993).
Bronchiectasis is a common complication of
endobronchial TB (Lee et al. 1991). It is defined as
the irreversible dilatation of a bronchial tree. It is
typically secondary to pulmonary destruction and
fibrosis (traction bronchiectasis). It may also result
from central bronchostenosis (Fig. 23.27c). Since the
vast majority of post primary TB affects apical and
Radiology of Pulmonary Tuberculosis 373

a b

Fig. 23.27. a Endobronchial tuberculosis: chest x-ray showing mul-

tiple small nodular opacities (arrows). b Endobronchial TE. Chest
x-ray showing consolidation of the left upper lobe with large
cavity with irregular outline and an intracavitary mass, mycetoma
(arrowheads). Multiple macronodular shadows are seen bilater-
ally (arrows) indicating bronchogenic spread. Note hyperinflation
of the right upper lobe and the paraspinal tuberculous abscess
(curved arrows). c Endobronchial tuberculosis: chest x-ray demon-
strates extensive bronchiectasis and cavitation in the right upper
lobe with multiple bilateral basal nodular and bronchial linear
shadows of indicating bronchogenic spread (arrows) c

posterior segments of the upper lobes, facilitating
bronchial drainage, bronchiectasis is usually asymp-
tomatic. Chest radiographic findings include tram
line opacities (dilated thickened wall bronchi with
loss of normal tapering) and ring shadows, occasion-
ally with air-fluid levels (Fig. 23.9). Bronchography
was until recently the investigation of choice for the
diagnosis of bronchiectasis (Fig. 23.29).
CT, particularly HRCT, has replaced bronchogra-
phy in the diagnosis of bronchiectasis in the recent
years (Park et al.1999), with sensitivity up to 82-97%
(Munroe et al. 1990; Grenier et al. 1986). Features on
HRCT include uniform bronchial dilatation extending
to the lung periphery without tapering (Fig. 23.28a),
bronchial wall thickening and ring shadows repre-
senting markedly dilated bronchi (Fig. 23.30).
HRCT is able to demonstrate the lobular distribu-
tion of airway abnormalities in early bronchiectasis
not seen using conventional x-ray or conventional
CT. Features include subpleural nodules and branch-
ing linear structures ("tree-on-bud" appearance)
corresponding to impacted bronchioles (Fig. 23.28c)
(Davis et al. 1993; Moon et al. 1997). Hypertrophy of
bronchial arteries, secondary to bronchiectasis, is
demonstrated on contrast enhanced CT.
With effective treatment, most of the changes of
endobronchial spread resolve over time; residual
fibrosis, bronchiectasis and emphysematous changes
are invariably seen. The latter is believed to result
from traction by adjacent fibrosis, from paracicatri-
cial emphysema, and from bronchial stricture (1m
et al. 1993). CT after treatment will demonstrate
the residual changes of fibrosis, bronchiectasis and
emphysema. Lobular emphysematous changes will
give rise to a mosaic pattern (Fig. 23.31) (Eber et al.
1993; Martin et al. 1986).
 374
a b
Fig.23.28. a Endobronchial tuberculosis: HRCT scan show-
ing irregular bronchial wall thickening, loss of tapering
(long arrow), multiple scattered linear branching densities,
tree-on-bud (small arrow), and macronodular shadows. b CT
of the same patient at lower level showing widespread mac-
ronodular and linear branching shadows of endobronchial
spread (arrowheads). Note solid consolidation in the left lower
lobe and massive adenopathy (arrow). c CT scan of the same
patient at lung bases demonstrating impaction of the periph-
eral bronchioles with infective material resulting in extensive
linear and nodular shadows (arrowheads)
Fig.23.29. Bronchogram demonstrating both cystic (arrow-
heads) and tubular (arrows) dilatation of the lower lobe
bronchial tree
M. Al Shahed et al.
c
23.3.3.4
Tuberculoma
A tuberculoma is a focus of acid-fast bacilli that is
encapsulated by connective tissue. It can be a mani-
festation of either primary or secondary TB, and it is
believed to represent localized parenchymal disease
that alternately activates and heals. Tuberculoma
always carry a potential risk of cavitation and dis-
semination. They are seen in 3-6 % of cases, can be
multiple and may be associated with satellite nodules
(Leung 1999). Tuberculoma, by radiographic investi-
gation, appears as a well-defined, round or oval mass-
like lesion (Fig. 23.32a). Most tuberculoma are <3 cm
in size (range 0.5-5 cm) (Lee et al. 1991). The major-
ity of these lesions remains stable for a long time;
20-30% may calcify. CT shows them as low attenua-
tion masses that do not enhance with contrast media
(Gawne-Cain and Hansell 1996) (Fig. 23.32b-d).
 Radiology of Pulmonary Tuberculosis 375

Fig. 23.30. a CT chest showing multiple ring shadows of cystic bronchiectasis in the superior segment of the right upper lobe.
b CT scan of patient presented with post-primary TB showing extensive cystic bronchiectatic changes as complication of endo-
bronchial disease. Note fluid levels (arrows). Pleural effusion was due to cardiac failure

Fig. 23.31. a CT scan demonstrating change of fibrosis with residual bronchiectatic changes m the left upper lobe. Note the mosaic
pattern in the lung fields due to lobular emphysematous changes (arrowheads). b CT demonstrating end result of endobronchial
disease. There is evidence of traction bronchiectasis (arrowheads), bullae (open arrow) and tubular bronchial dilatation (arrow).
Note pleural thickening in the right apical zone. Note mosaic appearance in the left upper lobe from lobular emphysema

23.3.3.5 bronchopleural fistula. CT demonstrates the pleural

Pleural Disease in Post-primary Tuberculosis
The pleural effusion accompanying post primary TB
is usually an empyema, and it carries a worse progno-
sis than does effusion in primary TB. When present,
the condition is invariably associated with parenchy-
mal disease. Radiologically, it appears as lobulated
pleural collections associated with pleural calcifica-
tion (Figs. 23.13, 23.33). Presence of air fluid level
(hydropneumothorax) usually indicates presence of
disease to better advantage than ordinary radiogra-
phy (Fig. 23.34) (Hulnick et al.1983). CT is also more
sensitive in demonstrating associated parenchymal
lesions and peripheral subpleural cavitary lesions,
which could be the cause of empyema formation
(Fig. 23.26b); CT is also more sensitive in dem-
onstrating associated complications such as bron-
chopleural fistula, rib destruction and cold abscess
(empyema necessitatis) (Fig. 23.35). Fistula can also
be demonstrated by fistulogram (Fig. 23.36).
 376 M. Al Shahed et al.

Fig.23.32. a Chest x-ray of an adult showing well-defined

round lesion with central lucency representing tuberculoma
(arrow). Two other nodular opacities are seen in the right
lower lobes representing multiple tuberculomas. b CT of the
same patient in prone position showing the tuberculoma of
soft tissue density and central lucency (arrow). c CT at lower
level demonstrating two other tuberculomas presenting as
homogenous round tissue opacities (arrowheads). d CT chest
of another patient showing single tuberculoma in the left
lower lobe with central focal calcification (small arrowhead)
Note the fibrotic band extending to an area of focal pleural
d thickening (large arrowhead)
 Radiology of Pulmonary Tuberculosis 377

Fig. 23.33. Chest x-ray demonstrates extensive pleural disease

caused by previous tuberculous infection. There is consider-
able pleural thickening and calcification causing fibrothorax
(arrowhead). Note compensatory emphysema of the right
upper and middle lobe

Fig. 23.34. a CT of the chest demonstrating tuberculous

empyema (arrow) manifesting as localized collection of high
density fluid associated with pleural thickening (arrowheads).
b CT of the chest showing a large empyema with associated
thickening of parietal pleura (large arrowheads). Note fluid
bronchogram in collapsed consolidated apical segment of the
left upper lobe (small arrowheads)

a b

Fig. 23.35. a CT chest showing pleural thickening (short arrows), localized collection with irregular calcification (large arrow)
associated with rib destruction and localized cold abscess appearing as low density area within the intercostal muscles (arrow-
heads). Note localized pleural thickening with calcification on the left. b Chest x-ray of the same patient later after completion of
antituberculous treatment showing resolution of the pleural pathology demonstrated on a and residual changes in the affected
ribs (arrowheads)
378 M. Al Shahed et al.

ally tuberculous in origin, but can also develop in sar-

Fig.23.36. Fistulogram demonstrate pleuro bronchial fistula
caused by tuberculous changes in the apical segment of the
left upper lobe (arrowheads)
23.3.3.6
Mycetoma Formation
Pulmonary mycetomas are masses of Aspergillus
hyphae matted together to form a fungus ball. The
fungi colonize a pre-existing cavity in the lung, usu-
coidosis, emphysematous bullae, histoplasmosis and
in cavitary bronchial carcinoma. Approximately 15%
of tuberculous cavitary lesions larger than 25 mm in
diameter become colonized by Aspergillus Fumiga-
tus (Research Committee 1970). Fungal ball devel-
ops in both tuberculous cavities as well as in areas
of cystic bronchiectasis. The majority of mycetomas
are found in the upper lobes and superior segments
of the lower lobes. The majority are asymptomatic.
Haemoptysis, although uncommon, being the most
important complication that may warrant surgical
resection or bronchial artery embolization. When
haemoptysis is a presenting symptom, differentia-
tion between mycetoma formation, reactivation of
tuberculosis and development of bronchial carci-
noma becomes difficult. The radiographic findings
include demonstration of a mass within a cavity. Air
crescentic sign, an air rim between the wall of the
cavity and the fungal ball, is highly suggestive but
not specific (Fig. 23.37). The wall of the cyst is usually
thickened, and an associated adjacent pleural reac-
tion is not uncommon; osteomyelitis of the adjacent
rib can also occur. Right and left lateral decubitus
films may show the mobile nature of the fungal
ball. Immature mycetoma is seen as an irregular
spongework of soft tissue containing air spaces and
obliterating the cavity (Rottenberg and Shaw 1996).
HRCT aids in the diagnosis of mycetoma as well as in
pre-operative assessment (Kuhlman et al. 1990).

Fig.23.37. a Mycetoma. Chest x-ray demonstrates a large cavity in

the left upper lobe with a fungal ball (arrow). (Courtesy of Dr. M.
Mutairi) b Localized view of this region shows the characteristic
appearance of crescentic sign (arrow). (Courtesy of Dr. M. Mutiari) b
Radiology of Pulmonary Tuberculosis
23.4
Assessment of Activity of Pulmonary T8
Chest radiography plays an important role in the
diagnosis of pulmonary TB, both assessing the activity
and monitoring the response to treatment. Inactivity
cannot be reliably diagnosed on a single radiograph,
and serial radiographs are required. Tuberculous foci,
which appear inactive over an extended period, such
as extensive calcification and pulmonary fibrosis, can
harbour active disease with the potential to break down
under certain favorable circumstances. Features that
suggest active disease are ill-defined air space shad-
owing, miliary nodulation and cavitation. CT is more
sensitive in the detection of miliary disease, cavitation
and lymphadenopathy. HRCT has been proven to playa
role in the early detection of endobronchial spread and
in the determination of disease activity. The presence
of branching centrilobular lesions without evidence of
fibrotic bronchovascular distortion are highly sugges-
tive of active disease (1m et al. 1995).
Conventional radiographs are used to follow up
patients once treatment has started. Radiographic
resolution of pulmonary changes can take from 6 to
24 months. Complete resolution is seen in two-thirds
of patient (Lee et al. 1991). Radiographic worsening
can be seen despite appropriate therapy "transient
radiographic progression" and should be differenti-
ated from true worsening (see text). In patients with
HIV, assessment of disease activity is less sensitive
as the possibility of secondary infection and other
pulmonary complications are common and should
always be considered (Barnes et al. 1991).
23.5
Non-tuberculous Mycobacteria
One to three percent of mycobacteria other than M.
tuberculosis have been identified as causes of pul-
monary infection (Woodring and Vandiviere 1990).
These include M. kansasii, M. avium-intercellular
complex, M. fortuitum and M. gordonae. These are
free-living sporocyte. Infection is acquired by inhala-
tion or ingestion from the environment. Non-tuber-
culous mycobacterial infections have a predilection
for individuals with pre-existing chronic obstructive
pulmonary disease, underlying debilitating disease, a
previous tuberculous infection and patients who are
immunocompromised. Males are more commonly
affected than females, and tuberculin skin tests are
usually negative or weakly positive. Radiological fea-
379
tures of non-tuberculous mycobacterial infections are
in general indistinguishable from post primary TB
and have similar site preferences: posterior segments
of the upper lobes and the superior segments of the
lower lobes. Cavitation is a distinct feature, occurring
in up to 96% of cases. Cavities tend to be multiple
(Abelda et al. 1985; Erasmus et al. 1999). CT findings
of non-tuberculous mycobacteria ranges from small
to large nodular infiltrates, consolidation, cavitations
and bronchiectasis (Hartman et al. 1993).
23.6
Mycobacterial Infection and Human
Immunocompromised Virus (HIV)
23.6.1
Background
Mycobacterial infection is an important cause of
morbidity and mortality in immunocompromised
patients, particularly those infected with HIY. The
association between HIV and tuberculosis has been
observed since 1984 (Barnes et al. 1991; Greenberg et
al. 1994). Pulmonary TB occurs in 4-21 % of patients
with acquired immunocompromised syndrome
(AIDS) (Davis et al.1993), and it was declared to be an
AIDS-defining disease in 1993. The increased suscepti-
bility of these patients to tuberculosis is best explained
by their altered immunity function. The T-helper lym-
phocyte (CD 4 Tcell) and to lesser extent macrophage
depletion and dysfunction process are central to the
HIV disease rendering these patients particularly sus-
ceptible to tuberculous infection. Both M. tuberculo-
sis and non-tuberculous mycobacteria have become a
serious problem in patients infected with HIY. Non-
invasive diagnostic methods, such as sputum stain and
culture, are less effective in the immunocompromised
host. Moreover, other infections and non-infectious
disease processes are common in these patients and
there is considerable overlap in radiological mani-
festation, making accurate diagnosis more difficult.
Clinicians are often forced to choose between the use
of invasive techniques to determine the exact cause
of the pulmonary disease or an empirically chosen
therapy (McLoud 1989; McLoud and Naidich 1992).
The role of radiology in this setting is important and
involves not only detection of abnormalities on chest
radiographs or CT, but also analysis of these changes
with regard to the possible diagnosis and choice of an
appropriate interventional technique, e.g. percutane-
ous needle biopsy and aspiration. The role of radiology
380
also includes monitoring the responses to therapy and
the development of complications.
23.6.2
Mycobacterial Tuberculosis
Reactivation is the primary mechanism of infection.
Rapid progression of newly acquired infection has
also been described (Small et al. 1994). Clinical and
radiological features depend on the degree of cellular
immune compromise, especially CD 4 T-cell count. In
early stages of HIV disease (CD 4 count>200 cell/mm\
radiological findings are typical of post primary TB,
especially apical cavitary disease and bronchogenic
spread. Skin testing to tuberculin is usually positive
at this stage. With a greater degree of immunocom-
promise (CD 4 count <200 cell/mm\radiographs typi-
cally show a pattern more consistent with primary TB,
including hilar and mediastinal adenopathy, non-cavi-
tary infiltrates equally distributed in both upper and
lower lobes, multiple nodules, miliary disease and uni-
lateral and bilateral pleural effusion (Figs. 23.38,23.39).
However, normal radiographs have been reported in
up to 15% of cases (Greenberg et al. 1994). Skin test-
ing at this stage is positive in only 20-40% of patients,
similarly bronchoalveolar lavage is positive in only
20% of cases. In contradistinction, chest radiographic
abnormalities can be identified in up to 85% of cases,
making accurate interpretation imperative, as therapy
with antituberculous drugs are usually effective
(Barnes et al.1991; Leung 1999; Greenberg et al. 1994).
HRCT is helpful, especially in the evaluation of miliary
Fig.23.38. Atypical mycobacteria. Chest x-ray of immunocom-
promised patient presented with miliary pattern. Note the diffuse
fine nodular opacities in both lung fields from MAl infection
M. Al Shahed et al.
disease and in the identification and characterization
of lymphadenopathy. HRCT is also helpful in demon-
strating parenchymal lesions in patients whose chest
radiographs are normal (Leung etal. 1996).
23.6.3
Non-Tuberculous Infections
23.6.3.1
Mycobacterial Avium-intercellular Complex (MAl)
The incidence of (MAl) in patients with AIDS has risen
since pneumocystis prophylaxis has been widely insti-
tuted (Hoover et al. 1993). Unlike infection with M.
tuberculosis, pulmonary disease is relatively uncom-
mon in patients infected with MAL Significant radio-
graphic abnormalities occur in approximately 5% of
cases and then only late in the course of HIV disease
(Horsburgh 1991). Infection with MAl typically causes
disseminated extrapulmonary disease, presumably
due to the fact that the main route of entry is the
gastrointestinal tract. There are no distinctive pul-
monary radiographic abnormalities in patients with
MAl infection and AIDS; moreover, other associated
opportunistic diseases are common. The radiographic
pulmonary findings described in the literature range
from normal chest x-rays to mediastinal adenopathy,
pulmonary infiltrates, diffuse nodules, miliary spread
and pleural effusion (Fig. 23.40). Localized pulmonary
disease is uncommon, occurring in less than 5% of
patients (Horsburgh 1991), and cavitation is rare. MAl
is very resistant to antituberculous drugs, and no treat-
Fig.23.39. Post renal transplant patient presented with respira-
tory failure. Chest x-ray show extensive consolidation, mac-
ronodular pattern and massive mediastinal adenopathy (arrow-
heads). Lung biopsy confirmed MAl infection
Radiology of Pulmonary Tuberculosis 381

Fig.23.40. a Chest x-ray of an adult patient with atypical

mycobacterium MAl infection demonstrating consolida-
tion, patchy infiltration and bronchiectatic changes in the
right middle lobe (straight arrow). Note cavitation (curved
arrow) and pleural effusion (arrowhead). (Courtesy of Dr. M.
Enani). b CT scan of the same patient showing macro-nodular
opacities in the right lower lobe (arrows), consolidation in the
middle lobe c pleural thickening and effusion (small arrow-
head), and adenopathy (large arrowhead). (Courtesy of Dr. M.
Enani). c CT scan of the same patient at higher cuts showing
bronchiectatic changes in the right middle lobe and medial
basal segment of the right lower lobe (arrows) c

ment regimen has proven entirely satisfactory. Because 23.6.3.3

of the poor outcome of treatment, prophylaxis is rec-
ommended in patients with CD4 count <50 celI/mm 3•
23.6.3.2
Mycobacterial Kansasii
M. Kansasii is second only to M. tuberculosis in the
frequency with which it causes pulmonary disease.
Radiographic manifestation mimics post primary
tuberculosis, including upper lobe cavitary infil-
trates, endobronchial lesions and lymphadenopathy
(Bamberger et al.1994; Christensen et al.1978, 1981;
Connolly et al. 1993). Normal chest radiographs
have also been reported (Fishman et al. 1997).
Other Non-Tuberculous Infections
Virtually all atypical mycobacterial infections
that complicate HIV infection have been caused
by MAl and M. Kansasii. M. gordonae, M. fortui-
tum, M. chelonei, M. haemophilum, M. malmoense
and M. terrae are extremely rare causes and usu-
ally manifest as disseminated disease (Fakih et al.
1996; Saubolle et al. 1996; Carbonara et al. 2000).
Radiological manifestations are nonspecific, rang-
ing from normal radiographs to focal and diffuse
infiltrates. Prognosis is generally poor, as most of
these organisms are resistant to antituberculous
treatment.
 382
Fig. 23.41. Post enhanced CT of the chest for a 67-year-old female patient with hairy cell leukemia. a Mediastinal setting at the
level of aortic arch showing multiple enlarged lymph nodes (arrowheads). b Lung window setting at the same level demonstrates
ill-defined increased parenchymal densities and macronodular shadows (arrows)
23.6.3.4
Non-Tuberculous Bacteria in Immunodeficiency
States Other than HIV
Malignancy and immunosuppressive therapy for
organ transplantation (steroid and chemotherapy)
can predispose patients to infection with non-tuber-
culous mycobacteria. Among malignancies, there is a
particular strong association with hairy cell leukemia
(Fig. 23.41) (Libshitz et al. 1981; Weinstein et al. 1981).
A review of the literature points to the prevalence of
MAl infection in these patients and to the fact that the
disease is more common than was previously thought.
Pulmonary involvement, according to multiple reports,
is most often a manifestation of disseminated infec-
tion rather than a primary lung disorder. Reported
radiographic manifestations include normal chest
radiographs, single or multiple pulmonary nodules
and masses, lymphadenopathy and patchy alveolar
infiltrates with or without cavitation (Albelda et al.
1985; Bamberger et al. 1994; Long et al. 1991).
References
Abernathy RS (1989) Tuberculosis in children and its manage-
ment. Semin Respir Infect 4:232-242
Agrons G A, Markowitz RI, Kraner SS (1993) Pulmonary tuber-
culosis in children. Semin RoentgenoI2:158-172
Akira M. Sakatani M, Ishikawa H (2000) Transient radio-
graphic progression during initial treatment of pulmo-
nary tuberculosis: CT findings. J Comput Assist Tomogr
24:426-431
M. Al Shahed et al.
Albelda SM, Kern JA, Marinelli DL, Miller WT (1985) Expand-
ing spectrum of pulmonary disease caused by nontubercu-
lous mycobacteria. Radiology 157:289-296
Amodio J, Abramson S, Berdon W (1986) Primary pulmo-
nary tuberculosis in infancy: a resurgent disease in urban
United States. Paediatr RadioI16:185-189
Amorosa JK, Smith PR, Cohen JR (1978) Tuberculous medias-
tinallymphadenitis in the adult. Radiology 126:365-368
Auerbacho (1949) Tuberculosis of trachea and major bronchi.
Am Rev Tuberc 60:604-606
Bamberger DM, Driks MR, Gupta MR et al. (1994) Myco-
bacterium kansasii among patients infected with human
immunodeficiency virus in Kansas city. Kansas city AIDS
Research Consortium. Clin Infect Dis 18:395-400
Barnes PF, Bloch AB, Davidson PT, Snider DE (1991) Tuber-
culosis in patients with human immunodeficiency virus
infection. N Engl J Med 324:1644-1650
Bass JB, Farer LS, Hopewell PC et al (1986) Treatment of tuber-
culosis and tuberculosis infection in adults and children.
Am Rev Respir Dis 134:355-363
Bobrowitz D (1980) Reversible roentgenographic progression
in the initial treatment of pulmonary tuberculosis. Am Rev
Respir Dis 121:735-742
Buckner CB, Walker CW (1990) Radiologic manifestation of
adult tuberculosis. J Thorac Image 5:28-37
Campbell lA, Dyson AJ (1977) Lymphnode tuberculosis: a
comparison of various methods of treatment. Tuberculo-
sis 58:171-179
Carbonara S, Tortoili E, Costa D et al (2000) Disseminated
Mycobacterium terrae infection in a patient with advanced
human immunodeficiency virus disease. Clin Infect Dis 30:
831-835
Christensen EE, Dietz GW, Ahn CH et al (1978) Radiographic
manifestations of pulmonary mycobacterium kansasii
infections. AJR 131:985-993
Christensen EE, Dietz GW, Ahn CH et al (1981) Initial roent-
genographic manifestation of pulmonary M. tuberculo-
sis, M. kansassi, M. intercellularis infections. Chest 80:
132-136
Radiology of Pulmonary Tuberculosis
Choyke PL, Soskman HD, Curtis AM et al (1983) Adult onset
tuberculosis. Radiology 148:357-362
Collins J, Blankenbaker D, Stern E (1998) CT patterns ofbron-
chiolar disease: what is tree-in-bud? AJR 171:365-370
Connolly MG Jr, Baughman RP, Dohn MN (1993) Mycobac-
trium kansasii presenting as an endobronchial lesions. Am
Rev Respir Dis 148:1405-1407
Davis SD, Yankelevits DF, Williams T et al (1993) Pulmonary
tuberculosis in immunocompromized hosts: epidemiologi-
cal, clinical and radiological assessment. Semin Roentgenol
28:119-130
Derham RJ (1956) Post primary intrathoracic tuberculosis
in childhood with special reference to its sequelae. Taxes
J Med 52:583-585
Eber CD, Stark P, Bertozzi P (1993) Bronchiolitis obliterans on
high resolution CT: a pattern of mosaic oligemia. J Comput
Assist Tomogr 17:853-856
Erasmus JJ, McAdams HP, Farrell MA et al (1999) Pulmonary
non-tuberculous mycobacterial infection: Radiologic
manifestations. Radiographics 19:1487-1503
Fakih M, Chapalamadugu S, Ricart A, Corriere N Amsterdam
D (1996) Mycobactrium malmoense bacteremia in two
AIDS. J Clin MicrobioI34:731-733
Fishman JE, Schwartz DS, Sais GJ (1997) Mycobacterium kan-
sasii pulmonary infection in patients with AIDS: spectrum
of chest radiographic findings. Radiology 204:171-175
Fitzgerald JM, Gizybouski S, Allen EA (1991) The impact of
human immunodeficiency virus infection on tuberculosis
and its control. Chest 100:191-200
Fraser RG, Pare JAP, Pare PD et al (1989) Infectious diseases
of the lung. In: Diagnosis of diseases of the chest, vol 2, 3rd
edn. Saunders, Philadelphia pp 882-940
Gawne-Cain ML, Hansell DM (1996) The pattern and distribu-
tion of calcified mediastinal lymph nodes in sarcoidosis
and tuberculosis: A CT study. Clinical Radio 51:263-267
Gepport EF, Left A (1979) The pathogenesis of pulmonary and
miliary tuberculosis. Arch Intern Med 139:1381-1383
Greenberg SD, Frager D, Suster B, Walker S, Stavropoulos
C, Rothpeal A (1994) Active pulmonary tuberculosis in
patients with AIDS: spectrum of radiographic findings
(including a normal appearance). Radiology 193:115-119
Grenier P, Maurice F, Musset D et al (1986) Bronhiectasis:
assessment by thin section CT. Radiology 161:95-99
Harishigani M, Mcloud T, Shepard J, Ko J et al (2000) Tubercu-
losis from head to toe. Radiographics 20:469-470
Hartmann TE, Swensen SJ, Williams DE (1993) Mycobactrium
avium-intercellulare complex: evaluation with CT. Radiol-
ogy 187:23-26
Hoover DR, Saah AJ, Bacellar H et al (1993) Clinical manifesta-
tion of AIDS in the era of pneumocytis prophylaxis. N Engl
J Med 329:1922-1926
Horsburgh CR Jr (1991) Mycobacterium avium complex infec-
tion in the acquired immunodeficiency syndrome. N Engl
J Med 324:1332-1338
Hulnick DH, Naidich PD, McCauley DL (1983) Pleural tuber-
culosis evaluated by computed tomography. Radiology
1489:759-765
Hung S, Chang S (1999) New pulmonary lesions during therapy
for extrapulmonary tuberculosis. Chest 116: 1794-1797
1m JG, Hoh H, Shim YS, Lee JH,Ahn J, Han MC, Noma S (1993)
Pulmonary tuberculosis: CT findings early active disease
and sequential change in antituberculous therapy. Radiol-
ogy 186:653-660
383
1m JG, Itoh H, Han MC (1995) CT of pulmonary tuberculosis.
Semin Ultrasound CT/MR 16:420-434
Iwai K, Kiao C, Tsuchiya S, Kita R, Kinoshita T (1979) Histo-
logical observation on the tuberculous lesions aggravated
during rifampicin treatment (in Japanese). Kekkaku 54:
473-478
Joffe N (1960) Cavitating primary pulmonary tuberculosis in
infancy. Br J Radiol 33:430-433
Kuhlman JE, Deutsch JH, Fishman EK et al (1990) CT fea-
tures of thoracic mycobacterial disease. Radiographics 10:
413-431
Lamont AC, Cremin BJ, Pelleret RM (1986) Radiological pat-
terns of pulmonary tuberculosis in paediatric age group.
Pediatr RadioI16:2-7
Lee KS, 1m JG (1995) CT in adults with tuberculosis of the
chest: characteristic findings and role in management. AJR
164:1361-1367
Lee KS, Kim YH, Kim WS, Hwang SH et al (1991) Endobron-
chial tuberculosis: CT features. J Comput Assist Tomogr
15:424-428
Lee KS, Song KS, Lim TH, Kim PN et al (1993) Adult onset
pulmonary tuberculosis; findings on chest radiographs
and CT scans. AJR 160:753-758
Leung AN, Millet NL, Pired PR et al (1992) Primary tubercu-
losis in childhood. Radiographic manifestations. Radiology
182:87-91
Leung NA (1999) Pulmonary tuberculosis: the essentials. Radi-
ology 210:307-322
Leung NA, Brauner MW, Gamus G et al (1996) Pulmonary
tuberculosis: comparison of CT findings in HIV-sero-
positive and HIV-seronegative patients. Radiology 198:
687-691
Libshitz HI, Shuman LS, Gresik MV (1981) Pneumonia in
hairly cell leukemia Radiology 139:19-24
Long R, Maycher B,Scaicini M et al (1991) The chest roentgen-
ogram in pulmonary tuberculosis patients seropositive for
human immunodeficiency virus type 1. Chest 99:123-127
Marshall BG, Chambers MA (1988) Central nervous system
tuberculosis: the paradox of host immune response. J
Infect 36:3-4
Martin KW, Sagel SS, Siegel BA (1986) Mosiac oligemia stimu-
lation pulmonary infiltrates on CT. AJR 147:670-673
Matthay RA, Neff TA, Iseman MD (1974) Tuberculous pleural
effusion developing during chemotherapy for pulmonary
tuberculosis. Am Rev Respir Dis 109:469-472
McAdams HP, Erasmus J, Winter JA (1995) Radiologic mani-
festations of pulmonary tuberculosis. Radiol Clin North
Am 33:655-678
McGuinness F, Naidich DP, Jagirdar J, Leitman B, Mc Cauley DL
(1992) High resolution CT findings in miliary lung disease.
J Comput Assist Tomogr 16:384-390
McLoud TC (1989) Pulmonary infections in the immunocom-
promised host. Radiol Clin North Am 27:1059-1062
McLoud TC, Naidich DP (1992) Thoracic disease in the
immunocompromised patient. Radiol Clin North Am 30:
525-554
Miller WT (1994) Tuberculosis in the 1990s. Radio Clin North
Am 32:649-661
Miller WT, Miller WT (1993) Tuberculosis in the normal host:
radiological findings. Semin Roentgenol 28:109-118
Moon Wk, 1m JG, Yeon KM et al (1997) Tuberculosis of the
central airways: CT findings of active and fibroxic disease.
AJR 169:649-653
384
Munroe NC, Cooke JC, Currie DC et al (1990) Comparison of
thin section computed tomography with bronchography
for identifying bronchiectatic segments in patients with
chronic sputum production. Thorax 45:135-139
Nagakura Y (1960) Pathologic-anatomic studies or resected
lungs of younger patients with pulmonary tuberculosis.
Clin Resp Organs (Tokyo) 15:462
Onwubalili JK, Scott GM, Smith H (1986) Acute respiratory
distress related to chemotherapy of advanced pulmonary
tuberculosis; a study of two cases and review of the litera-
ture. Q J Med 59:599-610
Palmer PES (1979) Pulmonary tuberculosis usual and unusual
radiographic presentations. Semin RoentgenoI14:204-242
Park S, Hong YK, Joo SH, Choe KO, Cho SH (1999) CT find-
ings of pulmonary tuberculosis presenting as segmental
consolidation. JComput Assist Romogr 23:736-742
Research committee of the British Thoracic and Tuberculosis
Association (1970) Aspergilloma and residual tuberculous
cavities: the result of survey. Tubercle 51 :227- 245
Rottenberg GT, Shaw P (1996) Radiology of pulmonary tuber-
culosis. Br J Hosp Med 56: 195-199
Saubolle MA, Kiehn TE, White MH, Rudinsky MF, Armstrong
D (1996) Mycobacterium haemophilum: microbiology and
expanding clinical and geographic spectra of disease in
humans. Clin Microbiol Rev 9:435-447
Small PM, Hopewell PC, Singh SP et al. (1994) The epidemiol-
ogy of tuberculosis in San Francisco. A population based
M. Al Shahed et al.
survey using conventional and molecular methods. New
Engl J Med 330:1703-1709
Smith LS, Schillaci RF (1987) Endobronchial tuberculosis:
serial fibroptic bronchoscopy and natural history. Chest
91:644-647
Spencer D, Yagan R, Blinkhorn R, Spagnuolo PJ (1990) Anterior
segment upper lobe tuberculosis in the adult. Occurrence
in primary and reactivation disease. Chest 97:384-388
Stansberry SD (1990) Tuberculosis in infants and children. J
Thorac Imag 5:17-27
Stead WW, Kerby GR, Schuleter DP et al (1968) The clinical
spectrum of primary tuberculosis in adults: confusion with
reinfection in the pathogenesis of chronic tuberculosis.
Ann Intern Med 68:731-737
Weber AI, Bird KT, Janower ML (1968) Primary tuberculosis
in children with particular emphasis on changes affecting
the tracheobronchial tree. AJR 103:123-132
Weinstein RA, Golomb HM, Grumet G et al (1981) Hairy cell
leukemia: association with disseminated atypical mycobac-
terial infection. Cancer 48:380
Weizman Z, Shvi Y, Ayalon A (1980) Primary cavitary tubercu-
losis in an infant chest 77:578-581
Woodring JH, Vandiviere HM (1990) Pulmonary disease caused
by non tuberculous mycobacteria. J Thorac Imag 5:64-76
Woodring JH, Vandiviere HM, Fried AM et al (1986) Update:
the radiographic feature of pulmonary tuberculosis. AJR
146:497-506
24 Pulmonary Function Test and Tuberculosis
MAJDY M. IDREEs, SIRAJ 0 WALl, ABDULLA AL-AMOUDI

CONTENTS Before the availability of chemotherapy, pulmo-

nary function assessment was frequently performed
24.1 Pulmonary Function Abnormalities in patients with pulmonary tuberculosis. In 1846,
in Tuberculosis 385
24.1.1 Historical Aspect 385 Hutchinson suggested a correlation between the sever-
24.2 TB and Restrictive Lung Disease 385 ity of the reduction in vital capacity and the extent of
24.2.1 Parenchymal Lung Disease 385 the parenchymal disease.
24.2.2 Pleural Disease and Fibrothorax 386 There are no characteristic pulmonary function
24.2.3 Endobronchial Tuberculosis 386
abnormalities in patients with pulmonary tuberculosis.
24.2.4 Surgical and Collapse Procedures 386
24.3 TB and Obstructive Lung Disease 387 Restrictive, obstructive or mixed pattern can present,
24.3.1 Endobronchial Tuberculosis 387 depending on the extent, chronicity and the location of
24.3.2 Bronchiectasis 388 the disease (see below). However, in most patients with
24.4 TB and Mixed Pattern of Obstructive mild to moderate disease, the pathological abnormality
and Restrictive Abnormality 389
is transient and complete recovery in pulmonary func-
24.5 TB and Respiratory Failure/Hemodynamics 389
24.6 Preoperative Evaluation 390 tion tests takes place after completion of therapy.
24.7 Case Illustration 391 As an infection control measure, in our lab a pul-
24.8 Discussion 392 monary function test is not performed until three
References 392 sputum samples have stained negative for acid-fast
bacilli (AFB).

24.1
Pulmonary Function Abnormalities in
Tuberculosis
24.1.1
Historical Aspect
Tuberculosis (TB) is an ancient disease. It has been
found in the spines of Egyptian mummies. Drawings
of persons with Potts's disease have been found on the
walls of Egyptian tombs. Hippocrates (460-370 B.C.)
called the disease "phthisis," and Varro (116-28 B.C.)
conceived the notion that organisms too small to be
seen by the naked eye might be causing the disease.
M. M. IDREEs, MD, FRCP (C), FCCP
Head, Pulmonary Function Laboratory, Division of Pulmonary
Medicine, Department of Medicine, Riyadh Armed Forces
Hospital, ClIO, P.O. Box 7897, Riyadh 11159, Saudi Arabia
S. O. WAll, MD, FRCP (C), FCCP
Head, Division of PulmonaryMedicine,Department of Medicine,
King Khalid Hospital, National Guard, Jeddah, Saudi Arabia
A. AL-AMOUDI, MD, FRCP (C)
Deputy Head, Department of Medicine, King Faisal Specialist,
Hospital & Research Center, Jeddah, Saudi Arabia
24.2
T8 and Restrictive Lung Disease
Both complications and therapeutic measures for
controlling tuberculosis can result in restrictive lung
disease.
24.2.1
Parenchymal Lung Disease
The pulmonary tissue response to the inhalation
of Mycobacterium tuberculosis is characterized by
hypersensitivity leading to necrosis that is sur-
rounded by inflammatory cells. These cells form a
fibrotic tissue encompassing the area involved in the
inflammatory process, isolating it from the remain-
ing lung parenchyma. Limitation of the infection to
this stage has no significant effect on pulmonary
function or gas exchange abnormality.
Decreased static and dynamic lung volume, reduced
forced expiratory volume in one second (FEV!) propor-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
386
tionally to forced vital capacity (FVC) and reduction in
the diffusion capacity (DLco) can result from necrotiz-
ing tuberculous pneumonitis. In massive pneumonia,
severe hypoxemia and widening of the alveolar-arterial
oxygen gradient (A-a 02) take place. This is related
to the low V/Q ratio and the development of venous
admixture. In the absence of coexisting significant
airway obstructions, hypercapnia is rarely observed.
Tuberculosis is an uncommon, but treatable,
cause of Acute Lung Injury I Acute Respiratory
Distress Syndrome ALIIARDS that might develop
as a complication of tuberculous pneumonitis. The
exact pathogenesis of ALII ARDS in TB is unclear.
Massive release of mycobacteria into pulmonary cir-
culation with secondary hyperimmune and inflam-
matory reaction leading to infiltrative and oblit-
erative endarteritis, basal membrane thickness and
vascular and endothelial leakage is the postulated
mechanism. Whether partial or complete resolution
of the lung function abnormalities will take place
depend on the extent of scarring and fibrous tissue
deposition after the acute injury. Most patients who
survive the acute attack will have a mild degree
of parallel reduction in both FEVI and FVC three
months after hospital discharge that will continue to
improve during the first year. A subset of survivors
might end with permanent residual abnormality
manifested by low diffusion capacity.
In a group of 73 patients with cavitary disease and
27 patients with non-cavitary disease, Marcus and
colleagues reported an average vital capacity of 67%
and a diffusing capacity of 63% of that predicted at
baseline (Marcus et al. 1963). In contrast, Malik and
Martin reported an average arterial oxygen tension
(Pa02) of 76 mmHg at baseline in 104 patients with
tuberculosis, but the number of cavities was not
specified (Malik and Martin 1969).
More recently, Long and coworkers reported
milder changes in pulmonary function in these
patients. They prospectively studied 25 patients with
pulmonary tuberculosis. Pulmonary function tests
were performed at baseline, 1 and 6 month intervals.
Patients with non-cavitary disease (15 cases) had
virtually normal lung function, while those with
cavitary disease (10 cases) had very mild restrictive
changes (vital capacity and diffusing capacity aver-
aging 88 and 85%, respectively, of that predicted) and
trivial hypoxemia with Pa02 averaging 85 mmHg. In
the same study, some correlation of lung function
with structure (number of cavities) was found (Long
et al. 1998). This may explain why patients in earlier
studies had inferior values, i.e. they had more disease
or pre-existing lung function abnormalities.
M. M. Idrees et aI.
24.2.2
Pleural Disease and Fibrothorax
Pleural involvement is very common in tuberculous
infection. It usually occurs through direct extension
from a subpleural focus. Tuberculous pleural effusion
and residual pleural thickness may occur in up to
50% of patients (Barbas et al.I991). The exact patho-
genesis behind the development of pleural thickness
is probably related to a delayed hypersensitivity reac-
tion, rather than to an inflammatory response to the
infection. The role of corticosteroids in reducing
the risk of pleural thickness is controversial. Recent
reports suggest no beneficial effect (Lee et al. 1988;
Senderovitz and Viskum 1994). The coexistence
of empyema thoracis may complicate the clinical
picture, leading to grossly thickened pleura with
loculated pleural fluid. Once extensive, this process
can result in trapped lung syndrome and a variable
degree of volume loss leading to significant restric-
tive lung disease. This is manifested by low TLC and
a proportional reduction in both FEVI and FVC.
DLco may also be reduced, resulting in a variable
degree of gas exchange abnormalities. Shunting and
venous admixture may take place at the atelectatic
lobes, leading to a significant degree of widening of
A-a 02 gradient and hypoxemia.
24.2.3
Endobronchial Tuberculosis
Endobronchial tuberculosis will be discussed in
detail as a cause of obstructive lung disease (see
below). It should be noted, however, that a restrictive
pattern is not uncommon in patients with endobron-
chial TB. Post obstructive pneumonitis, atelectasis or
coexisting parenchymal or pleural disease are likely
causes for restrictive lung disease in this patient
population.
24.2.4
Surgical and Collapse Procedures
The need for collapse procedures such as thora-
coplasty, pneumothorax, pneumoperitoneum and
phrenic nerve crush has been virtually eliminated by
the great success of anti-TB chemotherapy. However,
excisional surgery can still be called upon to deal
with complications, such as bronchopleural fistula,
persistent empyema, significant hemoptysis from
bronchiectasis or aspergilloma or excision of a lobe
Pulmonary Function Test and Tuberculosis
or segment infected by multi-drug resistant organ-
isms. These "volume reduction" procedures result in
a variable degree of restriction. In addition, a mixed
obstructive-restrictive pattern has been observed.
The obstructive element in these patients may be
related to the development of compensatory hyper-
inflation of the lung postoperatively. It has also been
suggested that airway obstruction may be related
to the alteration of the bronchial caliber, coexist-
ing cigarette smoking or endobronchial location of
specific inflammation (Marino and Giuliano 1998).
DLco may be normal when compared to residual
alveolar volume. The net effect on gas exchange and
pulmonary circulation hemodynamics depend upon
the integrity of the residual pulmonary tissues.
24.3
18 and Obstructive Lung Disease
Obstructive airway disease came to be recognized as
a complication of advanced pulmonary tuberculosis
within a few years after the advent of effective che-
motherapy.Anno and Tomaschefskyin 1955 reported
pulmonary function studies in 25 selected patients
with proven pulmonary tuberculosis. They dem-
onstrated that airway obstruction and pulmonary
hyperinflation were more common in far advanced
than in less advanced disease (Anno and Tomaschef-
sky 1955). Hallet and Martin (1961) reported on 710
patients with pulmonary tuberculosis admitted to
the Firland Sanitarium. The obstructive pattern was
found in 34% of these patients and occurred more
frequently in men than in women, although the
difference was not statistically significant. The inci-
dence of the obstructive abnormality increased with
advanced age. The duration of smoking appeared to
influence the incidence of obstructive disease; how-
ever, when the age factor was controlled, the smok-
ing relationship was no longer significant (Hallet and
Martin 1961; Martin and Hallet 1961).
The failure to show a very strong association
between smoking and airway obstruction in patients
with pulmonary tuberculosis is an interesting obser-
vation that is difficult to explain. It may be related to
the predominant effect of tuberculosis on the air-
ways and the subsequent damage, irrespective of the
smoking status of the patient. However, current data
suggest an additive effect of smoking with tuberculo-
sis in producing airway obstruction.
The mechanisms of airway obstruction in pul-
monary tuberculosis remain largely speculative.
387
Obstructive lung disease might be due to tuberculous
involvement of the bronchial tree. Sub-mucosal tuber-
culous involvement and nonspecific inflammation are
frequently observed in resected specimens of bronchi
from tuberculous patients (Thompson and Kent 1958).
Bronchial obstruction and dilatation can complicate
lymph node involvement of the primary infection. In
chronic post primary tuberculosis, severe bronchiec-
tasis may result from the involvement of the tuber-
culous lesion in specific segments of the bronchial
tree (see below). Chronic bronchitis, emphysema and
bronchial asthma may develop as purely coincidental
phenomena, especially in patients with atopy or a his-
tory of heavy smoking.
The specific pathological entities involved in pulmo-
nary tuberculosis presenting with obstructive pulmo-
nary insufficiency will be discussed in detail below.
24.3.1
Endobronchial Tuberculosis
Endobronchial tuberculosis (EBTB) is most often a
complication of primary pulmonary tuberculosis in
children, although it may also occur in adults. The
prevalence of EBTB varies among studies, ranging
from 10-50%, probably reflecting the heterogeneity
of the studies' population and design. In the authors'
experience of 98 adults with proven pulmonary TB,
12 (12%) were found to have endobronchial disease
(unpublished data); 10 (10%) of them had submu-
cosal infiltration and hyperemia and 2 (2%) had a
polypoid tumor.
Bronchoscopically, EBTB may present as an endo-
bronchial mass lesion (either polypoid or ulcerative
granuloma), submucosal infiltration, hyperemia with
edema and fibrostenosis. In one study, 50 EBTB cases
were diagnosed by bronchial biopsy, bronchial fine
needle aspiration and washing. Mass lesions were
found in 31 cases (62%),21 of which were ulcerous
granulomas and 10 polypoid masses, while 11 cases
(22%) were of submucosal infiltrative, 16 (32%) fibro-
stenosis and 5 (10%) hyperemia and edema (Altin et
al.1997).
The pathogenesis of EBTB is not yet completely
understood, and the course of EBTB differs accord-
ing to the type. The pathogenesis, presentation, treat-
ment and prognosis of EBTB are discussed in detail
in the chapter on Endobronchial TB. The prognosis
of the actively caseating and edematous-hyperemic
types is poor, resulting in fibrostenosis in two-thirds
of patients. Granular and non-specific bronchitic
types have a more favorable prognosis (Shim 1996).
388 M. M. Idrees et al.

Clinically,cough, mostly nonproductive,is invariably FIoWI~J FNex

present and usually persistent. General symptoms, such
as fever, anorexia and weight loss, usually predominate
over specific pulmonary symptoms. However, dyspnea,
wheeze and strider may be the prominent features and
mimic other conditions such as asthma.
A variety of lung function abnormality patterns 6 8
are recognized by PFT in patients with EBTB. They
include:
iP Pure, restrictive defect as manifested by a propor- 10
tional reduction of FVC and FEV1, with normal
FEVlIFVC % ratio. (Post obstructive pneumonitis,
atelectasis or coexisting parenchymal or pleural Fig. 24. I. FIV loop: variable extrathoracic obstruction
disease are likely causes for restrictive lung dis-
ease in EBTB.)
AawIVa) FNex
lIIil Pure obstructive pattern defined as FEV1/FVC
% ratio is less than 70%, with a low FEVI value. 10

(This is usually caused by the generalized spread

of the disease endobronchially to large and small
airways.)
(I Possible increased airway hyperresponsiveness.

One study, however, looked at this particular issue

by comparing the provocation concentrations
of histamine required to reduce FEVI by 20%
(PC 20) of the pre-challenge baseline between
patients with EBTB, symptomatic asthmatics and
normal healthy controls. PC 20 in EBTB patients
and normal controls was significantly higher
than in asthmatics. Furthermore, PC 20 was not
affected by disease location within the bronchial
tree and did not correlate with FEVI or FVC (Park
et al. 1995).
.. Abnormal inspiratory curve of the flow volume
loop, indicating a variable upper airway (Le. extra-
thoracic) obstruction (Fig. 24.1) leading to inspira-
tory flow reduction. This might occur secondary to
the involvement of the extrathoracic upper airway,
typically found in extensive laryngeal tuberculosis
(LTB). LTB can result from the direct spread of
contaminated sputum to the larynx, or it can be a
secondary manifestation of hematological spread.
.. Fixed large airway obstruction, characterized by
equally reduced inspiratory and expiratory flow
(Fig.24.2). This can be a direct result of large
airway stenosis secondary to EBTB. Hoheisel and
coworkers found that 7% of their patients had
a total occlusion of their airways secondary to
EBTB, 13% had less than total, 13% had signifi-
cant,45% minor and 22% had no residual stenosis
(Hoheisel et al. 1994).
5
10
FN~
Fig. 24.2. FIV loop: fixed obstruction
24.3.2
Bronchiectasis
The pathogenesis of bronchiectasis in tuberculosis is
multifactorial. Bronchial stenosis secondary to EBTB
can result in post obstructive pneumonitis with a sec-
ondary bacterial infection. Exogenous compression
of the bronchial lumen by enlarged lymph node pro-
duces results similar to endobronchial obstruction.
In both situations, the inflammatory destruction of
the bronchial lumen and the subsequent dilatation
of the bronchi is mainly caused by a mixed bacterial
infection and purulent bacterial secretion, with an
exaggerated local immune reaction, rather than by
the local spread of M. tuberculosis.
Pulmonary function abnormalities in this situ-
ation can show pure obstruction, pure restriction
or a combination of both abnormalities, depending
on the type and extent of the involvement and the
degree of residual damage. In extensive disease, both
DLco and lung compliance are reduced, indicating a
significant degree of gas exchange abnormality, with
venous admixture and increased lung stiffness. The
Pulmonary Function Test and Tuberculosis 389

A-a 02 gradient is widened, and, at a more advanced 13 patients with pulmonary tuberculosis requiring
stage, chronic hypoxemia and respiratory failure mechanical ventilation. Seven patients had miliary
can develop. Chronic corpulmonale may develop tuberculosis and six had tuberculosis pneumonia.
secondarily to a persistent increase in pulmonary Eight had ARDS and another two had probable
artery pressure. The physical destruction of the ARDS. The mortality rate of these patients was 69%,
pulmonary vessels by the inflammatory process may which is almost twice that of the hospital mortality
play an important role in the development of pul- for non-tuberculosis pneumonia requiring mechani-
monary hypertension, but this is unlikely to be the cal ventilation (36%) (Penner et al. 1995).
major cause. Hypoxemia, active immune modulators Circulatory disturbance characterized by septic
and cytokines release are more likely to be the most shock is a rare syndrome which has been reported
important factors in the development of full-blown in patients with disseminated tuberculosis (Gachot
pulmonary hypertension and corpulmonale. et al. 1990; Piqueras et al. 1987; Sandoval et al. 1991).
In recent years, the occurrence of extra-pulmonary
and disseminated tuberculosis has been increasing,
especially among HIV patients and intravenous drug
24.4 users, and the incidence of septic shock complicating
T8 and Mixed Pattern of Obstructive disseminated tuberculosis is expected to be on the
and Restrictive Abnormality rise as well.
Pulmonary hypertension (PHT) is a known sequela
In miliary TB, the reported physiological changes are of complicated pulmonary tuberculosis. The exact
of an obstructive-restrictive pattern. In a group of 33 pathophysiology for PHT is not clear. However, various
patients with miliary TB, pulmonary function test- mechanisms have been attributed to the development
ing was performed in 31 patients and arterial blood of PHT and corpulmonale in tuberculosis, including
gases were analyzed in 13 patients (Shama et al.1992). chronic hypoxia and a reduced pulmonary capillary
The physiologic abnormalities observed suggested a bed, caused either by the disease itself or as a result
mild to moderate restrictive defect (vital capacity of surgical therapy (Ershov et al. 1996). Sandoval and
of 56% of that predicted, ±14), with impairment of colleagues studied the hemodynamic behavior of pul-
pulmonary diffusing capacity (68±2l) and arterial monary circulation in both anatomical and functional
hypoxemia (Pa02 of 75 mmHg ±12), and widening settings. Six patients with miliary tuberculosis were
of the A-a 02 gradient. In addition, there was a mild studied. Cardiac catheterization, with subsequent
reduction in the flow rates, based on reduced mid- hemodynamic measurements, was performed at rest
expiratory flow and increased residual volume-to- and during exercise. The mean values of the baseline
total lung capacity. This suggests peripheral airway pulmonary function testing, expressed as percentage
involvement and subsequent gas trapping. These of the predicted value, showed a vital capacity of 62%,
changes are consistent with the diagnosis of inter- FEVI of 50%, FEVlIFVC 76% and TLC of 79.5%. The
stitial lung disease and peripheral airway dysfunc- resting pulmonary artery pressure, cardiac index,
tion. With treatment, however, there was significant pulmonary vascular resistance and the PAd-PWP
improvement in gas exchange (mean Pa02 increased (pulmonary diastolic minus pulmonary wedge pres-
significantly, P value of <0.03), but no change in pul- sure) were all within normal limits. As a group, the
monary function abnormality (Shama et al. 1992). hemodynamic response to exercise was characterized
Other pathological entities causing mixed obstruc- by an increase in both the mean pulmonary artery
tive-restrictive abnormalities are EBTB and post pressure and cardiac index, while pulmonary vascular
resectional surgeries (see above). resistance remained within the normal range. At rest,
hypoxemia (Pa02 59±3 mmHg) without hypercapnia
and normal pH were found in the group as a whole.
The alveolar arterial oxygen gradient and the dead
24.5 space (VDIVT) ratio increased. During exercise, no
T8 and Respiratory Failure/Hemodynamics significant changes occurred in the gas exchange vari-
ables, except for oxygen consumption, respiratory rate
Tuberculosis is an uncommon primary cause of and minute ventilation (Sandoval et al.199l).
respiratory failure requiring mechanical ventila- The correlation between survival and pulmonary
tion. In a retrospective study that extended over a hypertension in patients with sequelae of pulmonary
lO-year period, Penner and colleagues reported only tuberculosis was also studied. Twenty-one patients
390
with a past history of treated pulmonary tubercu-
losis and a mean Pa02 of 59 mmHg and PaC02 of
51 mmHg were included in the study. Spirometry
showed a mean vital capacity of 44% of that pre-
dicted and a FEVI of 66%. Twenty patients had an
mPAP greater than 30 mmHg, and 18 received home
oxygen therapy. Among those patients who required
home oxygen therapy, blood gases and pulmonary
hemodynamics did not vary significantly between
those who died within 2 years (short-term survivors)
and those who lived for more than 5 years (long-term
survivors). However, lung function parameters (FVC
and FEV!) were significantly lower in the short-term
vs.long-term survivors, suggesting that the principle
prognostic determinant for long-term survival in
patients with post tuberculosis sequelae is the degree
of restrictive pulmonary abnormality, and not PHT
or the severity of gas exchange abnormalities (Sasaki
et a1.1998).
24.6
Preoperative Evaluation
Preoperative pulmonary function testing emerged
in the 1950s as an appealing approach for predicting
complications after thoracotomy.
The goals of preoperative evaluation are to:
• Identify patients at risk
• Assess the degree of risk for post operative com-
plications
• Identify factors associated with increased risk
• Keep both morbidity and mortality low
In a retrospective analysis of 454 patients who
underwent thoracic surgery, age (greater than 65
years), obesity and hypocapnia (PaC02 less than
31 mm Hg) were found to be associated with increased
morbidity and mortality (Beaufils et al. 1992).
In patients undergoing lung resection for pul-
monary tuberculosis, mortality was as high as 40%
among those with maximum voluntary ventilation
(MVV) less than 50% of the predicted and FVC less
than 70%. Other parameters associated with poor
outcome included age older than 70 and FEVI less
than 21 (Gaensler et aI.1995).
More recent studies confirmed the correlation
between MVV and poor postoperative outcome.
Olsen and coworkers (1975) reported that 5 out of
M. M. Idrees et aI.
6 patients deemed inoperable by pulmonary artery
pressure and perfusion scan-predicted FEVI mea-
surement had MVV less than 50% of the predicted
outcome, whereas only 2 out of 23 patients who were
deemed operable had low MVV.
FEV1 is also frequently measured to assess the risk
of resection surgery for multi-drug resistant tubercu-
losis or other indications, such as lung cancer. Preop-
erative FEVI has been used in conjunction with per-
fusion lung scan to estimate postresectional FEVl.
Based on data extrapolated from COPD patients,
estimated postoperative FEVI (FEVI-PPO) less than
0.8 I, or 30% of the predicted, has been considered
the threshold below which resection is considered
prohibitive. When this criterion was used to deter-
mine operability in high-risk lung resection patients,
denying surgery to patients with an FEVI-PPO less
than 0.81 defined a group having a 20% postoperative
mortality rate (Olsen et aI, 1975).
More recently, Melendez and Fischer divided the
preoperative tests into three predictive value groups:
lljl Spirometry, blood gas analysis and stair climbing
tolerance are helpful for screening but not very
discriminating. This group had a poor outcome
predictive value.
" An intermediate predictive value was achieved
using diffusion capacity, exercise-induced decrease
in oxygen saturation and exercise pulmonary vas-
cular resistance.
G A high predictive value was achieved by the full car-
diopulmonary exercise test for the measurement of
the oxygen consumption (V02) at 40 watts of exer-
cise, or V02 max (Melendez and Fischer 1997).
Richter and coworkers also evaluated the best pre-
dictor of postoperative morbidity and mortality for
thoracic resection surgery. In all, they prospectively
evaluated 97 patients. All patients had preoperative
maximal exercise testing and dynamic spirometry.
Logistic regression showed maximum preoperative
workload to be the only predictor of cardiopulmo-
nary complications. Maximum oxygen consumption
(V02 max) was predictive of cardiopulmonary death.
V02 max of less than 50% of the predicted value was
associated with a high risk of death from cardio-
pulmonary causes. However, both FEV 1 and V02
max were predictive of postoperative complications.
Finally, V02 max was correlated with long-term sur-
vival, while spirometric variables were not (Richter
et al. 1997).
Pulmonary Function Test and Tuberculosis 391

Recently, the role of cardiopulmonary exercise
testing in the preoperative evaluation for thoracic
resection surgery has become the subject of great
interest. It has been recommended that patients
with FEV1-PPO of less than 40% of predicted values
should be exercised to assess oxygen transport and
consumption. Patients with FEVI-PPO greater than
30% of predicted values and peak exercise V02
greater than 15 mllkg/min should be offered surgery
with the goal of resecting the smallest volume of
tissue that would be compatible with a cure (Gilbreth
and Weisman 1994).
monary illness and no contact with sick people. She
has no pets, nor had she recently traveled.
She was given a three-day course of azithromycin
by her GP with no significant response. Clinically, she
looked ill and mildly distressed. Temperature was
37.8 0c. She had very audible wheezes all over the
lung fields. Chest-x-ray showed mild degree hyperin-
flation, but was otherwise unremarkable.

24.7
Case Illustration

N.S.S is a 54-year-old woman, with no previous ill-

ness aside from mild rhinitis. She presented with a
3-week history of low-grade fever, cough, shortness
of breath and wheeze. FNa
She is a non-smoker and had no family history of
atopy. There was no occupational risk factor for pul-

Pulmonary Function Test -- -- --

R tot 0.30 1.11 371.2
RIN
REX
SR tot 0.96
1.00
1.34
4.20 436.6
4
[fl '0 2

TLC 4.44 4.67 105.1

ITGV 2.54 3.42 134.8
1IIIe(IJ
RV 1.88 2.97 158.4
RVO/OTLC
ITGVO/O
41.74
55.82
63.71
73.33
152.6
131.4
0 1 2 3 4 5
• 7 I

FVC 2.23 1.76 78.5 1.62 -7.9

7-
FEV 1 1.85 1.26 68.0 1.56 24.0
FEVI0/0F 71.57 96.35 34.6
IC 1.67 1.60 95.8 1.62 0.7 2
FET 2.65 1.26 -52.3
PEF 5.41 3.02 55.9 3.39 12.1 1
FEF 25 4.92 1.66 33.7 2.39 44.1 /'
FEF 50 3.28 0.86 26.3 1.34 55.2 1
FEF 75 1.06 0.39 36.2 0.56 45.5
VCIN 2.34 1.87 80.0 1.84 -1.3 -1
PIF 1.74 1.85 6.3

............
FIF 50 1.74 1.85 6.3
~
FE%FIF 49.66 72.53 46.1
392
Arterial blood gases on room air were:
iii PH 7.41
PaC02 5.40 kPa
I'll P02 8.64 kPa
HC03 25.2 mmolll
• 02 Sat 91%
24.8
Discussion
Pulmonary function testing showed a moderate
degree of obstruction. The borderline FEVl/FVC
ratio was related to incomplete exhalation and under-
estimation of FVC. Total lung capacity was normal,
but both residual volume (RV) and RV% TLC were
high, indicating gas trapping. There was a significant
response to bronchodilator.
Arterial blood gases showed a mild degree of
hypoxemia and widening of the A-a 02 gradient.
The patient was treated initially as a case of bron-
chial asthma/bronchitis. Steroids, bronchodilator and
antibiotics were started with partial initial response
followed by deterioration. Three sputum samples for
AFB were negative.
Bronchoscopy was done and an endobronchial
mass lesion in the left main stem bronchus was
found. This was associated with generalized airway
hyperemia. Anti-TB chemotherapy was started and
the patient subsequently improved.
References
Altin S, Cikrikcioglu S, Morgul M et al (1997) 50 endobronchial
tuberculosis cases based on bronchoscopic diagnosis. Res-
piration 64:162-164
Anno H, Tomaschefsky JF (1955) Studies on the impairment
of respiratory function in pulmonary tuberculosis. Am Rev
Tuberc 71:333
Barbas CSV, Cukier A, de Varvalho CRR et al (1991) The
relationship between pleural fluid finding and the devel-
opment thickening in patients with pleural tuberculosis.
Chest 100:1264-1267
Beaufils LC, Brachet A, Manuelian M et al (1992) Preoperative
evaluation of respiratory function in thoracic surgery. Do
reliable predictive criteria exists in that type of surgery.
Agressologie 33:15
M. M. Idrees et al.
Ershov AL, Evstaf'ev IuA, Sobkin AL et al (1996) The signifi-
cance of exacerbations of lung diseases in the development
of chronic corpulmonale and their treatment. Probl Tuberk
4:14-16
Richter LK, Svendsen UG, Milman N et al (1997) Exercise
testing in the preoperative evaluation of patients with
bronchogenic carcinoma. Eur Respir J 10:1559
Gachot B, Wolff M, Clair B (1990) Severe tuberculosis in
patients with human immunodeficiency virus infection.
Intensive Care Med 16:491-493
Gaensler EA, Cugell DW, Lindgren I et al (1995) The role
of pulmonary insufficiency in mortality and invalidism
following surgery for pulmonary tuberculosis. J Thorac
Cardiovasc Surg 29:163
Gilbreth EM, Weisman 1M (1994) Role of exercise stress testing
in preoperative evaluation of patients for lung resection.
Clin Chest Med 15:389
Hallet WY, Martin q (1961) The diffuse obstructive pulmo-
nary syndrome in a tuberculosis sanatorium. Etiological
factors. Ann Intern Med 54:1146
Hoheisel G, Chan BK, Chan CH et al (1994) Endobronchial
tuberculosis: diagnostic features and therapeutic outcome.
Respir Med 88:593-597
Lee CH, Wang WJ, Lan RS et al (1988) Corticosteroids in the
treatment of tuberculous pleurisy: a double blind, placebo-
controlled, randomized study. Chest 94:1256-1259
Long R, Maycher B, Dhar A et al (1998) Pulmonary tubercu-
losis treated with directly observed therapy. Chest 113:
933-943
Malik SK, Martin q (1969) Tuberculosis, corticosteroid therapy,
and pulmonary function. Am Rev Respir Dis 100:13-18
Marcus H, Yoo OH,Akyol T et al (1963) A randomized study of
the effect of corticosteroid therapy on healing of pulmonary
tuberculosis as judged by clinical, roentgenographic and
physiologic measurements. Am Rev Respir Dis 88:55-64
Marino de Rosa, Giuliano C (1998) Respiratory function
impairment in pulmonary tuberculosis. Rays 23:87-92
Martin q, Hallet WY (1961) The diffuse obstructive pulmo-
nary syndrome in a tuberculosis sanatorium. Incidence
and symptoms. Ann Intern Med (Chic) 54:1156
Melendez JA, Fischer ME (1997) Preoperative pulmonary
evaluation of the thoracic surgical patient. Chest Surg Clin
North Am 7:641
Olsen GN, BlockAJ, Swenson EW et al (1975) Pulmonary func-
tion evaluation of the lung resection candidate: a prospec-
tive study. Am Rev Respir Dis 111:379
Park CS, Kim KU, Lee SM et al (1995) Bronchial hyperreac-
tivity in patients with endobronchial tuberculosis. Respir
Med 89:419-422
Penner C, Robert D, Kunimoto D et al (1995) Tuberculosis as a
primary cause of respiratory failure requiring mechanical
ventilation. AJRCCM 151:867-872
Piqueras AR, Marrueces L, Artugas A (1987) Miliary tuber-
culosis and adult respiratory distress syndrome. Intensive
Care Med 13:175-182
Sandoval J, Cicero R, Seoane M et al (199l) Behavior of the
pulmonary circulation at rest and during exercise in mili-
ary tuberculosis. Chest 99:152-154
Sasaki Y, Yamagishi F, Suzuki K et al (1998) Survival and
Pulmonary Function Test and Tuberculosis
pulmonary hemodynamics in patients with sequelae of
pulmonary tuberculosis who received anti-tuberculosis
chemotherapy and home oxygen therapy. Nihon Kokyuki
Gakkai Zasshi 36:934-938
Senderovitz T, Viskum K (1994) Corticosteroids and tubercu-
losis. Respir Med 88:561-565
Shama SK, Pande JN, Singh YN et al (1992) Pulmonary func-
393
tion and immunologic abnormalities in miliary tuberculo-
sis. Am Rev Respir Dis 145:1167-1171
Shim YS (1996) Endobronchial tuberculosis. Respirology 1:
95-106
Thompson JR, Kent G (1958) Occult tuberculous endobronchi-
tis in surgically resected lung specimens. Am Rev Tuberc
77:931
25 Thoracic Surgery for Tuberculosis
YUJI SHIRAISHI
CONTENTS
25.1 History of Thoracic Surgery for Tuberculosis 395
25.1.1 Role of Surgery in the Pre-Rifampin Era 395
25.1.2 Post-Rifampin Era and Emergence
of Drug-resistant Tuberculosis 396
25.2 Surgery for Multidrug-Resistant Pulmonary
Tuberculosis 396
25.2.1 Background 396
25.2.2 Preoperative Medical Treatment 396
25.2.3 Indications for Surgery 399
25.2.4 Preoperative Evaluation 399
25.2.5 Surgical Technique 399
25.2.6 Postoperative Care 400
25.2.7 Types of Operative Procedures 400
25.2.8 Operative and Late Mortality 401
25.2.9 Postoperative Complications 401
25.2.10 Cure Rate of Surgical Treatment 402
25.2.11 Surgical Treatment Other than Pulmonary
Resection 402
25.3 Surgery for Tuberculous Infection
of the Pleural Space 403
25.3.1 Clinical Manifestations and Diagnosis 403
25.3.2 Medical Therapy 403
25.3.3 Surgical Treatment 405
References 407
25.1
History of Thoracic Surgery for Tuberculosis
25.1.1
Role of Surgery in the Pre-Rifampin Era
Thoracic surgery began when the specialty was
"chest" surgery for treating tuberculosis. Our hospital
was established as a sanatorium owned by the Japan
Anti-Tuberculosis Association. The Section of Chest
Surgery began to operate on patients with pulmonary
tuberculosis in 1948. At first, most patients under-
Y. SHIRAISHI, MD
Head, Section of Chest Surgery, Fukujuji Hospital, 3-1-24 Mat-
suyama Kiyose, Tokyo 204-8522, Japan
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
went thoracoplasty. However, pulmonary resection,
such as lobectomy, pneumonectomy or segmentec-
tomy, soon became the predominant operation. The
number of pulmonary resections performed annu-
ally reached around 240 in 1954. On the other hand,
the number of cases treated with collapse therapy
decreased (Fig. 25.1).
Until the advent of antituberculous chemotherapy
in the 1940s, pulmonary resection frequently resulted
in high mortality because of bronchial failure or
poor wound healing due to uncontrolled tuberculous
infection. Therefore, rather than resectional surgery,
collapse therapy, especially thoracoplasty, was the
primary operation for the management of tuberculo-
sis (Pomerantz and Mault 2000). The concept behind
collapse therapy involves compressing the cavities,
which prevents oxygen from entering them, caus-
ing aerobic tubercle bacilli to die from suffocation
(Pomerantz and Brown 1995). Other examples of col-
lapse therapy, which are less frequently performed,
include plombage, therapeutic pneumothorax, pneu-
moperitoneum and phrenic nerve crush.
Under the auspices of adequate chemotherapy,
pulmonary resection became safer and thus was
performed more frequently in the 1950s and 1960s.
Collapse therapy eventually yielded to pulmonary
resection. However, with the introduction of rifampin,
almost all patients with tuberculosis could be success-
fully treated with medication alone. Even the removal
of cavities persisting after medical treatment proved
unnecessary. The need for surgery, therefore, markedly
diminished and sanatoriums in many countries were
closed down. In our hospital, the number of operations
performed for tuberculosis decreased year by year.
Eventually, fewer than 20 operations were performed
for tuberculosis in 1975 (Fig. 25.1). In the late 1970s
and early 1980s, indications for the surgical treatment
of tuberculosis were limited to the following: massive
hemoptysis, bronchopleural fistula, bronchial stenosis,
a trapped lung and the need to rule out cancer (Pomer-
antz and Mault 2000).
396
250 , . . . - - - - - - - - - - - - - - - - - - - ,
N 200 f----+---..:>"7----'",,..,~----___1
u 150 f----~-----____'\_--------1
m
b 100
e
f---~--------_\_-------1
r
50 1---.----.--+-_----------::--1.-----"-=:=-----j
"•
0'-- -----"---------"'-------'-----'---.::....::...---"
1945 1950 1955 1960 1965 1970
Year
- - Pulmonary resection - ..-. Thoracoplasty
Fig. 25.1. The number of operations performed for pulmonary
tuberculosis annually at Fukujuji Hospital between 1948 and
1975. At first, the number of thoracoplasty was higher than that
of pulmonary resection, but soon the situation was reversed.
Pulmonary resections were most frequently performed in the
late 1950s and early 1960s
25.1.2
Post-Rifampin Era and Emergence
of Drug-resistant Tuberculosis
Unfortunately, during the last five decades of the
chemotherapy era, certain strains of Mycobacterium
tuberculosis have acquired resistance to various anti-
tuberculous drugs. Particularly, multidrug-resistant
strains, which are defined as resistant to isoniazid
and rifampin, have emerged. These organisms are
often resistant to other first-line drugs (ethambutol,
streptomycin, and pyrazinamide) and also to some
of the other less efficacious drugs. The emergence of
multidrug-resistant tuberculosis has posed a major
problem in antituberculous therapy (Iseman 1993).
Medical therapy alone has been associated with high
rates of relapse or treatment failure and a mortality
rate of approximately 50% (Goble et al. 1993). Patients
with such resistant strains invariably have either a
thick-walled cavitary disease or a destroyed lobe or
lung (Figs. 25.2-25.4). Cavities contain from 107 to
4
109 organisms compared with 10 2 to 10 in nodules
(Canetti 1965). Thus, the removal of this heavy bac-
terial burden by pulmonary resection may enhance
the efficacy of medical treatment. A destroyed lobe
or lung acts similar to a cavitary disease.
Therefore, the surgeon's role has recently come
back to the armamentarium in the treatment of
tuberculosis (Pomerantz and Brown 1995). Cur-
rently, multidrug-resistant tuberculosis is the most
common indication for surgery for tuberculosis
in many countries. Examples are Italy (Rizzi et al.
1995, 1997), Japan (Nakajima 1997), South Africa
Y. Shiraishi
(van Leuvan et al. 1997), South Korea (Sung et al.
1999), Turkey (Kir et al. 1997), and the United States
(Treasure and Seaworth 1995; Pomerantz et al. 2001).
Exceptions are France (Mouroux et al. 1996; Soui-
lamas et al. 2001) and Hungary (Furak et al. 2001),
where surgery for tuberculosis is seldom performed
for multidrug-resistant strains.
1975
25.2
Surgery for Multidrug-Resistant
Pulmonary Tuberculosis
25.2.1
Background
A team at the National Jewish Center for Immunol-
ogy and Respiratory Medicine in the United States at
Denver, Colorado, is a pioneer of the surgical treat-
ment for multidrug-resistant tuberculosis. Between
the 1970s and early 1980s, pulmonary resection was
still infrequently performed for multidrug-resistant
tuberculosis at the National Jewish Center. Only
nine (5.3%) of 171 patients with multidrug-resistant
tuberculosis underwent surgery from 1973 to 1983
(Goble et aI.1993). However, the center has employed
resectional surgery aggressively over the last two
decades, due to the fact that even the best medical
regimens available have been associated with unac-
ceptably high rates of treatment failure, whereas
surgical treatment has proved safe and efficacious
(Iseman et al. 1990). Beginning with lobectomies
in patients having abundant respiratory reserve,
the National Jewish Center group has progressed to
pneumonectomies or bilateral resections in patients
with very marginal cardiopulmonary reserve. In con-
trast with the earlier series (Goble et al. 1993), out
of 109 patients treated between 1983 and 1993 and
eligible for long-term follow-up, 62 (57%) underwent
pulmonary resections (Iseman 2000).
25.2.2
Preoperative Medical Treatment
Patients with multidrug-resistant tuberculosis initially
should be administered the best antituberculous che-
motherapy possible based on drug sensitivity tests. The
recommended regimen consists of at least four drugs
to which the organism is sensitive, two of which have
not been used previously in the patient. Resectional
surgery should be performed only after approximately
 Thoracic Surgery for Tuberculosis 397

Fig. 25.2a, b. A 59-year-old man had bilateral cavitary diseases in the both upper lobes (a, b). His tubercle bacilli were resistant to
isoniazid, rifampin, and streptomycin. The patient was treated with chemotherapy employing ethambutol, pyrazinamide, cyclo-
serine, kanamycin, ethionamide, and levofloxacin. After six months of chemotherapy, cavities shrunk, but still remained (c). He
underwent staged bilateral upper lobectomies (d). Latissimus dorsi muscle flaps were used to cover the both bronchial stumps

a b

Fig. 25.3a, b. A 48-year-old man had multiple cavitary diseases in the upper and lower lobes of the right lung (a, b). The left lung
was intact. His tubercle bacilli were resistant to seven drugs: isoniazid, rifampin, ethambutol, streptomycin, cycloserine, para-
aminosalicylcclic acid, and ethionamide. After being treated with isoniazid, rifampin, kanamycin, ethionamide and levofloxacin,
he underwent a right pneumonectomy. The bronchial stump was reinforced with the latissimus dorsi
 398 Y. Shiraishi

a b

c d

Fig. 25.4a-d. A 59-year-old man having the destroyed left upper lobe (a, b) was infected with tuberculosis resistant to isoniazid,
rifampin, ethambutol, streptomycin, cycloserine, ethionamide, and levofloxacin. He was treated with pyrazinamide, kanamycin,
and para-aminosalicylcclic acid, and then underwent a left upper lobectomy with a superior segmentectomy of the lower lobe.
The latissimus dorsi was used, but air leakage prolonged (c). Thoracoplasty removing from the second to the seventh ribs was
performed on postoperative day 35, resulting in the obliteration of residual space (d)

three months of intensive medical treatment, and
when the lowest bacterial count possible is achieved
in the sputum smear and culture results. This may
decrease the incidence of postoperative complications.
Attempting resection without establishing antituber-
culous control poses an increased risk of bronchial
stump or chest wall wound dehiscence.
It is ideal if the sputum smear and culture of the
patient becomes negative at the time of operation.
This has been possible, however, in only about half
of the patients with multidrug-resistant tuberculosis,
even those treated with appropriate antituberculous
regimens (Pomerantz et al. 2001). The analysis of
the 1973-1983 series at the National Jewish Center
revealed that the median time for cultures to become
negative was two months, with the great majority
becoming negative at four months (Goble et al.I993).
Therefore, if the patient remains culture positive
after three or four months of chemotherapy, the
patient is unlikely to be cured with medication alone
Thoracic Surgery for Tuberculosis
and should be considered for surgery if feasible.
Improvement in pulmonary hygiene by pulmonary
rehabilitation is essential during the preoperative
period. Nutritional support is also important during
both the medical treatment phase before surgery and
the postoperative period.
25.2.3
Indications for Surgery
The following indications for surgery proposed by
Iseman (2000) have been widely accepted. (A) The
patient is persistently culture positive for multidrug-
resistant tuberculosis despite extended drug retreat-
ment. (B) Patterns of drug resistance are so extensive
as to be likely to be associated with the risk of treatment
failure or relapse with additional resistance. (C) The
patient has localized cavitary or necrotic/destructive
disease in a lobe or region of the lung that is amenable
to resection without producing respiratory insuffi-
ciency and/or severe pulmonary hypertension.
Surgery is sometimes performed on patients who
have achieved negative sputum cultures with chemo-
therapy but are at high risk of relapse because of cavi-
tary diseases or a destroyed lung and extensive patterns
of resistance. Pulmonary resection is also performed on
patients with anticipated noncompliance and on those
with little or no medical options if the current treatment
fails. Contraindication to surgery is bilateral disease that
is so extensive that resection is not feasible.
25.2.4
Preoperative Evaluation
Patients undergoing pulmonary resection for multi-
drug-resistant tuberculosis should be carefully evalu-
ated in order to keep morbidity and mortality rates
low. Preoperative evaluation should include standard
chest x-rays, computed tomography (CT) scan of
the chest, pulmonary function tests and a ventila-
tion perfusion scan. These studies are necessary to
assess the extent of the disease and to determine the
acceptable resection based on anticipated postopera-
tive pulmonary reserve. Bronchoscopy should also be
performed if there is hemoptysis or a possibility of
bronchostenosis. When pulmonary hypertension is
suspected, right heart catheterization should be per-
formed to determine whether the patient is suitable
for resectional surgery. A mean pulmonary artery
pressure of 25 mmHg or greater often precludes
pneumonectomy (Pomerantz and Brown 1997).
399
The purpose of surgery for patients with multi-
drug-resistant tuberculosis is to remove all gross
lesions. This requires the removal of all cavitary dis-
eases as well as the destroyed lung. Scattered nodular
lesions may remain because their mycobacterial
burden is low. Indications for a pneumonectomy,
in contrast to a lesser procedure, include destruc-
tion of an entire lung, multiple cavities in one lung
(Fig. 25.3), stenosis of the main stem bronchus, infec-
tion involving more than one lobe of the lung and
tuberculosis associated with a chronic empyema.
25.2.5
Surgical Technique
At the beginning of the surgery, bronchoscopy is
performed to cleanse the airway in order to prevent
spread to other portions of the lung. A double-lumen
endobronchial tube, which prevents intraoperative
contamination of the contralateral lung, is routinely
used to achieve lung isolation. A posterolateral tho-
racotomy incision is routinely used, and the fifth or
sixth rib is usually removed. When the chest cavity is
entered, dense adhesions are usually found over the
upper lobe and often over the superior segment of
the lower lobe. Dissection of adhesions may be done
intrapleurally or extrapleurally. However, extrapleu-
ral dissection has been recommended whenever the
pleural space is obliterated to prevent the operative
field from being contaminated with tubercle bacilli
(Brown and Pomerantz 1995; Conlan et al. 1995).
During extrapleural dissection, hemostasis should be
achieved using electrocautery and/or bipolar scissors
(PowerStar Bipolar Scissors; Ethicon, Somerville, NJ)
to minimize blood loss. Special care must be taken
to avoid spillage of the contents of cavities into the
operative field. Frequently, once pneumolysis is com-
pleted, the hilar structures can be isolated without
difficulty. However, in the case of a completion pneu-
monectomy, pulmonary arteries and veins may have
to be dissected within the pericardium. The bronchial
stump is covered with muscle flaps or omentum if
indicated. Bronchial closure can then be done either
by suture or with staples.
The liberal use of muscle and omental flaps in combi-
nationwith pulmonaryresection for multidrug-resistant
tuberculosis has been advised in order to decrease the
incidence of postoperative complications (Pomerantz
et al. 1991; Treasure and Seaworth 1995). Indications
for using these flaps include a positive sputum culture
preoperatively, a preoperative bronchopleural fistula,
polymicrobial contamination of the pleural space and
400
anticipated space problems after lobectomy (Pomerantz
and Brown 1997). The choice of muscle flap is the latis-
simus dorsi. The technique for harvesting the latissimus
dorsi has been well described previously (Pairolero et
al.1983). The flap is sutured to the mainstem bronchial
stump, in the case of a pneumonectomy, and used to
cover the hilum. When a lobectomy is performed, the
flap is sutured to the bronchial stump and also placed
at the top of the remaining lung.
Other flaps less frequently used include the pecto-
ralis major muscle, the serratus anterior muscle and
the intercostal muscle. An omental flap is used when
a previous thoracotomy precludes the mobilization
of a muscle flap (Pomerantz and Brown 1997). It is
also used when a left or right pneumonectomy is
performed in the presence of massive contamina-
tion. These flaps may be unnecessary despite positive
sputum in middle lobe and lingula resections, lower
lobectomies and segmental resections.
Hemostasis then should be achieved carefully,
occasionally with the argon beam coagulator. The
chest cavity is thoroughly irrigated and drained. If
a muscle flap has been used, a suction drain should
be placed below the skin flap and left until drainage
becomes less than 25 mllday. On completion of the
surgery, bronchoscopy is usually done to cleanse the
tracheobronchial tree of secretions that may result
from manipulation of the lung during resection.
Pomerantz (2000) has advocated that an Eloesser's
procedure (Eloesser 1935) be done for patients with
extensive contamination after completion of a pneu-
monectomy. The Eloesser is closed 4 to 6 weeks later
with a Clagett procedure (Clagett and Geraci 1963).
25.2.6
Postoperative Care
Postoperative care is usually the same as that for pul-
monary resection for lung cancer. It should be empha-
sized that surgery is an adjunct to medical therapy.
Table 25.1. Types of pulmonary resections for multidrug-resistant tuberculosis
Reference No. of Pneumonectomy
operations
Treasure and Seaworth 19 8 9
(1995)
Kir et aI. (1997) 27 20 (right 5, left 15)
Van Leuvan et aI. (1997) 62 35 (right 7, left 28)
Nakajima (1997) 40 14
Sung et aI. (1999) 27 9 16
Pomerantz et al. (2001) 180 82 (right 22, left 60)
Y. Shiraishi
Therefore, the patient should be kept on antitubercu-
lous medication for 12 to 24 months postoperatively,
depending on the bacteriologic and radiographic find-
ings. Stopping antituberculous drugs too soon leads to
further resistance or mycobacterial spread.
25.2.7
Types of Operative Procedures
Since it is crucial to resect all grossly diseased areas of
the lung, the rates of pneumonectomy have been high
in some reports (Table 25.1). Kir and colleagues (1997)
reported an extremely high rate of pneumonectomy.
Out of 27 operations, 20 (74%) were pneumonecto-
mies. Van Leuvan and associates (1997), Pomerantz
and coworkers (2001) and Treasure and Seaworth
(1995) reported pneumonectomy rates of 56 to 42%.
When a lobectomy or an operation that is more than
a lobectomy, but less than a pneumonectomy, is
performed, the right upper lobe is most commonly
involved. In a report by Pomerantz and associates
(2001), 55 of 93 patients undergoing lobectomy had
the right upper lobe, or the right upper lobe in combi-
nation with another portion of the lung, resected.
In patients with pulmonary tuberculosis, the
destruction of the lung requiring a pneumonectomy
occurs more frequently on the left than on the right
side. This is revealed by the fact that 73% of all pneu-
monectomies were on the left in a series by Pomerantz
and colleagues (2001). Van Leuven and associates
(1997) reported a similar incidence (80%) ofleft-sided
pneumonectomies. Similarly, in a report by Kir and
colleagues (1997), 75% of pneumonectomies were
performed on the left side. Ashour (1997) reported
that 16 (80%) of 20 pneumonectomies were done on
the left side when pneumonectomy was performed
for patients with post-tuberculosis lung destruction.
The reasons for this left-sided predominance remain
unclear. Ashour and associates (1990) postulated that
the predilection for left-lung destruction might be
Lobectomy Segmental
resection
2
7
26 1
24 2
2
93 5
Thoracic Surgery for Tuberculosis
related to the smaller caliber of the left main stem
bronchus, a tight mediastinum through which the left
main stem bronchus traverses and the more horizontal
course of the left main stem bronchus.
25.2.8
Operative and Late Mortality
In general, pulmonary resection for multidrug-
resistant tuberculosis can be performed with low
mortality. Operative mortality rates ranging from
zero to 3.3% have been reported (Table 25.2). Causes
of operative deaths described in the reported series
include postpneumonectomy pulmonary edema,
pulmonary embolism, bronchopleural fistula with
respiratory failure, myocardial infarction and a
cerebral vascular accident. Operative survivors have
seldom died from the progression of multidrug-resis-
tant tuberculosis. Pomerantz and colleagues (2001)
reported that 11 late deaths occurred but only three
were caused by multidrug-resistant tuberculosis. Of
the three patients, one patient had his antitubercu-
lous medication stopped inexplicably in the postop-
erative period, a second had cavitary disease on the
other side and refused a second operation and a third
died from progressive tuberculosis in spite of con-
tinued chemotherapy. Other causes included respira-
tory failure, a self-induced drug overdose, myocardial
infarction, renal failure and unknown causes.
25.2.9
Postoperative Complications
Although operative mortality has been acceptably low,
postoperative complication rates have been high. Mor-
bidity rates have ranged from 12 to 26% (Table 25.2).
Complications described after pulmonary resection
for multidrug-resistant tuberculosis are summarized
in Table 25.3. These complications occur more com-
Table 25.2. Morbidity and mortality after pulmonary resec-
tions for multidrug-resistant tuberculosis
Reference No. of Mortality Morbidity
operations (No.) (No.)
Treasure and Seaworth 19 0% (0) 21% (4)
(1995)
Kir et al. (1997) 27 0% (0) 19% (5)
Van Leuvan et al. (1997) 62 1.6% (1) 23% (14)
Sung et al. (1999) 27 0% (0) 26% (7)
Pomerantz et al. (2001) 180 3.3% (6) 12% (20)
401
Table 25.3. Complications after pulmonary resections for mul-
tidrug-resistant tuberculosis
Reference Complications (no. of patients)
Kir et al. (1997) Revision because of hemorrhage (2);
bronchopleural fistula (2);
apical residual space (1)
Van Leuvan et al. (1997) Postoperative hemorrhage (5);
postpneumonectomyempyema (4);
wound infection (2);
pneumonectomy stump fistula (2);
postpericardiotomy syndrome (1);
respiratory failure (1);
space problem (1)
Sung et al. (1999) Prolonged air leakage (3);
reoperation due to bleeding (2);
bronchopleural fistula (1);
reversible blindness (1)
Pomerantz et al. (2001) Respiratory failure (6);
bronchopleural fistula (5);
wound infection (3);
postoperative hemorrhage (3);
recurrent nerve injury (2);
intrathoracic bowel herniation (1)
monly after surgery for tuberculosis because of the
adhesive nature of the disease process, as well as the
debilitated catabolic condition of the patient.
A bronchopleural fistula is a dreaded complication
after surgery for mycobacterial infections. Factors
adversely affecting the incidence of bronchopleu-
ral fistula include right pneumonectomy, positive
sputum at the time of surgery, significant polymi-
crobial contamination, diabetes and prior chest wall
irradiation (Pomerantz 2000). Although some inves-
tigators have not used muscle flaps to cover the bron-
chial stump (Kir et al. 1997; van Leuvan et al. 1997;
Sung et al. 1999), the use of muscle flaps has been
advocated to decrease the incidence of bronchopleu-
ral fistula (Treasure and Seaworth 1995; Pomerantz et
al. 2001). Bronchopleural fistula is rarely encountered
after resection of the middle lobe, lingula, lower lobe
or segment.
When resections less drastic than pneumonecto-
mies are performed, the inability of the remaining
lung to fill the pleural space has occasionally resulted
in a chronic space problem (Fig. 25.4). Thoracoplasty
has been used to treat space problems (Pomerantz
and Mault 2000). However, Pomerantz and coworkers
(1991) have recommended using muscle flaps to pre-
vent this complication and to avoid thoracoplasty.
While operative morbidity and mortality are the
obvious disadvantages of pulmonary resection, over-
all mortality is reduced, compared with simple medi-
cal treatment, and the patient's opportunity to rejoin
society outweighs the risk.
402
25.2.10
Cure Rate of Surgical Treatment
The combination of pre- and postoperative antitu-
berculous medications and pulmonary resection has
resulted in high cure rates in patients with multidrug-
resistant tuberculosis. Iseman and colleagues (l990)
noted that 92% of patients remained sputum-cul-
ture-negative for tuberculosis a mean of 39 months
after resection. Recent studies around the world
have also demonstrated that cure can be obtained
in more than 80% of the surgical cases (Table 25.4).
Although no randomized study has been done to
compare the cure rates of medical treatment plus
surgical treatment with those of medical treatment
alone, the overall cure rate during the surgical era is
substantially higher than that in the previous era (81
vs. 56%) (Iseman 2000).
In a recent report from Denver, the largest of all
reported series, 172 patients with multidrug-resis-
tant tuberculosis underwent 180 pulmonary resec-
tions over a 17-year period (Pomerantz et al. 2001).
Despite the fact that one half (911172) of the patients
had positive sputum at the time of the operation, the
sputum remained positive in only four patients (2%)
after the operation. In another report, also from the
United States, 17 (89%) of 19 patients with drug-resis-
tant tuberculosis remained sputum-negative up to 12
months after resection (Treasure and Seaworth 1995).
A study from Turkey achieved an overall success
rate of 89% in patients undergoing surgery (Tahaoglu
et al. 2001). Of 158 patients with multidrug-resistant
tuberculosis, 122 were treated with chemotherapy
alone and 36 underwent resectional surgery in addi-
tion to chemotherapy. Of the 36 patients undergoing
surgery, 21 (58%) had cures and 11 (31 %) had prob-
able cures. In contrast, the success rate for patients
with chemotherapy alone was 73%. In another study
from Turkey, 27 human immunodeficiency virus-
negative patients with multidrug-resistant tuber-
culosis underwent resectional surgery (Kir et al.
1997). In all patients but one, negative cultures were
obtained prior to operation, and only one patient
(4%) developed a relapse after operation.
A group in South Africa performed pulmonary
resections on 62 patients (van Leuven et al. 1997). Of
these, 38 patients were sputum negative at the time
of surgery. Of the remaining 24 patients with posi-
tive sputum culture at the time of surgical treatment,
18 (75%) converted to sputum negative immediately
afterward. For all patients who were sputum negative
after the surgery, the actuarial relapse-free rate was
80% beyond 36 months postoperatively. Failure to
Y. Shiraishi
Table 25.4. Success rates of pulmonary resections for multi-
drug-resistant tuberculosis
Reference No. of patients Success
(no. of operations) rate
Treasure and Seaworth (1995) 19 89%
Kir et aI. (1997) 27 96%
Van Leuvan et aI. (1997) 62 80%
Nakajima (1997) 37 (40) 89%
Sung et al. (1999) 27 96%
Tahaoglu et al. (2001) 36 89%
Pomerantz et al. (2001) 172 (180) 98%
convert to negative sputum cultures was consider-
ably less likely after pneumonectomy than it was after
lobectomy or segmentectomy.
In a report from our hospital, 37 patients with
multidrug-resistant tuberculosis underwent 40 pul-
monary resections (Nakajima 1997). Excluding two
patients who died postoperatively, 38 cases were
available for evaluation. Bacteriological relapses were
confirmed in seven cases of the 38. Of these seven
cases, two underwent completion pneumonectomy,
achieving complete cure. Thus, the ultimate cure rate
after pulmonary resection was 89% (31135).
The South Korea group demonstrated that sputum
negative conversion was achieved in 22 patients
(81.5%) initially, out of 27 patients undergoing pul-
monary resection (Sung et al. 1999). Furthermore,
continued postoperative chemotherapy could con-
vert to negative in another four patients (14.8%) and
an overall success rate of 96%.
25.2.11
Surgical Treatment Other than Pulmonary
Resection
As mentioned above, pulmonary resection in combi-
nation with chemotherapy is an effective treatment
for patients with multidrug-resistant tuberculosis.
However, not all patients are eligible for resectional
surgery. Extensive bilateral diseases and/or poor
cardiopulmonary function may preclude pulmonary
resection. For these high-risk patients, several alterna-
tive surgical approaches have been recommended.
Jouveshomme and colleagues (l998) revisited
the efficacy of collapse therapy with plombage for
the treatment of patients with multidrug-resistant
mycobacteria. Out of seven patients undergoing this
surgery, four were infected with multidrug-resistant
tuberculosis. Collapse therapy with insertion of five
to nine spheres resulted in long-standing bacterio-
Thoracic Surgery for Tuberculosis
logical conversion in all four patients. Jouveshomme
and colleagues concluded that collapse therapy with
plombage is a safe alternative therapy in patients
with multidrug-resistant tuberculosis at high risk for
drug treatment failure but considered unsuitable for
pulmonary resection.
Tseng and associates (2000) used cavernostomy
combined with intrathoracic muscle flap transposi-
tion to treat patients with fibrocavernous pulmonary
tuberculosis. Using this relatively simple operative
procedure, they intended to prevent massive blood
loss during pulmonary resection and to preserve
pulmonary function. Ten patients were treated with
this technique, including three patients with multi-
drug-resistant tuberculosis. However, cavernostomy
combined with intrathoracic muscle transposition
resulted in failure in all three patients with multidrug-
resistant tuberculosis. They concluded that patients
with multidrug-resistant tuberculosis were probably
not suitable for this procedure.
25.3
Surgery for Tuberculous Infection
of the Pleural Space
25.3.1
Clinical Manifestations and Diagnosis
Tuberculous infection of the pleural space can occur
in patients with a primary parenchymal infection
and in patients with postprimary tuberculosis.
Pleural tuberculosis may also occur as sequelae of
collapse therapy for tuberculosis. Pleural involve-
ment by tuberculosis presents numerous and
diverse clinical aspects. Barker and Shields (1994)
classified pleural tuberculosis into four manifesta-
tions: (1) pure pleural tuberculosis, (2) pure pleural
disease with mixed tuberculous/pyogenic infection,
(3) pleuroparenchymal tuberculosis and 4) mixed
tuberculous/pyogenic pleural disease with paren-
chymal infection. They recommended that the term,
"tuberculous empyema:' be discarded. However, the
term "tuberculous empyema" is still used by many
investigators (Elliott et al. 1995; Ali et al. 1996; Bai
et al. 1998; Sahn and Iseman 1999; AI-Kattan 2000;
Massard and Wihlm 2000).
Usually, patients with chronic tuberculous empy-
ema have been asymptomatic for years before coming
to clinical attention, because the markedly thickened
pleura virtually isolates the tubercle bacilli into the
empyema cavity. Some patients may have constitu-
403
tional symptoms, such as fatigue and weight loss, and
also manifest low-grade fever and increasing dyspnea
with or without chest pain. However, once a broncho-
pleural fistula has developed, the patients become
extremely ill and sometimes moribund. They may
present with acute fever, dyspnea and production of
abundant purulent sputum. This not only identifies
the disease, it also increases the risk of spread of the
tubercle bacilli through contaminated sputum. In
addition, the pleural space is frequently contami-
nated with one or more pyogenic organisms, which
results in a mixed tuberculous/pyogenic empyema.
The typical finding of tuberculous empyema on
chest radiography is a moderate to large pleural effu-
sion and calcified pleura (Figs. 25.5,25.6).An air-fluid
level is often observed, representing a bronchopleural
fistula (Fig. 25.6). CT scan shows a thick fibrous peel,
calcification of parts of, or all of, the rim of the fibrous
peel and loculated pleural fluid between the parietal
and the visceral peel (Figs. 25.5,25.7). The underlying
lung is usually entrapped. A destroyed lobe or lung
is also observed in many cases (Fig. 25.7). Thora-
centesis yields grossly purulent fluid whose smear is
positive for acid-fast bacilli and whose culture is sub-
sequently positive for Mycobacterium tuberculosis.
The fluid should be routinely examined for aerobic,
anaerobic and fungal cultures.
25.3.2
Medical Therapy
Like parenchymal disease, antituberculous chemo-
therapy forms the basis of the therapy for pleural
tuberculosis (Sahn and Iseman 1999). When tuber-
culous infection is associated with mixed pyogenic
infection of the pleural space, proper antibiotic
therapy should be added. Drug-sensitive tuberculous
strains should be treated with multidrug regimens
containing isoniazid, rifampin and ethambutol to
prevent acquired resistance. Resistant organisms may
require as many as five drugs, including streptomy-
cin, pyrazinamide, cycloserine, and so on, to achieve
bacteriologic control. Antituberculous chemotherapy
should be continued for at least three months, and
sputum conversion should be achieved before con-
sidering any major surgical intervention. It should
be noted, however, that in chronic empyema associ-
ated with thickening and calcification of the pleura,
the inability to achieve therapeutic drug levels in the
pleural fluid can lead to drug resistance.
Iseman and Madsen (1991) reported five patients
with chronic pleural tuberculosis and bronchopleu-
 404 Y. Shiraishi

Fig. 25.5a, b. A 62-year-old man had tuberculous infection of the pleural space. The pleural fluid was positive for Mycobacterium
tuberculosis. Although he was treated with isoniazid, rifampin, and ethambutol, pleural tuberculosis resulted in an entrapped
lung (a). The underlying lung was not destroyed, and no bronchopleural fistula was observed (b). He underwent decortication
and transposition of the latissimus dorsi

a b

Fig.25.6a-c. A 69-year-old man who had a history of pleu-

ritis had drug-sensitive pleuroparenchymal tuberculosis. The
pleural fluid was positive for Mycobacterium tuberculosis, and
a regimen employing isoniazid, rifampin, and ethambutol was
started. During chemotherapy, he developed a bronchopleural
fistula. An apparent air-fluid level was seen on the chest radio-
graph (a). At first, he was treated by open window thoracos-
tomy (b). Purulent contents in the pleural space were positive
for Aspergillus flavus. After 261 days of open drainage, he
underwent decortication, transposition of the latissimus dorsi
and the serratus anterior, and thoracoplasty (c). He has had no
relapse of tuberculosis and empyema since the operation c
 Thoracic Surgery for Tuberculosis 405

ANT
R L

Fig. 25.7a-d. A 72-year-old man having undergone therapeutic pneumothorax 48 years ago had a chronic empyema complicated
with minor bronchopleural fistula (a). He presented with increasing hemoptysis. The left lung was destroyed (b), and blood flow
going into the left lung was markedly diminished in ventilation perfusion scan (c). He underwent extrapleural pneumonectomy
and had the bronchial stump reinforced with a latissimus dorsi muscle flap (d). Operative time was 8 h, 5 min. Intraoperative
blood loss was 1,180 mI, and four units of packed red blood cells were transfused. His postoperative course was uneventful

ral fistula in whom chemotherapy was complicated 25.3.3

by the evolution of drug resistance. Thickened and
calcified empyema walls might have limited the pen-
etration of drugs into the infected empyema space,
resulting in suboptimal drug concentrations and the
acquisition of drug resistance. Elliott and associates
(1995) demonstrated a patient with drug-resistant
chronic tuberculous empyema in whom substantial
differences between achievable serum and pleural
fluid drug concentrations were displayed. The case
strongly suggested that subtherapeutic drug concen-
trations in the pleural fluid might have contributed to
the acquisition of drug resistance.
Surgical Treatment
The initial step for the surgical treatment of pleu-
ral tuberculosis is adequate pleural space drainage
(Sahn and Iseman 1999). This can be achieved either
with chest tube thoracostomy or with open window
thoracostomy. Once the pleural space has been
cleansed, a definitive operation for empyema should
be performed if feasible. The goal of the operation is
either to obliterate the pleural space completely or
to remove the empyema space with the underlying
diseased lobe or lung. These goals cannot usually
406
be achieved with drainage alone, due to thickened
and calcified pleura and an entrapped lung. Bai and
associates (1998), however, reported that 19 of 27
patients with tuberculous empyema were treated
successfully with chemotherapy and repeated tho-
racentesis, or with closed-tube thoracotomy. Ali and
coworkers (1996) reported interesting cases with
tuberculous empyema, where, using open drainage
and antituberculous chemotherapy, "entrapped"
lungs which had totally collapsed expanded to fill
the entire pleural space.
Definitive operations include decortication,
decortication limited to the parietal sides of the
empyema wall, thoracoplasty, transposition of extra-
thoracic muscle flaps and resection of the empyema
along with a lobe or a lung. Choosing the most suit-
able operation based on the extent of the disease is
of paramount importance, and this often requires
a thoracic surgeon's expertise. CT scan of the chest
is useful in differentiating parenchymal and pleu-
ral involvement and in evaluating the status of the
underlying lung, such as whether it contains cavitary
lesions, bronchiectasis or has been destroyed.
When the underlying lung is expected to be
re-expandable, the procedure of choice may be
decortication (Thurer 1996; Katariya and Thurer
1998; Massard and Wihlm 2000; AI-Kattan 2000).
In patients with pure pleural tuberculosis, a pleural
residual occupying one-fourth to one-third of the
hemithorax three months after antituberculous che-
motherapy requires that decortication be carried out
to avoid undue delays in resolution (Langston et al.
1967; Barker and Shields 1994) (Fig. 25.5). Massard
and associates (1995) performed decortication liber-
ally after collapse therapy for tuberculosis, even on
patients with late empyema, and reported favorable
outcomes. The postoperative course after decortica-
tion may be complicated, due mainly to the pleural
space problem. Prolonged air leaks eventually seal
with prolonged drainage, provided that the lung is
completely re-expanded. Residual pleural spaces
after decortication may be managed either with
transposition of muscle flaps or with thoracoplasty
(Massard and Wihlm 2000).
Chronic tuberculous empyema is commonly
associated with the underlying lung that cannot be
re-expanded to fill the pleural space. Should CT scan
demonstrate cavitary disease, large cystic bronchi-
ectasis or destroyed lesions in the underlying lung,
re-expansion of the lung obviously cannot be antici-
pated. In this instance, the use of a modified decor-
tication procedure has been advocated. Iioka and
colleagues (1985) described a technique involving
Y. Shiraishi
decortication of the visceral peel and obliteration of
the dead space by collapsing the parietal wall without
rib resection (Fig. 25.8). The goal of treatment may
also be achieved using either muscle or omental flaps,
or using thoracoplasty. Garda-Yuste and colleagues
(1998) demonstrated a dual-procedure technique,
involving preparatory open window thoracostomy
and subsequent thoracomyoplasty, which achieved
satisfactory results. Using this technique, once the
pleural cavity is stabilized, intrathoracic transposi-
tion of the extrathoracic skeletal muscle is performed
to obliterate the space (Fig. 25.6). Other investigators
have reported the use of thoracoplasty for the treat-
ment of tuberculous empyema (Hopkins et al. 1985;
Horrigan and Snow 1990; Peppas et al. 1993).
When the underlying lung has been extensively
destroyed, extrapleural pneumonectomy should be
considered (Sarot 1949). Surgical success is antici-
pated on exchanging an actively infected pleural
space for a merely contaminated one. However,
extrapleural pneumonectomy for a destroyed lung
associated with empyema is considered a high-risk
procedure (Okano and Walkup 1958; Langston et al.
1967; Odell and Henderson 1985; Massard et al.1995;
Massard and Wihlm 2000). In a report by Massard
and colleagues (1996), the incidence of empyema
or bronchopleural fistula after pneumonectomy (or
chronic infection is higher in patients with sequelae
of tuberculosis. Halezeroglu and associates (1997)
also reported that the combined morbidity and
mortality rate after pneumonectomy for a destroyed
lung is significantly higher in patients with preopera-
tive empyema. Rather than performing extrapleural
pneumonectomy, Odell and Henderson (1985)
invented the technique of pneumonectomy through
an empyema. As reported by Odell and Buckels
(1999), however, postpneumonectomy empyema
developed in 45.7% of the survivors. Blyth (2000)
also performed pneumonectomy through empyema,
resulting in a similar incidence (44%) of postpneu-
monectomy empyema.
Even though there have been admonitions against
using extrapleural pneumonectomy, our group (Shi-
raishi et al. 2000) has performed this high-risk proce-
dure on 94 patients with chronic empyema (Fig. 25.7).
In our study, operative mortality was 8.5%, and 89%
of the operative survivors were free of empyema at
5 years. Langston and coworkers (1967) found that
complications following extrapleural pneumonec-
tomy were more likely to occur in patients who had
had drainage used as a preliminary measure and who
still had a sinus tract at the time of the operation. As
a result of this finding, Barker and Shields (1994) as
 Thoracic Surgery for Tuberculosis 407

a b

c d

Fig. 25.8a-d. A 26-year-old man infected with drug-sensitive tuberculosis (a) was complicated with secondary pneumothorax
(b). The pleural fluid was positive for Mycobacterium tuberculosis. He was treated with isoniazid, rifampin, and ethambutol.
Despite chest tube drainage, the complete re-expansion of the right lung could not be achieved (c). Rupture of cavities might
have been responsible for air leakage. He underwent decortication of the visceral peel and obliteration of the dead space by
collapsing of the parietal wall without rib resection (d)

a rule do not recommend preoperative pleural drain- References

age prior to extrapleural pneumonectomy. In con-
trast, our study demonstrated that preoperative chest
tube drainage and/or open-window thoracostomy
did not adversely affect the incidence of postopera-
tive empyema. It must be emphasized, however, that
surgery for pleural tuberculosis is extremely compli-
cated and formidable, and should only be performed
by experienced thoracic surgeons.
AI-Kattan KM (2000) Management of tuberculous empyema.
Eur J Cardio Thorac Surg 17:251-254
Ali SM, Siddiqui AA, McLaughlin JS (1996) Open drainage of
massive tuberculous empyema with progressive reexpan-
sion of the lung: an old concept revisited. Ann Thorac Surg
62:218-224
Ashour M et al. (1990) Unilateral post-tuberculous lung
destruction: the left bronchus syndrome. Thorax 45:
210-212
408
Ashour M (1997) Pneumonectomy for tuberculosis. Eur J
Cardio Thorac Surg 12:209-213
Bai K-J et al (1998) Tuberculous empyema. Respirology 3:
261-266
Barker WL, Shields TW (1994) Tuberculous and fungal infec-
tions of the pleura. In: Shields TW (ed) General thoracic
surgery, 4th edn. Williams and Wilkins, Baltimore, pp
701-709
Blyth DF (2000) Pneumonectomy for inflammatory lung dis-
ease. Eur J Cardio Thorac Surg 18:429-434
Brown J, Pomerantz M (1995) Extrapleural pneumonectomy
for tuberculosis. Chest Surg Clin North Am 5:289-296
Canetti G (1965) Present aspects of bacterial resistance in
tuberculosis. Am Rev Respir Dis 92:687-703
Clagett aT, Geraci JE (1963) A procedure for the management
of postpneumonectomy empyema. J Thorac Cardiovasc
Surg 45:141-145
Conlan AA et al (1995) Elective pneumonectomy for benign
lung disease: modern-day mortality and morbidity. J
Thorac Cardiovasc Surg 110:1118-1124
Elliott AM et al (1995) Failure of drug penetration and acqui-
sition of drug resistance in chronic tuberculous empyema.
Tuberc Lung Dis 76:463-467
Eloesser L (1935) An operation for tuberculous empyema.
Surg Gynecol Obstet 60:1096-1097
Funik J et al (2001) Surgical intervention for pulmonary tuber-
culosis: analysis of indications and perioperative data relat-
ing to diagnostic and therapeutic resections. Eur J Cardio
Thorac Surg 20:722-727
Garcia-Yuste M et al (1998) Open-window thoracostomy and
thoracomyoplasty to manage chronic pleural empyema.
Ann Thorac Surg 65:818-822
Goble M et al (1993) Treatment of 171 patients with pulmo-
nary tuberculosis resistant to isoniazid and rifampin. N
Engl J Med 328:527-532
Halezeroglu S et al (1997) Factors affecting postoperative
morbidity and mortality in destroyed lung. Ann Thorac
Surg 64:1635-1638
Hopkins RA et al (1985) The modern use of thoracoplasty. Ann
Thorac Surg 40:181-187
Horrigan TP, Snow NJ (1990) Thoracoplasty: current applica-
tion to the infected pleural space. Ann Thorac Surg 50:
695-699
Iioka S et al (1985) Surgical treatment of chronic empyema:
a new one-stage operation. J Thorac Cardiovasc Surg 90:
179-185
Iseman MD (1993) Treatment of multidrug-resistant tubercu-
losis. N Engl J Med 329:784-791
Iseman MD (2000) Drug-resistant tuberculosis. In: Iseman
MD (ed) A clinician's guide to tuberculosis. Lippincott
Williams and Wilkins, Philadelphia, pp 323-354
Iseman MD, Madsen LA (1991) Chronic tuberculous empyema
with bronchopleural fistula resulting in treatment failure
and progressive drug resistance. Chest 100:124-127
Iseman MD et al (1990) Surgical intervention in the treatment
of pulmonary disease caused by drug-resistant Mycobacte-
rium tuberculosis. Am Rev Respir Dis 141:623-625
Jouveshomme S et al (1998) Preliminary results of collapse
therapy with plombage for pulmonary disease caused by
multidrug-resistant mycobacteria. Am J Respir Crit Care
Med 157:1609-1615
Katariya K, Thurer RJ (1998) Surgical management of empy-
ema. Clin Chest Med 19:395-406
Y. Shiraishi
Kir A et al (1997) Role of surgery in multi-drug-resistant
tuberculosis: results of 27 cases. Eur J Cardio Thorac Surg
12:531-534
Langston HT, Barker WL, Graham AA (1967) Pleural tubercu-
losis. J Thorac Cardiovasc Surg 54:511-519
Massard G, WihIm J-M (2000) Tuberculous and fungal infec-
tions of the pleura. In: Shields TW, loCicero J III, Ponn RB
(eds) General thoracic surgery, 5th edn. Lippincott Wil-
liams and Wilkins, Philadelphia, pp 717-727
Massard G et al (1995) Decortication is a valuable option for
late empyema after collapse therapy. Ann Thorac Surg 60:
888-895
Massard G et al (1996) Pneumonectomy for chronic infec·
tion is a high-risk procedure. Ann Thorac Surg 62:
1033-1037
Mouroux J et al (1996) Surgical management of pleuro-
pulmonary tuberculosis. J Thorac Cardiovasc Surg Ill:
662-670
Nakajima Y (1997) Surgical treatment of multidrug resistant
pulmonary tuberculosis cases (in Japanese). Kekkaku 72:
25-34
Odell JA, Buckels NJ (1999) Techniques of pneumonectomy:
pneumonectomy through an empyema. Chest Surg Clin
North Am 9:369-378
Odell JA, Henderson BJ (1985) Pneumonectomy through an
empyema. J Thorac Cardiovasc Surg 89:423-427
Okano T, Walkup HE (1958) Some problems in extrapleural
pneumonectomy for tuberculous empyema and destroyed
lung. J Thorac Surg 35:523-531
Pairolero PC, Arnold PG, Piehler JM (1983) Intrathoracic
transposition of extrathoracic skeletal muscle. J Thorac
Cardiovasc Surg 86:809-817
Peppas G et al (1993) Thoracoplasty in the context of current
surgical practice. Ann Thorac Surg 56:903-909
Pomerantz BJ et al (2001) Pulmonary resection for multi-
drug resistant tuberculosis. J Thorac Cardiovasc Surg 121:
448-453
Pomerantz M (2000) Surgery for the management of Myco-
bacterium tuberculosis and nontuberculous mycobacterial
infections of the lungs. In: Shields TW, LoCicero JIII, Ponn
RB (eds) General thoracic surgery, 5th edn. Lippincott Wil-
liams and Wilkins, Philadelphia, pp 1066-1075
Pomerantz M, Brown J (1995) The surgical management of
tuberculosis. Semin Thorac Cardiovasc Surg 7:108-111
Pomerantz M, Brown JM (1997) Surgery in the treatment of mul-
tidrug-resistant tuberculosis. Clin Chest Med 18:123-130
Pomerantz M, Mault JR (2000) History of resectional surgery
for tuberculosis and other mycobacterial infections. Chest
Surg Clin North Am 10:131-133
Pomerantz M et al (1991) Surgical management of resistant
mycobacterial tuberculosis and other mycobacterial pul-
monary infections. Ann Thorac Surg 52:1108-1111
Rizzi A et al (1995) Results of surgical management of tuber-
culosis: experience in 206 patients undergoing operation.
Ann Thorac Surg 59:896-900
Rizzi A et al (1997) Modern morbidity following pulmonary
resection for postprimary tuberculosis. World J Surg 21:
488-491
Sahn SA, Iseman MD (1999) Tuberculous empyema. Semin
Respir Infect 14:82-87
Sarot IA (1949) Extrapleural pneumonectomy and pleurec-
tomy in pulmonary tuberculosis. Thorax 4: 173-223
Shiraishi Yet al (2000) Morbidity and mortality after 94 extra-
Thoracic Surgery for Tuberculosis
pleural pneumonectomies for empyema. Ann Thorac Surg
70:1202-1206
Souilamas R et al (2001) Surgical treatment of active and
sequela forms of pulmonary tuberculosis. Ann Thorac
Surg 71:443-447
Sung S-W et al (1999) Surgery increased the chance of cure in
multi-drug resistant pulmonary tuberculosis. Eur J Cardio
Thorac Surg 16:187-193
Tahaoglu K et al (2001) The treatment of multidrug-resistant
tuberculosis in Turkey. N Engl J Med 345:170-174
409
Thurer RJ (1996) Decortication in thoracic empyema. Chest
Surg Clin North Am 6:461-490
Treasure RL, Seaworth BJ (1995) Current role of surgery in Myco-
bacterium tuberculosis. Ann Thorac Surg 59:1405-1407
Tseng Y-L et al (2000) Intrathoracic muscle flap transposi-
tion in the treatment of fibrocavernous tuberculosis. Eur
J Cardio Thorac Surg 18:666-670
Van Leuven M et al (1997) Pulmonary resection as an adjunct
in the treatment of multiple drug-resistant tuberculosis.
Ann Thorac Surg 63:1368-1372
26 Radionuclides in Pulmonary
and Extra-Pulmonary Tuberculosis
DAVID HAMILTON and JAWDA AL-NABULSI

CONTENTS 26.1
Introduction
26.1 Introduction 411
26.2 Pulmonary Tuberculosis 411
26.2.1 67Gallium Citrate (67 Ga) 412 Radionuclide investigation is a sensitive, but generally
26.2.2 201Thallium Chloride e01Tl) 414 non-specific, indicator of the presence and the extent
26.2.3 99Technetiumm (99Tc m ) of tuberculosis; proving valuable in differentiating
Radiopharmaceuticals 414 active from inactive disease. It also provides a method
26.2.4 lllIndium (lllIn) Radiopharmaceuticals 415
of revealing reactivation of the disease, estimating the
26.2.5 Sodium 123Iodide (1231) 415
26.2.6 18Fluorodeoxyglucose Positron Emission therapeutic response, and assessing consequential
Tomography 415 alteration of normal physiology. Experience extends
26.2.7 Alteration of Normal Physiology 415 over 40 years but in the last 10 years, with the resur-
26.3 Soft Tissue Tuberculosis 416 gence of the disease, new radiopharmaceuticals and
26.3.1 67Gallium Citrate (67 Ga) 416
imaging techniques have been added to the arma-
26.3.2 Leukocytes 418
26.3.3 99Technetiumm (99Tc m ) mentarium. Also in this period, radionuclides have
Radiopharmaceuticals 418 increasingly been focused on detecting the opportu-
26.3.4 18Fluorodeoxyglucose Positron Emission nistic infection complications associated with human
Tomography 418 immunodeficiency virus (HIV}-infected patients.
26.3.5 Alteration of Normal Physiology 419
The non-specificity of imaging using radionu-
26.4 Renal Tuberculosis 419
26.5 CNS Tuberculosis 420 clides, particularly in this disease, must be empha-
26.5.1 67Gallium Citrate (67 Ga) 420 sized. In vivo distribution of radiopharmaceuticals
26.5.2 99Technetiumm (99Tc m ) demonstrates characteristics that are common to
Radiopharmaceuticals 420 many pathologies and interpretation must be under-
26.5.3 201Thallium Chloride e01Tl) 420
taken with extremely careful consideration of the
26.5.4 18Fluorodeoxyglucose Positron Emission
Tomography 421 clinical environment.
26.5.5 Partition Test 421 Most experience in tuberculosis has been gained
26.6 Bone and Joint Tuberculosis 422 using planar imaging, and this remains adequate for
26.6.1 99Tc m Polyphosphonate 422 the majority of investigations. With the increasing
26.6.2 67Gallium Citrate (67 Ga) 425
sophistication of single photon emission tomogra-
26.6.3 lllIn Leukocytes 427
26.6.4 18Fluorodeoxyglucose Positron Emission phy (SPET) technology, however, this type of acquisi-
Tomography 427 tion is being increasingly utilized to improve detec-
References 428 tion, particularly for small, deep lesions. Positron
emission tomography (PET) has recently provided
an alternative perspective to the functional imaging
of this disease.

D. HAMILTON, PhD, FIPEM
Department of Nuclear Medicine, Riyadh Armed Forces
Hospital, P.O. Box 7897, Riyadh 11159, Saudi Arabia
J. AL-NABULSI, DMRD, MSc NM
Department of Nuclear Medicine, Riyadh Armed Forces
Hospital, P.O. Box 7897, Riyadh 11159, Saudi Arabia
26.2
Pulmonary Tuberculosis
A number of radiopharmaceuticals have been used in
the evaluation of pulmonary tuberculosis. Their main
contribution to clinical investigation is the ability to

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
412
differentiate active from inactive disease, thus enabling
evaluation of therapeutic response and revelation of
reactivation. In patients with acquired immune defi-
ciency syndrome (AIDS), in particular, nuclear medi-
cine procedures are more sensitive than morphologic
imaging modalities in localizing sites of pulmonary
tuberculous infection (Abdel-Dayem et al. 1997).
26.2.1
67Gallium Citrate (67Ga)
Early studies were performed primarily using 67Ga
(Siemsen et al. 1978), and this remains the most
widely used (Gulaldi et al. 1995; Onsel et al. 1996;
Utsunomiya et al. 1997; Degirmenci et al. 1998) and
best radiopharmaceutical for imaging mycobacterial
infections (McAfee 1996). It is a "catch-all" radio-
pharmaceutical that demonstrates accumulation in
infection, inflammation, and neoplasm, but requires
a delay of 24-72 h between administration and imag-
ing (Gulaldi et al.1995; Onsel et al.1996; Utsunomiya
et al. 1997; Degirmenci et al. 1998; O'Doherty et al.
1997; Lin and Hsieh 1999) because of prolonged
blood clearance and biliary excretion (Gulaldi et
al. 1995). However, the relatively long half-life of the
radionuclide facilitates protracted imaging protocols
that accommodate slow accumulation in chronic and
indolent infections. An advantage of using 67Ga to
investigate suspected infection is its ability to image
the whole body. This may give rise to incidental
findings (Grieff and Lisbona 1991; Lin and Hsieh
1999), and also facilitates determination of the most
appropriate biopsy site by exposing involved acces-
sible peripheral nodes (Goldfarb et al. 1997).
The accumulation of gallium in tuberculous
lesions is complex, but is thought to result from
transcapillary exudation of transferrin-bound gal-
lium with subsequent binding to leukocytes or
bacteria (Moody and Delbeke 1992; Brophey et al.
1995b). It is affected by malnutrition (Walsh et al.
1985), but not by the impaired cellular immunity in
patients with AIDS (Lee et al. 1994). The distribution
pattern of the radiopharmaceutical alone can be
useful to differentiate different types of pulmonary
infections (Lee et al. 1994). It can, however, be similar
to the pattern demonstrated by other pathologies
and therefore, differentiation may be difficult in the
absence of a consideration of the clinical environ-
ment (Abdel-Dayem et al. 1995).
Generally, the distribution tends to be asymmetric
and irregular (Fogelman et al. 1994), mainly at the
upper lobes {Goldfarb et al.1995).As with lymphoma,
D. Hamilton and J. AI-Nabulsi
focal uptake at the hila, mediastinum, and through-
out the lung parenchyma is also expected (Ganz and
Serafini 1989; Lee et al. 1994). As noted in sarcoidosis,
hilar or diffuse lung involvement is most often seen
(Fogelman et al. 1994). Pneumocystis pneumonia
(PCP) usually shows bilateral and diffuse uptake, and
bacterial pneumonia tends to show segmental or lobar
uptake (Lee et al. 1994). Atypical mycobacterial infec-
tions present more frequently with extra hilar nodes
(Ganz and Serafini 1989). In miliary disease, 67Ga
primarily demonstrates a diffuse bilateral pattern
(Winzelberg 1981) that is similar to the pattern dem-
onstrated by diffuse pulmonary metastases (Abdel-
Dayem et al. 1995). The intensity of the accumulation
is usually similar to or slightly less than that in the liver.
Intensity greater than that in the liver is reported in
only 12% of patients (Kao et al. 1993). Localized accu-
mulation occurs less frequently (Moody and Delbeke
1992; Kao et al. 1993; Malhotra et al.1985).
In pulmonary tuberculosis associated with AIDS,
typical distribution patterns differ significantly from
those seen in "classical" tuberculosis. The traditional
upper lobe presentation is the exception rather than
the rule; lower lobe disease and adenopathy are more
common (Goldfarb et al. 1995, 1997; Abdel-Dayem
et al. 1997). Parenchymal lesions are more heteroge-
neously distributed than would be expected in tradi-
tional pulmonary tuberculosis (Goldfarb et al. 1995),
and it becomes difficult to differentiate other oppor-
tunistic infections on the basis of the 67Ga distribu-
tion alone. Differentiation with PCP would usually
be possible by relying on the characteristic diffuse
distribution demonstrated by this infection (Abdel-
Dayem et al. 1997). However, this pattern can vary
in AIDS patients and become more heterogeneous,
ill-defined, and peri-hilar, causing problems in inter-
pretation. Also, there tends to be localized uptake
in bacterial pneumonia, and ill-defined, peri-hilar
uptake in cytomegalovirus (CMV) infection (Kramer
et al. 1989). Lymph node accumulation occurs in the
majority of cases, however (Goldfarb et al. 1995;
Abdel-Dayem et al. 1997). This forms the basis of a
more definite differentiation from other opportunis-
tic infections. The presence of lymph node uptake, in
the presence or absence ofparenchymal uptake, raises
the possibility of tuberculous involvement; whereas a
lack of uptake in the lymphoid regions suggests that
the disease is unlikely (Goldfarb et al. 1995, 1997;
Parmett et al. 1995; Santin et al. 1995; Abdel-Dayem
et al. 1997). Such lymph node accumulation does,
however, make differentiation from lymphoma more
difficult because of the similar presentation of these
two entities (Kramer et al. 1989).
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
Imaging with 67Ga is a very sensitive indicator of
the presence of active tuberculosis (Siemsen et al.
1978; Moody and Delbeke 1992; Kao et al. 1993) and
seems more accurate than physical examination of
the neck or plain X-rays of the chest (Goldfarb et al.
1997). Active disease results in a positive image in up
to 97% of patients (Siemsen et al. 1976, 1978; Walsh
et al. 1985; Lin and Hsieh 1999). 67Ga is extremely
valuable in differentiating active disease from estab-
lished scarring, since inactive disease results in a
negative image in up to 100% of patients (Siemsen
et al. 1978; Kao et al. 1993). False negative results
do occur, however. Conditions that can cause this
include anti-tuberculous treatment (Kao et al. 1993)
and malnutrition (Walsh et aI.1985). Less commonly,
a false negative image can arise in association with
drug interference or leukopenia as a result of chemo-
therapy (Kao et aI.1993).
In patients with AIDS, 67Ga scintigraphy is also a
very sensitive technique for localizing sites of tuber-
culous infection (Palestro et al. 1991; Santin et al.
1995; Abdel-Dayem et al. 1997; Goldfarb et al. 1997).
A normal 67Ga distribution is very suggestive of an
absence of active disease (Lee et al. 1994; Santin et al.
1995), but cannot exclude the possibility completely
(Utsunomiya et al. 1997). In terms of opportunistic
infections other than tuberculosis, anti-tuberculous
treatment can increase the false negative rate of the
67Ga investigation (Goswami et al. 2000).
67Ga scintigraphy is complementary to chest X-ray
for determining the activity of disease (Gulaldi et
al. 1995; Onsel et al. 1996) because this is difficult to
diagnose solely morphologically, in the presence of
extensive fibrotic areas (Loken 1987). There is a cor-
relation between the extent of 67Ga accumulation and
the extent of radiographic abnormalities but often, the
67Ga uptake is greater than the extent of disease shown
on the chest X-ray (Siemsen et al. 1978; Walsh et al.
1985). Also, abnormal patterns of 67Ga accumulation
may be seen earlier than radiographic abnormalities
(Walsh et al. 1985; Moody and Delbeke 1992). In mili-
ary disease, both 67Ga scintigraphy and chest X-rays
are mandatory for reliable diagnoses, especially in
high-risk patients. Some lesions may be missed when
only one of the modalities is used (Kao et al. 1993).
67Ga images may be positive in the presence of normal
chest radiographs (Walsh et al. 1985) even when the
accumulation is localized (Kao et al. 1993; Grieff and
Lisbona 1991). In AIDS, 67Ga scintigraphy and chest
X-rays are again complementary and mandatory
techniques (Parmett et al. 1995; Goldfarb et al. 1997),
achieving a combined sensitivity of 96% (Abdel-
Dayem et al. 1997). 67Ga is much more sensitive than
413
chest X-rays when only lymph node accumulation is
present (Goldfarb et al. 1997). Chest X-rays are even
less likely to demonstrate abnormality when there is
only peripheral node 67Ga uptake than when intratho-
racic uptake is present (Santin et al. 1995).
Imaging with 67Ga is better than conventional
chest radiography in establishing reactivation of the
disease, again, because 67Ga uptake is rarely observed
in inactive tuberculosis (Siemsen et al. 1978; Moody
and Delbeke 1992). Because of the high sensitivity
and specificity, abnormal 67Ga accumulation, even
in the presence of a normal chest X-ray and nega-
tive sputum smears, suggests active disease. This
helps to justify further diagnostic procedures and
early implementation of empiric therapy (Walsh et
al. 1985; Grieff and Lisbona 1991). Such a regimen is
also true in HIV-positive patients, especially in areas
with a high prevalence of tuberculosis infection, and
can be very useful in those patients from whom a
biopsy is difficult to obtain. Because the diagnosis
of tuberculosis is unlikely when the 67Ga scan does
not show nodal uptake (Santin et al. 1995); in those
patients suspected of having active tuberculosis, but
whose chest X-ray is not suggestive of this, the pres-
ence of intrathoracic lymphadenopathy on 67Ga scin-
tigraphy heightens suspicion of the disease (Parmett
et al. 1995; Santin et al. 1995).
67Ga scintigraphy is better than morphological
radiology for assessing the response of the disease
to therapy (Siemsen et al. 1978; Kao et al. 1993). 67Ga
uptake decreases with the duration of therapy, indi-
cating a decrease in disease activity that is generally
consistent with clinical findings. The uptake increases
as the disease worsens (Utsunomiya et al. 1997). Suc-
cessful treatment results in a marked reduction of
67Ga accumulation within a few months of initiation
(Siemsen et al. 1978). In patients with AIDS, as well as
providing an indicator of therapeutic response, 67Ga
scintigraphy can be used to help determine the point at
which medication may be stopped, as it is more reliable
than clinical symptoms or the findings on chest radiog-
raphy (Bekerman et al.1980; Abdel-Dayem et al.1997).
67Ga scintigraphy suffers from three main disad-
vantages in its use. It produces images of poorer qual-
ity than some other radiopharmaceuticals; it requires
a protracted delay of 48-72 h between administration
and imaging; and it may require cleansing of the
bowel to ensure accurate diagnostic information. Also,
the specificity of 67Ga is poor because it is a "catch-all"
radiopharmaceutical, which demonstrates increased
accumulation in infection, inflammation, and neo-
plasm (Walsh et al. 1985). The specificity in patients
with AIDS, in particular, is poor because other oppor-
414
tunistic infections such as PCP or CMV can cause diffi-
culty in the interpretation (Abdel-Dayem et al.1997).
There are a number of radiopharmaceuticals that
can be used as alternatives to 67Ga and that show
similar patterns of distribution. Unlike 67Ga, scinti-
grams obtained with these radiopharmaceuticals can
be obtained within a few hours of administration and
require little patient preparation.
26.2.2
201Thallium Chloride (20'TI)
20ln scintigraphy, undertaken soon after administra-
tion, might be more useful than 67Ga scintigraphy for
the evaluation of active disease (Utsunomiya et al.
1997; Degirmenci et al. 1998). 20lTI was first used for
this purpose following incidental imaging of active
pulmonary tuberculosis on cardiac studies (Lee et
al. 1994; Burke 1995) and observation of diffusely
increased accumulation in miliary tuberculosis
(Abdel-Dayem et al.1995). The uptake of tracer in an
infected area is variable immediately after injection
(Lee et al. 1994), but a high lesion to normal tissue
ratio is normally obtained at 15 min (Utsunomiya et
al. 1997). Patients in whom pulmonary tuberculosis is
inactive should have negative findings on 20ln scin-
tigraphy (Lee et al. 1994). Even patients who showed
improvements on other investigations but in whom
subsequent inadequate treatment allowed recurrence
of tuberculous lesions showed slight 20lTI uptake. In
view of these observations, the possibility of recur-
rence should be considered in patients who show any
abnormal 20ln accumulation, even in the presence of
normal clinical data (Utsunomiya et al. 1997).
20ln scintigraphy is also possibly more useful than
67Ga for the evaluation of therapeutic response. 20lTI
accumulation decreases with duration of therapy,
indicating a decrease in disease activity that is con-
sistent with clinical findings. The uptake increases
as the condition worsens. The retention of 201n, over
several hours seems to suggest whether healing is
underway rather than reflecting the level of disease
activity. It decreases significantly in patients showing
improvement after therapy and increases in patients
experiencing deterioration (Utsunomiya et al.1997).
As with 67Ga, the specificity of 20lTI imaging early
after administration is poor. Tuberculosis cannot be
differentiated from certain tumors (Utsunomiya et
al. 1997) for which 201T1 also shows high avidity (Lee
et al.1994).A major problem in the follow up of AIDS
patients using nuclear medicine is the inability to dif-
ferentiate among mycobacterial infections, lymphoma,
D. Hamilton and J. Al-Nabulsi
and Kaposi sarcoma (KS) (Lee et al.1994). The specific-
ity of scintigraphy can be improved using sequential
20lTI and 67Ga imaging. Both infection and lymphoma
generally result in a positive 67Ga image, whereas a
negative 67Ga image in the presence of a positive early
(less than 15 min after injection) 20lTI image has a
high specificity for the diagnosis of KS (Lee et al. 1994;
Abdel-Dayem et al. 1996). Differentiation with lym-
phoma, which demonstrates the same concurrent 67Ga
and early 20ln accumulation as tuberculosis, remains
difficult. In infection, the increased 20ln accumula-
tion decreases relatively rapidly, whereas lymphoma
retains 20lTI much longer. Delayed 201Tl scintigraphy, at
3-4 h following administration, therefore, enables dif-
ferentiation between the two entities. Local mismatch
on sequential delayed 201 Tl and 67Ga images is reported
to be highly specific for AIDS-related mycobacterial
infections (Lee et al. 1994); but correct results are not
always achieved (Abdel-Dayem et al. 1994, 1996; Gomez
et al.1996). Other opportunistic infections may show a
better clearance of 2ol TI than tuberculosis. In pCP,201n
accumulation is seen only in the early images; delayed
images show significant clearance. In miliary tubercu-
losis, however, although the increased accumulation is
seen predominantly in the early images, the clearance
in the delayed images is not as significant as in PCP
and both early and delayed 20lTI images showed dif-
fuse uptake with no significant clearance in the delayed
images (Abdel-Dayem et al. 1995).
26.2.3
99Technetium rn (99Tcrn ) Radiopharmaceuticals
Tuberculosis generally shows considerable methoxy-
isobutylisonitrile (MIBI) uptake, which is most
probably related to disease activity. It can be used to
evaluate therapeutic response, showing a decrease in
accumulation with regressing disease and an increase
with worsening disease, and can reveal reactivation
of previously treated disease. The extent of abnor-
mal MIBI accumulation in positive images generally
exceeds the area of chest X-ray abnormalities. The
main problem with MIBI is its inability to differen-
tiate tuberculosis from other benign and malignant
disease processes. Another drawback is its extensive
usual distribution, which can compromise disease
detection. For instance, normal skeletal muscle
uptake of the thorax and lower neck can mask abnor-
mal accumulation in the apical regions, particularly
for small lesions (Onsel et al. 1996).
Tetrofosmin (Myoview) behaves in a manner similar
to MIBI. It may be more sensitive than sputum smear
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
tests and gives clinically useful information until results
of sputum cultures are available. It can contribute to
the evaluation of the response to therapy. As with MIBI,
specificity of tetrofosmin is poor; differential diagnosis
between tuberculosis and tumor is difficult because of
marked uptake in both (Degirmenci et al.1998).
Pentavalent dimercaptosuccinic acid [(V) DMSAj
may perform similarly to 67Ga for assessment of the
extent and activity of tuberculosis. Citrate does not
appear useful in this disease (Gulaldi et al. 1995).
Although glucoheptonate (GH) is less expensive and
more convenient to use than 67Ga, and although the
results are similar using both radiopharmaceuticals,
GH is not considered an adequate substitute for 67Ga
(Vorne et al. 1988; Braga et al. 1998). Two radiophar-
maceuticals that have been used successfully in infec-
tion imaging, human polyclonal immunoglobulin G
(HIG) and nanocolloid, do not seem appropriate in
this disease. Both showed an absence of accumulation
in patients with active disease in whom MIBI demon-
strated positive results. They also showed an absence
of uptake in patients previously treated and in patients
with a suspicion of relapse (Onsel et al. 1996).
26.2.4
llllndium (111 In) Radiopharmaceuticals
Somatostatin analogue (octreotide and pentetreo-
tide) scintigraphy may be of value in detecting the
presence and assessing the extent of active disease
(Ozturk et al. 1994; Vanhagen et al. 1994) because
of the presence, in granulomatous lesions, of acti-
vated lymphocytes that have somatostatin receptors
(Ozturk et al. 1994). Accumulation is not specific to
tuberculosis, however, since patients with sarcoidosis
also show increased uptake (Vanhagen et al. 1994),
albeit not focal (Ozturk et al. 1994). Also, normal
as well as activated lymphocytes and macrophages
have been shown to express somatostatin receptors
(Vanhagen et al. 1994).
Leukocytes radiolabeled with III In have been used
extensively in the imaging of infection but do not
seem appropriate in this disease, particularly when it is
associated with AIDS (Fineman et al. 1989; Palestro et
al. 1991). Although experimental studies with radiola-
beled monoclonal antibodies appear promising for the
specific localization of tuberculosis (Lee et al. 1992), no
human trials have been undertaken thus far (Onsel et al.
1996). The high cost of these studies, the development
of the HAMA response, and the lack of experience in
clinical studies have limited routine application (Degir-
menci et al.1998).
415
26.2.5
Sodium 12310dide (1231)
1231 is not suggested as a radiopharmaceutical for
detecting tuberculosis; it may, however, be compro-
mised in its ability to detect malignant disease by
previously treated pulmonary tuberculosis (Bakheet
et al. 1999).
26.2.6
18Fluorodeoxyglucose Positron Emission
Tomography
Focal accumulation of 18fluorodeoxyglucose (l8FDG)
occurs in active tuberculosis (Patz et al. 1993; Knight
et al. 1996; Bakheet et al. 1998; Goo et al. 2000),
albeit sometimes mildly (Patz et al. 1993), and this
has caused false-positive results in the detection of
malignant disease (Bakheet et al.1998).As with previ-
ous radiopharmaceuticals, specificity is poor, show-
ing little differentiation with malignancy (Goo et al.
2000). In comparison with 67Ga, 18FDG seems more
sensitive for the detection of opportunistic infections
other than tuberculosis in the presence of anti-tuber-
culous therapy (Goswami et al. 2000).
26.2.7
Alteration of Normal Physiology
Alteration of normal physiology, as a result of tuber-
culous infection, can be demonstrated using several
radiopharmaceuticals.
Changes to the distribution of pulmonary blood
flow, consequential to altered ventilation, can be
documented using 99Tc ffi macroaggregated albumin.
Typically, a patchy distribution of reduced accumula-
tion is obtained.
When compared with X-ray findings, previous
tuberculosis may lead to disproportionately large
defects on lung perfusion scintigraphy (Fogelman
et al. 1994). The same perfusion pattern would also
be obtained in pulmonary embolism, however. This
low specificity can be improved by imaging the
ventilation pattern as well. A concurrent pattern of
hypoventilation, indicated as areas of reduced tracer
accumulation, suggests a primary ventilation defect
causing a perfusion abnormality rather than an
embolism. In the presence of corroborative clinical
information, such a concurrence can implicate an
infective process. Regional patterns of hypoventila-
tion are readily demonstrable using radioactive gases
 416
a b
m
Fig. 26.1a,b. a Posterior lung perfusion using 99Tc macroaggregated albumin. The image shows multiple large areas of hypo-
perfusion in the right lung of a patient with pulmonary tuberculosis. These areas correspond with fibrotic changes seen on
the chest X-ray (b)
Fig. 26.2. Posterior lung perfusion, undertaken 22 months later,
on the patient presented in Fig. 26.1. The image shows a worsen-
ing in the perfusion of both lungs, with disease progression
or radio-aerosols. Of all the options, 81Krm is the best,
allowing ventilation images to be obtained during
equilibrium breathing concurrently with the perfu-
sion imaging and in the same imaging orientations.
The specificity of the full ventilation-perfusion test,
using 81 Kr m for differentiating between airways dis-
ease and pulmonary embolism, is extremely high.
D. Hamilton and J. Al-Nabulsi
The alternative gas, xenon in the form of 127Xe or
133Xe, does not produce as good an image as 81 Kr m
and suffers from further degradation of image quality
because of scatter from the perfusion phase, if this is
performed concurrently. This gas requires the patient
to perform breathing maneuvers during imaging to
obtain a ventilation image. These restrict the ventila-
tion views, usually to only the posterior orientation.
Radio-aerosols allow multiple images of ventilation
to be obtained without imaging maneuvers having to
be performed. They are, however, labeled with 99Tc m
which, being the same radionuclide as is used in
the perfusion phase, precludes concurrent imaging.
Sequential imaging, in which the perfusion phase is
undertaken approximately 2 h after the radio-aerosol
study, is usual.
26.3
Soft Tissue Tuberculosis
26.3.1
67Gallium Citrate (67 Ga)
67Ga is presently the best radiopharmaceutical for
imaging mycobacterial infections in soft tissue
(McAfee 1996). The technique is very sensitive (Sarkar
and Ravikrishman 1978; Sarkar et al. 1979; Lin and
Hsieh 1999; Yang et al. 1992) and has proven useful
 Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
a
Fig. 26.3a,b. Posterior lung ventilation using 81Kr"', undertaken at the same time as (a) the image in Fig. 26.1a and (b) the image
in Fig. 26.2. The images show multiple large areas of hypoventilation corresponding with the areas of hypoperfusion previously
described
in the identification of distant sites of tuberculous
involvement (Kattapuram et al. 1979; Moody and
Delbeke 1992), presenting as a single focus or dif-
fuse area of increased accumulation, or as multiple
foci arranged in a random distribution. The progres-
sion of the disease can be followed by identifying an
enlargement of a focus or an increase in number of
foci (Shih et al. 1986; Schmidt and Rebarber 1994;
Lin and Hsieh 1999). 67Ga scintigraphy is also useful
in assessing the response to therapy (Pettengell et al.
1990; Moody and Delbeke 1992).
A number of cases of abnormal 67Ga accumulation
occurring at a single site in the abdomen have been
reported (Lin and Hsieh 1999). These have included a
focus in the intestine (Yang et al. 1992; Lin and Hsieh
1999), two foci in the liver (Ohta et al. 1996), and three
foci in the spleen (Kao et al. 1996). Diffusely increased
tracer accumulation throughout the abdomen has
been shown in tuberculous peritonitis (LaManna et al.
1984; Sumi et al. 1999). Decreased hepatic accumula-
tion is also associated with this. The scintigraphic
findings of tuberculous peritonitis are not specific;
many intra-abdominal processes present with simi-
lar features. Nevertheless, if 67Ga scintigraphy shows
diffuse abdominal uptake with decreased hepatic
accumulation, tuberculous peritonitis should be con-
sidered in the differential diagnosis (Sumi et al. 1999).
Other examples of diffuse areas of abnormality at a
single site have included infection in the peritoneum at
the base of the pelvis detected on whole body imaging
417
as diffuse abdominal uptake and linear uptake along
the bottom of the pelvic cavity (Hashimoto et al. 2000)
and tuberculous enteritis (Pettengell et al. 1990).
After 24 h, the fecal route predominates, thus
making the interpretation of uptake in the abdominal
region somewhat difficult. In spite of this drawback,
the possibility of involvement of the GI tract by infec-
tion or tumor may be suspected if excessive GI activ-
ity is seen in the early scans (e.g., in a 24 h image), or
if the intensity of GI uptake on the 48-72 h image is
equal to or greater than the liver activity (Beckerman
and Bitran 1988). Bowel cleansing and changes in the
location and/or configuration of the intestinal activity
from one scan to another have been used for distinc-
tion of physiologic bowel excretion from pathologic
changes (Beckerman and Bitran 1988; Yang et al.1992;
Lin and Hsieh 1999). However, uptake due to diffuse
inflammation or due to neoplastic involvement of the
intestinal wall may change location and configuration
in two consecutive scans, thus diminishing the reliabil-
ity of such a "change shape/location" sign (Beckerman
and Bitran 1988).
Asymmetric 67Ga accumulation in lymph nodes
or salivary glands is an unusual finding and suggests
the need for more aggressive diagnostic investigation
(Kattapuram et al. 1979; Bihl and Maier 1987; Moody
and Delbeke 1992). Disseminated tuberculosis pres-
ents as abnormal 67Ga accumulation in numerous
organs (Moody and Delbeke 1992), for instance, in
the kidney and peritoneum (Sarkar and Ravikrish-
418
man 1978; Sarkar et al. 1979). Lymph node accumu-
lation has been reported in the mediastinum, left
lung hilus and retroperitoneum (Prat et al. 1991); in
esophageal and supraclavicular regions (Young et al.
1996); and in tuberculous salpingitis, endometriosis,
and both kidneys (Leventhal et al.1981).
The differential diagnosis of increased 67Ga uptake
includes both infective/inflammatory and malignant
processes, and 67Ga alone is thus not specific enough
to provide pathognomonic signs for tuberculosis
(Bihl and Maier 1987; Dhekne et al. 1987; Gomez et
al. 1996). In tuberculosis associated with AIDS, it is
reported that sequential delayed 20ln and 67Ga scin-
tigraphy allows differentiation between infection and
malignancy because, in infection, there is clearance
of 20lTI on delayed images (Lee et al. 1994). This find-
ing has been disputed, however. In a patient with HIV
and tuberculous lymphadenopathy, both delayed
20lTI and 67Ga images showed increased uptake.
This suggests that caution must be exercised when
differentiating between malignancy and infection/
inflammation in HIV-positive patients using delayed
20ln and 67Ga scans, and that careful microbiological
and clinical investigation is still necessary to exclude
AIDS-related mycobacterial infections (Gomez et al.
1996).
26.3.2
Leukocytes
Although increased accumulation is seen on both
lllIn leukocyte and 67Ga images (Schmidt and Rebar-
ber 1994), generally leukocytes are less sensitive for
the detection of tuberculosis (Brophey et al. 1995b).
In tuberculosis enteritis there is a slight, statistically
significant, superiority of67Ga over leukocyte imaging,
partly because it does not rely on vigorous leukocyte
migration to the area of disease for its success. There
is positive scintigraphy with leukocytes despite pro-
longed treatment, indicating that leukocyte recruit-
ment may continue for several weeks despite optimum
anti-tuberculous treatment. However, it is not certain
that serial scans are of benefit in the assessment of
treatment (Pettengell et al. 1990).
Mixed results have been obtained with leukocytes
labeled with 99Tc ffi • On the positive side, leukocyte
scintigraphy showed increased ileocecal accumula-
tion in disseminated tuberculosis (Prat et al. 1991).
On the negative side, an absence of uptake occurred
in a paravertebral abscess (Hovi et al. 1993) and in
a number of patients with extensive tuberculous
lesions (Brophey et al. 1995b).
D. Hamilton and J. Al-Nabulsi
26.3.3
99'fechnetium m(99'fem) Radiopharmaceuticals
Methoxyisobutylisonitrile (MIBI) has been shown
to accumulate in a tuberculous lesion in the female
breast and indicates that tuberculosis must be added
to the differential diagnosis of breast uptake of MIBI
(Ohta et al. 1998).
In comparison with 67Ga, (V) DMSA will reveal
an inflammatory lesion as a wider area of less
well-defined and lower uptake because it detects
the relatively hypervascular and hypermetabolic
tissue surrounding such a lesion. Thus, intense 67Ga
accumulation in association with moderate, poorly
defined, and wider (V) DMSA uptake suggests a
marked inflammatory reaction in a lesion. Such a
pattern was shown in solitary muscular involvement
by tuberculosis. Otherwise, in a large proportion of
malignant soft tissue tumors, 67Ga accumulation is
relatively fainter than that of (V) DMSA (Kobayashi
et al. 1995).
As in pulmonary tuberculosis, HIG does not seem
appropriate in this disease. Neither increased accu-
mulation nor cold lesions were revealed in a patient
with spondylitis and a paravertebral abscess (Hovi et
al. 1993).
Hydroxymethylene diphosphonate, generally
regarded as purely a skeletal imaging pharmaceuti-
cal, can reveal tuberculous lesions in soft tissue. It
showed a large and well-delineated area of increased
uptake in extraskeletal tissue corresponding to a
tuberculous abscess (Tamgac et al. 1995). Pyrophos-
phate is used as a myocardial radiopharmaceutical for
evaluation of myocardial infarcts. It also seems to be
useful for the detection of early tuberculous pericar-
ditis and showed abnormal uptake along the lateral
portion of the left ventricle in a case of tuberculous
pericarditis. Although no calcification was proven on
biopsy, the abnormal uptake of pyrophosphate along
the left ventricle was considered to be caused by early
calcifying inflammatory changes in the pericardium
(Ishino et al. 1992).
26.3.4
18Fluorodeoxyglucose Positron Emission
Tomography
As in pulmonary disease, moderate or intense
accumulation of 18FDG occurs in active soft tissue
tuberculosis. It presents as focal uptake at single
(Bakheet et al. 2000) or multiple sites (Bakheet et al.
1998) or as diffuse accumulation (Braga et al. 2001).
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
Moderate, diffuse, and homogeneous accumulation
was documented as corresponding to tuberculous
involvement of the spinal cord (Braga et al. 2001). In
one patient, PET imaging showed very intense focal
18PDG accumulation in the breast, which responded
to anti-tuberculous drugs and was regarded as tuber-
culous mastitis (Bakheet et al. 2000). Intense multi-
focal 18PDG accumulation was documented in two
patients with widespread tuberculous lymphadeni-
tis (Bakheet et al. 1998). However, specificity is poor
and PET imaging alone cannot differentiate acute or
chronic infection or inflammation from malignancy
(Bakheet et al. 1998,2000; Cook et al. 1996).
26.3.5
Alteration of Normal Physiology
Alteration of normal physiology can be detected
during active infection or in inactive disease as a
result of altered anatomy.
99Tc ffi pertechnetate sialogram studies in two
patients with unilateral tuberculosis in parotid and
submandibular glands showed normal radionuclide
uptake. One patient showed blocked excretion in the
distended parotid after administration of a gustatory
stimulus. The other patient demonstrated a normal
excretion phase (Bihl and Maier 1987).
Superior vena cava (SVC) obstruction developed
from progressive enlargement of mediastinal tuber-
culosis, which was demonstrated on consecutive 67Ga
scans. An initial 99Tc ffi pertechnetate SV cavagram
showed no definite evidence of obstruction and an
initial sulfur colloid scintigraphy showed normal
tracer distribution throughout the liver. A follow-up
radionuclide SV cavagram, 4 months after the initial
study, demonstrated interruption of flow in the SVC
and development of collateral circulation in the right
anterior chest wall. A follow-up 99Tc ffi sulfur colloid
scintigraphy showed two hot spots in the liver (Shih
et al. 1986) 3 weeks later.
An abdominal cerebrospinal fluid (CSP) pseudo-
cyst associated with tuberculous peritonitis involving
the peritoneal end of a ventriculoperitoneal shunt
was detected using intrathecal administration of III In
diethylenetriamine pentaacetic acid (DTPA) to assess
the patency of the shunt. A study of the shunt revealed
tracer flowing freely through the shunt but a loculated
collection of activity was seen at the peritoneal end
of the shunt catheter. Tuberculous peritonitis was
considered to be the cause of the failure of CSP resorp-
tion from the peritoneal cavity and the formation of a
thick-walled cystic lesion (Suga et al. 2000).
419
Alteration of normal physiology in the liver and
spleen can be revealed using 99Tc ffi sulfur colloid.
Photopenic areas demonstrate an absence of func-
tion, such as multiple discrete defects in the spleen
(Kattapuram et al. 1979) or as diffuse hepatocellular
dysfunction with marked splenomegaly (Winzelberg
1981). Two patients with pseudotumor of the liver
showed 67Ga uptake, which corresponded to a rela-
tively larger cold defect seen on a simultaneously
obtained liver-spleen scan (Dhekne et al. 1987).
Three focal areas increased 67Ga accumulation at the
left upper abdominal quadrant showed three photon-
deficient areas on sulfur colloid scintigraphy in the
spleen corresponding to the 67Ga distribution (Kao
et al. 1996).
The effect of tuberculous thyroiditis can produce
the same 131Iodine (1 31 1) distribution as occurs in
malignancy. In a patient with this disease, scintig-
raphy initially indicated a reduced 131 1 uptake with
enlarged gland and well-defined area of absent
accumulation. Both uptake and distribution showed
significant improvement 3 months following initia-
tion of anti-tuberculous therapy. This improvement
continued at 6 months, by which time the thyroid
uptake had returned to within normal limits (Wang
et al. 1972).
26.4
Renal Tuberculosis
Renal tuberculosis has been particularly extensively
studied since the 1960s, with almost all of the early
contribution originating from Russia. The function
of the kidney can be investigated from three perspec-
tives: static parenchymal imaging, dynamic renogra-
phy, and absolute function.
Scarring is readily demonstrable by parenchymal
imaging using DMSA. Lesions appear as areas of
reduced tracer accumulation, indicating non-func-
tioning parenchyma. Small scars are sometimes only
detectable using SPET.
The effect of such scarring on perfusion, function,
and excretion throughout the kidneys can be revealed
using renography with hippuran, DTPA, or betiatide
MAG3. Due to its dynamic nature, this imaging is
undertaken from one orientation only, usually poste-
rior.Abnormalities will present as reduced perfusion,
function, or excretion, but activity-time curves may
be needed to accurately assess any such deteriora-
tion. Because progressive fibrosis may occur, follow-
up renograms are recommended initially at about
420 D. Hamilton and J. Al-Nabulsi

1 month and then at 3-6 month intervals depending

on the site of tuberculosis in the renal tract (Sweny
et al. 1989). Confirmation of the non-functioning of
a kidney and the compensatory hypertrophy of the
contralateral kidney with dilated upper ureter was
shown on a DTPA renogram (Taher 1998).
The absolute effect of the disease on renal func-
tion can be assessed by estimating glomerular fil-
tration rate (GFR), using SlCr edetic acid or 99Tc rn
DTPA. Abnormalities will be revealed by a reduction L
(in mllmin) compared with normal ranges. The test
is one of overall renal function and cannot provide
an indication of whether one kidney is more severely
affected than the other. GFR is the best single param-
eter for the assessment of absolute renal function
(Brochner-Mortensen 1978; Groth 1984; Picciotto et
al. 1992; Piepsz et al. 1994). It is measured in mllmin
corrected for body surface area, using blood samples
usually taken between 2 hand 4 h after adminis- Fig. 26.4. 99Tc m diethylenetriamine pentaacetic acid scintigra-
phy demonstrating non-functioning left kidney
tration. Its sensitivity can be appreciated by the
example of a 65-year-old man who presented with
tuberculosis of the urinary tract. The initial GFR was
66 mllmin/1.73 m 2 compared with a normal range of 26.5.2
73-116. This had reduced to 29/min/1.73 m 2 com- 99'fechnetium rn (99'fc rn ) Radiopharmaceuticals
pared with a normal range of 71-112 with disease
progression, 16 months later. Autologous leukocytes labeled with 99Tcrn can be used
to diagnose tuberculous meningitis (Kim et al.1995).
Data on perfusion imaging in tuberculous meningitis
are lacking but this has recently been studied using
26.5 ethylene cystine dimer (ECD). The ECD studies were
eNS Tuberculosis more frequently abnormal compared with computed
tomography (CT) but did not correlate with the
Meningitis caused by tuberculosis has been studied, severity or stage of meningitis, or clinical outcome
most recently, using imaging techniques which are very (Misra et al. 2000).
different from the older non-imaging partition test. In
contrast to structural imaging techniques, functional
methods can detect brain alterations before morpho- 26.5.3
logical damage occurs (Villringer et al. 1995). If the 201Thallium Chloride (l01TI)
intracranial lesions are large and relatively superficially
located, planar imaging might be adequate to make the Because brain involvement due to opportunistic
diagnosis, but for most lesions the use of SPET is neces- infections or primary CNS lymphoma is a common
sary (Kim et al. 1995; Lee et al. 1999). complication of HIV infection and early diagnosis is
important for therapy strategies and prognosis, much
work has been concentrated in this area (Villringer
26.5.1 et al. 1995).
67Gallium Citrate (67Ga) In patients with AIDS, 201Tl can be used to dif-
ferentiate lymphoma from tuberculosis in lesions
Although useful for most presentations of tuberculo- identified on CT/magnetic resonance imaging (MRI)
sis, 67 Ga scintigraphy is of limited use in the detection (Lorberboym et al. 1998; Lee et al. 1999). This radio-
of tuberculous meningitis. It is postulated that the pharmaceutical concentrates in many tumors, but
blood-brain barrier mechanism of the central ner- more importantly, its concentration in infectious and
vous system (CNS) plays a role in this low detection inflammatory foci is relatively low (Lee et al. 1999).
rate (Lin and Hsieh 1999). The absence of accumulation on images, early after
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis 421

administration, at the site of a CT/MRI abnormal- pharmaceutical but DTPA has the advantage of easier
ity excludes the diagnosis of lymphoma with a high availability (von Wenzel et al. 1989).
degree of confidence (Lorberboym et al.1998). In the Following intravenous or oral administration
presence of abnormal early 20lTl uptake, it is essen- (Wiggelinkhuizen and Mann 1980), the ratio of the
tial to perform delayed imaging, at approximately activity in the blood to that in the CSF is measured at
3 h after injection, to distinguish tuberculosis from 48 h for 82Br (Girgis et al. 1990) and at 24 h for DTPA
lymphoma (Lorberboym et al. 1998; Lee et al. 1999). (von Wenzel et al. 1989). In normal subjects, the con-
This is because an infectious process can sometimes centration in plasma is 2.5-3.5 times that in lumbar
show early accumulation, which then clears relatively CSF and 3-5 times that in ventricular CSF. The con-
quickly in comparison with neoplastic lesions (Lee centration gradient between plasma and CSF appears
et al. 1999). Although most reports have referred to to be maintained by an active transport system in the
extracranial lesions to differentiate between tuber- choroid plexus, and proximity to the plexus probably
culosis and KS or infarct in patients with AIDS, accounts for the differences in the levels in lumbar
sequential delayed 20lTl and 67Ga imaging can be and ventricular CSF. The radiopharmaceuticals cross
undertaken. Both lymphomas and infection are 67Ga the blood-brain barrier to a greater extent in patients
avid. Lymphoma would show a positive 20lTl pattern with meningitis, more so in tuberculous than in viral
whereas tuberculosis would show an absence of 20lTI (Wiggelinkhuizen and Mann 1980). Both tracers thus
accumulation. KS and infarct show an absence of 67Ga yield decreased serum to CSF concentration ratios in
accumulation. An absence of 201Tl uptake would sug- patients with tuberculous meningitis (von Wenzel et
gest infarct. A pulmonary lesion with a 20lTI positive al. 1989; Girgis et al. 1990). A low ratio, below 1.6, is
and 67Ga negative pattern would indicate KS. Since strong support for a diagnosis of tuberculous men-
KS is extremely rare in the CNS, such a pattern would ingitis and one above 1.6 is against the diagnosis
be unexpected. In practice, therefore, 67Ga scanning (Wiggelinkhuizen and Mann 1980; Daniel 1987). The
is probably unnecessary if brain lesions are 20lTl avid depression of the ratio is independent of the stage
(Lee et al.1999). of the untreated disease and the prognosis, however
(Wiggelinkhuizen and Mann 1980).
The sensitivity of the test approximates to 90%
26.5.4 (Wiggelinkhuizen and Mann 1980; Daniel 1987). The
18Fluorodeoxyglucose Positron Emission results for DTPA are not quite as good as 82Br. The
Tomography accuracy for distinguishing among tuberculous and
viral and septic meningitis, was found to be 86.9%
Primary CNS lymphomas are known to be meta- using DTPA compared with 90.9% using 82Br (von
bolically very active tumors, even if the patients Wenzel et al. 1989). It is possible that the number of
are receiving steroids. This high metabolism of CNS false-negative results could be kept to a minimum
lymphomas was also found in HIV-infected patients by performing the test as soon as anti-tuberculous
(Villringer et al.1995). Many infectious foci are 18FDG therapy is started and by avoiding, if possible, the
avid, which is useful in detecting both infections and use of corticosteroids until the test is complete
malignancy, but it is of limited value in distinguish- (Wiggelinkhuizen and Mann 1980). These reintegrate
ing between them (O'Doherty et al. 1997), although the blood-brain barrier and cause very dramatic
differentiation has been reported in one patient with changes to the penetration of compounds into the
tuberculoma (Villringer et al. 1995). spinal fluid (Buchanan 1981). The test has also been
reported to be an indicator of the development of
permanent sequelae (Girgis et al.1990).
26.5.5 A low ratio, of 1.6 or less, is not specific for tubercu-
Partition Test lous meningitis (Wiggelinkhuizen and Mann 1980).
Although the partition test can differentiate among
The partition test is particularly useful in distinguish- those patients with encephalitis and those with men-
ing early tuberculous meningitis from viral meningitis ingitis (Buchanan 1981), false-positive results occur
(Wiggelinkhuizen and Mann 1980). The partition of in both pyogenic (Wiggelinkhuizen and Mann 1980)
certain ions between serum and CSF after a loading and aseptic meningitis and, although the specific-
dose reflects the integrity of the blood-brain barrier ity approximates to 90%, concern remains that the
(Daniel 1987). Both ammonium 82Bromide (82 Br) and test may not have the very high specificity that is
99Tcffi DTPA can be used. 82Br is the traditional radio- necessary for detecting this disease (Daniel 1987).
422
In particular, the differentiation between tuberculous
meningitis and partially treated bacterial meningitis
remains elusive (Buchanan 1981). Little is known
about the duration of the depression of the ratio in
treated tuberculous meningitis, but specific anti-
tuberculous therapy does not affect the ratio in the
first few weeks (Wiggelinkhuizen and Mann 1980).
The rise in serum to CSF ratio following therapy
could be used, however, to distinguish bacterial from
tuberculous meningitis (Girgis et al. 1990). There is
no relationship between the ratio and the severity
of the meningeal inflammatory response (Wig-
gelinkhuizen and Mann 1980).
26.6
Bone and Joint Tuberculosis
Radioisotope imaging is usually more sensitive than
plain radiography in the detection of bone and joint
tuberculosis (Dickinson et al. 1996), as an uncertain
time interval can exist between the onset ofthe clinical
presentation and the radiological diagnosis (Everaert
et al. 1997). There is a possibility of false negative
results however and, if there is a strong clinical sus-
picion of skeletal tuberculosis, further investigations
should be performed even if the nuclear medicine
studies are normal (Pui et al. 1986). Nuclear medicine
techniques may be more useful when the diagnosis
has been made and when an assessment of the extent
of involvement is required, particularly to detect
asymptomatic lesions (Dickinson et al. 1996), many
of which are unsuspected clinically (Rust et aI.1981).
A whole body survey in a patient with extra-pulmo-
nary tuberculosis is important because multiple
lesions are not uncommon. Both phosphonate bone
and 67 Ga scans facilitate such whole body screening
to detect multiple sites of involvement (Muradali et
al. 1993; Lin et al.1998; Lin and Hsieh 1999) and they
reveal sites for further detailed evaluation by other
modalities (Lin et al. 1998; Lin and Hsieh 1999).
26.6.1
"Tem Polyphosphonate
Tuberculous infection of bone has been investigated
extensively since the 1960s, when radioactive stron-
tium was used. The present mainstay of investigation
is the polyphosphonate scan, which was introduced
in the early 1970s (Fanning et al. 1974). Bone scintig-
raphy is an excellent test for detecting disseminated
D. Hamilton and J. AI-Nabulsi
tuberculosis of the skeleton (Boumpas et al. 1987; Lin
et al. 1998; Dickinson et al. 1996), both in making the
diagnosis and in localizing further areas of increased
uptake (Dickinson et al. 1996). It can be used as a
relatively inexpensive investigation to decide which
patients are likely to benefit from MRI, particularly at
the stage when X-rays are normal (Desai 1994).
The investigation comprises three phases: perfu-
sion, blood pool, and late. Although infections are
usually apparent on all three, in indolent infections
like tuberculosis, often only the third phase is posi-
tive. The lesions appear either as a single focus or
as multiple foci in a random pattern. Pulmonary
involvement need not be present and involvement is
not limited to the axial skeleton (Muradali et al.1993).
The lesions usually show increased radionuclide
accumulation, reflecting increased osteoblastic activ-
ity. Less commonly, they show decreased uptake due
to interference with the blood supply by inflamma-
tory products (Pui et al. 1986) or fibrous replacement
of the osteoid upon healing, which results in lack of
bone imaging agent uptake and a consequential area
of photopenia (Rust et al.1981).
Single focus examples include abnormality at
tibia and pubic bone (Abdelwahab et al. 1991), rib
(Muradali et al. 1993), and hip joint, more intense at
the periphery, affecting both the femoral head and
acetabulum (Greenspan and Stadainik 1995). Many
examples ofmultiple,scattered,focal areas ofaccumu-
lation are reported (Boumpas et al. 1987; Dickinson
et al. 1996). Particular sites have included: ribs, tho-
racic and lumbar spine, shoulder girdle (Rust et al.
1981); ribs, spine, appendicular skeleton (Nocera et
al. 1983); sternum, clavicle, ribs, thoracic and lumbar
spine, iliac crest, tibia, sacroiliac joints, acetabulum
(Muradali et al. 1993); elbow, lumbar and thoracic
spine, shoulder, maxilla, temporomandibular joint,
as well as faint sternal uptake (Hardoff et al. 1995).
Bone scintigraphy is a sensitive technique for
detecting active skeletal tuberculosis (Lisbona et al.
1993; Desai 1994). False normal bone imaging can
occur, however, particularly in low-grade, indolent
(Dickinson et al. 1996; Pui et al. 1986), or severely
destructive osteomyelitis (Pui et al. 1986; Lisbona
et al. 1993). The results of bone scintigraphy must
be interpreted in the context of the clinical situa-
tion; a negative result does not exclude tuberculous
bone infection (Muradali et al. 1993). The technique
is reasonably sensitive in spinal tuberculosis (Desai
1994), but false negative results do seem to be found
more frequently in this presentation. The reasons
for this are unknown (Lisbona et al. 1993) but are
likely to be due to the indolent nature of the infection
 Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis 423

Fig.26.5a-c. 99Tc rn methylene diphosphonate bone scintigra-

phy in a patient with multi-focal tuberculous osteomyelitis.
a Posterior blood pool showing a focus of slightly increased
accumulation in the region of the left first rib. b Anterior skull
showing focus of increased accumulation in the right frontal
bone. c Right anterior oblique head and neck showing foci of
increased accumulation at C3 and the left first rib

Fig. 26.6. Lateral cervical spine X-ray, correspond-

c ing to Fig. 4c, showing destruction of C3
 424 D. Hamilton and J. AI-Nabulsi

Fig. 26.7. A case of spinal tuberculosis involving thoracic vertebrae 8-11. a Late
phase 99Tc rn bone scintigraphy showing a complete absence of accumulation in
T9 with increased accumulation in T8 and TIO and slightly increased uptake in
TIL Loss of height at 11 is apparent. b Blood pool image showing only slightly
reduced activity in the region of T9 and only slightly increased activity in the
region of T8 and TIO-l1. c Lateral magnetic resonance image with contrast,
showing destruction at the T9level that is enhancing. Also shown is a paraspinal
c abscess not revealed by the bone scintigraphy

(Pui et al. 1986). Bone scintigraphy is usually more
sensitive in detecting tuberculous osteomyelitis and
shows abnormality earlier than conventional radio-
graphs (Nocera et al. 1983; Pui et al. 1986; Lisbona
et al. 1993; Muradali et al. 1993). It is not specific for
tuberculous lesions (Lin et al. 1998) and will demon-
strate non-specific findings, which may be seen in
other conditions (Sharif et al. 1989; Greenspan and
Stadainik 1995). The presentation may mimic malig-
nancy (Rust et al. 1981; Nocera et al. 1983; Boumpas
et al. 1987; Muradali et al. 1993; Hardoff et al. 1995)
because multiple asymmetrical areas of increased
uptake are often regarded as being virtually diagnos-
tic of metastases (Dickinson et al. 1996). Experience
indicates, however, that any confusing bone lesion
in an "at risk" patient may prove to be tuberculous
(Abdelwahab et al. 1991) and may create confusion
about the diagnosis (Nocera et al. 1983).
 Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis 425

Fig. 26.8. a Diffusely increased accumulation at the distal half of the

right femur on 99Tc m bone scintigraphy. Two months after anti-tuber-
culous therapy, b the accumulation in the femur has reduced but c
there is increased uptake at the distal end of the left humerus

b c

Serial bone scintigraphy can be used to assess 26.6.2

the response to treatment (Dickinson et al. 1996).
It can remain positive for a prolonged period after
initiation of anti-tuberculous therapy (Sarkar
et al. 1979), but generally returns to normal in
3-6 months (Rust et al. 1981; Nocera et al. 1983;
Boumpas et al. 1987). The assessment can encom-
pass the effect of interrupted treatment (Rust et
al. 1981).
67Gallium Citrate (67Ga)
67 Ga scintigraphy has an advantage over 99Tcffi phos-
phonate bone scintigraphy in that it detects soft
tissue as well as bone infections (Pui et al. 1986; Lis-
bona et al. 1993; Hardoff et al. 1995; Lin et al. 1998).
In the spine, in particular, it highlights distant septic
foci in the soft tissues or skeleton, which are more
 426
c and the left first rib
amenable to biopsy than the spine itself (Lisbona et
al. 1993). Skeletal uptake of 67Ga is related to both
reticuloendothelial activity and to bone turnover.
Lesions appear as focal areas of increased accumu-
lation, and may be solitary or appear as multiple
foci in a random distribution. Although the pattern
is not characteristic only of tuberculosis, it must be
considered when this pattern is observed (Hardoff et
al.1995). Pulmonary involvement need not be present
(Lisbona et al. 1993).
The detection sensitivity of 67Ga scintigraphy is
high for skeletal tuberculosis (Lin et al. 1998; Lin
and Hsieh 1999). It seems more sensitive than 99Tc ffi
phosphonate bone scintigraphy (Lisbona et al. 1993).
It also seems more sensitive in spinal lesions (Sarkar
D. Hamilton and J. Al-Nabulsi
Fig. 26.9a-c. 99Tcm methylene diphosphonate bone scintigraphy
undertaken 11 months after the scintigraphy presented in
Fig. 26.5, showing disease regression. a Anterior blood pool
showing no evidence of the focus of slightly increased accu-
mulation in the region of the left first rib. b Anterior skull
showing no evidence of the focus of increased accumulation
in the right frontal bone. c Right lateral head and neck show-
ing no evidence of the foci of increased accumulation at C3
and Ravikrishman 1978; Sarkar et al. 1979; Lisbona
et al. 1993; Hardoff et al. 1995; Everaert et al. 1997;
Lin and Hsieh 1999). As with 99Tc ffi phosphonate
bone scintigraphy, false-negative results do occur
in spinal tuberculosis and with a greater frequency
than in other parts of the skeleton (Pui et al. 1986;
Lin et al. 1998; Lin and Hsieh 1999) and 67Ga citrate
images are insensitive in low-grade, indolent, or
severely destructive osteomyelitis (Pui et al. 1986;
Lin et al. 1998). Sequential 67Ga scanning offers the
opportunity to monitor response to therapy (Lisbona
et al. 1993), particularly in the spine (Everaert et al.
1997); since abnormally increased 67Ga accumulation
resolves within months of initiation of anti-tubercu-
lous therapy (Pui et al. 1986).
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
67Ga scintigraphy is frequently performed in AIDS
patients with fever of unknown origin to identify a
site for a more invasive investigation. In these patients,
however, disseminated mycobacterial infection also
alters the usual 67Ga distribution in the bone marrow.
This is assessed by measuring the skull uptake, chosen
because of the absence of overlapping structures,
which may be a sign of peripheral marrow activa-
tion and could reflect the presence of expanded bone
marrow. Abnormal skull uptake appears to be a sensi-
tive and specific indicator of disseminated tuberculous
infection in HIV-infected patients (Gomez et al.1995).
67Ga scintigraphy is more sensitive than X-rays
for the detection of active tuberculous osteomyelitis
(Lisbona et al. 1993). Specificity of 67Ga scintigraphy
is poor, the distribution pattern being characteristic
not only of tuberculosis. It is unable to differentiate
between tuberculous and pyogenic osteomyelitis
(Lisbona et al. 1993), an ' ',creased uptake occurs in
malignancy (Hardoff et, '995).
Combined 99Tcffi phos) mate bone and 67Ga scin-
tigraphy can improve th{ ecificity for evaluation of
skeletal tuberculosis, 67C..l scintigraphy ascribing a
septic origin to the less specific increased uptake seen
on 99Tc ffi phosphonate bone scintigraphy (Lisbona et
al. 1993; Dickinson et al. 1996). However, there have
been few reports where both radiopharmaceuticals
are used (Lin et al. 1998). Both 99Tcffi phosphonate
bone and 67Ga scintigraphy generally show concurrent
accumulation in tuberculous osteomyelitis (Brophey
Fig. 26.10. 67Ga scintigraphy undertaken 6 weeks after the
99Tc m methylene diphosphonate bone scintigraphy shown in
Fig. 26.9. Right lateral showing slight residual increased accu-
mulation in the region of C3
427
et al. 1995a). However, the uptake is more intense with
67Ga, and its distribution extends beyond the 99Tcffi
phosphonate distribution, consistent with the soft
tissue component of the lesions (Hardoff et al. 1995).
A combination protocol improves sensitivity in com-
parison with single radiopharmaceutical investigation
(Lin et al. 1998).
Abnormal 67Ga accumulation can resolve within
months ofinitiation of anti-tuberculous therapy (Sarkar
et al. 1979), but significant uptake can persist. 67Ga scin-
tigraphy, 2 months after initiating anti-tuberculous
therapy, demonstrated abnormal residual uptake in
both bone and soft tissue (Boumpas et al. 1987).
20lTl is reported to accumulate avidly in tubercu-
lous osteomyelitis, but it is not clear whether this is on
early or delayed scintigraphy. Comparative scintigra-
phy showed marked localized hyperemia and intense
focal osteoblastic activity with 99Tcffi phosphonate, but
only a moderate degree of 67Ga uptake. The combined
scan appearance favored a primary bone neoplasm;
chronic osteomyelitis being less likely since 20lTl avid-
ity in the lesion was so much greater than that of 67Ga,
but a biopsy confirmed tuberculous osteomyelitis with
no evidence of malignancy (Mansberg et al.1997).
Other radiopharmaceuticals have been used, in com-
bination with 99Tcffi phosphonate bone scintigraphy. In
two patients with tuberculous spondylitis, 99Tcffi phos-
phonate bone scintigraphy gave positive results. How-
ever, scintigraphy with HIG and monoclonal antibodies
gave false-negative findings, as did the blood pool phase
of the phosphonate scintigraphy (Sciuk et al. 1991). In
tuberculous lesions, l1lIn chloride bone marrow scin-
tigraphy shows areas of decreased accumulation cor-
responding to areas of increased 99Tcffi phosphonate
bone uptake. It is postulated that the photopenic areas
represent bone marrow replacement by the tuberculous
inflammatory process (Nocera et al.1983).
26.6.3
1111n Leukocytes
In tuberculous osteomyelitis, I11In labeled leuko-
cytes show increased accumulation, concurrent with
increased 67Ga uptake (Schmidt and Rebarber 1994).
26.6.4
1sFluorodeoxyglucose Positron Emission
Tomography
18 FD G
is useful for assessing the process of inflam-
matory activity in tuberculous osteomyelitis by
428
quantifying the pathologic increase in the glucose
metabolism of the inflammatory processes (Kalicke
et al. 2000; Schmitz et al. 2000). Compared with 99Tc ffi
phosphonate bone scintigraphy, soft tissue changes
can also be observed, including lung involvement.
The reason for this is the high spatial resolution com-
pared with single photon scintigraphy, which enables
differentiation between osteomyelitis and infection
of surrounding soft tissue (Schmitz et al. 2000).
There is significantly increased 18FDG uptake in
malignant tissue, as a result of which differentiation
between infection and metastases is difficult. How-
ever, in fractures and pseudarthroses only very low
18FDG uptake is observed, which means that it is pos-
sible to differentiate these from infection (Kalicke
et al. 2000). Although it is thought to act mainly
as a tumor-imaging agent (O'Doherty et al. 1997)
because malignant cells demonstrate higher glucose
metabolic activity than benign lesions do (Goo et al.
2000), 18FDG PET also shows the increased glucose
metabolism of activated inflammatory cells, such as
leukocytes, granulocytes and macrophages (Bakheet
et al. 1998; Kalicke et al. 2000; Schmitz et al. 2000).
Therefore, 18FDG uptake in tuberculosis is not unex-
pected (Bakheet et al.1998). There is overlap between
18FDG uptake of malignant lesions and some infec-
tious processes (Knight et al. 1996) mainly due to the
presence of macrophages, but early reports indicate
that 18FDG is a promising technique to differentiate
tumors from infections (O'Doherty et al.1997).
Early after surgical intervention, however, it is
not possible to differentiate between post-surgical
reactive changes and further infection because of
non-specific tracer uptake (Kalicke et al. 2000). The
progress of therapy can be clearly shown (Kalicke
et al. 2000). In comparison with 67Ga, it is likely that
the superior resolution of PET cameras and the bio-
kinetic distribution of 18FDG will provide enhanced
diagnostic capability, and the time saved by getting
a diagnostic result may be put to appropriate use for
the treatment or further investigation of the patient.
The other area that is of potential use is the absolute
quantification of uptake, which could be used in the
assessment of response to therapy at an early stage of
treatment (O'Doherty et al. 1997).
References
Abdel-Dayem HM et al (1994) Sequential thallium and gallium
chest imaging in AIDS patients with non-KS pulmonary
complications (abstract). Eur J Nucl Med 21:S77
D. Hamilton and J. Al-Nabulsi
Abdel-Dayem H et al (1995) Diffuse TI-201 uptake in the lungs.
Etiologic classification and pattern recognition. Clin Nucl
Med 20:164-172
Abdel-Dayem HM et al (1996) Evaluation of sequential thal-
lium and gallium scans of the chest in AIDS patients. J Nucl
Med 37:1662-1667
Abdel-Dayem HM et al (1997) Sites of tuberculous involve-
ment in patients with AIDS. Autopsy findings and evalua-
tion of gallium imaging. Clin Nucl Med 22:310-314
Abdelwahab IF et al (1991) Atypical skeletal tuberculosis mim-
icking neoplasm. Br J RadioI64:551-555
Bakheet SM et al (1998) F-18-FDG uptake in tuberculosis. Clin
Nucl Med 23:739-742
Bakheet SM et al (1999) Radioiodine uptake in inactive pulmo-
nary tuberculosis. Eur J Nucl Med 26:659-662
Bakheet SM et al (2000) F-18 FDG uptake in breast infection
and inflammation. Clin Nucl Med 25:100-103
Bekerman C, Bitran J (1988) Gallium-67 scanning in the clini-
cal evaluation of human immunodeficiency virus infection:
indications and limitations. Semin Nucl Med 8:273-276
Bihl H, Maier H (1987) Unilateral gallium-67 uptake in pri-
mary tuberculosis of the major salivary glands. Clin Nucl
Med 12:650-653
Boumpas DT et al (1987) Skeletal tuberculosis resembling
metastatic disease on bone scintigraphy (letter). J Nucl
Med 28:1507-1509
Braga FJHN et al (1998) 99mTc-gluco-heptonate scintigraphy
in lung tuberculosis patients: initial results (abstract). Eur
J Nucl Med 25:1125
Braga FJHN et al (2001) Neural tuberculosis detected by F-18
FDG positron emission tomography. Clin ~ucl Med 26:
706
Brochner-Mortensen J (1978) Routine methods and their reli-
ability for assessment of glomerular filtration rate in adults.
Dan Med Bull 25:181-202
Brophey M et al (1995a) The scintigraphic presentation of
Pott's disease. Clin Nucl Med 20:191-193
Brophey Met al (1995b) Prominent small bowel Ga-67 uptake
associated with Yersinial and tuberculous enterocolitis.
Clin Nucl Med 20:107-110
Buchanan N (1981) Radioactive bromide partition test in the
diagnosis of tuberculous meningitis (letter). J Pediatr 99:506
Burke T (1995) TI-201 uptake in pulmonary Mycobacterium
avium complex infection. Clin Nucl Med 20:835-836
Cook GJ, Fogelman I, Maisey MN (1996) Normal physiological
and benign pathological variants of 18-fluoro-2-deoxyglu-
cose positron-emission tomography scanning: potential
for error in interpretation (review). Semin Nucl Med 26:
308-314
Daniel TM (1987) New approaches to the rapid diagnosis of
tuberculous meningitis. J Infect Dis 155:599-602
Degirmenci B et al (1998) Technetium-99m-tetrofosmin
scintigraphy in pulmonary tuberculosis. J Nucl Med 39:
2116-2120
Desai SS (1994) Early diagnosis of spinal tuberculosis by MRI.
J Bone Joint Surg [Br] 76B:863-869
Dhekne RD et al (1987) Tuberculous pseudotumor of the liver.
Clin Nucl Med 12:816-819
Dickinson FL, Finlay DB, Belton IP (1996) Multifocal skeletal
tuberculosis: bone scan appearances. Nuc1 Med Commun
17:957-962
Everaert H et al (1997) Ga-67 uptake in a case of tuberculous
spondylitis. Clin Nucl Med 22:403-404
Radionuclides in Pulmonary and Extra-Pulmonary Tuberculosis
Fanning A, Dierich H, Lentle B (1974) Bone scanning with
technetium 99mTc polyphosphate in tuberculous osteomy-
elitis. Tubercle 55:227-230
Fineman DS et al (1989) Detection of abnormalities in febrile
AIDS patients with In-Ill-labeled leukocyte and Ga-67
scintigraphy. Radiology 170:677-680
Fogelman I,Maisey MN, Clarke SEM (1994) An atlas of clinical
nuclear medicine, second edition. Dunitz, London
Ganz WI, Serafini AN (1989) The diagnostic role of nuclear
medicine in the acquired immunodeficiency syndrome. J
Nucl Med 30:1935-1945
Girgis NI et al (1990) The use of the bromine partition test
in the diagnosis and prognosis of tuberculous meningitis.
East Afr Med J 67:404-406
Goldfarb CR et al (1995) Ga-67 scintigraphic appearances of
tuberculosis in the 'new' tuberculosis (abstract). J Nucl
Med 36:23P
Goldfarb CR et al (1997) Gallium scanning in the 'new' tuber-
culosis. Clin Nucl Med 22:470-474
Gomez MV et al (1995) Abnormal gallium-67 skull uptake:
a sign of peripheral marrow activation in HIV-positive
patients with disseminated mycobacterioses. J Nucl Med
36:2211-2213
Gomez MV et al (1996) Identification of AIDS-related tuber-
culosis with concordant gallium-67 and three-hour delayed
thallium-201 scintigraphy. Eur J Nucl Med 23:852-854
Goo JM et al (2000) Pulmonary tuberculoma evaluated by means
ofFDG PET: findings in 10 cases. Radiology 216:117-121
Goswami GK et al (2000) Discrepancy between Ga-67 citrate
and F-18 fluorodeoxyglucose positron emission tomo-
graphic scans in pulmonary infection. Clin Nucl Med 25:
490-491
Greenspan A, Stadainik RC (1995) Increased uptake around
the hip joint. Semin Nucl Med 3:283-286
Grieff M, Lisbona R (1991) Detection of miliary tuberculosis
by Ga-67 scintigraphy. Clin Nucl Med 16:910-912
Groth S (1984) Calculation of 51Cr-EDTA clearance in children
from the activity in one plasma sample by transformation
of the biexponential plasma time-activity curve into a
monoexponential with identical integral area below the
time-activity curve. Clin PhysioI4:61-74
Gulaldi NC et al (1995) The visualization of pulmonary tuber-
culosis with Tc-99m (V) DMSA and Tc-99m citrate in com-
parison to Ga-67 citrate. Clin Nucl Med 20:1012-1014
Hardoff R, Efrat M, Gips S (1995) Multifocal osteoarticular
tuberculosis resembling skeletal metastatic disease. Evalu-
ation with Tc-99m MDP and Ga-67 citrate. Clin Nucl Med
20:279-281
Hashimoto T et al (2000) Peritoneal linear uptake of Ga-67 caused
by tuberculous peritonitis. Clin Nucl Med 25:214-215
Hovi I et al (1993) Technetium-99m-HMPAO-labeled leu-
kocytes and technetium-99m-labeled human polyclonal
immunoglobulin G in diagnosis of focal purulent disease.
J Nucl Med 34:1428-1434
Ishino Y, Shiozaki H, Nakata H (1992) Positive Tc-99m pyro-
phosphate myocardial scintigraphy in a patient with tuber-
culous pericarditis. Clin Nucl Med 17:515-517
Kalicke T et al (2000) Fluorine-18 fluorodeoxyglucose PET
in infectious bone diseases: results of histologically con-
firmed cases. Eur J Nucl Med 27:524-528
Kao C-H et al (1993) Usefulness of gallium-67 citrate scans
in patients with acute disseminated tuberculosis and com-
parison with chest x-rays. J Nucl Med 34:1918-1921
429
Kao P-F et al (1996) Accumulation of Ga-67 citrate in a tuber-
culous splenic abscess. Report of a rare case. Clin Nucl
Med 21:49-52
Kattapuram SV, Lee VW, Shapiro JH (1979) Primary tuberculo-
sis diagnosed by gallium scan. Clin Nucl Med 4:238-239
Kim DG et al (1995) 99mTc_HMPAO labeled leukocyte SPECT
in intracranial lesions. Surg Neurol 44:338-345
Knight SB et al (1996) Evaluation of pulmonary lesions with
FDG-PET. Comparison of findings in patients with and
without a history of prior malignancy. Chest 109:982-988
Kobayashi H et al (1995) Solitary muscular involvement by
tuberculosis: CT, MRI and scintigraphic features. Comput
Med Imag Graph 19:237-240
Kramer EL et al (1989) Diagnostic implications of Ga-67 chest-
scan patterns in human immunodeficiency virus-seroposi-
tive patients. Radiology 170:671-676
LaManna MM et al (1984) Gallium localization in peritonitis.
Two case reports. Clin Nucl Med 9:25-27
Lee JD et al (1992) Immunoscintigraphy in the detection of
tuberculosis with radiolabelled antibody fragment against
Mycobacterium bovis bacillus Calmette-Guerin: a prelimi-
nary study in a rabbit model. Eur J Nucl Med 19:1011-1015
Lee VW et al (1994) Pulmonary mycobacterial infections in
AIDS: characteristic pattern of thallium and gallium scan
mismatch. Radiology 193:389-392
Lee VW et al (1999) Intracranial mass lesions: sequential
thallium and gallium scintigraphy in patients with AIDS.
Radiology 211:507-512
Leventhal WD,Gordon L,Hotchkiss S (1981) Ga-67-citrate and
ultrasonographic visualization of genitourinary tuberculo-
sis. Clin Nucl Med 6:504-505
Lin W-Y, Hsieh J-F (1999) Gallium-67 citrate scan in extrapul-
monary tuberculosis. Nuklearmedizin 38:199-202
Lin W-Y et al (1998) Diagnostic value of bone and Ga-67 imag-
ing in skeletal tuberculosis. Clin Nucl Med 23:743-746
Lisbona R et al (1993) Gallium-67 scintigraphy in tuberculous
and nontuberculous infectious spondylitis. J Nucl Med 34:
853-859
Loken MK (1987) Pulmonary nuclear medicine. Appleton and
Lange, California
Lorberboym M et al (1998) Thallium-201 retention in focal
intracranial lesions for differential diagnosis of primary
lymphoma and nonmalignant lesions in AIDS patients. J
Nucl Med 39:1366-1369
Malhotra CM et al (1985) Ga-67 studies in a patient with
acquired immunodeficiency syndrome and disseminated
mycobacterial infection. Clin Nucl Med 10:96-98
Mansberg VJ, Murray IP, Rossleigh MA (1997) Complementary
scintigraphy in tuberculous osteomyelitis. Clin Nucl Med
22:776-778
McAfee JG (1996) Editorial comment on Brophey et al1995b.
The Year Book of Nuclear Medicine. Mosby, St Louis
Misra UK, Kalita J, Das BK (fOOO) Single photon emission
computed tomography in tuberculous meningitis. Post-
grad Med J 76:642-645
Moody EB, Delbeke D (1992) Nuclear medicine case of the day.
Case 1: miliary tuberculosis. Am J Roentgenol 158:1382-1386
Muradali D et al (1993) Multifocal osteoarticular tuberculosis:
report of four cases and review of management. Clin Infect
Dis 17:204-209
Nocera RM et al (1983) Tc-99m MDP and indium-Ill chlo-
ride scintigraphy in skeletal tuberculosis. Clin Nucl Med
8:418-420
430
O'Doherty MI et al (1997) PET scanning and the human
immunodeficiency virus-positive patient. I Nucl Med 38:
1575-1583
Ohta H et al (1996) Liver tuberculoma detected by Ga-67 imag-
ing. Clin Nucl Med 21:577
Ohta H et al (1998) Tc-99m sestamibi uptake in tuberculosis
of the breast. Clin Nucl Med 23:106-108
Onsel C et al (1996) Technetium-99m-MIBI scintigraphy in
pulmonary tuberculosis. I Nucl Med 37:233-238
Ozturk E et al (1994) The visualization of granulomatous
disease with somatostatin receptor scintigraphy. Clin Nucl
Med 19:129-132
Palestro CJ et al (1991) Relative efficacy of In-III leukocyte
and Ga-67 imaging in HIV (+) patients (abstract). I Nucl
Med 32:1003
Parmett SR et al (1995) Utility of Ga-67 scintigraphy in detect-
ing radiologically negative tuberculosis in HIV-positive
patients (abstract). I Nucl Med 36:207P-208P
Patz EF et al (1993) Focal pulmonary abnormalities: evalua-
tion with F-18-Fluorodeoxyglucose PET scanning. Radiol-
ogy 188:487-490
Pettengell K et al (1990) Radionuclide scintigraphy in tubercu-
lous enteritis. Gastrointest RadioI15:148-150
Picciotto G et al (1992) Estimation of chromium-51 ethylene
diarnine tetra-acetic acid plasma clearance: a comparative
assessment of simplified techniques. Eur I Nucl Med 19:30-35
Piepsz A, Pintelon H, Ham HR (1994) Estimation of normal
chromium-51 ethylene diamine tetra-acetic acid clearance
in children. Eur I Nucl Med 21:12-16
Prat L et al (1991) Ga-67 citrate and Tc-99m HMPAO leukocyte
scanning in extrapulmonary tuberculosis. Clin Nucl Med
16:865-866
Pui MH, Chin-Sang HR, Rubenstein ID (1986) False-normal
bone imaging in spinal tuberculosis. Clin Nucl Med II:
245-248
Rust RI, Park HM, Robb IA (1981) Skeletal scintigraphy in
miliary tuberculosis: photopoenia after treatment. Am I
Radiol 137:877-879
Santin M et al (1995) Utility of the gallium-67 citrate scan
for the early diagnosis of tuberculosis in patients infected
with the human immunodeficiency virus. Clin Infect Dis
20:652-656
Sarkar SD, Ravikrishman KP (I978) Gallium-67 citrate scan-
ning in extra-pulmonary tuberculosis (abstract). I Nucl
Med 19:734
Sarkar SD et al (1979) Gallium-67 citrate scanning-a new
adjunct in the detection and follow-up of extrapulmo-
nary tuberculosis: concise communication. I Nucl Med
20:833-836
Sciuk I et al (1991) Comparison of technetium-99m polyclonal
human immunoglobulin and technetium-99m monoclonal
antibodies for imaging chronic osteomyelitis. First clinical
results. Eur I Nucl Med 18:401-407
Schmitz A et al (2000) Use of fluorine-18 fluoro-2-deoxy-D-
glucose positron emission tomography in assessing the
process of tuberculous spondylitis. I Spinal Disord 13:
541-544
D. Hamilton and I. Al-Nabulsi
Schmidt U, Rebarber IF (1994) Tuberculous pericarditis iden-
tified with gallium-67 and indium-III leukocyte imaging.
Clin Nucl Med 19:146-147
Sharif HS et al (1989) Brucellar and tuberculous spondylitis:
comparative imaging features. Radiology 171:419-425
Shih WI et al (I986) Progression of mediastinal tuberculosis
and superior vena caval obstruction demonstrated by gal-
lium-67 citrate and radiocolloid liver scintigraphy. Clin
Nucl Med II:126-128
Siemsen IK et al (1976) Gallium-67 scintigraphy of pulmonary
diseases as a complement to radiography. Radiology II8:
371-375
Siemsen IK, Grebe SF, Waxman AD (1978) The use of gallium-
67 in pulmonary disorders. Semin Nucl Med 8:235-249
Suga K et al (2000) Findings on radionuclide cisternogra-
phy in a large abdominal cerebrospinal fluid pseudocyst
associated with tuberculous peritonitis. Clin Nucl Med 25:
1063-1065
Sumi Yet al (1999) Tuberculosis peritonitis: gallium-67 scinti-
graphic appearance. Ann Nucl Med 13:185-189
Sweny P, Farrington K, Moorhead IF (1989) The kidney and its
disorders. Blackwell, Oxford
Taher MA (1998) Nuclear scan and sonogram in renal tuber-
culosis (letter). I Assoc Physic India 46:401-402
Tamgac F et al (1995) Extraskeletal accumulation of Tc-99m
HMDP in a tuberculous cold abscess. Clin Nucl Med 20:1092
Utsunomiya K et al (1997) Clinical significance of thallium-
201 and gallium-67 scintigraphy in pulmonary tuberculo-
sis. Eur I Nucl Med 24:252-257
Vanhagen PM et al (1994) Somatostatin analogue scintigraphy
in granulomatous diseases. Eur I Nucl Med 21:497-502
Villringer K et al (1995) Differential diagnosis of CNS lesions
in AIDS patients by FDG-PET. I Comput Assist Tomogr 19:
532-536
Von Wenzel KS, Klopper IF, Wasserman HI (I989) The techne-
tium-99m DTPA partition test in the diagnosis of tubercu-
lous meningitis. South Afr Med I 20:488-489
Vorne M, Sahlstrom K, Alanko K (1988) Poor accumulation of
technetium-99m glucoheptonate in sarcoidosis and other
diffuse infiltrative lung diseases as compared with gallium-
67 citrate. Clin Nucl Med 13:107-109
Walsh TI et al (1985) The value of gallium-67 scanning in pul-
monary tuberculosis. Am Rev Respir Dis 132:746-747
Wang Y, Sabow LT, Dee WF (1972) 131-1 study of thyroid
tuberculosis mimicking thyroid carcinoma. CRC Crit Rev
Radiol Sci 3:101-103
Wiggelinkhuizen I, Mann M (1980) The radioactive bromide
partition test in the diagnosis of tuberculous meningitis in
children. I Pediatr 97:843-847
Winzelberg GG (1981) Radionuclide evaluation of miliary
tuberculosis. Clin Nucl Med 6:330-331
Yang S-O et al (1992) Detection of extrapulmonary tuberculo-
sis with gallium-67 scan and computed tomography. I Nucl
Med 33:2II8-2123
Young T-H et al (1996) Esophageal tuberculosis with supracla-
vicular lymph node involvement demonstrated by Ga-67
imaging. Clin Nucl Med 21:344
27 Tuberculosis of the Heart and Pericardium
ERNESTO E. SALCEDO and AHMAD S. OMRAN

CONTENTS cal mycobacterium infections should now be included,

by the astute clinician, in the differential diagnosis of
27.1 Introduction 431 chronic infections or obscure clinical processes.
27.2 Tuberculous Pericarditis 431
27.2.1 Clinical Pathophysiology 431
The heart and pericardium can be infected by
27.2.2 Acute Pericarditis 432 the tuberculous mycobacterium via hematogenous
27.2.2.1 Diagnosis 432 spread occurring during the primary tuberculous
27.2.2.2 Management 432 infection, but also by direct, peribronchial or lym-
27.3 Pericardial Effusion and Cardiac Tamponade 433
phatic dissemination (Table 27.1).
27.3.1 Clinical Pathophysiology and Epidemiology 433
27.3.1.1 Diagnosis 433
The most common form of cardiac involvement
27.3.1.2 Management 434 by tuberculosis is as pericarditis, with the long-term
27.4 Constrictive Pericarditis 435 development of calcific constrictive pericarditis.
27.4.1 Pathophysiology and Epidemiology 435 Other forms of cardiac involvement in tuberculosis
27.4.1.1 Diagnosis 436 are quite rare and include myocarditis and endocar-
27.4.1.2 Diagnostic Tests 436
27.4.1.3 Management 439
ditis. In this chapter we review the important issues
27.5 Tuberculous Endocarditis 439 regarding the clinical presentation, diagnostic pro-
27.5.1 Clinical Pathophysiology 439 cedures and contemporary medical management of
27.5.1.1 Diagnosis 440 patients with cardiac tuberculosis.
27.5.1.2 Management 440
27.6 Tuberculous Myocarditis 440
27.6.1 Clinical Pathophysiology 440 Table 27.1. Mycobacterium routes to heart and pericardium
27.6.1.1 Diagnosis 441 Hematogenous
27.6.1.2 Management 441 During primary tuberculous infection
References 441 Lymphatic
Lung
Bronchial
Mediastinal nodes
Peribronchial
27.1 Contiguous
Introduction Mediastinal nodes
Pleural
Tuberculosis remains a considerable public health
problem in non-industrialized countries. In industri-
alized countries, with the recent epidemic of human
immunodeficiency virus (HIV) infection and immu- 27.2
nocompromised hosts, tuberculosis has resurfaced as Tuberculous Pericarditis
an important disease process. Typical as well as atypi-
27.2.1
Clinical Pathophysiology
E. E. SALCEDO, MD
Head Non-invasive Laboratories, King Abdulaziz Cardiac Pericardial involvement by mycobacterium repre-
Center, National Guard Hospital, P.O. Box 22490, Riyadh 11426, sents the most frequent form of cardiac tuberculo-
Saudi Arabia sis. Clinically, tuberculous pericarditis can present
A. S. OMRAN, MD
Consultant Cardiologist, King Abdulaziz Cardiac Center,
as acute pericarditis, pericardial effusion, cardiac
King Abdulaziz Medical City, National Guard, Riyadh, tamponade and constrictive pericarditis (Table 27.2).
P.O. Box 22490, Riyadh 11426, Saudi Arabia Tuberculosis remains an uncommon cause of large

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
432
Table 27.2. Classification of tuberculous pericardial involve-
ment
Acute pericarditis
Pericardia I effusion
Without tamponade and with tamponade
Constrictive pericarditis
Acute
Sub acute
Chronic
Pericardial calcification
Without hemodynamic consequences
With hemodynamic consequences
pericardial effusion and cardiac tamponade is dis-
tinctly rare. However, constrictive pericarditis is a
common long-term sequelae of tuberculosis.
27.2.2
Acute Pericarditis
The infectious pericarditis include viral pericardi-
tis, bacterial suppurative pericarditis, fungal peri-
carditis and tuberculous pericarditis. Tuberculous
pericarditis is a rare form of acute pericarditis, but
needs to be considered in the differential diagnosis
in immunocompromised patients and in areas of the
world where tuberculosis is still endemic. Subclinical
pericarditis has also been recognized in patients with
pulmonary tuberculosis (Kishk et al. 1998).
27.2.2.1
Diagnosis
In contradistinction with acute viral or bacterial peri-
carditis, where there is usually an abrupt initiation of
symptoms, tuberculous pericarditis has a much more
insidious beginning. The usual symptoms of an acute
febrile episode followed by chest pain with pleuritic
exacerbation and constitutional symptoms are usu-
ally lacking. A more common clinical scenario is that
of presentation as fever of unknown origin. Immuno-
compromised patients and children do present with
the more typical picture of acute pericarditis.
The presence of a pericardial friction rub is the
most reliable clinical marker for the diagnosis of
pericarditis. During the acute episode of pericarditis,
the pericardial friction rub may come and go, and its
recognition may require having the patient sit up and
lean forward. With chronic pericarditis, such as with
tuberculosis, the pericardial friction rub may also wax
and wane depending on the degree of inflammation
and the presence and size of a pericardial effusion.
E. E. Salcedo and A. S. Omran
The characteristic electrocardiogram findings of
pericarditis can also be expected to wax and wane in
tuberculous pericarditis, depending on the severity
of the inflammatory process. These ECG changes
include diffuse ST elevation without reciprocal
changes, T wave inversion, and PR segment devia-
tions.
Acute phase reactants such as sedimentation rate
and C-reactive protein, markers of an acute inflam-
matory process, are less likely to be elevated in
tuberculous pericarditis than in the more common
forms of acute viral pericarditis. Elevation of cardiac
enzymes during acute pericarditis is the result of
associated myocarditis or extension of the inflamma-
tory process into the myocardium from the visceral
pericardium. In tuberculous pericarditis, elevation
of cardiac enzymes is distinctly rare and its presence
should raise the question of associated myocarditis.
The chest X-ray, besides demonstrating the pul-
monary evidence of tuberculosis, offers little help
in the diagnosis of acute pericarditis. A left pleural
effusion is commonly associated with pericarditis,
but is a non-specific finding. The cardiac silhouette
remains normal, unless there is associated pericar-
dial effusion. The typical calcification of the pericar-
dium does not appear until late in the course of the
illness, and is not expected to be present during the
acute tuberculous pericarditis phase.
The echocardiogram, as will be discussed later, is
of great value in assessing the presence and size of
pericardial effusion, the presence and degree of peri-
cardial thickening and in demonstrating the charac-
teristic hemodynamic changes of constriction and
tamponade by Doppler techniques. During the acute
pericarditis phase, when there is no or little pericar-
dial effusion, the echocardiogram may be normal.
When a small fibrinous pericardial effusion is pres-
ent during the acute pericarditis process, a partially
echo-free space may be seen surrounding the heart.
In the presence of a pericardial effusion, the echocar-
diogram can be used to guide a pericardiocentesis
needle for diagnostic purposes.
Although nuclear imaging with Gallium-67 has been
found to be useful in demonstrating the inflammatory
process in acute pericarditis (Spodick 1997), its use in
tuberculous pericarditis has not been described.
27.2.2.2
Management
The management of acute tuberculous pericarditis
is no different than that of the more common forms
of acute pericarditis. The treatment is directed to the
Tuberculosis of the Heart and Pericardium 433

control of symptoms and, as much as possible, to Table 27.3. Etiology of moderate-large pericardia! effusions
the eradication of the etiological agent. Ibuprofen is Cheema Colombo Sagrista-Sauleda
the preferred nonsteroidal anti-inflammatory drug
at doses of 300-800 mg every 6-8 h. Steroids are n 43 25 322
Tamponade 23% 44% 37%
contraindicated in acute tuberculous pericarditis, Idiopathic 40% 32% 20%
unless the patient is well covered with appropri- Neoplastic 2% 36% 13%
ate antimycobacterial therapy. The effectiveness of Uremia 26% 20% 6%
corticosteroids to prevent the eventual development Iatrogenic 0% 0% 16%
of constriction is at most limited (Chen et al. 1996). Post-acute myocardial 0% 8% 8%
infarction
The management of pericardial effusion in patients Pyogenic 9% 0% 0%
with tuberculous pericarditis is discussed in the fol- Collagen disease 0% 0% 5%
lowing section. Tuberculosis 20% 0% 2%
Other 3% 4% 30%

Adapted from Cheema et al. (1999); Colombo et a!. (1988);

27.3
Pericardial Effusion and Cardiac
Tamponade
27.3.1
Clinical Pathophysiology and Epidemiology
Acute pericarditis from infective or inflamma-
tory processes, or from physical and immunologic
origin, can lead to pericardial effusion. A variety of
other conditions such as myxedema, renal failure,
pregnancy and malignancy can produce pericardial
effusion (Table 27.3). The cause of the pericardial
effusion can frequently be inferred from the under-
lying clinical problem (Sagrista-Sauleda et al. 2000).
If the pericardial effusion is found as an incidental
finding and it is small, no further work-up or treat-
ment may be necessary. However, if the pericardial
effusion is moderate or large, every effort should be
made to identify its cause.
As has been previously discussed, in industrialized
countries the incidence of tuberculous pericarditis
has declined over the past several decades. However,
in Africa, Asia and East Europe where tuberculosis is
still endemic, tuberculous pericarditis and effusion
are more common.
As can be seen from Table 3, the etiology of a mod-
erate to large pericardial effusion varies from coun-
try to country. In the patients reported by Cheema
(Cheema et al. 1999) from Saudi Arabia, the incidence
of pericardial effusion related to tuberculosis was
much higher (20%), than those reported by Sagrista-
Sauleda (2%) from Spain and by Colombo (O%) from
the United States.
In addition, tuberculous pericarditis is more preva-
lent in immunocompromised patients and, in particu-
lar, in patients with the acquired immunodeficiency
syndrome. In one series from Africa, of 37 patients
Sagrista-Sauleda et al. (1988)
with effusive pericarditis, 32 (86%) were associated
with tuberculosis and 30 (81 %) were HIV positive
(Haas and Des Prez 1995)
Tuberculosis remains an uncommon cause of car-
diac tamponade. It has been reported in disseminated
tuberculosis with myocarditis and pericarditis (Afzal
et al. 2000). In addition, in a series of 231 patients
with pericardial disease, three of nine patients with
tuberculous pericarditis initially manifested as tam-
ponade, and only 7% of all tamponade patients had
tuberculous pericarditis (Permanyer-Miralda et al.
1985).
27.3.1.1
Diagnosis
The diagnosis of pericardial effusion is straightfor-
ward when using two-dimensional echocardiogra-
phy. An "echo-free" space is visualized surrounding
the heart; the size of this echo-free space is related
to the volume of the pericardial fluid. Normally
there is only about 20-30 cc of pericardial fluid
in the pericardial space. In this situation, there is
virtually no separation between the visceral and
parietal pericardium layers, and there is absence
of the echo-free space surrounding the heart. With
a small pericardial effusion (less than 100 cc) there
is about 1 cm of echo-free space around the heart.
With a large effusion (>500 cc), the echo-free space
increases to over 2 cm.
The confirmation for the clinical diagnosis of car-
diac tamponade is based on the echocardiographic
demonstration of a pericardial effusion, usually but
not necessarily large in size (Fig.27.l), and by the
characteristic hemodynamics shown by Doppler
 434 E. E. Salcedo and A. S. Omran

Fig.27.1. Two-dimensional parasternal

long axis (a) and short axis (b) views
in a patient with very large pericardial
effusion and cardiac tamponade

a b

echocardiography or catheterization. The most useful

echocardiographic signs of cardiac tamponade are
inspiratory shift of the ventricular septum toward the
left ventricle (Fig. 27.2), right ventricular diastolic col-
lapse and right atrial compression. Doppler echocar-
diography provides additional information regarding
increased intrapericardial pressure (Fig. 27.3). There is
a significant accentuation of the normal transvalvular
flow velocities. The normal inspiratory decreases of
about 10-20% of mitral and aortic flow are accentu-
ated to 30-40% with tamponade. Parallel to this there
is a dramatic increment of tricuspid and pulmonary
flow velocities by up to 80%.
Although the demonstration of the presence
of pericardial effusion and tamponade is fairly
straightforward by echo-Doppler techniques, the Fig.27.2. M-mode echocardiogram in same patient as in
demonstration that tuberculosis is the cause of the Fig. 27.7 showing inspiratory shift of the ventricular septum
pericardial effusion is more complex. Less than half toward the posterior ventricular wall. This is characteristic
of patients with tuberculous pericarditis have evi- of cardiac tamponade
dence of pulmonary tuberculosis (Sagrista-Sauleda
et al. 1988).
Because of the slow growth of Mycobacterium
tuberculosis there is usually a delay of about 5 weeks 27.3.1.2
between admission and diagnosis (Sagrista-Sauleda Management
et al. 1988). Sputum cultures are positive in about
50%, gastric aspirate cultures and acid fast stains are Antibiotic therapy is similar to that used in pulmo-
positive in about 15-20% and in 30-75% of pericar- nary tuberculosis. Resolution occurs in 2-3 months
dial aspirate cultures (Gooi and Smith 1978). in about 80% of patients. A form of sub-acute con-
Other tools that assist in the diagnosis of tuberculo- striction occurs in 20% of patients, resolving in sev-
sis as the cause of pericardial effusion include serodi- eral months in half of them, but requiring eventual
agnosis (Ng et al.1995), the polymerase chain reaction, pericardiectomy in about 10% (Strang 1997).
testing for adenosine deaminase activity, and activity Corticosteroids can be used to control fluid accu-
of interferon level (Woods and Goldsmith 1989; Seino mulation, to reduce the need for repeated pericardio-
et al.1993; Koh et al.1994; Shah et al.1998). centesis and to control impending cardiac tampon-
Tuberculosis of the Heart and Pericardium 435

especially in areas of the world where tuberculosis is

Fig. 27.3. Mitral (a) and tricuspid (b) inflow pattern obtained by
transthoracic echo in a patient with cardiac tamponade. There
is reduction in the mitral inflow velocity after inspiration
and the velocity increases after expiration. Opposite changes
are seen in the tricuspid inflow as the velocity increases with
inspiration
still endemic.
Cardiac tamponade in the early clinical stage of
Tuberculous pericarditis has been found to be the most
predictive factor in the subsequent development of con-
strictive pericarditis (Suwan and Potjalongsilp 1995).
Constrictive pericarditis is caused by pericardial
fibrosis and calcification with typical histological fea-
tures (Figs. 27.4, 27.5). These processes compromise
the filling of the heart during diastole and eventually
produce right-sided heart failure. The rigid pericar-
dium restricts the inflow of blood into the ventricles;
a high driving pressure across the atrio-ventricular
valves results in early rapid diastolic filling and an
abrupt increase in ventricular pressure. Flow is termi-
nated early in diastole, and there is marked elevation
of pressure in all four heart chambers. As a conse-
quence of this, a dip and plateau pattern is seen on the
ventricular pressure curves, and a restrictive filling
pattern is seen in the Doppler tracings. These findings
will be illustrated on the next section.
Figure A Causes of Constrictive Pericarditis
5Or-----

--~
ade (Strang et al. 1988). Additionally, large doses of
prednisolone (60 mg/day for 4 weeks and 15 mg/day
35 --
for 2 weeks) has been shown to decrease mortality
from 11 % to 4% (Strang et al.I987).
30 ----
25
In the rare instances of overt tuberculous cardiac
20
tamponade, emergent pericardiocentesis may be life
15
saving. If there is recurrence of pericardial effusion
10
with unstable hemodynamics, a subxiphoid pericar-
5
dial window may be required.
1aIop81hic ....... Pe<ic:ard11s _lIOn ~11On Mhnlldlls 00'*0
ourget\'

(Adapted Irom Ling LH, 1999)

27.4
Constrictive Pericarditis Figure B Etiology of Infective myocarditis and Pericarditis

80
27.4.1
Pathophysiology and Epidemiology 70- - - -- ----
50

Constrictive pericarditis is a serious sequelae of

50
tuberculous pericarditis and an important determi-
nant of patient outcome. Since surgery may be the 00

only viable therapeutic option, its recognition is of 30

great importance.
20
The clinical spectrum of constrictive pericarditis
has changed over the years (Ling et al. 1999), and its 10

etiology has shifted from infectious diseases to post-

operative and post-radiation causes (Fig. 27.A). Tuber- VwsI !leclorlal MycOPUIM QQmylia Myccb.clorlO

culosis remains an important cause of constriction; {Adapled from Fair1ey CK, 1996}
436
Fig. 27.4. Histological pericardial section obtained at time of
pericardiectomy. The pericardium is thickened and chroni-
cally inflamed. The inner surface is covered with a thick layer
of fibrin, several granulomas are found
Fig. 27.5. Higher magnification histological specimen from the
same patient as Fig. 27.1. A granuloma composed of epitheli-
oid cells and also showing caseation necrosis and multinucle-
ated giant cells of the Langhans type are demonstrated
27.4.1.1
Diagnosis
History and Physical Findings
A high degree of suspicion is required to consider
constrictive pericarditis in the differential diagnosis
of patients with right heart failure. This is particularly
the case when the signs and symptoms for right heart
failure are out of proportion to the degree of pulmonary
disease or left sided pathology. Tuberculous constrictive
pericarditis needs to be considered if there is evidence
E. E. Salcedo and A. S. Omran
of tuberculosis in other organs or the patient lives, or
comes from, an endemic area of tuberculosis. As the
survival of immunocompromised patients improves, it
is expected that long-term complications of tubercu-
losis, such as constriction, will also increase. Again, in
these patients a great degree of suspicion will assist the
clinician in arriving at the correct diagnosis.
Patients will seek medical attention because of
peripheral edema, ascites with abdominal discomfort
and fatigue. Shortness of breath with dyspnea on exer-
tion and orthopnea are not uncommon. These symp-
toms, more commonly seen with "left heart failure;' are
seen in patients with constrictive pericarditis because
of a fixed cardiac output, associated pleural effusions,
and elevated diaphragms from the ascites.
On physical exam, jugular venous distention with
sharp x and y descents are quite noticeable. Paradoxi-
cal systolic retraction of the apical impulse is another
hallmark of constrictive pericarditis. Other physical
findings include the Kussmaul's sign (inspiratory
swelling of the neck veins) and a pericardial knock,
which occurs at the time of a third heart sound, but is
of higher frequency and intensity.
27.4.1.2
Diagnostic Tests
Chest X-Ray
The cardiac silhouette on chest X-ray is usually
normal, and the pulmonary vasculature is indistinct.
Evidence of pulmonary tuberculosis mayor may not
be evident.
Bilateral pleural effusions are common, and fre-
quently there is upper pulmonary flow redistribution
and Kerley B lines may be present. Dilation of the
superior vena cava and azygos vein may be evident.
Calcification of the pericardium is the hallmark
of tuberculous constrictive pericarditis (Fig. 27.6).
Although sometimes it is apparent in the PA view,
calcification in the pericardium is much easier to
visualize from the lateral projection and is very
evident with fluoroscopy. A rim of calcium appears
surrounding the heart, with preference over the
right atrium, right ventricle, and the diaphragmatic
wall. Calcification of the pericardium occurs in
about 50% of patients with constrictive pericarditis
(Rienmuller et a1. 1993). Other potential causes of
cardiac calcification need to be considered in the dif-
ferential diagnosis. These include calcified left ven-
tricular aneurysm, calcified intracardiac thrombus,
coronary calcifications and valvular calcifications.
Therefore, although non pathognomonic, calcifica-
Tuberculosis of the Heart and Pericardium 437

Fig. 27.6. PA and lateral chest X-ray

in a patient with tuberculous calcific
constrictive pericarditis. The cardiac
silhouette is somewhat generous. The
pulmonary vasculature is normal. A rim
of calcium is seen affecting mainly the
apex and the distal posterior wall. Note
that the calcium is better appreciated in
the lateral projection

tion around the pericardium is a very useful and

cost-effective finding in patients suspected of having
tuberculous constrictive pericarditis.

Chest Computed Tomography

Computed tomography (CT) is the diagnostic tool

of choice to evaluate pericardial thickness (Fig. 27.7)
and presence and severity of pericardial calcification.
High spatial resolution sectional pictures that pro-
vide exquisite anatomic detail have now replaced the
blurred images initially obtained with conventional
CT imaging. Associated findings, such as dilated
superior vena cava and lateral bowing of the inter-
atrial septum, can be clearly demonstrated by high
resolution CT imaging.
The pericardial thickness measured by CT has
been found to correlate extremely well with patho-
logic specimens (Grover-McKay et al. 1991). In addi-
tion, CT can localize areas of the pericardium that
are more or less thickened, and as such serve as a
tool for surgical planning by assisting the surgeon in
choosing the optimal surgical approach. The value of
CT in providing prognostic information in patients
with constrictive pericarditis has been described
(Rienmuller et al. 1985). These authors found that
all 5 of 16 patients with non-detectable poster lateral
wall died at or soon after pericardiectomy. They con-
cluded that myocardial fibrosis or atrophy provoked
the non-visualization of myocardium, and defined
patients at high surgical risk.
Magnetic Resonance Imaging
Magnetic Resonance Imaging (MRI), also allows
for excellent imaging of the pericardium. The normal
pericardium appears on MRI as a low intensity line
not exceeding 3 mm in width (Stark et al. 1984). MRI
can readily demonstrate the presence and extent of
pericardial thickening. However, it should be remem-
Fig. 27.7. High resolution chest computed tomography in the
same patient as in Fig. 27.3. There is a small pericardial effu-
sion (PE) and the thickened pericardium is seen surround-
ing both the left ventricle (LV) and right ventricle (RV). The
thickness of the pericardium varies from a few millimeters
to 1 em
bered that a thickened pericardium could be seen in
conditions other than constrictive pericarditis. Peri-
cardial thickening has been noted by MRI after open
heart surgery (Sechtem et al. 1986) and in uremic or
infectious pericarditis.
MRI has also been used to differentiate constric-
tive pericarditis from restrictive cardiomyopathy
(Soulen et al. 1985). In contrast to restrictive cardio-
myopathy, in constrictive pericarditis, MRI will dem-
onstrate a thickened pericardium and an associated
conical or tubular narrowing of non-hypertrophied
ventricles (White 1998).
MRI allows for demonstration of asymmetric
pericardial thickening and compression, assisting
the surgeon in planning for the best surgical window
438
(D'Silva et al. 1992). The disadvantages of MRI over
CT for the evaluation of pericardial thickening and
constriction include its longer scan time and its
inability to reliably detect pericardial calcification
(Sechtem et al. 1986).
Cardiac Catheterization
The classical hemodynamic findings of constrictive
pericarditis include a large and rapid Y descent in
the right atrial waveform and diastolic equaliza-
tion of pressures in the four cardiac chambers
(Fig. 27.8). Unfortunately, these abnormalities are
not exclusively seen in constrictive pericarditis. For
example, a large Y descent is seen in patients with
right ventricular infarction. Diastolic equalization on
the cardiac chambers can also be seen in restrictive
cardiomyopathy. Since constrictive pericarditis usu-
ally requires pericardiectomy as the definite form
of therapy, and surgery is of no use (and actually is
detrimental on patients with restrictive cardiomy-
opathy), a clear differentiation between both disease
processes is essential.
The most valuable hemodynamic method to dif-
ferentiate constrictive physiology from restrictive
physiology employs the principle of discordance of
right ventricular and left ventricular systolic pres-
sure during respiration. In constrictive pericarditis,
at end inspiration the left ventricular systolic pres-
sure decreases whereas the right ventricular systolic
pressure increases. In restrictive cardiomyopathy,
both the right ventricular and left ventricular sys-
tolic pressure decrease at end-inspiration (Hurrell
et al. 1996).
-t ----------- .--_tee_----------------
PACW PA
2e 2e
E. E. Salcedo and A. S. Omran
Echocardiography and Doppler Studies
M-mode echocardiography in constrictive pericar-
ditis demonstrates normal left ventricular size and
systolic function and mild left atrial enlargement.
Abnormalities of septum and posterior wall motion
are also frequently present. A sudden anterior dis-
placement of the septum followed by a brisk poste-
rior rebound is noted. Evaluation of pericardial thick-
ness by m-mode echocardiography is limited because
the pericardial signal is significantly affected by gain
and gray scale settings.
On two-dimensional echocardiography, the thick-
ened pericardium can be somewhat better appreci-
ated than by m-mode echocardiography. In addition,
one will observe small ventricular cavities, normal
systolic function, bilateral atrial enlargement and
increased vena cava diameter and hepatic veins. The
variation in intracardiac flow velocities produces
an expiratory septal bulge toward the left ventricle,
which is fairly characteristic of constrictive physiol-
ogy.
Transesophageal echo (TEE) provides an adequate
means of detecting pericardial thickening (Fig. 27.9).
An excellent correlation between pericardial thick-
ness by TEE and CT has been described (Ling et al.
1997). TEE provides the means of evaluating not only
the pericardial thickness but also the means to fully
evaluate, through Doppler methods, the hemody-
namics of constriction.
Doppler echocardiography via transthoracic or
transesophageal methods allows for detailed char-
acterization of the hemodynamic alterations seen
in constrictive pericarditis (Hurrell et al. 1996). Fur-
Fig. 27.8. Pressure tracings in a patient with
constrictive pericarditis. There is diastolic
equalization of pressures in the left atrium
as seen through the capillary wedge pres-
sure (PACW), the pulmonary artery (PA),
and the right ventricle (RV). In addition, a
dip and plateau or "square root" sign can
be appreciated in the RV tracing
Tuberculosis of the Heart and Pericardium
Fig. 27.9. A transesophageal echocardiogram in a patient with
constrictive pericarditis showing a thickened calcified peri-
cardium surrounding the right atrium (RA), and the right
ventricle (RV). The left atrium (LA) is not well seen from this
projection
thermore, Doppler is of great value in distinguishing
constrictive pericarditis from restrictive cardiomy-
opathy (Hatle et al.I989).
The characteristic Doppler findings of constric-
tive pericarditis include an expiratory increase in the
mitral inflow E velocity of equal or greater than 25%
than on inspiration. In addition, there is a reciprocal
decrease in the tricuspid velocity of 40% or more
(Fig. 27.10). Further, the isovolumic relaxation time
decreases during expiration by a mean of 20%.
Respiratory alterations in pulmonary venous flow
velocities by TEE have permitted the differentiation
of constrictive pericarditis from restrictive cardio-
myopathy (Klein et al.I993).
Using the combination of pulmonary venous
systolic/diastolic flow ratio of 0.65 or more in inspira-
tion, and a percentage increase of peak diastolic flow
during expiration of 40% or more, Klein was able to
correctly distinguish constrictive pericarditis from
restrictive cardiomyopathy in 86% of 31 patients with
diastolic dysfunction.
27.4.1.3
Management
The medical management of tuberculous constric-
tive pericarditis is directed toward eradication of the
infective organism by appropriate antibiotic therapy
to decrease the inflammatory process when present
and to the management of right heart failure. Most
patients will require increasing doses of diuretics as
the hemodynamic compromise worsens.
439
Fig.27.10. Transesophageal Doppler flow velocities showing
respiratory variation of mitral (A), and tricuspid (B) inflow
pattern in a patient with constrictive pericarditis. During
inspiration there is a noticeable drop in the mitral inflow
velocity with reciprocal changes in the tricuspid inflow
In the management of patients with known or sus-
pected constrictive pericarditis, it is imperative to make
a clear distinction between the constrictive process and
a possible restrictive component. This differentiation,
at times, is not easy and requires detailed echo/Doppler
hemodynamic evaluation as well as right heart cath-
eterization, as previously discussed.
Eventually most patients with significant constric-
tive pericarditis will require surgical intervention, as
discussed in a following chapter.
27.S
Tuberculous Endocarditis
27.5.1
Clinical Pathophysiology
Endocardial tuberculosis is extremely rare, found in
only 0.14% (19 of 13,658 autopsies) of cases by Rose
(1987). The endocardial involvement can be either
miliary with multiple small tuberculi (less than 3 mm
440
in diameter) or nodular formed by confluent foci of
tuberculi.
Endocardial involvement by tuberculous infection
may affect the valvular surfaces creating the picture
typical of valvular infective endocarditis, or it can
involve the endothelial surfaces of the cardiac cham-
bers or great vessels.
When granulomas due to Mycobacterium tubercu-
losis are found in valvular structures, they are usually
seen in immunocompromised patients with dissemi-
nated tuberculosis. However, valvular tuberculous
endocarditis in an immunocompetent patient has
been described (Klinger et al. 1998). This particular
patient had severe mitral regurgitation secondary to
perforation of the anterior mitral leaflet.
Another potential valvular involvement from
tuberculosis includes the formation of subvalvular
left ventricular aneurysms. Deshpande reported 19
subvalvular aneurysms seen in 16 patients. There
were 12 isolated subaortic aneurysms, three isolated
submitral aneurysms and one patient with a subaor-
tic and submitral aneurysm. He described an asso-
ciation between infective endocarditis and the sub-
valvular aneurysms and showed a strong association
between the submitral aneurysms and tuberculosis
(Desphande et al. 2000).
Extension of tuberculoma into the left ventricular
free wall with pseudoaneurysm formation and suc-
cessful repair of rupture has been described (Halim
et al. 1985).
Other clinical presentations of endocardial tuber-
culosis include pulmonary vein obstruction from left
atrial tuberculoma, right ventricular outflow tract
obstruction, and superior vena cava obstruction
(Chang et al.1999).
27.5.1.1
Diagnosis
As described above, endocardial involvement by
tuberculosis may present in different forms and a
clinical diagnosis is certainly difficult. This is made
even more problematic because endocardial tubercu-
losis is, as previously noted, quite rare. Nevertheless,
in any patient with tuberculosis (especially miliary)
or in any immunocompromised patient with cardio-
vascular symptoms, the possibility of endocardial
tuberculosis needs to be raised.
Physicians should perform a detailed cardiovascu-
lar history and examination in patients with tubercu-
losis, in searching for symptoms suggestive of cardiac
arrhythmias or heart failure, and in the presence of
new or changing heart murmurs. If endocarditis is
E. E. Salcedo and A. S. Omran
suspected, an echocardiogram should be performed
in search of vegetations, regurgitant lesions and
subvalvular aneurysms. Occasionally, a large tuber-
culoma with the appearance of a cardiac tumor may
be formed in one of the cardiac chambers or great
vessels. In these cases, a cardiac MRI may further
clarify the nature of the cardiac mass.
A heightened degree of clinical suspicion and
echocardiography remain the main tools for the
diagnosis of tuberculous endocarditis.
27.5.1.2
Management
As with all forms of tuberculosis, appropriate anti-
microbial therapy remains the fundamental form
of therapy. Because of this, a prompt and correct
diagnosis is mandatory.
A case of miliary tuberculosis with aortic valvuli-
tis, which resolved with antituberculous therapy, has
been reported (Cope et al. 1990). This case empha-
sizes the central role of antimicrobials in the manage-
ment of tuberculous endocarditis.
In addition to the antimicrobial therapy, patients
with valvular tuberculous endocarditis may benefit
from valve surgical procedures with valvular repair
or replacement.
A successful surgical repair of tuberculous aortic
insufficiency has been described (Soyer et al.I981).
It has been noted that tuberculosis is a rare com-
plication of valve replacement. The majority of myco-
bacterial infections occurring after valve surgery
have, for some unknown reason, the nontuberculous
forms of M. chelonae and Mfortuitum (Grange 1992).
These mycobacteria have a very poor response to
therapy, so prevention by avoidance of contamina-
tion of all surgical materials is of great importance.
Surgical excision of large endocardial tuberculo-
mas needs to be considered if they produce obstruc-
tion to normal flow or have the potential of major
embolic event.
27.6
Tuberculous Myocarditis
27.6.1
Clinical Pathophysiology
As with tuberculous pericarditis, the myocardium
can be reached by the mycobacteria by either direct
extension, lymphangitic spread or by the hema-
Tuberculosis of the Heart and Pericardium 441

togenous route. In contrast to the common peri- (Diaz-Peromingo et al. 2000) This represents another
cardial involvement with tuberculosis, myocardial potential source of lethal arrhythmias in patients
involvement is quite rare. Before the introduction with cardiac tuberculosis.
of specific antituberculous medications, myocardial
involvement by mycobacteria was reported in only 27.6.1.2
less than 0.30% of patients dying from tuberculosis Management
(Horn and Saphir 1935).
Although rare, it is important to recognize tuber- Since the ante mortem diagnosis of tuberculous
culosis as a cause of myocarditis because it has the myocarditis is rare, the chances of managing
potential of presenting as sudden death (Dada et al. patients with this process are also rare. A high index
2000; Alkuja and Miller 2001), particularly with mili- of suspicion is required, especially in the presence
ary tuberculosis (Chan and Dickens 1992). of immunocompromised patients or in the pres-
The organisms reported to cause infective myo- ence of miliary tuberculosis. The main goal is to
carditis and pericarditis as reported by the Public provide prompt and effective antimicrobial ther-
Health Laboratory Service of the United Kingdom apy. Darwish et al. (1998) reported a patient with
are shown in Figure B, and demonstrate the paucity tuberculous pancarditis who presented with atrial
of cardiac involvement by mycobacteria (Fairley et flutter and converted to normal sinus rhythm after
al. 1996). initiation of antituberculous therapy. Resolution of
ventricular tachycardia and endocardial tubercu-
27.6.1.1 loma following antituberculous therapy has been
Diagnosis reported (O'Neill et al. 1991).
Other considerations that need to be considered
Since most cases of myocardial tuberculosis are clini- when treating patients with known or suspected
cally silent, the recognition of myocardial involvement myocardial tuberculosis are the avoidance of any
in patients with tuberculosis is frequently made only medications that may prolong the QT interval of the
during post-mortem examination. Rose (1987), electrocardiogram. As described above, prolongation
encountered myocardial tuberculosis in 19 patients of the QT interval has been described in tuberculous
(0.14%) at autopsy over a 27 year period. Myocardial myocarditis, and any situation or medication that
tuberculosis was diagnosed ante mortem in only further prolongs the QT interval should be avoided.
one patient. There have been occasional reports of If there is associated pericarditis and nonsteroidal
pre-mortem diagnosis, but this represents more the anti-inflammatory drugs are to be used, ibuprofen
exception than the rule. It is possible that with the would be preferred over indomethacin, as the former
increased clinical use and the high resolution of MRI increases coronary flow and the latter decreases it.
and TEE, more patients with either nodular or miliary Patients with tuberculous myocarditis have potential
forms of tuberculous myocarditis will be recognized foci for arrhythmias and they would be placed in fur-
ante mortem. ther jeopardy by decreasing their coronary flow with
Dahar (1998) described the chest X-ray and chest indomethacin.
CT of a patient in whom at autopsy the apical myocar-
dium revealed myocardial tubercula, and at histology
granulomas with giant cells. The chest X-ray showed
a right pleural effusion, increased cardiac silhouette References
and perihilar edema. A chest CT showed thickened
pericardium and a small pericardial effusion. Afzal A, Keohane M, Keeley E et al (2000) Myocarditis and
There have been isolated reports on electrocar- pericarditis with tamponade associated with disseminated
tuberculosis. Can TCardioI16:519-521
diographic abnormalities and cardiac arrhythmias
Alkuja S, Miller A (2001) Tuberculosis and sudden death: a
in patients with myocardial involvement in tuber- case report and review. Heart Lung 30:388-391
culosis. Doherty et al. (1996) described a patient Chang BC, Ha TW, Kim TT et al (1999) Intracardiac tubercu-
that developed infection of a pacemaker pulse-gen- loma Ann Thorac Surg 67:226-228
erator pocket and died from ventricular fibrilla- Chan AC, Dickens P (1992) Tuberculous myocarditis present-
ing as sudden cardiac death. Forensic Sci Int 57:45
tion. At autopsy, she was found to have Widespread
Cheema MA, Ghalib MB, Shatoor AS, Suliman FA, Al-Hroub
miliary tuberculosis, including myocardial tissue SS, Kardash M, Ahmed MK (1999) Pattern of pericardial
involvement. Prolonged QT syndrome in a patient disease in the Asir region of Saudi Arabia. Ann Saudi Med
with tuberculous myocarditis has been described 19:171-173
442
Chen WT, Chen CC, Yu FC et al (1996) Clinical response of
tuberculous pericarditis to medical treatment: a retrospec-
tive survey. Chin Med J 58:7-11
Colombo A, Olson HG, Egan J et al (1988) Etiology and prog-
nostic implications of a large pericardial effusion in men.
Clin Cardiol11:389-394
Cope AP, Heber M, Wilkins EG (1990) Valvular tuberculous
endocarditis: a case report and review of the literature. J
Infect 21:293-296
Dada MA, Lazarus MG, Kharsany AB, Sturn AW (2000) Sudden
death caused by myocardial tuberculosis: case report and
review of the literature. Am J Forensic Med Pathol 21:
385-388
Darwish Y, Mushannen B, Hussain KM, Dadkah S, Atkinson
J, Zar F, Kogan A (1998) Pancardiac tuberculosis-a case
report. Angiology 49:151-156
Deshpande J, Vaideeswar P, Sivaraman A (2000) Subvalvular
left ventricular aneurysms. Cardiovasc Pathol 5:267-271
Diaz-Peromingo JA, Marino-Callejo AI, Gonzalez-Gonzalez C,
Garcia-Rodriguez et al (2000) Tuberculous myocarditis pre-
senting s long QT syndrome. Eur J Intern Med 11:340-342
Doherty JG, Rankin R, Kerr F (1996) Miliary tuberculosis
presenting as infection of a pacemaker pulse-generator
pocket. Scott Med J 41:20-21
D'Silva SA, Nalladaru ZM, Dalvi BV et al (1992) MRI as guide
to surgical approach in tuberculous pericardial abscess.
Case report. Scand J Thorac Cardiovasc Surg 26:229-231
Fairley CK, Ryan M, Wall PG, Weinberg J (1996) The organisms
reported to cause infective myocarditis and pericarditis in
England and Wales. J Infect 32:223-225
Gooi HL, Smith JM (1978) Tuberculous pericarditis in Burm-
ingham. Thorax 33:94-96
Grange JM (1992) Mycobacterial infections following heart
valve replacement. J Heart Valve Dis 1:102-109
Grover-McKay M, Burke S, Thompson SA et al (1991) Measure-
ment of pericardial thickness by cine computed tomogra-
phy. Am J Card Imaging 5:98-103
Haas DW, Des Prez RM (1995) Tuberculous pericarditis. In:
Mandell GL, Bennett JE, Dolin R (eds) Mandell, Douglas
and Bennett's principles and practice of infectious disease,
vol 2, 4th edn
Halim MA, Mercer EM, Guinn GA (1985) Myocardial tuber-
culoma with rupture and pseudoaneurysm formation-suc-
cessful surgical treatment. Br Heart J 54:603-604
Hatle LK, Appleton CP, Popp RL (1989) Differentiation of
constrictive pericarditis and restrictive cardiomyopathy
by Doppler echocardiography. Circulation 79:357-370
Horn H, Saphir 0 (1935) The involvement of the myocardium
in tuberculosis: a review of the literature and report of
three cases. Am Rev Tuberc 32:492-506
Hurrell DG, Nishimura RA, Higano ST et al (1996) Value of
dynamic respiratory changes in left and right ventricular
pressures for the diagnosis of constrictive pericarditis.
Circulation93:2007- 2013
Kishk YT, Allam MH, Abdel-Wahab AM (1998) Subclinical
pericarditis in 289 patients with pulmonary tuberculosis;
further indirect evidence by analysis of 53 patients with
pericardial disease. J Am Coil CardioI31:220-230
Klein AL, Cohen Gl, Pietrolungo JF et al (1993) Differentiation
of constrictive pericarditis from restrictive cardiomyopa-
thy by Doppler Transesophageal echocardiographic meth-
ods of respiratory variations of pulmonary venous flow. J
Am Coil CardioI22:1935-1943
E. E. Salcedo and A. S. Omran
Klinger K, Brandi 0, Doerfler M, Schluger N, Rom WN (1998)
Valvular endocarditis due to Mycobacterium tuberculosis.
Int J Tuberc Lung Dis 2:435-437
Koh KK, Kim EJ, Cho CH et al (1994) Adenosine deaminase
and carcioembrionic antigen in pericardial effusion
diagnosis, specially in suspected tuberculous pericarditis.
Circulation 89:2728-2735
Ling LH, Oh JK, Tei C et al (1997) Pericardial thickness mea-
sured with Transesophageal echocardiography: feasibility
and potential clinical usefulness. J Am Coil Cardiol 29:
1317-1323
Ling LH, Oh JK, Shaff HV, Danielson JK, Mahoney WV, Seward
JB, Tajik AJ (1999) Constrictive pericarditis in the modern
era. Evolving clinical spectrum and impact on outcome
after pericardiectomy. Circulation 100:1380-1386
Ng TTC, Strang JIC, Wilkins EGL (1995) Serodiagnosis of
pericardial tuberculosis. Q J Med 88:317-320
O'Neill PG, Rokey R, Greenberg S, Pacifico A (1991) Resolution
of ventricular tachycardia and endocardial tuberculoma
following antituberculosis therapy. Chest 100:1467-1469
Permanyer-Miralda G, Sagistra-Sauleda J, Soler-Soler J (1985)
Primary acute pericardial disease: a prospective study of
231 consecutive patients. Am J CardioI56:623-630
Rana BS, Jones RA, Simpson IA (1999) Recurrent pericardial
effusion: the value of polymerase chain reaction in the
diagnosis of tuberculosis. Heart 82:246-247
Rienmuller R, Doppman JL, Lissner J et al (1985) Constrictive
pericardial disease:prognostic significance of a non visual-
ized ventricular wall. Radiology 156:753-755
Rienmuller R, Gurgan M, Erdman E et al (1993) CT and MR
evaluation of pericardial constriction: a new diagnostic
and therapeutic concept. J Thorac Imaging 8:108-121
Rose AG (1987) Cardiac tuberculosis: a study of 19 patients.
Arch Pathol Lab Med III :442-446
Sagrista-Sauleda J, Permanyer-Miralda G, Soler-Soler J (1988)
Tuberculous pericarditis: ten year experience with a pro-
spective protocol for diagnosis and treatment. J Am Coil
Cardiol11:724-728
Sagrista-Sauleda J, Merce J, Permanyer-Miralda G et al (2000)
Clinical clues to the causes of large pericardial effusion.
Am J Med 109:95-101
Sechtem U, Tscholakoff D, Higgins CB(1986) MRI on the
abnormal pericardium. Am J RoentgenoI147:245-252
Seino Y, Ikeda U, Kawaguchi K et al (1993) Tuberculous peri-
carditis presumably diagnosed by PCR analysis. Am Heart
J 126:249-251
Shah S, Miller A, Mastellone A et al (1998) Rapid diagnosis of
tuberculosis in various biopsy and body fluid specimens by
the AMPLICOR Mycobacterium tuberculosis polymerase
chain reaction test. Chest 113:1190-1194
Soulen RL, Stark DD, Higgins CB (1985) Magnetic resonance
imaging of constrictive pericardial disease. Am J Cardiol
55:480
Soyer E, Brunet E, Chevalier B, Leroy J, Morere M, Redonnet
M (1981) Tuberculous aortic insufficiency: report of a case
with successful surgical treatment. J Thorac Cardiovasc
Surg 82:254-256
Spodick DH (1997) Acute, clinically noneffusive ("dry") peri-
carditis. In: Spodick DH (ed) The pericardium: a compre-
hensive textbook. Dekker, New York, pp 94-113
Stark DD, Higgins CB, Lanzer P et al (1984) Magnetic reso-
nance imaging of the pericardium: normal and pathologic
findings. Radiology 150:469-474
Tuberculosis of the Heart and Pericardium
Strang JIG (1997) Tuberculous pericarditis. J Infect 35:
215-219
Strang JIG, Kakasa HHS, Gibson DG et al (1987) Controlled trial
of prednisolone as adjuvant in treatment of tuberculous con-
strictive pericarditis in Transkei. Lancet 2:1418-1422
Strang JIG, Kakasa HHS, Gibson DG et al (1988) Controlled
clinical trial of complete open surgical drainage and of
prednisolone in the treatment of tuberculous pericardial
effusion in Transkei. Lancet 2:759-764
443
Suwan PK, Potjalongsilp S (1995) Predictors of constrictive
pericarditis after tuberculous pericarditis. Br Heart J 73:
187-189
White RD (1998) Magnetic resonance imaging. In: Topol
EJ (ed) Texbook of cardiovascular medicine. Lippincott
Raven, Philadelphia
Woods GL, Goldsmith JC (1989) Fatal pericarditis due to
mycobacterium avium intracellulare in acquired immuno-
deficiency. Chest 1355-1357
28 Mycobacterial Lymphadenitis
M. MONIR MADKOUR and RASHID AL-KUHAYMI

CONTENTS incidence of lymph node tuberculosis remained

constant (Rieder et al. 1990; Cowie and Sharpe 1997;
28.1 Epidemiology 445
Geldmacher et al. 2002). The clinical features of
28.2 HIV and Tuberculous Lymphadenitis 446
28.3 Pathogenesis 446 tuberculous lymphadenitis may mimic other benign
28.4 Clinical Features 447 and malignant diseases and may present as a diag-
28.5 Diagnosis 448 nostic dilemma.
28.6 Histology and Cytology
and Microbiology Findings 448
28.7 Polymerase Chain Reaction 450
28.8 Imaging Findings 450
28.9 Treatment of Mycobacterial Lymphadenitis 451 28.1
References 452 Epidemiology

Tuberculosis and its extrapulmonary manifestations,

particularly of the lymph nodes, are re-emerging as a
Cervical tuberculous lymphadenitis was, perhaps, major problem in the developed western industrial-
first described by the ancient Egyptians in 1534 B.C. ized world. Several contributing factors to such a rise
The medical papyrus of Ebers (Ebers papyrus 1875) in incidence may include the immigration of indi-
comprises no pages and contains the description of viduals from endemic areas, overcrowding, the rise
two cases of tuberculosis of the cervical lymph nodes. in homelessness, the relative difficulties with acces-
These medical papyri were purchased in Luxor by sibility to medical care facilities and infection with
Edwin Smith in 1862 and are now in the university human immunodeficiency virus (HIV) (Bloch et al.
library of Leipzig (see Chap. 1). Cervical lymph node 1987; Rieder et al. 1990; Shriner et al. 1992; Cantwell
tuberculosis is commonly known as scrofula, a Latin et al. 1994; Butt 1997).
term for "glandular swelling" or from the French "full In the United States, tuberculous lymphadenitis
necked sow". occurs in 2-5% of all patients with tuberculosis and in
It was also called "king's evil" during the European 30.9% of extra-pulmonary disease (Appling and Miller
Middle Ages or Dark Ages and was believed to be 1981; Shikhani et al' 1989; Rieder et al' 1990). Rieder
cured by the "king's touch". Surgical excision, as a and colleagues (1990) noted the increased number of
treatment method, was first described by Abu'l-Qasim patients (22,764) with tuberculosis in the USA in 1986,
Khalaf Ibn-Abbas AI-Zahrawi, known as Albucasis as reported by the Centers for Disease Control (CDC).
(936-1013), in his book entitled, Practica - Surgery Ofthese patients, 17.5% had extra-pulmonary manifes-
and Instruments. (see Chap. 2 - Historical Aspects) tations, and the lymph nodes were involved in 30.9%.
Tuberculosis of the lymph nodes is the most The demographic data of patients with extra-pulmo-
common extra-pulmonary manifestation of the nary tuberculosis were related to age, ethnic origin
disease. Despite the reduction in the incidence of and gender. Age groups younger than 15 years were
pulmonary tuberculosis in developed countries, the more likely to have tuberculous lymphadenitis. Racial
and ethnic minorities, foreign-born, black and Asians
were more likely than non-Hispanic white patients to
M. M. MADKOUR, MD, DM, FRCP have tuberculous lymphadenitis. The author noted
Consultant, Department of Medicine, Riyadh Armed Forces female predominance among those with adenitis.
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
R. AL KUHAYMI, FRCS
In Great Britain, Thompson et al. (1992) reported 67
Director of Surgery, Riyadh Armed Forces Hospital, patients with peripheral tuberculous lymphadenitis
P.O. Box 7897, Riyadh 11159, Saudi Arabia who attended Leicestershire group of teaching hospi-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
446
tals between 1979 and 1989. The majority of patients
[54 (81%)] were from the Indian subcontinent (India,
Pakistan or Bangladesh) and only 13 (19%) were of
white ethnic origin. Female predominance was noted
in both ethnic groups and the incidence of family his-
tory was higher in the Indian group.
In Germany, Geldmacher et al. (2002) reviewed
the data of 60 hospitalized patients with tuberculous
lymphadenitis aged 12 years or more over an 8-year
period (1992-1999). Tuberculous lymphadenitis
occurred in 5.1 % of all patients with tuberculosis
hospitalized during the same period. Immigrants
from Afghanistan, India and Pakistan were more
affected with adenitis than native German patients
(70% versus 30%). Female predominance was noted
in both groups. The mean age of native patients
was significantly higher than of the immigrants
(53.8 years versus 35.5 years). Infection with HIV
was present in only one patient.
In Western Australia, Pang (1992) reported 172
patients with culture-positive mycobacterial lymph-
adenitis over an 18-year period (1972-1989). Tuber-
culosis lymphadenitis was found in 53 patients and
non-tuberculous mycobacterial lymphadenitis in
119. Female predominance was noted in both groups.
Tuberculous adenitis was found in 71 % of adult Asian
migrants while non-tuberculous adenitis was found
in 92% of non-Aboriginal Australian children.
28.2
HIV and Tuberculous Lymphadenitis
Several authors have indicated that extra-pulmonary
tuberculosis was more frequent in patients with HIV
than pulmonary involvement and, therefore, extra-
pulmonary tuberculosis became synonymous with
HIV infection (Shafer et al. 1991; Barnes et al. 1991;
Onorato and McCray 1992; Rosenblum et al. 1994).
Shriner et al. (1992) compared the incidence of myco-
bacterial infection of lymph nodes in 11 HIV-sero-
positive patients with 29 HIV-seronegative patients.
They noted that mycobacterial lymphadenitis was
an uncommon manifestation of HIV infection.
The incidence of mycobacterial adenitis primarily
depends upon the endemicity of M. tuberculosis. In
high endemic areas, in HIV-infected patients; lymph
node infection is caused by M. tuberculosis. In USA-
born, HIV-infected patients, adenitis is more likely
caused by non-tuberculous mycobacteria.
In Canada, the high frequency of extra-pulmo-
nary tuberculosis was not attributable to HIV infec-
M. M. Madkour and R. AI-Kuhaymi
tion. Cowie and Sharpe (1997) from Alberta, Canada
reported their findings in a prospective population-
based study on 351 patients with tuberculosis diag-
nosed during a 5-year period (1990 to 1994).
Extra-pulmonary tuberculosis was diagnosed
in 160 (46%) patients, of whom 79 had superficial
lymph node involvement. Patients of Asian ethnic
origin accounted for 82% of those with adenitis with
female predominance (71 %) and immigration from
an Asian country was considered a major risk factor
for tuberculosis lymphadenitis. Although lymph
node tuberculosis is recognized by many authors to
occur in a younger age, in this study, the Asian-born
group was older than other groups. The prevalence of
HIV infection was reported as 1.4%, and the authors
concluded that it was not contributing to the high
incidence of extra-pulmonary tuberculosis.
The high incidence of extra-pulmonary tuberculo-
sis in the Indian subcontinent patients was suspected
to be due to the reduced immunocompetence against
tuberculosis as a result ofVitamin D deficiency caused
by reduced exposure to sunlight (Davies 1985).
In India, Arora and Kumar (1999) reported a pro-
spective study of 190 patients with HIV co-infected
with tuberculosis, and extra-pulmonary tuberculo-
sis was found in 46 patients (24.2%). Tuberculous
lymphadenitis was found in 27 (58.7%) patients with
extra-pulmonary tuberculosis.
Tuberculous cervical lymphadenitis was found to
be the most common cause of cervical lymph node
enlargement. In Singapore, Chao et al. (2002) reported
72 patients with inflammatory cervical lymphadenitis,
and 25 (33.8%) of these were due to tuberculosis.
In Saudi Arabia, tuberculous lymphadenitis was
found to be the most common cause of cervical
lymph node enlargement in approximately 45% of
patients (Thabet et al. 1984; Abba et al. 2002). The
incidence of tuberculous lymphadenitis in patients
with extra-pulmonary disease in Saudi Arabia
ranged between 22.2% and 54.4% (Froude and
Kingston 1982; Mokhtar and Salman 1983). It is more
common among the age group of 20-30 years with
female predominance (Thabet et al. 1984; Ibekwe et
al.1997; Abba et al. 2002).
28.3
Pathogenesis
Tuberculous lymphadenitis is the most common
manifestation of non-respiratory tuberculosis world-
wide. It is often caused by M. tuberculosis but M. bovis
Mycobacterial Lymphadenitis 447

was found to be the cause in 2.8% of a series of 2272 nodes within a group. Adhesion to the overlying skin
mycobacterial isolations from tuberculous lymph- may occur, resulting in induration and development
adenitis in South East England (Grange et al. 1982). of a purplish discoloration.
It is considered that tuberculous lymphadenitis is a The center ofthe enlarging mass lesion becomes soft
local manifestation of a systemic disease (Kent 1967). and caseous; material may rupture into the surround-
Non-tuberculous lymphadenitis is a localized disease ing tissue or through the skin with a sinus formation.
caused by environmental mycobacteria found in the If untreated, discharging sinus may remain unhealed
soil, animals, raw milk, natural and artificial water for years, but healing may take place with scarring, and
systems (Dunn and Hodgson 1982; Pang 1992). The calcification may also occur. Tuberculous mediastinal
infection is transmitted by drinking untreated water, lymph adenitis may enlarge and cause compression of
milk, via oropharyngeal mucosa, salivary glands, ton- major blood vessels, phrenic or recurrent laryngeal
sils or by skin injuries from contaminated objects or nerves or cause erosion of a bronchus or the pericar-
animals, particularly in children and in immuno- dium (see Chap. 17 and 20 - Primary & Post-Primary
compromised adults, particularly with HIV infection Pulmonary Tuberculosis). Asymptomatic intestinal or
(Shriner et al. 1992; Pang 1992). M. avium complex hepatic tuberculosis may spread via lymphatic drain-
is more often the cause than other non-tuberculous age to the mesenteric, hepatic or peripancreatic lymph
mycobacteria such as M. malmoense, M. kansasii, M. nodes (Brizi et al.1998).
chelonei, M. fortuitum and M. scrofulaceum. Axillary lymph node tuberculosis may occur in up
Tuberculous lymphadenitis may occur during to 1% of children following recent vaccination with
primary tuberculous infection or as a result of reacti- BCG (Mori et al.1996). Inguinal lymph node tuberculo-
vation of a dormant foci or by direct extension from sis may occur as a result of tuberculosis of the external
a contiguous focus. In primary pulmonary tubercu- genitalia and may present with discharging inguinal
losis, the bacilli enters the body via inhalation and sinuses (see chapter on tuberculosis of the skin).
bacteremia may occur. Hilar, mediastinal and para-
tracheal lymph nodes are the first site of spread of
infection from the lung parenchyma.
The infection may then spread via the lymphatics to 28.4
the nearest cervical lymph nodes. However, sub-clinical Clinical Features
pulmonary disease with no apparent imaging sequelae
may be a focus of dissemination to the surrounding The main presenting features of tuberculous lymph-
lymphatic. Supraclavicular lymph node involvement adenitis as reported by most authors are palpable
may reflect the lymphatic drainage routes for lung masses in approximately 75% of patients without
parenchyma mycobacterial disease (Shriner et al. constitutional symptoms in about two-thirds of them
1992). Cervical tuberculous lymphadenitis may repre- (see Table 28.1). Cervical lymph nodes are the most
sent a spread from the primary focus of infection in the common site of involvement and reported in 60% to
tonsils, adenoids, sinonasal or osteomyelitis of the eth- 90% of patients (Manolides et al. 1993). The incidence
moid bone (Wadman et al. 1981; Dandapat et al. 1990; of fever, weight loss and night sweating varied widely
Johnson et al.1995; Jha et al. 2001; Jang et al. 2001). in different series and ranged from 10% to 100% of
Tuberculous lymphadenitis most frequently patients. The duration of symptoms before diagnosis
involves the cervical lymph nodes followed in fre- may range from a few weeks to several months (Geld-
quency by mediastinal, axillary, mesenteric, hepatic macher et al. 2002; Aggarwal et al. 2001; Memish et al.
portal, peripancreatic and inguinal lymph nodes 2000; Fain et al. 1999; Chen et al. 1992; Thompson et al.
(Krishnaswamy et al. 1972; Thompson et al. 1992; 1992; Pang 1992).
Freixinet et al. 1995; Mori et al. 1996; Baran et al. Lymph node enlargement is usually painless with
1996; Brizi et al. 1998; Arora and Kumar 1999; Ayed matting due to periadenitis and adhesion to a sur-
and Behbehani 2001; Jang et al. 2001; Geldmacher rounding structure at a single site in 50-70% of patients
et al. 2002). In the initial stage of superficial lymph (Memish et al. 2000; Dandapat et al.1990; Krishnaswamy
node involvement, progressive multiplication of the et al.1972; Deital et al.1989; Hooper 1972). Multiple sites
tubercle bacilli, the onset of delayed hypersensitivity involvement may occur in up to 35% of patients (see
is accompanied by marked hyperemia and swelling, Table 28.1). Complications may be noted at the time of
necrosis and caseation of the centers of the nodes. presentation in 10% to 20% of patients, including skin
This may be followed by marked perinodal inflam- inflammation, abscess formation or cutaneous dis-
mation, progressive swelling and matting with other charging sinus (Pang 1992; Aggarwal et al. 2001).
448 M. M. Madkour and R. Al-Kuhaymi

Table 28.1. Symptoms and sites of tuberculous lymphadenitis (in various series). ND not described

Series Number of Symptoms % Sites %

patients
Fever Weight Sweating Cervical Mediastinal Axillary Inguinal Multiple sites

Geldmacher et al. (2002) 60 NO 100 100 63.3 26.7 8.3 1.7 35

Fain (1999) 59 30.3 47.5 22 73.1 NO 15.4 9.6 15.3
Thompson et al. (1992) 67 11.9 14.9 10.4 85 NO 10.4 4.5 NO
Chen et al. (1992) 71 14.1 9.9 NO 91.5 NO 12.7 7 25.4
Pang (1992) 53 NO NO NO 80 3.7 13.2 3.7 24.5
Oandapat et al. (1990) 80 40 85 37 70 NO 6 9 15
Ayed and Behbehani (2001) 34 all 41 35 NO NO 34 NO NO NO
mediastinal

The clinical features of non-tuberculous mycobac-
teriallymphadenitis are localized to the site involved
and are more rapidly growing with rarity of associ-
ated systemic manifestations (Spyridis et al. 2001;
Flint et al. 2000; Pang 1992). Localized complications
at the site of lymph nodes involvement, such as skin
inflammations, abscess formation and discharging
cutaneous sinuses, are much more frequent than in
tuberculous lymphadenitis. Pang (1992) found these
complications in 52 of the 119 patients (43.7%) with
non-tuberculous mycobacterial lymphadenitis.
28.5
Diagnosis
Tuberculous lymphadenitis remains the most
common cause of cervical lymphadenopathy in
developing countries and reported in about 40%
of cases (Thabet et al. 1984; Abba et al. 2002). The
differential diagnosis is extensive and includes infec-
tion (bacteria, virus, fungi), neoplasms (lymphoma,
metastatic carcinoma, Hodgkin's disease, sarcoma),
sarcoidosis, non-specific reactive hyperplasia and
drug reactions. Medical and social history, particu-
larly of contact with tuberculous patients, and chest
radiograph may be helpful. Tuberculin skin testing
may be positive in over 85% of patients (Ibekwe et al.
1997; Lee et al.1992; Dietel et al.I989). Tuberculin test
may be negative in patients with non-tuberculous
mycobacterial lymphadenitis or in patients with HIV
co-infection. Most patients with cervical tuberculous
lymphadenitis have normal chest radiographs (58%
to 86%) (Ibekwe et al. 1997; Lee et al. 1992; Thomp-
son et al. 1992; Dietel et al. 1989). Histopathological
or cytological diagnosis can be obtained by lymph
nodes tissue biopsies or by fine-needle aspiration
(FNA). The detection of epithelioid cells and granu-
loma is highly suggestive of tuberculosis, but other
diseases may produce similar findings. The confirma-
tion of the diagnosis can be achieved by identifying
the acid-fast bacilli (AFB) by Ziehl-Neelsen (Z-N)
staining of lymph nodes aspirates or biopsies or by
positive cultures from these specimens. Positive cul-
tures of surgical biopsies are reported in 62-79% in
different series (Krishnaswamy et al. 1972; Campbell
and Dayson 1977; Dandapat et al. 1990). The diag-
nosis of tuberculous lymphadenitis can be achieved
with diagnostic accuracy in 76-97% of cases by the
combination of FNA cytology, AFB smear and culture
(Bailey et al. 1985; Das et al. 1990; Das Gupta et al.
1994). They will also differentiate between lymph-
adenitis caused by M. tuberculosis and those caused
by non-tuberculous mycobacterial infection. Lymph
node biopsies may be necessary if FNA was not help-
ful. Mycobacterial culture required for the confirma-
tion of the diagnosis is time consuming. A rapid and
accurate laboratory method of diagnosis with ability
to predict the presence of multi-drug resistance may
be achieved by polymerase chain reaction (PCR) and
PCR-single strand conformational polymorphism
(SSCP) (Gong et al. 2002).
28.6
Histology and Cytology and Microbiology
Findings
Excisional biopsy of lymph nodes remains a common
method of obtaining tissue specimen in developing
countries for histological and microbiological exami-
nations. Abba et al. (2002) from Saudi Arabia retrospec-
tively reviewed the clinicopathological features in 258
patients with cervical lymphadenopathy. All patients
had excisional biopsies from the lymph nodes and
tuberculous granulomata were found in 98 (37.9%). In
western Australia, Pang (l992) performed excisional
biopsies on 151, by needle aspirations in seven and
Mycobacterial Lymphadenitis
by sterile swab obtained at incisional drainage in ten.
All had culture-positive results as well as histological
features of granulomas. Al-Hadrani et al. (2000) from
Yemen reported 302 patients with peripheral tubercu-
lous lymphadenitis prospectively studied between Jan-
uary 1995 and December 1997. All patients underwent
lymph node biopsy by surgical excision for histological
examination, and 284 (94%) had histological findings
consistent with tuberculosis. The authors noted that
mycobiological examinations were not useful as only
5 cases of 17 had positive cultures. During the past two
decades, FNA has been used more by several authors as
the initial technique for the diagnosis of mycobacterial
lymphadenitis, and if cytological findings are uncer-
tain, tissue biopsies by surgery are usually advisable
(Geldmacher et al. 2002; Pang 1992; Lau et al. 1990,
1991; Kamboj et al. 1994; Shaha et al. 1986; Khan et al.
1994; Chao et al. 2002; Dandapat et al. 1990; Qadri et
al. 1991). The sensitivity and specificity of FNA cytol-
ogy in the diagnosis of tuberculous lymphadenitis
were 88% and 96%, respectively, as reported by Chao
et al. (2002). Lau et al. (1990) reported the results of a
consecutive series of 1349 FNA biopsies from the head
and neck region of 1193 patients and found that 108
patients showed granulomatous changes. Only 68 of
them had subsequent surgery with histological diag-
nosis, and tuberculous lymphadenitis was confirmed
in 63. The authors indicated that the specificity of FNA
in diagnosing granulomatous lymphadenitis in their
series was 63/68 (93%). Smears from 52 patients stained
with the Z-N technique showed AFB in 19 (37%). Of
the AFB-positive smears, 13 and six were associated
with granulomatous and non-specific inflammatory
cytological findings, respectively. Of the smears of the
same 52 patients, nine had positive culture of myco-
bacterial tuberculosis. The authors indicated that the
overall sensitivity of FNA in diagnosing tuberculous
cervical lymphadenopathy was 77%. Cultures of the
specimen obtained by excisional biopsy showed Myco-
bacterium tuberculosis in 80% of cases. In Saudi Arabia,
the specificity of FNA was found to be 93%. Qadri et
al. (l991) described the sensitivity and specificity of
FNA in the diagnosis of mycobacterial lymphadenitis.
The authors reported positive smear staining in 23%
while cultures were positive in 53%. Khan et al. (1994)
from Saudi Arabia had similar results of FNA micro-
biological findings and reported AFB-positive smear
staining in 25% and cultures in 62%. Dandapat et al.
(1990) from India reported that FNA cytology was
positive in 83%, culture of surgical biopsies were posi-
tive in 65% and noted that "In our practice, biopsy in
always undertaken to confirm the diagnosis". In India,
Prasoon (2000) used FNA cytology for evaluation of
449
783 patients who presented with lymphadenopathy due
to various causes over a 4-year period (1994 and 1998),
and 213 (27.2%) were tuberculous. Cytological patterns
included epithelioid cell granulomas alone with co-
existence of necrosis, AFB or both, and necrosis with
AFB. The author noted that the chance of finding AFB
was highest in patients presenting with cold abscess.
The presence of granuloma and of AFB had an inverse
relationship.
Geldmacher et al. (2002) found that both FNA and
lymph node excisional biopsies were similarly effec-
tive methods in obtaining sufficient lymph node tissue
for histological and microbiological examinations.
The cytological and histological analyses were equally
positive and revealed epithelioid cells in 88.3%, giant
cells in 55% and caseation in 43.3%. Microbiological
confirmation by cultures was equally positive in 40%
by FNA or biopsies. Z-N staining was positive in 13.3%
in both FNA and excisional biopsies.
Lymph node samples obtained by FNA can provide
sufficient collection of material for cytological and
bacteriological examination and is considered as an
alternative and easy procedure. Gupta et al. (1993) from
Kuwait reported the cytodiagnostic and microbiologi-
cal confirmation findings by FNA on 102 patients with
tuberculous lymphadenitis. Patients were diagnosed
by FNA cytological examination with tuberculous
lymphadenitis. They were subjected to mycobacterial
examination by Z-N staining and by culture of direct
or of concentrated smear materials. Z-N staining for
AFB was positive in 29.5%. The cultures for mycobac-
teria were positive in 49%. The authors concluded that
cytodiagnosis supplemented with AFB smear staining
and culture helped in establishing a definite diagnosis
in 56.9% of cases.
There are only few comparative studies on the diag-
nostic characteristics of mycobacterial lymphadenitis
among patients co-infected with HIV and those with-
out HIY. Shriner et al. (1992) from California reported
a 16-year experience of 40 patients with tuberculous
lymphadenitis of whom 11 were HIV-seropositive
and 29 were HIV-seronegative. Positive AFB stain was
significantly higher in HIV-seropositive than sero-
negative patients (72.7% versus 14.8% respectively).
Positive FNA culture was significantly higher in HIV-
seropositive than seronegative patients (100% versus
12.5% respectively). Similar comparative diagnostic
studies were reported by Finfer et al. (199l) from
New York. In a total of 30 patients with tuberculous
lymphadenitis; 13 were with AIDS or ARC and 17 were
without AIDS. Positive smear stain for AFB was found
in 9 of 13 patients with AIDS and ARC, but found in
only 1 of 10 non-AIDS patients.
450
Isolated tuberculous mediastinal and hilar lymph-
adenopathy in the absence of parenchymal lung
disease may present as a diagnostic challenge (Ayed
and Behbehani 2001; Domingo 1996; Julia et al. 1990;
Dhand et al. 1979; Liu et al. 1978). Several diagnostic
methods that require procedures such as fiberoptic
bronchoscopy, transthoracic image guided FNA using
computed tomography (CT), magnetic resonance
imaging, fluoroscopy and ultrasound are well estab-
lished for the diagnosis of mediastinallymphadenop-
athy. Mediastinoscopy or thoracotomy may rarely be
performed when other procedures are not diagnostic
or if lymphoma is suspected (Khan et al. 1994; Ayed
and Behbehani 2001; Zwischenberger et al. 2002).
Ayed and Behbehani (2001) reported 34 patients with
isolated mediastinal tuberculous lymphadenitis with
no parenchymal lung and all had bronchoscopy and
mediastinoscopy lesions. Bronchoscopy showed endo-
bronchial hyperemia in seven patients (21 %); of them,
three had positive culture for Mycobacterium tubercu-
losis, while granulomatous inflammation was found in
all. Mediastinoscopy confirmed that diagnosis in all
34 [100% patients, granuloma in 100%, and positive
microbiological features in 30 patients (88.2%)].
Baran et al. (1996) reported their findings on
17 patients with isolated tuberculous mediastinal
lymphadenitis without parenchymal lung lesions.
All patients had bronchoscopy and 15 had endo-
bronchial abnormalities and biopsies were obtained.
Bronchoscopy yielded the diagnosis in nine patients
(53%). Mediastinoscopy or thoracotomy was per-
formed on six patients that were not diagnosed by
bronchoscopy and the diagnosis was confirmed in
100% of them.
28.7
Polymerase Chain Reaction
Detection of Mycobacterium tuberculosis DNA from
FNA material using polymerase chain reaction (PCR)
assays has been extensively reported (Kim et al.I996).
Conventional utilization of Z-N staining requires a
high bacillary load of 10,000-100,000 per ml of aspi-
rate material to be able to detect it. Various series have
reported a wide range of AFB smear positivity, from
0% to 75% (Arora and Arora 1990; Das Gupta et al.
1994; Ellison et al. 1999; Harrison and Jayasundera
1999; Ibekwe et al. 1997; Kumar et al. 1998).
The molecular basis of mycobacterial resistance to
the antituberculous agents has been recently elucidated
using the SSCP method, PCR-SSCP (Banerjee et al.
M. M. Madkour and R. Al-Kuhaymi
1994; Morris et al.1995; Gong et al. 2002). It is also pos-
sible that PCR can differentiate between lymphadenitis
caused by Mycobacterium tuberculosis and that caused
by nontuberculous mycobacterial infection. Gong et al.
(2002) used one-step PCR and PCR-SSCP for the pur-
poses of diagnosis and detection of rifampicin-resistant
strain on FNA sediment material. Sixty-three patients
with positive bacteriologic confirmation and/or clini-
cal evidence of tuberculosis and a control group of 55
cases of non-tuberculosis were studied over a 24-month
period. The bacilli were detected by PCR in 49 of 63
cases of tuberculosis (77.8%) while by Z-N stain in 25
of 63 (39.7%). Twenty-three cases were positive by PCR
and Z-N staining. Twenty-six cases had positive PCR
but negative Z-N stain. Two cases had negative PCR
but positive Z-N stain. Twelve cases had negative PCR
and Z-N stain. The rpo B gene mutation was observed
in seven of twenty-two cases of tuberculous lymphad-
enitis (31.8%). The authors concluded that FNA cytol-
ogy supplemented by PCR methods and conventional
examinations could confirm the diagnosis and predict
the presence of multi-drug-resistance bacilli.
The comparison between the use of conventional
methods ofdiagnosis by FNA and the use ofPCR and its
practical application in developing countries has been
reported by Goel et al. (2001). The authors reported 142
patients with peripheral lymphadenitis due to various
causes. Tuberculosis was diagnosed in 78 (54.9%),
based on cytology alone in 71.8% and increased to
83.8% when supplemented by AFB smear stain. The
correct diagnosis was achieved in 94.8% of patients
when the four conventional methods were combined
(FNA cytology, AFB smear, culture and biopsy). The
peR was done on 52 cases and was positive in 39, of
them 13 were non-tuberculous. Tuberculosis was con-
firmed by PCR in 37 of39 patients (94.6%). The authors
recommended that PCR should only be reserved for
the problem cases in developing countries.
28.8
Imaging Findings
Most patients with tuberculous cervical lymphad-
enitis have normal chest radiography. However, an
associated abnormal chest radiograph suggestive of
pulmonary tuberculosis was present in 42% (Chen
et al. 1992). Power Doppler sonography of tubercu-
lous cervical lymphadenitis may show avascularity,
displaced hilar vessels and low intranodal vascular
resistance, yet remains non specific (Ahuja et al. 2001a,
b). Lymph node calcifications may be depicted in less
Mycobacterial Lymphadenitis 451

than 10% of patients (Geldmacher et al. 2002; Khan et tomycin (SM) and isoniazid (INH) is equally effec-
al. 2000). Hilar mediastinal and paratracheal adenopa- tive and had no greater relapse rate than an 18-month
thy can be depicted by chest X-rays (See imaging of regimen (Campbell and Dayson 1979).
pulmonary tuberculosis chapter). Enhanced CT scan In February 2003, the joint committee of the Ameri-
of the mediastinal tuberculous lymphadenitis may can Thoracic Society/Centers for Disease Control and
show peripheral ring enhancement and hypodensity Prevention/Infectious Diseases Society of America
of the central portion of the lymph nodes (see imag- published their recommendations on the treatment
ing of pulmonary tuberculosis chapter). Mesenteric, of pulmonary and extra-pulmonary tuberculosis (AT/
hepatic and peripancreatic tuberculous lymphadeni- CDC/IDSA 2003). In tuberculous lymphadenitis, a 6-
tis may be depicted by enhanced abdominal CT scan. month regimen of four drugs, INH, RIF, PZA and EMB,
The most common CT features include peripheral ring should be given in the initial phase for 2 months and
enhancement and hypodensity of the central part of the continuation phase of treatment should consist of
the lymph nodes. Homogeneous or inhomogeneous INH and RIF for a minimum of 4 months. In the initial
enhancement or lack of enhancement oflymph nodes 2-month phase of a 6-month regimen, INH, RIF, PZA
may be seen (Brizi et al.1998; Moon et al.1998; Pombo and EMB should be given daily throughout (regimen
et al. 1992; 1m et al. 1987). 1), daily for 2 weeks followed by two times weekly for
6 weeks (regimen 2), or three times a week (regimen
3). The continuation phase may be given daily, two
times weekly by DOT or three times weekly by DOT.
28.9 In the presence of INH resistance, there are fewer
Treatment of Mycobacterial Lymphadenitis failures and relapses using this regimen. If drug
susceptibility becomes available and the organisms
Most patients with mycobacterial lymphadenitis are were found to be susceptible to INH, then EMB can
initially treated with anti-tuberculous drugs. The dif- be discontinued in the initial2-month phase of treat-
ferentiation between tuberculous and non-tubercu- ment regimens.
lous forms of mycobacterial lymphadenitis cannot be The ATS/CDC/IDSA Committee (2003) indicated
ascertained by the histological evidence of caseating that therapeutic lymph node excision is not indicated
or necrotizing granulomata with or without positive except in unusual circumstances. Aspiration or inci-
microscopy for AFB, and the culture and PCR may be sion and drainage of large fluctuant lymph nodes
useful (Pang 1992). may be beneficial (Cheung et al.1988).
The guidelines for recommended treatment of In non-tuberculous mycobacterial lymphadenitis,
tuberculous lymphadenitis are similar to treatment most authors agree that effective treatment is often
regimens used for pulmonary tuberculosis. Regimens achieved by surgical excision of the affected lymph
may vary in various epidemiological and economical nodes (Jones and Campbell 1962; MacKellar 1976;
circumstances, particularly in high-incidence low- Harris et al. 1982; Joshi et al. 1989; Campbell 1990). In
income areas as compared with low-incidence high- Western Australia, Pang (1992) reviewed the records
income parts of the world. of 172 patients with culture-positive mycobacterial
Recommendations target, in general, countries in lymphadenitis over an 18-year period between January
which mycobacterial culture, drug susceptibility test- 1972 and December 1989. Tuberculous lymphadenitis
ing, radiographic facilities and second-line drugs are was found in 53 and non-tuberculous mycobacte-
not widely available as a routine. In endemic areas, rial lymphadenitis was found in 119 patients. All 119
particularly in Asian countries, rising prevalence of patients with non-tuberculous mycobacteriallymphad-
resistant M. tuberculosis makes it difficult to provide enitis had surgical intervention, including 91 with total
appropriate treatment when cultures are not available. excision, 15 with drainage plus total excision, five with
The World Health Organization and the International incisional drainage alone and eight patients had second
Union Against Tuberculosis and Lung Disease recom- excision for relapse or residual disease. Favorable out-
mendations of treatment are built around a national come of surgical treatment of non-tuberculous myco-
case management strategy, "Directly Observed Ther- bacterial lymphadenitis was confirmed by the authors
apy - Short course" (DOTS). These recommendations at 12 months in 86 patients (72.3%). The remaining
were based on and supported by the British Medical 33 patients (27.7%) were lost for follow-up after sur-
Research Council studies in 1973,1977,1985,1988. gery. Flint and colleagues (2000) from New Zealand
A 6- to 8-month regimen with addition of rifampin reported 57 children with non-tuberculous mycobacte-
(RIF) or pyrazinamide (PZA) to a base of daily strep- rial lymphadenitis that were treated surgically. Patients
452
who had excision achieved a significantly greater heal-
ing and lower re-operation rates than those treated with
incision and drainage, curettage or aspiration.
Spyridis et al. (2001) reported 50 children with
mycobacterial cervical lymphadenitis who were
examined between 1982 and 1997. Tuberculous
adenitis was found in 24 and non-tuberculous myco-
bacterial adenitis in 26 patients. All patients with
non-tuberculous mycobacterial adenitis were man-
aged by total excision of affected lymph nodes.
Isolated tuberculous mediastinal or peripheral
lymphadenitis in adults is usually treated with anti-
tuberculous drugs alone without surgical interven-
tion (Dandapat et al. 1990; Ayed and Behbehani
2001). In children, however, surgical intervention
may rarely be indicated if it is accompanied by
a severe tracheobronchial compression or when
bronchial lymph fistula is detected (Byrd et al. 1976;
Freixinet et al. 1995).
Reduction of the size of the lymph nodes swell-
ing without complications may occur in 70-90% of
patients during and after the completion of treat-
ment (Wong and Jafek 1974; Campbell 1990; Geldm-
acher et al. 2002). The relapse rate following standard
short-course regimen in 427 children with tubercu-
lous lymphadenitis was reported as 2.8% by McMas-
ter et al. (2000). Paradoxical response to treatment
with further enlargement or additional swelling of
new lymph nodes may occur in 6-30% of patients
within the first 2 months of initiating treatment
(Campbell 1990; Carter and Mates 1994; Memish et
al. 2000; Geldmacher et al. 2002). The pathogenesis of
this paradoxical response to treatment is not clearly
explained.
It is postulated to be due to hypersensitivity
response to various antigens of the tubercle bacilli
(Afghani and Lieberman 1994; Carter and Mates
1994; Lieberman 1994; Campbell and Dayson 1977;
Obeid et al. 1999; Geldmacher et al. 2002). It usu-
ally resolves spontaneously without alteration in
the course of treatment. In approximately 10% of
patients, local complications including ulceration,
fistulas or abscess formation may occur after initiat-
ing treatment.
The treatment of tuberculous lymphadenitis in
patients co-infected with HIV is essentially the same as
in those without HIY. The ATS/CDC/IDSA Committee
(2003) indicated there are two important exceptions to
this generalization in patients with HIV:
1. In the continuation phase, once weekly INH-RIF
should not be used
2. Twice weekly INH-RIF should not be used if
CD4+ lymphocyte counts less than 100/,.d.
M. M. Madkour and R. Al-Kuhaymi
All rifamycins are inducers of metabolic pathways
and may result in a decrease in serum concentration
of anti-retroviral agents to sub-therapeutic levels.
Rifapentine is less of an inducer than RIF and may
be used.
References
Abba AA, Bamgboye AE, Afzal M et al (2002) Lymphadenopa-
thy in adults. Saudi Med J 23:282-286
Afghani B, Lieberman JM (1994) Paradoxical enlargement or
development of intracranial tuberculomas during therapy:
case report and review. Clin Infect Dis 19:1092-1099
Aggarwal P, Wali JP, Singh S et al (2001) A clinico-bacteriologi-
cal study of peripheral tuberculous lymphadenitis. J Assoc
Physicians India 49:808-812
Ahuja A, Ying M, Yuen YH et al (2001a) Power Doppler
sonography to differentiate tuberculous cervical lymph-
adenopathy from nasopharyngeal carcinoma. AJNR Am J
NeuroradioI22:735-740
Ahuja A, Ying M, Yuen YH et al (2001b) Power Doppler sonog-
raphy cervical lymphadenopathy. Clin Radiol 56:965-969
AI-Hadrani AM, Aulaqi SM, Al Salami SF et al (2000) Manage-
ment strategies for peripheral tuberculous lymphadenopa-
thy. Saudi Med J 21:266-269
American Thoracic Society/Centers for Disease Control and
Prevention/Infectious Diseases Society of America (2003)
Treatment of tuberculosis. Am J Respir Crit Care Med 167:
603-662
Appling D, Miller RH (1981) Mycobacterial cervicallymphade-
nopathy update. Laryngoscope 91:1259-1266
Arora B, Arora DR (1990) Fine needle aspiration cytology in
diagnosis of tuberculous lymphadenitis. Ind J Med Res 91:
189-192
Arora VK, Kumar SV (1999) Pattern of opportunistic pulmo-
nary infections in HIV seropositive subjects: observations
from Pondicherry, India. Indian J Chest Dis Allied Sci 41:
135-144
Ayed AK, Behbehani NA (2001) Diagnosis and treatment of
isolated tuberculous mediastinal lymphadenopathy in
adults. Eur J Surg 167:334-338
Bailey TM, Akhtar M, Ali MA (1985) Fine needle aspiration
biopsy in the diagnosis of tuberculosis. Acta Cytol 29:
732-736
Banerjee A, Dubnau E, Quemard A et al (1994) InhA, a gene
encoding a target for isoniazid and ethionamide in Myco-
bacterium tuberculosis. Science 363:227-230
Baran R, Tor M, TahaoJu K et al (1996) Intrathoracic tubercu-
lous lymphadenopathy: clinical and bronchoscopic features
in 17 adults without parenchymal lesions. Thorax 51:87-89
Barnes PF, Bloch AB, Davidson PT et al (1991) Tuberculosis in
patients with human immunodeficiency virus infection. N
Engl J Med 324:1644-1650
Bloch AB, Rieder HL, Kelley GD et al (1987) The epidemiol-
ogy of tuberculosis in the United States. Clin Chest Med
10:297-313
British Medical Research Council (East African) (1973)
Controlled clinical trial of four short-course (6 month)
regimens of chemotherapy for treatment of pulmonary
tuberculosis. Second report. Lancet 1:1331-1338
Mycobacterial Lymphadenitis
British Medical Research Council (East African) (1974)
Controlled clinical trial of four short-course (6-month)
regimens of chemotherapy for treatment of pulmonary
tuberculosis. Third report. Lancet 2:237-240
British Medical Research Council (Hong Kong Chest Service)
(1977) Controlled trial of a 6-month and 9-month regimens
of daily intermittent streptomycin plus isoniazid plus pyra-
zinamide for pulmonary tuberculosis in Hong Kong. Am
Rev Respir Dis 115:727-735
British Thoracic Society Research Committee (1985) Short
course chemotherapy for lymph node tuberculosis: a con-
trolled trial. Br J Dis 290: 1106-11 08
British Thoracic Society Research Committee (1988) Short
course chemotherapy for lymph node tuberculosis: final
report at 5 years. Br J Dis Chest 82:282-284
Brizi MG, Celi G et al (1998) Diagnostic imaging of abdomi-
nal tuberculosis: gastrointestinal tract, peritoneum, lymph
nodes. Rays 23:115-125
Butt AA (1997) Nasal tuberculosis in the 20th century. Am J
Med Sci 313:332-335
Byrd RB, Griggs GA David G et al (1976) Surgical complica-
tions of cervical and mediastinal tuberculous adenitis in
an infant. Chest 70:544-546
Campbell IA (1990) The treatment of superficial tuberculous
lymphadenitis. Tubercle 71:1-3
Campbell lA, Dyson AJ (1977) Lymph node tuberculosis: a
comparison of various methods of treatment. Tubercle
548:171-179
Campbell lA, Dyson AJ (1979) Lymph node tuberculosis: a
comparison of treatments 18 months after completion of
chemotherapy. Tubercle 60:95-98
Cantwell MF, Snider DE, Cauther GM et al (1994) Epidemiol-
ogy of tuberculosis in the United States, 1985 through 1992.
Am Med Assoc 272:535-539
Carter EJ, Mates S (1994) Sudden enlargement of a deep cervi-
cal lymph node during and after treatment for pulmonary
tuberculosis. Chest 106: 1896-1898
Chao SS, Loh KS, Tan KK et al (2002) Tuberculous and nontu-
berculous cervical lymphadenitis: a clinical review. Otolar-
yngol Head Neck Surg 126:176-179
Chen YM, Lee PY, SU WJ et al (1992) Lymph node tuberculo-
sis: 7 year experience in Veterans General Hospital Taipei,
Taiwan. Tuber Lung Dis 73:368-371
Cheung WL, Siu KF, Ng A (1988) Tuberculous cervical abscess:
comparing the results of total excision against simple inci-
sion and drainage. Br J Surg 75:563-564
Cowie RL, Sharpe JW (1997) Extra-pulmonary tuberculosis: a
high frequency in the absence ofHIV infection. Int J Tuberc
Lung Dis 1:159-162
Dandapat MC, Mishra BM, Dash SP et al (1990) Peripheral lymph
node tuberculosis: a review of 80 cases. Br J Surg 77:911-912
Das DK, Pant IN, Chachra KL (1990) Tuberculous lymphad-
enitis: Correlation of cellular components and necrosis
in lymph node aspirate with AFB positivity and bacillary
count. Ind J Pathol MicrobioI33:1-10
Das Gupta A, Ghosh TN, Podder AK et al (1994) Fine needle
aspiration of cytology of cervical lymphadenopathy with
special reference to tuberculosis. J Ind Med Assoc 92:44-46
Davies PD~ (1985) A possible link between vitamin D defi-
ciency and impaired host defence to Mycobacterium
tuberculosis. Tubercle 66:301-306
Deitel M, Bendago M, Krajeden S et al (1989) Modern manage-
ment of cervical scrofula. Head Neck 1:60-66
453
Dhand S, Fisher M, Fewell JW (1979) Intrathoracic tuberculous
lymphadenopathy in adults. JAMA 241:505-507
Dietel M, Bendago M, Kraijden S et al (1989) Modern manage-
ment of cervical scrofula. Head Neck 11:60-66
Domingo P (1996) Isolated mediastinal tuberculous lymphad-
enitis. Arch Intern Med 156:1582
Dunn BL, Hodgson DJ (1982) ,Atyical' mycobacteria in milk. J
Appl Bacteriol 52:373-376
Ellison E, Lapuerta P, Martra SE (1999) Fine aspiration diagno-
sis of mycobacterial lymphadenitis: sensitivity and predic-
tive value in the United States. Acta CytoI43:153-157
Fain 0, Lortholary 0, Djouab M, Amoura I et al (1999) Lymph
node tuberculosis in the suburbs of Paris: 59 cases in adults
not infected by the human immunodeficiency virus. Int J
Tuberc Lung Dis 3:162-165
Finfer M et al (1991) Fine needle aspiration biopsy diagnosis
of tuberculous lymphadenitis in patients with and without
the acquired immune deficiency syndrome. Acta Cytol 35:
325-332
Flint D et al (2000) Cervical lymphadenitis due to non-tuber-
culous mycobacteria: surgical treatment and review. Int J
Pediatr OtorhinolaryngoI53:187-194
Freixinet J, Varela A, Rivero LL et al (1995) Surgical treatment
of childhood mediastinal tuberculous lymphadenitis. Ann
Thorac Surg 59:644-646
Froude JRL, Kingston M (1982) Extrapulmonary tuberculosis
in Saudi Arabia: a review of 162 cases. King Faisal Specialist
Hosp Med J 2:85-95
Geldmacher H, Taube C Kroeger C et al (2002) Assessment
of lymph node tuberculosis in Northern Germany. Chest
121:1177-1182
Goel MM, Ranjan V, Dhole TN et al (2001) Polymerase chain
reaction vs. conventional diagnosis in fine needle aspirates
of tuberculous lymph nodes. Acta CytoI45:333-340
Gong G, Lee H, Kang GH et al (2002) Nested PCR for diagnosis
of tuberculous lymphadenitis and PCR-SSCP for identifica-
tion of rifampicin resistance in fine needle aspirates. Cyto-
pathoI26:228-231
Grange JM, Collins C, Yates M (1982) Bacteriological survey
of tuberculous lymphadenitis in South-east England:
1973-1980. J Epidemiol Community Health 36:157-161
Gupta SK, Chugh TD, Sheikh ZA et al (1993) Cytodiagnosis of
tuberculous lymphadenitis: a correlative study with micro-
biologic examination. Acta CytoI37:329-332
Harris BH, Webb HW, Wilkinson AH Jr et al (1982) Mycobacte-
rial lymphadenitis. J Pediatr Surg 17:589-590
Harrison AC, Jayasundera J (1999) Mycobacterial cervical
adenitis in Auckland: diagnosis by the fine needle aspirate.
NZ Med J 112:7-9
Hooper AA (1972) Tuberculous peripheral lymphadenitis. Br
J Surg 59:353-359
Ibekwe AO, Al Shareef Z, Al Kindy S (1997) Diagnostic prob-
lems of tuberculous cervical adenitis (scrofula). Am J
OtolaryngoI18:202-205
1m JG, Song KS, Kang KS et al (1987) Mediastinal tubercu-
lous Iymphadneitis: CT manifestations. Radiology 164:
115-119
Jang YJ, Jung SW, Koo TW et al (2001) Sinonasal tuberculosis
associated with osteomyelitis of the ethmoid bone and cer-
vicallymphadenopathy. J Laryngol OtoI115:736-739
Jha BC, Dass A, Nagarkar NM et al (2001) Cervical tuberculous
lymphadenopathy: changing clinical pattern and concepts
in management. Postgrad Med J 77:185-187
454
Johnson IJM, Soames JV, Marshall HF (1995) Nasal tuberculo-
sis - an increasing problem? J Laryngol OtoI109:326-327
Jones PJ, Campbell PE (1962) Tuberculous lymphadenitis in
childhood: the significance of anonymous mycobacteria.
Br J Surg 50:302-314
Joshi W, Davidson PM, Jones PG et al (1989) Non-tuberculous
mycobacterial lymphadenitis in children. Eur J Pediatr 148:
751-754
Julia G, Rodriguez de Castro F, Fernandez JM et al (1990)
Transcarinal needle aspiration in the diagnosis of medi-
astinal adenitis in a patient infected with the immunodefi-
ciency virus. Thorax 45:414-415
Kamboj SS, Goel MM, Tandon P (1994) Correlation study of
histopathology and bacteriology in patients of tubercular
lymphadenitis. Ind J Chest Dis Allied Sci 36:187-191
Kent DC (1967) Tuberculous lymphadenitis: not a localized
disease process. Am J Med Sci 254:866-874
Khan J, Akhtar M, von Sinner WN, Bouchama A et al (1994)
CT-guided fine needle aspiration biopsy in the diagnosis of
mediastinal tuberculosis. Chest 106:1329-1332
Khan MQ, AI-Qahtani AQ, AI-Momen S et al (2000) Wide-
spread tuberculous calcification. Saudi Med J 21:386-389
Kim SS, Chung SM, Kim IN et al (1996) Application of
PCR from the fine needle aspirates for the diagnosis of
cervical tuberculous lymphadenitis. J Kor Med Sci 11:
127-132
Krishnaswamy H, Koshi G, Kulkarni KG et al (1972) Tubercu-
lous lymphadenitis in South India - a hospitalpathological
and bacteriological study. Tubercle 53:215-220
Kumar N, Tiwari MC, Verma K (1998) AFB staining in cyto-
diagnosis of tuberculosis without classical features: a
comparison of Ziehl Neelsen and f1uoroscent methods.
Cytopathology 9:208-214
Lau SK, Wei WI, Hsu C et al (1990) Efficacy of fine needle
aspiration cytology in the diagnosis of tuberculous cervi-
cal lymphadenopathy. J Laryngol Otoll04:24-27
Lau SK, Wie WI, Kwan S et al (1991) Combines use of fine
needle aspiration cytologic examination and tuberculin
skin test in the diagnosis of cervical tuberculous lymph-
adenitis. Arch Otolaryngol Head Neck Surg 117:87-90
Lee KC, Tami TA, Lalwani AK et al (1992) Contemporary man-
agement of cervical tuberculosis. Laryngoscope 102:60-64
Liu C, Fields WR, Shaw C (1978) Tuberculous mediastinal
lymphadenopathy in adults. Radiology 126:369-371
MacKellar A (1976) Diagnosis and management of atypical
mycobacterial lymphadenitis in children. J Pediatr Surg
11:85-89
Manolidis S, Frenkiel S, Yoskowitch A et al (1993) Mycobcterial
infections of the head and neck. Otolaryngol Head Neck
Surg 109:427-433
McMaster P, Ezeilo N, Vince JD (2000) A retrospective review
of paediatric lymph node tuberculosis in Port Moresby,
Papua New Guinea. Ann Trop Paediatr 20:223-226
Memish ZA, Mah MW, Mahmood SA et al (2000) Clinico-diag-
nostic experience with tuberculous lymphadenitis in Saudi
Arabia. Clin Microbiol Infect 6:137-141
Mokhtar A, Salman K (1983) Extrapulmonary tuberculosis.
Saudi Med J 4:317-322
Moon WK, 1m JG, Yeon KM et al (1998) Mediastinal tuber-
M. M. Madkour and R. AI-Kuhaymi
culous lymphadenitis: CT findings of active and inactive
disease. AJR 170:715-718
Mori T, Yamauchi Y, Shiozawa K (1996) Lymph node swelling
due to bacilli Calmette-Guerin vaccination with multi-
puncture method. Tuberc Lung Dis 77:269-273
Morris S, Bai GH, Suffys P et al (1995) Molecular mechanism
of multiple drug resistance in clinical isolates of Mycobac-
terium tuberculosis. J Infect Dis 171:954-960
Obeid T, Mousali Y, Wali S et al (1999) Paradoxical enlargement
of lymph nodes during therapy of central nervous system
tuberculosis. Ann Saudi Med 19:530-532
Onorato 1M, McCray E (1992) Prevalence of human immuno-
deficiency virus infection among patients attending tuber-
culosis clinics in the United States. J Infect Dis 165:87-92
Pang SP (1992) Mycobacterial lymphadenitis in Western Aus-
tralia. Tuberc Lung Dis 73:362-367
Pombo F, Rodriguez E, Mato J (1992) Patterns of contrast
enhancement of tuberculous lymph nodes demonstrated
by computed tomography. Clin RadioI46:13-17
Prasoon D (2000) Acid-fast bacilli in fine needle aspiration
smears from tuberculous lymph nodes. Where to look for
them. Acta CytoI44:297-300
Qadri SMH, Akhtar M,Ali MA (1991) Sensitivity of fine needle
aspiration biopsy in the detection of mycobacterial infec-
tions. Diagn CytopathoI7:142-146
Rieder HL, Snider DE Jr, Cauthen GM (1990) Extrapulmonary
tuberculosis in the United States. Am Rev Respir Dis 141:
347-351
Rosenblum LS, Castro KG, Dooley S et al (1994) Effect of HIV
infection and tuberculosis on hospitalizations and cost of
care for young adults in the United Sates, 1985 to 1990. Ann
Intern Med 121:786-792
Shafer RW, Kim DS, Weiss JP et al (1991) Extrapulmonary
tuberculosis in patients with human immunodeficiency
virus infection. Medicine 70:384-397
Shaha A, Webber C, Marti J (1986) Fine needle aspiration in
the diagnosis of cervical lymphadenopathy. Am J Surg 152:
320-423
Shikhani AH, Hadi UM, Mufarrij AA et al (1989) Mycobacterial
cervical lymphadenitis. Ear Nose Throat J 68:660-672
Shriner KA, Mathisen GE, Goetz MB (1992) Comparison of
mycobacterial lymphadenitis among persons infected with
human immunodeficiency virus and seronegative controls.
Clin Infect Dis 15:601-605
Spyridis P et al (2001) Mycobacterial cervical lymphadenitis in
children: clinical and laboratory factors of importance for
differential diagnosis. Scand J Infect Dis 33:362-366
ThabetJH, Ali Farzand, Ibrahim BB (1984) Incidence of tuber-
culous neck masses in the central zone of Saudi Arabia. J R
Coli Surg Edinb 29:226-228
Thompson MM, Underwood MJ, Sayers RD et al (1992) Periph-
eral tuberculous lymphadenopathy. Br J Surg 79:763-764
Wadman SR, Levine HL, Sebek BA et al (1981) Nasal tubercu-
losis: a forgotten entity. Laryngoscope 91: 11-16
Wong ML, Jafek BW (1974) Cervical mycobacterial disease.
Trans Am Acad Ophthalmol Otolaryngol 78:75-87
Zwischenberger JB, Savage C, Alpard SK et al (2002) Mediasti-
nal transthoracic needle and core lymph node biosy. Chest
121:1165-1170
29 Tuberculosis and Co-infection
with the Human Immunodeficiency Virus
ALIMUDDIN ZUMLA and JOHN M. GRANGE

CONTENTS 29.11.1 Barriers to Effective and Appropriate Therapy 471

29.11.2 Immune Reconstitution Syndrome 471
29.1 Historical Background 455 29.11.3 Increased Drug Side Effects 471
29.2 Risk Factors for Tuberculosis 456 29.11.4 Drug Resistance 472
29.3 Tuberculosis and HIV: 29.11.5 Recurrence After Treatment 472
Epidemiological Considerations 456 29.11.6 Post-Treatment Prophylaxis 472
29.4 The Distribution of the Burden 29.11.7 Prophylactic Therapy 473
of HIV-Related Tuberculosis 457 29.11.8 Malabsorption 474
29.5 Pathology of Tuberculosis in HIV-Infected 29.11.9 Adjunctive Treatments 474
Individuals 458 29.12 Issues Relating to Control of Tuberculosis
29.5.1 Anergic Response 458 in HIV-Endemic Areas 474
29.5.2 Hyporeactive Response 459 29.12.1 Quality of Health Care 474
29.5.3 Granulomatous Response 459 29.12.2 The WHO DOTS Strategy 475
29.6 Immune Interactions in HIV-Positive Patients 29.12.3 BCG Vaccination and HIV Serostatus 475
with Tuberculosis 459 29.12.4 Governmental Responses to the Threat Posed
29.7 The Clinical Features of HIV-Related by HIV and Tuberculosis 475
Tuberculosis 460 29.12.5 National Tuberculosis Control Programmes 476
29.7.1 Pulmonary Tuberculosis in HIV-Positive 29.13 Conclusions 476
Adults 461 References 476
29.7.2 Extrapulmonary Tuberculosis in HIV-Positive
Adults 463
29.7.3 Tuberculosis in HIV-Infected Children 464
29.7.4 The Dilemma of Multiple Aetiology
of Childhood Respiratory Diseases 465
29.8 Diagnosis of Tuberculosis in HIV-Positive 29.1
Individuals 466 Historical Background
29.8.1 The Sputum Smear in HIV-Positive Persons 467
29.9. Treatment of Tuberculosis 468
29.9.1 Drug Regimens for Treatment of Tuberculosis Owing to increased poverty in an overpopulated
in HIV-Negative Individuals 468 world, a lack of attention to tuberculosis services, and
29.9.2 Treatment of Tuberculosis in HIV-Positive the impact of the Human Immunodeficiency Virus
Individuals 468 (HIV)/Acquired Immune Deficiency Syndrome
29.10 Anti-Retroviral Therapy for Poor Developing
(AIDS) pandemic, there are more cases of tuber-
Countries 469
29.11 Management Considerations in HIV-Positive culosis today than at any previous time in human
Individuals 470 history. Historically, tuberculosis has been among the
most feared of all diseases and has been given names
such as the "Great White Plague" and "Captain of all
of these Men of Death". Most attempts to affect a cure
for tuberculosis in the pre-antibiotic era were either
A. ZUMLA, BSc, MBChB, MSc, PhD, FRCP (Lon), FRCP (Edin) harmful, for example blood-letting, or revolting, such
Professor of Infectious Diseases and International Health,
Director, Centre for Infectious Diseases and International
as the mixture of pigeon's faeces and weasel's blood
Health, Royal Free & University College Medical School, Wind- advocated by John of Gaddeston (1280-1361) in his
eyer Institute of Medical Sciences, Room G41, 46 Cleveland St, Curatio Scrophulorum.
London WIP 6DB, UK The introduction of effective anti-tuberculosis
J. M. GRANGE, MSc, MD (Lon) agents in the middle of the 20th century, commencing
Visiting Prof, Centre for Infectious Diseases & International
Health, Royal Free and University College Medical School,
withthe discoveryofstreptomycinbySelmanWaksman
Windeyer Institute of Medical Sciences, 46 Cleveland St, and his colleagues in 1944, opened the door to the
London WIP 6DB,UK potential conquest of this affliction through devel-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
456
opment of other potent anti-tuberculosis drugs.
Far from being conquered, however, tuberculosis is
currently one of the most prevalent infectious causes
of human suffering and death. The incidence and
prevalence of this disease continues to rise in devel-
oping countries while many developed countries
have witnessed a reversal of the downward trend
that had occurred since the beginning of the 20th
century (Report 2001). Among infectious diseases,
tuberculosis remains a leading cause of illness and
death, particularly in tropical countries. Due to the
relentless spread of tuberculosis and HIV through-
out the world, in 1993 the World Health Organization
(WHO) took the unprecedented step of declaring
tuberculosis and HIV I AIDS global emergencies
(World Health Organization 1994).
29.2
Risk Factors for Tuberculosis
The principal factors thought to predispose one to
tuberculosis are listed in Table 29.1. In the past two
decades, HIV infection has emerged as the most
important and widespread risk factor for progress
from latent tuberculosis infection to active tubercu-
losis and rapid progression to active diseases after
infection or re-infection by Mycobacterium tuber-
culosis. Infections such as measles, whooping cough
and chronic malaria and causes of lung damage,
particularly smoking and exposure to silicon and
other industrial dusts, are also risk factors. Other
predisposing conditions are those that compromise
the immune response and include malnutrition,
stress, alcoholism, diabetes, renal failure, treat-
ment with immunosuppressive drugs and steroids,
liver failure and cancers, especially haematological
malignancies. Transmission of infection is further
facilitated by overcrowding, poor ventilation and
low levels of ultra-violet light-conditions frequently
linked to poverty.
29.3
Tuberculosis and HIV:
Epidemiological Considerations
Infection with HIV is the greatest risk factor for reac-
tivation of tuberculosis among those who are dually
infected with Mycobacterium tuberculosis and HIV
(Chretien 1990). While a non-immunocompromised
A. Zumla and J. M. Grange
Table 29.1 Susceptibility factors for developing tuberculosis
Age Extremes of age
- Less than 5 years of age
- Elderly
5-15 years less likely
Race Asians
African
North American Indians
Immunological Immunosuppression
- HIV/AIDS
- Steroids
- Cytotoxic drugs
- Congenital immunodeficiencies
- Protein-energy malnutrition
Chronic medical conditions
- Diabetes mellitus
- Chronic liver or renal disease
- Malignancies
- Damaged lungs (e.g. silicosis)
Congenital immunodeficiency states
Chromosomal associations
- HLA-DQ allele
- NRAMP 1 gene
Stress Endocrine factors
Previous exposure BCG or exposure to environmental
to mycobacteria mycobacteria may offer some protection
Mycobacterial factors Virulence and ?strain differences
Dose of organism
person who has overcome primary infection with
Mycobacterium tuberculosis has about a 5% chance
of developing post-primary tuberculosis later in life,
the chance rises to 50% in an HIV-positive person.
The annual risk of a person co-infected with the
tubercle bacillus and HIV developing tuberculosis
is around 8%-over 20 times higher than in an HIV-
negative person. The chance of an HIV-positive
person developing tuberculosis following either pri-
mary infection or reinfection is very high, especially
in those with AIDS. In addition, the progression from
infection to overt disease can be very rapid, the time
scale being "telescoped" from several years to a few
months. These factors have led to a number of mini-
epidemics of tuberculosis in centres caring for AIDS
patients.
In the year 2001, there were an estimated 35 mil-
lion HIV-positive persons worldwide. As one third
of these were co-infected with tubercle bacilli and
therefore had an 8% chance of developing overt
tuberculosis annually, an additional million cases
of this disease would have occurred in that year as
a result of HIV infection. Tuberculosis is a common
cause of death in those with AIDS and accounted for
30% of the estimated 3 million AIDS-related deaths
in 2000 (Report 2001). The global incidence rates of
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
HIV-positive tuberculosis cases as estimated by the
WHO are shown in Fig. 29.1. The impact of the HIV
epidemic on the incidence of tuberculosis is most
evident in sub-Saharan Africa where, for the 10-year
period 1990-1999,3.9 million (almost a quarter) of
the expected 15 million incident cases of tuberculosis
were attributable to HIV infection. It has been fore-
casted that the number of new cases of tuberculosis
per year in this region will double by the end of the
decade (Cantwell and Binkin 1997).
The strongest association between HIV and tuber-
culosis has been recorded in sub-Saharan Africa
(Cantwell and Binkin 1997). HIV-seropositivity rates
among adults and children with tuberculosis range
from 20-70%, and there are some countries such as
Malawi and Zambia where HIV-seroprevalence rates
are consistently above 50%. In Zambia and Cote
d'Ivoire, HIV-seroprevalence rates in children aged
1 month to 14 years with tuberculosis are between
10-40%, with the highest overall age-specific HIV-
seroprevalence being found in children aged 1-4 years
(Luo et al. 1994; Lucas et al. 1996a). In Zambia, one
in four pregnant women are infected with HIV
(Fylkesnes et al. 1997). Furthermore, tuberculosis is
now among the top non-obstetric causes of maternal
death in that country (Ahmed et al. 1999). Two thirds
of Zambian children and adults with tuberculosis are
co-infected with HIV (Luo et al. 1994; Chintu and
Zumla 1997; Elliott et al. 1990). Tragically, HIV infec-
tion is spreading rapidly in the Indian subcontinent, in
China and in South East Asia where, owing to the huge
population, most of the world's cases of tuberculosis
_.
Fig. 29.1. Estimated incidence rates
of Human Immunodeficiency Virus-
positive tuberculosis. (Data for 1999
obtained from the World Health
Organization Global Tuberculosis
Report 2001)
457
occur. Molecular studies from the USA have shown
that, in almost two thirds of HIV-infected persons,
tuberculosis is due to recent infection rather than
reactivation of latent infection. Similarly, recent work
from sub-Saharan Africa has shown that a significant
number of new cases of tuberculosis, and recurrent
cases of tuberculosis, also result from recent transmis-
sion (Sonnenberg et al. 2001). Meta-analysis of several
studies of health care workers who had contact with
tuberculosis patients suggest that patients with tuber-
culosis co-infected with HIV are not more intrinsically
more infectious than are HIV-l negative tuberculosis
patients (Cruciani et al. 2001).
29.4
The Distribution of the Burden of HIV-Related
Tuberculosis
The burden of HIV-related tuberculosis is not evenly
distributed, as it is determined not only by the over-
all prevalence of infection by the tubercle bacillus
and HIV in the community, but by the social fac-
tors that encourage dual infection (Msamanga and
Fawzi 1997). In the USA, HIV-related tuberculosis
is largely restricted to socio-economically disad-
vantaged communities in New York City and other
large cities (Friedman et al. 1996; Coker 1998). It was
the occurrence of mini-epidemics of HIV-associated
tuberculosis in New York City in the early 1990s, and
the threat to the health of the general population
that led to an upsurge of interest and concern for
o
o
o 0
o
o
o
o o
,-
llIIII 11_ (per 1._)
0<0.1
0 0.1-0.9
o 1.0-9.9
1~
_ 100-249
_ ~2S0
o No I$linlIle
458
this condition. In the United Kingdom (UK), HIV-
related tuberculosis is relatively uncommon and is
principally restricted to populations who have immi-
grated from HIV-endemic areas (Rose et al. 2001).
Most cases of tuberculosis in the UK occur in the
more elderly of the majority white population and in
immigrants from countries where HIV infection is, so
far, uncommon (Watson et al. 1993; Yates et al. 1993).
The situation is, however, changing as HIV infection
is spreading to other regions endemic for tubercu-
losis and from which immigrants come to the UK.
Thus tuberculosis clinics are now being encouraged
to increase the use of HIV testing (Coker and Miller
1997). From the global perspective, the region most
affected by the dual epidemic of tuberculosis and
HIV as of 2001 is sub-Saharan Africa (Report 2001)'
although Asia and China and Eastern Europe are also
being increasingly affected. The more southerly Afri-
can countries, including Kenya, Tanzania, Zimbabwe,
Zambia, Malawi, South Africa and Botswana, are par-
ticularly affected, with around one in five adults being
infected with HIV (UNAIDS 1998). In these countries,
HIV is principally transmitted heterosexually, and
women are particularly at risk.
In Zambia at least one in four pregnant women are
HIV positive (Fylkesnes et al. 1997) and around 50%
are also infected with the tubercle bacillus, so that
one in eight are therefore co-infected. An example
of the societal effect of HIV-related tuberculosis in
some African countries is provided by the substantial
adverse effect on the health of hospital staff and on
children. In one South African hospital, the number
of cases of tuberculosis among its staff rose from
two in the period 1991-1992 to 20 in 1993-1996, and
12 of the 14 who were tested for HIV infection were
positive (Wilkinson and Gilks 1998). An increasing
number of cases of tuberculosis in African children
are HIV related. In Zambia, the HIV seroprevalence
rate among children admitted to hospital with tuber-
culosis rose from 18% to 67% over an 8-year period
up to 1995 while, over the same period, the HIV-
seroprevalence rate remained at a constant 10% in
children admitted for surgical conditions (Luo et al.
1994; Zumla and Chintu 1995).
29.5
Pathology of Tuberculosis in HIV-Infected
Individuals
An autopsy study from the Ivory Coast, West Africa,
showed that 38% of all HIV-positive cadavers and
A. Zumla and J. M. Grange
43-54% of those dying with AIDS-defining pathol-
ogy had active tuberculosis (Lucas and Nelson 1994).
At autopsy, HIV-positive patients with tuberculosis
showed fibrous and calcified lesions of tuberculosis
adjacent to recent active lesions containing acid-fast
bacilli. These new lesions may be due to reactivation
of primary complex lesions or may also be due to
re-infection. Relapse has been observed with differ-
ent genotypes of Mycobacterium tuberculosis. The
suggested mechanism of reactivation or increase in
susceptibility to mycobacterial infections in those
infected with HIV is the depletion of CD4+ lym-
phocytes and macrophages, although tuberculosis
has been seen in patients across a range of CD4
counts. Quantitative and qualitative defects in the
CD4+ cells lead to impaired granuloma forma-
tion (Zumla and James 1999). In patients infected
with HIV, the progressive decline in the number of
CD4+ cells provides a suitable model for the study
of the interaction between the mononuclear cells
and formation of epithelioid granulomas. In non-
HIV-infected patients with tuberculosis without any
obvious evidence of immunodeficiency, granuloma
formation and progression of human monocytes
through stages of differentiation are well charac-
terised and are typical of cell-mediated immune
responses found in other granulomatous disorders.
The monocyte differentiates through young macro-
phages, to mature macrophages, to epithelioid cells
and finally to giant cells of the Langhans type. In
HIV-infected patients, a range of histological fea-
tures is seen. Those individuals with well-preserved
CD4+ lymphocyte counts often show well-formed
epithelioid granuloma formation while poor epi-
thelioid cell formation and "sick" macrophages are
seen in patients with more severe forms of immu-
nodeficiency.
A wide spectrum of histopathological changes
are seen in HIV-infected patients with tuberculosis.
However, there are three identifiable histologic stages
of cellular immune responses, which may also cor-
relate well with the stage of HIV infection (Zumla et
aI.1999).
29.5.1
Anergic Response
In the late stages ofAIDS, disseminated anergic tuber-
culosis is seen at autopsy. While no relative decrease
in number of macrophages in the tuberculous lesion
is seen, there is decreased intensity of staining with
KP-l (CD68). Epithelioid macrophages are scanty,
Tuberculosis and Co-infection with the Human Immunodeficiency Virus 459

Langhans giant cells are absent, and granuloma

formation is absent (Fig. 29.2). There are few CD4+
T cells in the lesion. Caseous necrosis is replaced by
suppuration, coagulative necrosis and large amounts
of apoptotic debris. Large numbers of mycobacteria
are present within macrophages and in the necrotic
areas (Fig. 29.3).

29.5.2
Hyporeactive Response

With progressive immunosuppression and decline

of CD4+ T cell counts, loss of Langhans giant cells
and subsequently, epithelioid macrophages occur.
The proportion of macrophages with abundant
cytoplasm is also decreased. Poor intracellular kill-
ing of mycobacteria occurs and thus the number of
mycobacteria in the lesions also increases. The case-
ous centres enlarge centrifugally and lesions coalesce.
Necrosis is mixed suppurative and caseous.
Fig. 29.2. Histology of tuberculosis lesion in an immunosup-
pressed patient with low CD4+ lymphocyte counts. Epithelioid
macrophages are scanty, Langhans giant cells are absent, and
granuloma formation is absent. There are few CD4+ T cells
in the lesion. Caseous necrosis is replaced by suppuration,
coagulative necrosis and much apoptotic debris

29.5.3
Granulomatous Response

These patients have a relatively intact cellular

immune response and have a typical granulomatous
response (Fig. 29.4). Epithelioid macrophages and
Langhans giant cells are abundant and the numbers
of mycobacteria in the lesions are low (Nambuya et
al. 1988). Clusters of CD4+ T cells around epithelioid
macrophages are seen and Langhans giant cells are Fig. 29.3. Histology of tuberculosis lesion in an immunosup-
pressed patient with very low CD4+ lymphocyte counts. Large
present. The majority of macrophages have abundant
numbers of mycobacteria are present within macrophages and
cytoplasm and stain intensely with KP-1 (CD68) mac- in the necrotic areas
rophage markers.

29.6
Immune Interactions in HIV-Positive
Patients with Tuberculosis

It is recognised that HIV infection increases the risk

of tuberculosis and adversely affects the clinical
course of the disease (Bocchino et al. 2000). Con-
versely, there is evidence that co-infection with
Mycobacterium tuberculosis accelerates progression
of HIV infection towards AIDS (Whalen et al. 1995).
Mycobacterium tuberculosis and HIV are both viru-
Fig. 29.4. Histology of tuberculosis lesion in a Human Immu-
lent intracellular pathogens that enter and replicate nodeficiency Virus-infected patient with well-preserved CD4+
within macrophages that serve as their reservoirs. It is lymphocyte counts. Well-formed granuloma with epithelioid
now accepted that tuberculosis enhances local HIV-1 macrophages and Langhans giant cells
460
replication in vivo, although the exact mechanisms
operating are not known. Cytokines produced during
infection with Mycobacterium tuberculosis may result
in activation of latently HIV-infected cells with virus
expression and induction of virus replication (Zhang
et al. 1995). Other studies have demonstrated marked
increases in plasma HIV-1 viral load in tuberculosis
(Goletti et al. 1996). Increased IL-2, IL-6 and tumour
necrosis factor generated by infection with Myco-
bacterium tuberculosis may be responsible for these
increases in HIV viral load (Bellamy et al. 1998).
The concept that a Th-l type response without anti-
body correlates with protection from HIV disease has
been clouded by a dispute about whether a progressive
shift from Th-l towards Th-2 type response can be
shown to accompany AIDS or tuberculosis. In neither
early AIDS nor tuberculosis is the Th-l response lost,
but there is a superimposed Th-2 component that is
clearly inappropriate. HIV-exposed people who resist
progressive infection appear to have no detectable
antibody, whereas sero-positivity and peripheral blood
lymphocytes expressing IL-4 are associated with an
inexorable progression towards death. In tuberculosis,
the inappropriate Th-2 response leads to formation of
IgE antibody, high levels of IgG antibody and periph-
eral blood lymphocytes that express IL-4 (Schauf et al.
1993). Thus their Th-l response may be maintained,
but there is a Th-2 type response superimposed upon
it. In fact, in a mouse model, this mixed Th-l and
Th-2 responsiveness leads to a cytokine-mediated
tissue damage (Clerici and Shearer 1994). Thus the
basic immunity required for both diseases is quite
similar and a vaccine that promotes a Th-l response
will theoretically reduce the burden of both diseases.
Mycobacterium tuberculosis has been shown to aug-
ment chemokine receptor expression (CCR5) on mac-
rophage membranes and induce apoptosis of a portion
of infected macrophages (Sanduzzi et al. 2001).
29.7
The Clinical Features of HIV-Related
Tuberculosis
Pulmonary tuberculosis can occur in patients with
a wide spectrum of immunodeficiency and is not
entirely dependent on the degree of depletion of
CD4+ lymphocytes; thus tuberculosis may develop
at any stage of HIV infection (Harries 1998). Though
classified as an AIDS-defining condition (Centers for
Disease Control 1993), tuberculosis may develop early
in the course of HIV infection, often before other
A. Zumla and J. M. Grange
conditions suggestive of HIV infection or AIDS have
appeared. The nature, presentation and the clinical
and radiological features of tuberculosis depend on
the degree of immunosuppression. In those with rela-
tively good immunity, CD4+ lymphocyte counts and
a low viral load, the manifestations and presenting
symptoms of tuberculosis are essentially similar to
those in HIV-negative persons.
Approximately 70% of cases ofHIV-related tuber-
culosis are pulmonary and 30% are extrapulmonary.
If tuberculosis occurs in the early stages of HIV infec-
tion, when immunity is only partly compromised or is
intact, the clinical features are indistinguishable from
those of post-primary tuberculosis in HIV-negative
individuals. Thus tuberculosis is usually localised to
the apices of the lungs; there is lung destruction and
cavitation (Fig. 29.5) and light microscopy of Ziehl-
Neelsen stained sputum smears shows abundant
acid-fast bacilli. HIV-positive patients with more
advanced immunodeficiency present with atypical
pulmonary disease, sometimes resembling pri-
mary pulmonary tuberculosis and extrapulmonary,
including disseminated, forms of the disease. As the
immunocompetence decreases, there is an increasing
incidence of atypical forms of tuberculosis, and diag-
nostic difficulties are posed by the rather non-spe-
cific presenting features that may be confused with
those of other HIV-related infections (Whalen et al.
1995; Chaisson et al. 1987; Festenstein and Grange
1993; Huebner and Castro 1995). The atypical picture
is characterised by extensive pulmonary infiltrates
with no cavities (Figs. 29.6, 29.7), involvement of the
lower lobes of the lung (Fig. 29.8), enlargement of the
mediastinal lymph glands (Table 29.2) and sputum
smears that show no acid-fast bacilli.
Fig. 29.5. Chest X-ray of an Human Immunodeficiency Virus
-infected patient showing apical cavities. Mycobacterium
tuberculosis was isolated from the patient's sputum
Tuberculosis and Co-infection with the Human Immunodeficiency Virus 461

Fig. 29.6. Chest X-ray shows extensive infiltration of both lung

fields in an Human Immunodeficiency Virus-positive patient Fig.29.8. Chest X-ray of a Human Immunodeficiency Virus-
with tuberculosis positive patient with tuberculosis showing a cavitating lesion
in the left lower zone

Table 29.2. Clinical features of tuberculosis in Human Immuno-

deficiency Virus-infected patients

HIV- HIV-
positive negative

Respiratory symptoms +++ +++

Extrapulmonary disease +++ +
Sputum smear positivity (AAFB) + ++
False-negative tuberculin skin tests ++ ±
Atypical features on chest X-rays +++ ±
Cavitating lung lesions + +++
Adverse drug reactions ++ ±
Mortality rate +++ ±
Relapse rate following course of treatment ++ ±

acid-fast bacilli is a diagnostic challenge for the

Fig. 29.7. Extensive shadowing throughout the right lung
fields. Mycobacterium tuberculosis was isolated from patient's
sputum
29.7.1
Pulmonary Tuberculosis in HIV-Positive Adults
The proportion of patients with smear-negative pul-
monary tuberculosis is greater in those who are HIV
positive than in those who are HIV negative (Harries
1998) (Table 29.2). The diagnosis of tuberculosis in
a HIV-positive patient with a chronic cough, night
sweats, weight loss but negative sputum smears for
clinician (Harries and Maher 1996). In studies in
HIV-positive African patients with respiratory ill-
ness and negative sputum smears, specimens yield-
ing Mycobacterium tuberculosis were obtained by
use of bronchoscopy with broncho-alveolar lavage
or induction of sputum in about a third of patients
(Harries et al. 1997). HIV-infected patients may also
have other concomitant infections such as Pneumo-
cystis carinii pneumonia (PCP), bacterial pneumonia
due to a wide range of pathogens, Kaposi's sarcoma,
nocardiosis and fungal infections.
Even where facilities exist for more extensive
investigations, such as bronchoscopy with bron-
cho-alveolar lavage and biopsy, sputum culture and
462 A. Zumla and J. M. Grange

nucleic acid-based methods, the identification of the in some cases than in others. Patients with HIV are
Mycobacterium tuberculosis may be difficult. In most more likely to exhibit the radiographic features of
developing country health centres, the diagnosis of primary tuberculosis. These manifestations are
pulmonary tuberculosis is based on simple tech- mediastinal and hilar lymphadenopathy, pleural
niques only: sputum smear microscopy and, when effusions, middle and lower lung infiltrates and mili-
available, chest radiography. ary dissemination, with cavitation being uncommon.
Radiological findings in HIV-related pulmonary (The typical radiographic appearances in HIV-nega-
tuberculosis in many resemble those of typical tive individuals with post-primary tuberculosis are
pulmonary tuberculosis, but atypical appearances cavitation, calcification and upper lobe fibrotic
are common, notably as immune function declines. changes.) As these HIV-positive patients behave as
These include vague, spreading opacities suggestive ,immuno-naive' individuals they therefore develop a
of pneumonia, predominantly lower lobe disease, "childhood" pattern of tuberculosis.
pleural effusions, air-fluid levels and intrathoracic Alarge chest X-ray study of 963 HIV-infected adult
lymphadenopathy. Studies from HIV-positive tuber- Zairian and Zambian adults showed a significantly
culosis patients in the Tropics have confirmed the increased incidence of lymphadenopathy, pleural
different pattern of chest X-ray changes that occur effusions, parenchymal changes, consolidation and
(Harries 1998; Harries et al. 1997; Tshibwabwa- miliary disease, but significantly less cavitary dis-
Tumba et al. 1997; Batungwanayo et al. 1992; Pozniac ease and atelectasis (Tshibwabwa-Tumba et al.I997).
et al. 1992; Saks and Posner 1992; Awil et al. 1997) Similar trends were observed in a comparative radio-
(Table 29.3). These studies have drawn attention to logical study of 61 HIV-positive and 50 HIV-negative
the fact that HIV-related tuberculosis presents with South African tuberculosis patients (Saks and Posner
atypical radiological changes that are more dramatic 1992). Chest radiographs of the HIV-seropositive
group showed a significantly higher percentage of
Table 29.3. Chest X-ray patterns in tuberculosis Human Immu- hilar lymphadenopathy (50% vs 8%), pleural effu-
nodeficiency Virus-positive versus Human Immunodeficiency sions (38% vs 20%), and miliary lesions (8% vs 0%)
Virus-negative and interstitial changes (11% vs 4%). Cavitation
Authors and (country) Radiographic pattern HIV HIV (38% vs 82%) and atelectasis (31% vs 82%) were less
+ve (0/0) -ve (0/0) common in the HIV-seropositive group than in the
Saks and Posner Mediastinal and/or 50 8 seronegative group. Lymphadenopathy on chest X-
(1992) hilar adenopathy ray in previous studies has been reported in between
(South Africa) Pleural effusion 38 20 25-50% of HIV-infected adults with tuberculosis
Miliary disease 8 0 (Saks and Posner 1992). Classical miliary lesions are
Cavitation 38 82
Atelectasis 31 82 seen in a minority of HIV-positive patients with dis-
Batungwanayo Mediastinal and/or 30 0 seminated disease, but in most cases the formation
et aI. (1992) hilar adenopathy of the characteristic miliary granuloma is suppressed
(Rwanda) Pleural effusion 43 9 and the X-rays may appear deceptively normal. Fur-
Upper lobe infiltrate 16 55 ther details on the clinical aspects of HIV-related
Cavitation 39 91
Miliary disease 25 9
tuberculosis are available from WHO (1996).
Pozniak et aI. Mediastinal and/or 31 16 The available data to date suggest that the atypical
(1995) hilar adenopathy changes seen on chest X-rays of many HIV-infected
(Zimbabwe) Pleural effusion 26 13 patients with tuberculosis are uniform across coun-
Upper lobe infiltrate 43 67 tries in sub-Saharan Africa (Table 29.3). However, a
Cavitation 40 64
Awil et aI. Mediastinal and/or 26 6 wide range of chest X-ray features is seen in both
(1997) hilar adenopathy HIV-infected and HIV-non-infected individuals.
(Uganda) Pleural effusion 23 11 Approximately one-third of HIV-infected patients
Pneumonic infiltrate 46 26 show classical chest X-ray findings of tuberculosis.
Cavitation 18 57 The ability to form cavities or fibrotic changes do not,
Miliary disease 7 0
Tshibwabwa-Tumba Mediastinal and/or 26 13
as was once thought, appear to be related to the CD4
et aI. (1997) hilar adenopathy counts. The reasons why some HIV-infected individ-
(Zaire and Zambia) Pleural effusion 16 7 uals develop atypical radiological changes of tuber-
Miliary disease 10 5 culosis while others continue to show classic chest
Cavitation 33 78 X-ray features of tuberculosis are not clear. These
Atelectasis 12 24
differences have implications for the radiologist,
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
especially in developing countries where accurate
diagnosis of pulmonary infectious disease is often
not possible. The atypical chest X-ray appearances
of tuberculosis in a large proportion of HIV-infected
adults makes it imperative to keep this disease a
prime diagnostic consideration in HIV-infected
individuals with pulmonary lesions.
29.7.2
Extrapulmonary Tuberculosis
in HIV-Positive Adults
Manifestations of extra-pulmonary tuberculosis are
frequently seen in HIV-infected persons in sub-Saha-
ran Africa. The most common of these manifestations
are: (a) asymmetrical lymphadenopathy (Fig. 29.9),
(b) pericarditis, (c) pleurisy (Fig.29.l0), (d) bone
and skin lesions and (e) miliary disease (Nambuya
et al. 1988; Harries 1998; Harries et al. 1997; Gilks et
al. 1990). Patients usually present with non-specific
constitutional symptoms (fever, night sweats and
weight loss) and local symptoms and signs related
to the site of disease. These atypical forms are often
the result of a rapid progression of clinical disease
following infection and in a high proportion of cases,
multi-site and widely disseminated tuberculosis is
seen. For reasons that are not clear, involvement of
the central nervous and genitourinary systems is,
relative to the industrialised countries, uncommon
in Africa (Harries et al. 1998). This probably reflects
Fig.29.9. Asymmetrical cervical and superclavicular lymph-
adenopathy due to Mycobacterium tuberculosis in an Human
Immunodeficiency Virus-positive adult
463
Fig. 29.10. Chest X-ray showing pleural and pericardial effu-
sion in a 44-year-old Human Immunodeficiency Virus-posi-
tive patient
patient selection and differences in the availability
of diagnostic facilities, while strain differences in
the infecting organism is an as yet unproven pos-
sibility. More generalised dissemination of disease
may result in numerous minute lesions throughout
the body. This condition, termed "cryptogenic dis-
seminated tuberculosis", may be very difficult to
diagnose during life. Many unsuspected cases are
therefore diagnosed at autopsy and pose serious
risks to the pathologist. Tuberculin tests may be
negative, particularly in the more severely immuno-
suppressed, although the extent of induration is not
closely related to the CD4+ lymphocyte count and is
thus an independent marker of immune competence
(Diagbouga et al. 1998). Blood cultures are positive
in between one third and one half of patients with
disseminated HIV-related tuberculosis (Schaefer et
al. 1989; di Lonardo et al. 1995).
A high proportion of African patients with AIDS
develop severe wasting, a condition known locally
as "slim disease" (Fig.29.11). Many such patients
have chronic diarrhoea, and the condition was thus
thought to be due to enteropathy. However, in one
study in Africa, almost half the AIDS patients died
with "slim disease" and almost half of these were
found to have disseminated tuberculosis on autopsy,
compared to just over a quarter of those dying with-
out such wasting (Lucas et al. 1994). There was no
association between the occurrence of chronic diar-
rhoea and tuberculosis.
464
Fig.29.11. "Slim disease" (wasting due to Human Immuno-
deficiency Virus disease)
Lymphadenopathy is a frequent manifestation of
tuberculosis in HIV-infected individuals and may
present in a variety of ways (Bem 1997; Bem et al.
1993). While usually chronic and cryptic, it may also
occasionally be acute and resemble an acute pyogenic
infection. Diagnosis of lymph node tuberculosis may
be made by use of simple techniques such as needle
aspiration and staining with Ziehl Neelsen stain;
naked eye inspection of biopsied lymph nodes for
macroscopic caseation; and direct smears for acid
fast bacilli from the cut surface of a lymph node (Bem
1996). HIV-infected individuals with meningitis due to
Mycobacterium tuberculosis may be normal or near-
normal on examination of their cerebrospinal fluid,
and can easily be confused diagnostically with crypto-
coccal meningitis (a common presenting condition in
HIV infection; Mwaba et al. 2001) making the diagno-
sis very difficult. Sometimes empirical treatment has
to be started on clinical suspicion alone.
29.7.3
Tuberculosis in HIV-Infected Children
Accurate figures for burden of tuberculosis in HIV-
infected children are not readily obtainable (Donald
2000). This is due to the difficulty of accurately
diagnosing childhood tuberculosis, inadequate
health information systems in developing countries
and the lack of importance accorded to childhood
tuberculosis by disease control authorities. Although
A. Zumla and J. M. Grange
the relationship between HIV/AIDS and tuberculo-
sis in adults is well understood, a clear picture of
the relationship between HIV/AIDS and childhood
tuberculosis has yet to emerge. One of the principal
problems faced in paediatric tuberculosis is the dif-
ficulty in making an accurate diagnosis. It is gener-
ally recognised that, due to the overlap in symptoms
and signs of several common respiratory illnesses,
misdiagnosis is common, especially in HIV-infected
children where opportunistic infections such as PCP
may mimic tuberculosis. The diagnosis of this disease
in children is conventionally based on a combina-
tion of clinical and laboratory criteria. The chance
of seeing acid-fast bacilli in sputum and gastric fluid
aspirates is low and such specimens often cannot be
readily obtained from children. Even in countries
where advanced diagnostic facilities are available, the
diagnosis can only be confirmed by culture in about
50% of cases. Several of the signs and symptoms used
as clinical criteria for the diagnosis of childhood
tuberculosis (such as prolonged cough, weight loss,
failure to gain weight) are common features of other
pulmonary illnesses that are consequences of HIV-
induced immunosuppression. Radiographic changes
may be misleading and the distinction between
PCP and tuberculosis is often difficult, especially in
developing countries where advanced facilities for
the diagnosis of PCP are not available.
Tuberculin tests may be falsely negative and
sometimes uninterpretable in HIV-positive cases.
Against this background, it is apparent that some
children who do not have tuberculosis will satisfy the
clinical criteria for the diagnosis of this disease and
will receive a full course of anti-tuberculosis therapy.
Conversely, the diagnosis may be missed by the cur-
rent clinical criteria in others who have tuberculosis
but present with atypical signs and symptoms, and
these will not receive appropriate treatment. The
clinical and laboratory criteria for the diagnosis of
paediatric tuberculosis require re-evaluation. The
co-existence of tuberculosis with other fatal diseases
may confuse the clinical picture. Several studies
have shown that tuberculosis and HIV are having
an impact on childhood morbidity and mortality
(Chintu and Zumla 1997; Chintu et al. 1993, 1995;
Donald et al.1995; Donald 2002). Previous beliefs that
tuberculosis in children below the age of 2 years did
not occur because of duration of exposure (Lucas et
al. 1996a,b) is not borne out by a large autopsy study
of Zambian children dying of respiratory diseases.
The results of autopsy study shows that tuberculosis
in children can occur at any age, irrespective of HIV
status (Chintu et al. 2002).
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
The clinical differences in the presentation of dis-
ease between HIV-positive and HIV-negative children
are not as striking as in adults. In Africa, a large pro- The accurate diagnosis of respiratory infection in
portion of paediatric clinic attendances and hospital children is difficult, even where appropriate health
admissions are for pulmonary diseases. Distinguish-
ing tuberculosis from the other causes is never easy
and is usually based on clinical features, tuberculin
testing and a history of exposure to a source case.
Laboratory investigations are often unhelpful as
lesions are usually closed and sputum, even when it
can be obtained, is almost always negative for acid-fast
bacilli on microscopy. As mentioned above, even when
the most advanced diagnostic facilities are available,
the diagnosis can only be confirmed by culture in
about half the cases (Starke and Taylor-Watts 1989).
Positive cultures may be obtained from unusual sites:
in a study in South Africa 6 of 14 HIV-infected children
with culture-positive tuberculosis had otorrhoea, and
ear swabs were the source of the positive cultures in
three such cases (Schaaf et al. 1998).
As a result of the diagnostic difficulties experi-
enced in HIV-positive children, and the frequent
misdiagnosis, it is difficult to assess the magnitude of
the problem of HIV-related tuberculosis in children
(Chintu et al. 1993, 1995, 2002; Donald et al. 1995;
Donald 2002; Lucas et al. 1996b; Starke and Taylor-
Watts 1989; Schaaf et al. 1998; Starke 1993; Coovadia
1991).An idea of the true risk of tuberculosis in these
children may be obtained from autopsy studies. One
such study in West Africa suggested a low risk (Lucas
et al. 1996a), but autopsies performed on HIV-posi-
tive children in Bulawayo, Zimbabwe (Ikeogu et al.
1997), established a diagnosis of tuberculosis in 6 of
122 children (5%), and more recently a large autopsy
study of children dying of respiratory illnesses in
Zambia has shown that 20% of these children have
tuberculosis across a wide range of age groups and
HIV status.
29.7.4
The Dilemma of Multiple Aetiology
of Childhood Respiratory Diseases
The great overlap that exists between the clinical
presentation of several respiratory infections and
the occurrence of multiple pathology in children
who succumb to respiratory illnesses indicates that
efforts should be focused on tightening the current
management guidelines for respiratory infections.
Current WHO recommendations on the prevention,
early detection and treatment of respiratory infec-
tions in children in developing countries need to be
465
updated and should include specific guidelines on the
management of the HIV-positive child.
services exist. The presence of multiple diseases makes
management even more complicated and expensive
and may be responsible for estimates that up to three
quarters of children with HIV-infection in Africa will
die by the age of 5 years (Taha et al. 2000; Pillay et al.
2001). Even when treatment of PCP is optimal, the out-
come in children is poor. The only way forward, there-
fore, is to institute preventive measures (Abrams et al.
2001). The recommended antibiotic, co-trimoxazole,
may prevent both PCP and bacterial bronchopneumo-
nia. There are high HIV prevalence and vertical trans-
mission rates in sub-Saharan Africa and the probability
of death occurring by 12 months in sub-Saharan Africa
ranges from 0.23 to 0.35 and by 5 years is 0.57-0.68.
In many developing countries, health care facili-
ties are inadequate and expert care is unavailable.
Studies in Malawi, Zimbabwe and South Africa have
illustrated the difficulties associated with the early
diagnosis, treatment and management of children
with respiratory illnesses. Several diagnostic, thera-
peutic and epidemiologic dilemmas are created by
the inadequacy of the current clinical criteria for the
specific diagnosis of respiratory infections in chil-
dren (Vetter et al.1996; Zar et al. 2001). In a setting of
poor funding and staffing resources, it is very diffi-
cult to determine the precise cause of lung disease in
children, owing to the overlap in clinical features of
common opportunistic and non-opportunistic lung
infections, including tuberculosis. This emphasises
the importance of preparing national guidelines for
the deployment of available resources and of devel-
oping practical diagnostic, therapeutic and prophy-
lactic protocols in resource-poor African countries.
The early diagnosis, treatment and prevention of
respiratory diseases in children requires a concerted
effort on the part of governments and clinicians to
develop newer guidelines that will take into account
the changing epidemiology of paediatric lung disease
in light of the HIV-epidemic.
The UNAIDS (2000b) recommendations of co-tri-
moxazole prophylaxis for adults and children have
recently come under Widespread criticism, and several
issues need to be addressed before these recommenda-
tions are put into practice. Widespread prescription of
prophylactic antibiotics to prevent PCP and pyogenic
pneumonias has to be carefully weighed against
resources available, dangers of emergence of antibiotic
resistance and the persistent dilemma of investment of
scarce resources in a group of patients whose long term
466
prognosis under prevailing health services conditions
is poor. Furthermore, identification of HIV-positive
children is difficult without formal testing. Widespread
consultation based on available scientific evidence lead-
ing to development of recommendations for the man-
agement of respiratory infection in both HIV-positive
and HIV-negative children is urgently needed and long
overdue. Knowledge of the spectrum of respiratory ill-
nesses is important for the development of prevention,
diagnosis and treatment guidelines for all levels of care
in HIV-negative and HIV-positive children.
29.8
Diagnosis of Tuberculosis in HIV-Positive
Individuals
Co-infection with HIV in patients with tuberculosis
leads to an increase in the number of cases that are
difficult to diagnose because of atypical presentations
of pulmonary disease, extrapulmonary disease and
increase in the number of sputum-smear negative
cases. In many African countries, the time between
onset of symptoms and diagnosis of smear-positive
pulmonary tuberculosis is about 3-4 months. Even
longer delays may be found in patients diagnosed with
smear-negative pulmonary tuberculosis, probably
because of the lack of radiographic facilities. Delay in
the diagnosis and treatment of tuberculosis compro-
mises the chances of a cure in HIV-positive patients
by allowing the disease to advance unchecked and by
accelerating the decline in immunocompetence.
The diagnosis of smear-negative pulmonary
tuberculosis and extrapulmonary tuberculosis in
resource-poor countries is fraught with difficulty
because of a shortage of trained health personnel,
poor diagnostic facilities and lack of appropriate and
specific diagnostic guidelines.
The standard diagnostic tests are used for the
diagnosis of tuberculosis in HIV-positive individu-
als. These include:
a. Microscopy
b. Culture
c. Serology
d. Molecular methods
e. Chest X-rays
f. Tuberculin skin testing
Microscopy: Clinical specimens that are examined
by light microscopy for acid-fast bacilli after Ziehl-
Neelsen staining are: (a) sputum, (b) bronchoalveolar
lavage, (c) gastric washings, (d) cerebrospinal fluid,
A. Zumla and J. M. Grange
(e) pleural aspirates, (£) lymph node aspirates, (g)
bone marrow and (h) tissue biopsies. Ideally, three
sputum specimens collected on successive days
should be examined. In children, this may not be
possible and gastric aspirates may be useful. Laryn-
geal swabs are less sensitive than sputum or gastric
washings and swabbing is distressful for the child.
Differentiation of species is not possible on examina-
tion using light microscopy.
Culture: Isolation of Mycobacterium tuberculosis in
culture from the clinical specimen provides a defini-
tive diagnosis. Commonly used media are Lowen-
stein-Jensen (L]) or Kirshner broth containing a
mixture of antibiotics to prevent overgrowth of other
bacteria and fungi. Other methods such as BACTEC
460 radiometric system and Roche MB check system
rely on the growth of bacilli in a Middlebrook broth.
The BACTEC system incorporates a Cl4 -labelled sub-
strate and, during metabolism, Cl4 -labelled carbon
dioxide is produced and is monitored by the machine.
Problems of the disposal of radioactive waste limit
the use of this system and more modern automated
systems are based on the changing colour of dyes and
the unquenching of fluorescent compounds induced,
respectively, by the production of CO 2 and the con-
sumption of oxygen. In these systems, mycobacterial
growth is detected about 9 days earlier than on LJ
slopes.
Serological Tests: Despite several descriptions of
ELISA tests for tuberculosis, no universally appli-
cable test with acceptable sensitivity and specificity
exists. ELISPOT assays have shown promise under
controlled situations (Lalavani et al. 200 I) although
their usefulness in the field requires study. With the
introduction of microarray technology and the iden-
tification of newer mycobacterial antigens, there has
been renewed interest in development of serological
diagnosis.
Molecular Methods: Though the gold standard for
identification of Mycobacterium tuberculosis is still
culture, this is laborious in terms of time and demand-
ing on laboratory space. Newer techniques, predomi-
nately molecular, have therefore been developed. The
most validated of these have involved the amplification
of mycobacterial DNA (or occasionally RNA) by the
polymerase chain reaction (PCR) or more recently
the ligase chain reaction. These are proving rapid,
sensitive and specific in most circumstances; with
sensitivities approaching 100% and specificities from
85% in non-pulmonary samples to 95% for sputum
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
(Borun et al. 2001; Lindbrathen et al. 1997) and several
systems are now available commercially in kit form. likely to benefit from preventive therapy.
The technique also allows the species of mycobacteria
to be identified rapidly and accurately (Kox et al. 1997)
and the early identification of drug resistance may be
achieved by identifying specific resistance-associated
genetic loci (Goyal et al.1997). It may also be possible
to monitor some therapies with PCR to predict relapse
(Sonnenberg et al. 2001). Unfortunately, the sensitiv-
ity of these methods is decreased in samples that are
negative on microscopy mycobacteria, probably due to
inhibiting substances present in the clinical specimens.
In addition, problems with contamination and cost are
preventing implementation in those areas of the world
where tuberculosis is most prominent.
DNA Fingerprinting: Molecular techniques can dif-
ferentiate between isolates of Mycobacterium tuber-
culosis. This methodology is becoming useful in
differentiating between reactivation and reinfection
(Sonnenberg et al. 2001), tracing the source of nosoco-
mial and community outbreaks of the disease, defining
the molecular epidemiology of the disease and study-
ing outbreaks of multi-drug resistant tuberculosis.
Chest X-Ray: The chest X-ray is sensitive but non-
specific. It is therefore valuable in detecting pulmo-
nary lesions but does not confirm the diagnosis of
tuberculosis. As described earlier, a range of chest-
X-ray changes are seen in HIV-infected individuals
with tuberculosis and classical changes due to tuber-
culosis may not be present.
Tuberculin Skin Testing: Tuberculin skin testing using
purified protein derivative is widely performed in
industrialised countries as an aid to diagnosing
tuberculosis. The test is able to detect infection with
tuberculosis, but cannot distinguish latent infection
from disease. In most African countries the majority
of adults have been infected and the test is thus of
little diagnostic value. Furthermore, bacille Calmette-
Guerin (BCG) vaccination, which is widely admin-
istered in the tropics, results in positive tuberculin
skin tests that often persist into adulthood (Waddell
et al. 1999). A further limitation of skin testing in
areas with high HIV prevalence is that the immune
response, which forms the basis of a positive tuber-
culin skin test, becomes progressively impaired. For
these reasons, tuberculin skin testing is of little value
as a diagnostic test in adults with suspected tuber-
culosis in tropical countries, except for young adults
presenting with primary disease. It is, however, the
only test that can diagnose latent tuberculosis and is
467
therefore the method used to detect patients who are
Earlier studies showed that the rate of conversion
to tuberculin reactivity was comparable (Nunn et
al. 1994; Klausner et al. 1993) or lower (Elliott et al.
1993; Cauthen et al. 1996) among close household
contacts of HIV-infected than of HIV-non-infected
patients with pulmonary tuberculosis (Wallis and
Johnson 2001). Recent tuberculin testing data from a
community-based study in the Dominican Republic
of 803 household contacts of patients with newly-
diagnosed pulmonary tuberculosis revealed a lower
rate of transmission from HIV-infected tuberculosis
patients than HIV-non-infected patients to their
close contacts (61% tuberculin conversion vs. 76%;
P=O.OOOI) (Espinal et al. 2000). These data suggest
that HIV-infected patients with tuberculosis may be
less likely than HIV-non-infected patients to trans-
mit tubercle bacilli to their close contacts. This find-
ing notwithstanding, close contacts of HIV-infected
and HIV-non-infected patients with newly diagnosed
tuberculosis should be evaluated similarly.
29.8.1
The Sputum Smear in HIV-Positive Persons
The sputum smear is still the most frequently used
method for diagnosing HIV-associated tuberculosis
(Hudson et al. 2000). The sensitivity of sputum smear
examination can be improved, and the time taken to
examine the slide shortened, by use of fluorochrome
staining. Fluorescent microscopes are, however,
expensive and not widely available in tropical coun-
tries. A simple technique of sputum concentration
improves the yield of sputum smear in HIV infec-
tion (Bruchfeld et al. 2000). Sputum induction by
inhalation of hypertonic saline may yield positive
specimens if sputum is otherwise unobtainable or
smear negative (Parry et al. 1995). Smears of lymph
node needle aspirates have a high yield (Bern et al.
1993; Pithie and Chicksen 1992).
If the sputum or lymph node aspirate is smear-
negative, an assessment on the basis of clinical fea-
tures and chest radiograph remains the mainstay of
diagnosis, as bronchoscopy is not widely available in
the tropics and mycobacterial culture, when avail-
able, is slow. WHO has proposed a case definition
for smear-negative pulmonary tuberculosis based
on three negative sputum smears, radiographic
abnormalities consistent with active pulmonary
tuberculosis and no response to a course of broad-
spectrum antibiotics. Most patients fulfilling the case
468
definition in a Malawian study were HIV-seroposi-
tive (Hargreaves et al. 2001). The final diagnosis was
tuberculosis in 78%, but treatable conditions would
have been missed in 14%, indicating the need for
further refinement of this approach. Case defini-
tions for extrapulmonary tuberculosis also need to
be developed.
29.9.
Treatment of Tuberculosis
Highly effective anti-tuberculosis chemotherapy has
been available for many years, yet there are currently
more people with the disease than at any previous
period of human history and, unless there are major
changes in the global management of the disease, the
numbers will certainly increase further. Tuberculosis
is among the most cost-effective of all diseases to
treat (Murray et al. 1990; Foster et al.1997); a 6-month
course of anti-tuberculosis drugs costs approximately
only UK£18. Also, by curing one infectious patient,
transmission of disease to several others is prevented.
Several national and international organisations pub-
lish periodically updated comprehensive guidelines for
tuberculosis treatment that are valuable for clinicians
caring for patients with tuberculosis (Joint Tubercu-
losis Committee of the British Thoracic Society 1998;
Report 1992; Enarson et al. 2000; Centers for Disease
Control and Prevention 1998,2000; Burman and Jones
2001). New guidelines and updates are regularly pub-
lished by the WHO, Centers for Disease Control and
other regulatory bodies and these should be referred
to by the clinician (Centers for Disease Control and
Prevention 1998,2000).
29.9.1
Drug Regimens for Treatment of Tuberculosis
in HIV-Negative Individuals
Currently recommended anti-tuberculosis drug
regimens were developed from a large series of well-
controlled trials in large numbers of sputum-positive
subjects with pulmonary tuberculosis well before the
HIV epidemic. Modern anti-tuberculosis treatment is
based on a combination of several drugs that that rap-
idly sterilise the sputum and prevent the emergence
of drug resistance. At present, the minimum recom-
mended duration of treatment based on rifampicin-
containing regimens for drug-susceptible tuberculosis
is 6 months. Treatment consists of two phases. The
A. Zumla and J. M. Grange
initial 2-month intensive phase kills rapidly divid-
ing, metabolically active bacilli and is followed by
a 4- to 6-month continuation phase, which kills the
remaining slowly metabolising organisms. Four drugs
(isoniazid, rifampicin, ethambutol and pyrazinamide)
are recommended during the initial intensive phase
except where local drug resistance rates are known to
be low (less than 4%). The combined use of isoniazid,
rifampicin and pyrazinamide is critical for the success
of 6-month short-course regimens (with the addition
of ethambutol or streptomycin if initial resistance
to one agent is suspected). Rifampicin and isoniazid
should be given during the entire duration of 6-
month regimens for optimal efficacy. Pyrazinamide is
necessary only during the first 2 months of treatment.
Ethambutol can be stopped if the patient's bacilli are
found to be susceptible to isoniazid and rifampicin. In
developing countries, for economic reasons, the initial
2-month phase with quadruple therapy is followed by
isoniazid and ethambutol for an additional 6 months.
Directly observed short course chemotherapy (DOTS)
has been shown to be higWy effective and cost effective
in both industrialised and developing nations (China
Tuberculosis Control Collaboration 1996; Murray et al.
1991; Wilkinson et al.1996). Thus treatment should be
supervised whenever possible.
29.9.2
Treatment ofTuberculosis in HIV-Positive
Individuals
The treatment of tuberculosis in HIV-positive indi-
viduals follows the same well-established principles
used in the treatment of non-HIV-infected indi-
viduals (Table 29.4). The choice of chemotherapy
regimen may be limited by cost, toxicity or multiple
drug resistance. In general, the standard 6-month
rifampicin-based short-course chemotherapy regi-
mens described above have been found to be highly
effective for the treatment of HIV-infected individu-
als (Chaisson et al. 1987, 1996; Murray et al. 1999;
Jones et al. 1994). The U.S. Centers for Disease Con-
trol and WHO currently recommend a minimum
of 6 months of treatment for HIV-infected as well
as HIV-non-infected individuals. If the clinical or
microbiological response is slow, it is recommended
that treatment be extended for a total of 9 months
or at least 4 months after cultures become negative
(Centers for Disease Control and Prevention 1998).
Other countries, such as Brazil, have routinely
treated HIV-infected patients for 9 months (Brasil
Ministerio da Saude 1999).
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
Table 29.4. Principles of treatment in Human Immunodefi-
ciency Virus-infected patients with tuberculosis
1. Four drug regimen for initial empirical treatment
Isoniazid (INH), rifampicin (RIF) or rifabutin (RFB), pyrazin-
amide (PZA), and ethambutol (EMB) or streptomycin (SM)
Duration of treatment should be for 6 months at least
2. Directly Observed Treatment (DOTS) is preferred for all
patients
3. Concurrent treatment of tuberculosis and HIV
Rifampicin is a potent inducer of cytochrome CYP4S0 and
thus concurrent use of Protease inhibitors (PI's) and non-
nucleoside-reverse transcriptase inhibitors (NNRTIs) are
contraindicated. Options to consider are:
- Antiretroviral regimens which do not include PIs
or NNRTIs
- Anti-tuberculosis regimens which use streptomycin
instead of rifampicin or use of INH, ethambutol
and pyrazinamide for 24 months
- Rifabutin-based regimens with changes/adjustment
to dosages
While most HIV-infected patients with tuberculo-
sis initially respond well to standard 6-month short-
course chemotherapy, their relapse rates are higher
than among HIV-non-infected persons (Perriens et al.
1995; Chaisson et al. 1996; Soriano et al. 1988; Malkin eta
1. 1997; Mallory et al. 2000). HIV infection is associated
with progressive immune suppression; in the absence
of effective anti-retroviral therapy, other opportunistic
infections such as P. carinii pneumonia and cryptococ-
cal meningitis require lifelong suppressive or secondary
prophylactic therapy. These issues have raised concerns
regarding the optimal duration of the treatment of
tuberculosis in HIV-infected patients and the need for
an extended treatment of post-treatment prophylaxis.
Relapse in HIV-infected patients may be due to recur-
rence of disease with the original Mycobacterium tuber-
culosis strain after incomplete eradication of bacilli in
an immunocompromised patient or due to exogenous
reinfection with another Mycobacterium tuberculosis
strain being transmitted in the community (Small et al.
1993; Godfrey-Faussett et al. 1994).
As a general principle, in the absence of drug
resistance, the standard WHO-recommended short-
course regimen of four drugs (rifampicin, isoniazid,
pyrazinamide and either ethambutol or streptomy-
cin) for 2 months, followed by two drugs (rifampicin
and isoniazid) for a further 4 months is recom-
mended (Scientific Statement 1998). Some physicians
continue treatment for 9 months to further reduce
the risk of relapse (Murray 1998; Centers for Dis-
ease Control 1998). Unfortunately, as outlined below,
many patients die of other AIDS-related complica-
tions during or after completion of anti-tuberculosis
therapy. In one study, extending rifampicin-based
469
therapy from 6 to 12 months reduced the relapse rate
but did not improve survival (Perriens et a1. 1995).
In addition, owing to a range of human factors,
including perception of the nature of the disease, its
treatment and outcome and the stigmatising nature of
the diagnosis, HIV-positive patients in some regions are
less likely than seronegative patients to complete treat-
ment (Ackah et al. 1995). Thus, it is essential to adopt
strategies of DOT in order to ensure completion of ther-
apy (Scientific Statement 1998). This strategy may also
help to limit the emergence of drug-resistant tuberculo-
sis but, in several African countries, even the establish-
ment of national tuberculosis programmes employing
DOT-based strategies is failing to stem the rising tide of
tuberculosis. In Botswana, for example, the incidence of
tuberculosis rose by 120% between 1989 and 1996, par-
alleling that of the prevalence of HIV infection, despite
a decade of such control strategies and a low prevalence
of drug resistance (Kenyon et al.1999).
The response to anti-tuberculosis treatment is
best monitored by repeated sputum examination (at
least monthly until the sputum culture is negative)
by smear and, where facilities exist, culture. About
85% of patients treated with modern short-course
regimens containing isoniazid and rifampicin can be
expected to convert their sputum culture to negative
after 2 months of treatment (Report 1992). Patients in
whom treatment is found to be failing on clinically or
microbiological grounds should undergo careful reas-
sessment, including consideration of compliance, drug
absorption, and drug resistance. If a decision is made
to begin additional drugs, at least two drugs must be
added to which the patient's isolate may reasonably be
expected to be susceptible. The basic rule to follow is
never to add a single drug to a failing regimen.
29.10
Anti-Retroviral Therapy
for Poor Developing Countries
Anti-retroviral therapy is the adjunctive treatment
most likely to have a major effect in reducing deaths
from tuberculosis in HIV-positive individuals (Harries
et al. 2001a). After the 13th World AIDS Conference in
July, 2000, there has been a concerted push by indus-
trialised countries to improve access to anti-retroviral
drugs for less-developed countries. Despite this push, it
will be logistically very difficult to introduce anti-retro-
viral drugs in resource-poor settings with weak health
infrastructures. However, well-functioning control pro-
grammes have an established infrastructure with the
470
ability to monitor treatment and check on compliance,
and they could provide a good entry point for the pro-
vision on a wider basis of anti-retroviral drugs in this
setting. The use ofhigWy active anti-retroviral therapy
(HAART) in HIV-positive patients with tuberculosis
is problematic because of major interactions between
rifampicin and protease inhibitors (Burman and Jones
2001). However, dual or triple nucleoside analogues
or dual nucleoside analogues with a non-nucleoside
inhibitor are therapeutic possibilities that will need to
be tested first in controlled clinical trials (to determine
tolerability and safety) and then in district operational
studies (to establish feasibility, equity issues and drug
security). There is also the need for modelling and cost
exercises to look at long-term cost benefit.
When prescribing anti-retrovirals with anti-tuber-
culosis drugs, it is important to refer to the latest guide-
lines on the subject (Dean et al. 2002) since these are
updated frequently. Two pharmacokinetic issues com-
plicate treatment: (a) the possibility of malabsorption
of drugs and (b) the complex drug-drug interactions
between anti-retroviral and anti-tuberculosis drugs,
as described above. While HIV-infected patients with
tuberculosis commonly experience adverse drug inter-
actions, current recommendations are that HAART is
commenced early in patients with advanced HIV dis-
ease (CD4+ counts<100x106 cells/I) (Dean et al. 2002).
In clinically stable patients with CD4+ cells> 1OOxl 06/1,
HAART should be deferred until the continuation
phase of tuberculosis treatment, i.e. after 2 months of
anti-tuberculosis therapy. The current recommenda-
tion in this case is to replace rifampicin by rifabutin, a
much less powerful inducer of cytochrome enzymes,
and to commence or continue with the anti-retroviral
drugs (Centers for Disease Control 1998).
29.11
Management Considerations
in HIV-Positive Individuals
There are several specific issues which arise in the
treatment of HIV-infected individuals with tubercu-
losis. These patients overall tend to have:
a. Increased morbidity rates
b. Increased mortality rates
c. Increased number of drug side effects
d. Serious interactions between anti-retroviral drugs
and anti-tuberculosis drugs
e. The immune reconstitution syndrome
f. Increased recurrence rates after completion of treat-
ment
A. Zumla and J. M. Grange
Increased Morbidity Rates: Clinical response to
anti-tuberculosis treatment, clearing of chest X-ray
abnormalities and sputum conversion rates occur at
the same rates during treatment in both HIV-posi-
tive and HIV-negative patients with tuberculosis. The
HIV-positive tuberculosis patients do, however, often
run a stormy course while on anti-tuberculosis treat-
ment due to other complicating opportunistic infec-
tions and tumours. Common presentations of these
complicating conditions include recurrent fever,
chest infections, recurrent diarrhoea, oral candida,
bacteraemia, cryptococcosis and Kaposi's sarcoma.
This increased morbidity requires increased pre-
scriptions for antibiotics, antifungal agents, anti-diar-
rhoeal agents and analgesics, rendering the care more
expensive than is the case with HIV-negative patients.
There is also evidence that delay in the diagnosis
and treatment of tuberculosis may compromise the
chances of individual cure in HIV-positive patients.
Untreated tuberculosis in HIV-infected persons may
accelerate the decline in immunocompetence and the
progression to severe immunodeficiency.
Increased Mortality Rates: HIV-positive patients not
receiving anti-retroviral therapy have a much higher
mortality during and after anti-tuberculosis treatment
compared with HIV-negative patients (Luu et al. 1994;
Murray et al. 1999; Badri et al. 2001; World Health
Organisation 2001; Harries et al. 2001b; Mukadi et al.
2001). This is illustrated by the example of the Central
African country Malawi, where HIV infection was pres-
ent in 72% of adults with smear-positive pulmonary
tuberculosis in 1995 (Harries et al.1997). In 1999, WHO
reports 22% mortality among new smear positive cases
in Malawi with treatment success in 69% (World Health
Organisation 2001), which represents good control
despite an appallingly high mortality. In sub-Saharan
Africa, approximately 30% of HIV-positive smear-
positive tuberculosis patients die within 12 months
of commencing treatment, and about 25% of those
who survive die during the subsequent 12 months.
HIV-positive patients with smear-negative pulmonary
tuberculosis (possibly because of more severe immu-
nosuppression and diagnostic difficulties) fare even
worse (Cantwell and Binkin 1997). The reasons for the
high mortality are several and varied. There are the
wide social, cultural, and economic issues of poverty,
sex, illiteracy and stigma that contribute to the great
divide between the rich industrialised nations and the
poor less-developed nations of the world. These global
inequalities affect factors such as access to care, diagno-
sis and delivery of care, all of which have an important
bearing on illness and death.
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
The introduction of HAART has dramatically
reduced mortality rates in HIV-infected patients in
the USA and in Europe. Efforts are being made to
make HAART available to all HIV-infected patients
in developing countries, although the likelihood of
this becoming a reality in the foreseeable future is
small. Since opportunistic infections are a cause of
mortality in HIV-positive persons with tuberculosis,
alternative interventions with prophylactic antibiot-
ics may be useful. A placebo-controlled clinical trial
of co-trimoxazole prophylaxis in Cote d'lvoire in
HIV-positive smear-positive pulmonary tuberculosis
patients showed a 50% reduction in mortality in the
co-trimoxazole group. The UNAIDS used this data to
suggest that all tuberculosis programmes in sub-Saha-
ran Africa consider using co-trimoxazole prophylaxis
as an intervention strategy in an attempt to reduce
mortality. A study from Senegal, by contrast, found
no advantage in using co-trimoxazole prophylaxis
and this suggests that use of such prophylaxis may not
be universally applicable. The antibiotic susceptibility
patterns of local pathogens and also the costs, cost-
benefits, and consequences of introduction of prophy-
laxis regimens require careful consideration.
29.11.1
Barriers to Effective and Appropriate Therapy
It is not uncommon to find patients registered and
being treated for tuberculosis who have an additional
disease other than tuberculosis. To treat a patient who
has left ventricular failure with anti-tuberculosis drugs
is neither helpful to the patient nor to the outcome
of tuberculosis treatment. The numbers of patients
acquiring tuberculosis has increased by 300-400%
in high HIV-prevalent countries in the past decade
(Chretien 1990). A large increase in the numbers of
patients with tuberculosis without a commensurate
increase in resources reduces the quality of patient
care. In some African tuberculosis programmes,
patients are still admitted to hospital for 1-3 months
to receive the initial phase of treatment either because
there are difficulties in administering directly observed
treatment in the community or because patients are
too ill for treatment at home. Overcrowded wards
render good nursing and medical care difficult, and
there is a substantial risk of other nosocomial infec-
tions being transmitted to the patients.
HIV-positive patients treated with the older
regimens consisting of streptomycin, isoniazid, and
thiacetazone during the early stages of the epidemic
had higher death rates compared with those given
471
rifampicin-containing regimens (Nunn et al. 1991;
Elliott and Foster 1996). The latter may offer survival
advantages because of the broad-spectrum anti-
bacterial activity of rifampicin in preventing other
bacterial infections.
29.11.2
Immune Reconstitution Syndrome
The use of anti-retroviral therapy and subsequent
improvement in immunological parameters may
actually make some patients feel worse, a paradoxical
response, due to immune reactivation. Patients may
develop ascites, lymphadenopathy, fever, increase in
the size of cerebral lesions and pleural effusions. Cli-
nicians treating tuberculosis in HIV-positive patients
receiving anti-retroviral therapy need to be aware of
this phenomenon. Another problem encountered
with the simultaneous administration of anti-ret-
rovirals and anti-tuberculosis chemotherapy is the
temporary exacerbation of the symptoms and signs
of tuberculosis-the so-called paradoxical reac-
tions (Fishman et al. 2000). These, which have been
ascribed to hypersensitivity reactions to antigen
released by bacilli killed by the chemotherapy, mani-
fest as fever, enlargement of affected lymph nodes
and a worsening of the radiological appearance.
They are occasionally encountered in HIV-negative
patients but their incidence is higher in HIV-posi-
tive patients, particularly those given anti-retroviral
drugs, probably as a result of improving immune
responsiveness. These reactions are not indicative of
treatment failure and usually subside spontaneously.
Treatment should not be modified but short courses
of steroids may be required for severe paradoxical
reactions.
29.11.3
Increased Drug Side Effects
Adverse reactions to anti-tuberculosis drugs are
more frequent in HIV-positive patients leading to
interruption of treatment and occasional fatalities.
In HIV-positive patients who were given the old
treatment regimens in the mid-1980s, (streptomycin,
isoniazid and thiacetazone), adverse cutaneous reac-
tions occurred in 15-20%,and up to 6% of these reac-
tions were severe with Stevens Johnson syndrome or
toxic epidermal necrolysis. Cutaneous hypersensitiv-
ity reactions appeared to be even more common in
HIV-positive children, and severe reactions were
472
frequently associated with death (Coovadia 1991).
Thiacetazone, which was a useful and cheap anti-
tuberculosis drug in the pre-HIV era, was the
main cause of adverse cutaneous reactions. For this
reason, thiacetazone usage was abolished, in favour
of ethambutol. In HIV-positive patients given short
course regimens containing rifampicin, isoniazid and
pyrazinamide adverse reactions to anti-tuberculosis
drugs appear to be infrequent, although side effects
of these drugs do occur.
In one study in Kenya, thiacetazone toxicity was
18 times more frequent in HIV-positive than in HIV-
negative patients and the risk was directly related
to the degree of immunosuppression, suggesting an
immunological basis (Nunn et al. 1991). Thus, even
though they are more expensive, therapeutic regi-
mens containing rifampicin throughout (or rifabutin
if anti-retroviral therapy is concomitantly given)
should replace regimens containing thiacetazone
(Nunn et al. 1991; Elliott and Foster 1996).
29.11.4
Drug Resistance
Several outbreaks of multi-drug resistant tubercu-
losis (MDR-TB) have been reported from indus-
trialised countries amongst patients infected with
HIV (Moro et al. 2000). HIV infection itself does
not induce MDR-TB, but it fuels the spread of this
dangerous condition by increasing susceptibility to
infection and accelerating transmission between
individuals, especially in closed confined spaces such
as prisons. Recent data collated from WHO (Pablos-
Mendez et al. 1998) has shown that MDR-TB is high-
est in India, Eastern Europe, China and South-East
Asia, and lowest in sub-Saharan Africa. Nevertheless,
given the problems faced by many tuberculosis pro-
grammes in sub-Saharan Africa and given the vir-
tual absence of second-line anti-tuberculosis drugs
in these countries, MDR-TB is a real and potential
threat to tuberculosis control (Pape et al. 1993). The
advent of HIV, with large increases in the number of
cases of tuberculosis, has threatened to overwhelm
National Tuberculosis Control Programmes in sub-
Saharan Africa.
As several mini-epidemics of multi-drug resis-
tant HIV-related tuberculosis followed exposure
to source cases within hospitals and clinics, well-
defined policies for the prevention of transmission
of tuberculosis within such institutions, including
isolation of known and suspect infectious patients,
are essential.
A. Zumla and J. M. Grange
29.11.5
Recurrence After Treatment
After completion of treatment, the recurrence rates
of tuberculosis (defined as return of clinical features
of active tuberculosis, positive sputum smears for
acid-fast bacilli or positive sputum cultures for Myco-
bacterium tuberculosis) are increased in HIV-positive
patients. Recurrence rates have been observed at
between 18-22 per 100 person-years. The proportion
of recurrence of tuberculosis in HIV-infected patients
in sub-Saharan Africa due to disease re-activation or
re-infection is unknown. This information is impor-
tant when it comes to determining strategies to prevent
tuberculosis recurrence, such as using secondary iso-
niazid prophylaxis. Patients who relapse with smear-
positive pulmonary tuberculosis are treated with the
WHO-recommended retreatment regimen (Murrayet
al. 1999). Five drugs are given in the intensive phase
of this regimen because many such patients have
acquired resistance to isoniazid and streptomycin.
29.11.6
Post-Treatment Prophylaxis
The value of post-treatment prophylaxis has been
investigated in two studies conducted in countries
with a high prevalence of tuberculosis. In a large trial
in the Democratic Republic of the Congo, 335 HIV-
infected adults who completed 6 months of standard
short course chemotherapy (2 months of isoniazid,
rifampicin, pyrazinamide, and ethambutol followed by
4 months of isoniazid and rifampicin) were randomized
to receive an additional 6 months of twice weekly iso-
niazid and rifampicin or placebo (Perriens et al. 1995).
Fewer relapses occurred in the arm receiving extended
therapy. There was, however, no effect on survival.
In a randomized clinical trial in Haiti, relapse rates
in 142 HIV-infected and 91 HIV-non-infected adults
treated with 2 months of isoniazid, rifampicin, and
pyrazinamide followed by 4 months of isoniazid and
rifampicin were determined (Fitzgerald et al. 2000).
Patients were randomly assigned to 1 year of post-
treatment daily isoniazid or placebo after the initial
anti-tuberculosis treatment. Relapse rates were tenfold
higher among HIV-infected patients (10% vs 1%).
One year of post-treatment isoniazid decreased the
relapse rate from 7.8 per 100 person-years of observa-
tion (PYO) to 1.4 cases per 100 PYO among the HIV-
infected patients but had no effect on relapse in the
HIV-non-infected patients. The benefit of post-treat-
ment isoniazid was greatest among patients with symp-
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
tomatic HIV infection at the time of initial diagnosis of
tuberculosis. Post-treatment isoniazid had no effect on
mortality. This study did not include DNA fingerprint-
ing of relapse isolates to distinguish recurrent disease
from exogenous reinfection. Tuberculosis control pro-
grammes must balance the costs oflonger durations of
treatment and the likelihood of decreased compliance
with the potential for decreased relapse rates.
29.11.7
Prophylactic Therapy
In view of the very high risk of a co-infected person
developing active tuberculosis, and the adverse effect
of this disease on the immune status and survival of
the patient, there is a very good theoretical case for
provision of prophylactic treatment for those at risk.
In practice, serious problems have been encountered
in diagnosing dual infection, ruling out active tuber-
culosis and ensuring compliance with therapy with-
out breach of confidence of enhancement of stigma.
Studies in Mrica have shown that preventive treat-
ment with isoniazid for 6-12 months provides signifi-
cant protection against tuberculosis in HIV-infected
adults, at least in the short- to medium-term (Whalen
et al.1997). Protection seems to be greatest in those with
a positive tuberculin skin test, in whom death is also less
frequent. Feasibility studies on tuberculosis preventive
therapy in sub-Saharan Mrica have given disappointing
results, and no country in the region has yet adopted
chemoprophylaxis as a strategy for tuberculosis control
in HIV-positive individuals. While it may be difficult
to implement chemoprophylaxis as a country-wide
control strategy, it could be used safely and selectively
in certain situations, such as in occupational health ser-
vices for private businesses and factories, for personnel
working in international agencies and missions and
amongst high risk groups such as health care workers
and prisoners. Thus, isoniazid chemoprophylaxis may
be useful in selected groups of people, and it may also
be useful to prevent disease recurrence after an episode
of tuberculosis has been successfully treated.
In the initial studies, a 6- tol2-month course ofisoni-
azid was found to lower the incidence of tuberculosis in
co-infected persons (Fitzgerald et al. 2000; Whalen et al.
1997). Given that HIV-infected patients are at increased
risk of developing tuberculosis, the case for giving them
prophylactic anti-tuberculosis chemotherapy has been
the subject of intense debate (Wllkinson et al. 1998).
Prevention of tuberculosis in these individuals is a
logical public health aim. Before the HIV epidemic, iso-
niazid preventive therapy was being recommended for
473
contacts of patients with active disease. Concern was,
however, expressed about the increased side effects of
anti-tuberculosis drugs in HIV-positive individuals and
a number of clinical trials have now been undertaken
to examine the safety and efficacy of chemoprophylaxis
in this group of individuals. Studies of varying design
from Haiti, Zambia and Uganda (Fitzgerald et al. 2000;
Whalen eta 1. 1997; Wilkinson et al. 1998) have shown
that chemoprophylaxis in HIV-infected adults signifi-
cantly reduced the incidence of tuberculosis. The ques-
tion of how long the protection lasts and whether such
therapy can lead to the emergence of drug-resistant
strains of tuberculosis remain, however, to be answered.
Importantly, operational issues discussed below need to
be addressed. Tuberculosis infection in HIV-infected
individuals tends to reactivate after therapy stops.
Although prophylaxis leads to a reduction of the
risk of tuberculosis by approximately 60% in tubercu-
lin skin test positive adults with HIV infection, identi-
fying HIV-infected individuals is difficult in resource-
poor settings (Wilkinson et al.1998; Bucher et al.1999).
The development of voluntary counselling and testing
centres is seen as an effective tool to promote safer
sex through counselling and to offer those with HIV
infection interventions such as preventive therapy
for tuberculosis. Early experiences of this approach
in Uganda were disappointing, as high drop-out rates
were reported (Aisu et al.1995). Improved counselling
and the development of rapid on-site HIV testing are
yielding better results. It is nevertheless unlikely that
enough HIV-infected individuals can be identified and
preventive therapy should therefore be seen as benefit-
ing the individual rather than playing a major role in
controlling tuberculosis in tropical countries with a
high HIV prevalence. The relatively short-term benefit
of preventive therapy is a further problem (Quigley et
al. 2001; Johnson et al. 2001).
Subsequently, shorter combination regimens were
also shown to be effective. These include a 3-month
course of a rifamycin (rifampicin or rifabutin) plus
isoniazid and a 2-month course of a rifamycin plus
pyrazinamide (Wilkinson et al. 1998; Halsey et al.
1998) and,in tuberculin-positive HIV-infected patients
3-month regimens (isoniazid+rifampicin+pyrazin
amide) provide sustained protection up to 3 years
(Johnson et al. 2001). A study in Zambia revealed that
the 2-month combination regimens or 6 months of
isoniazid, administered twice weekly, reduced the inci-
dence of tuberculosis by about 40% compared with a
placebo group, although the overall mortality due to all
causes was not reduced (Mwinga et al. 1998). Preven-
tion was more effective in those with relatively limited
immunosuppression (positive tuberculin tests, high
474
lymphocyte counts and high haemoglobin levels). By
18 months, the incidence of tuberculosis in those who
had received prophylaxis was similar to that in the pla-
cebo group; indicating the need to consider repeated
courses or, perhaps, lifelong prophylactic treatment.
29.11.8
Malabsorption
There is some evidence that anti-tuberculosis drugs,
particularly rifampicin, may be poorly absorbed in
HIV-positive patients (Peloquin et al.1996). Rifampicin
is given at a dose of 10 mg/kg irrespective of whether
it is administered daily, three times a week, or twice a
week (Arthur et al. 2001). If rifampicin is malabsorbed,
drug concentrations in tissues and blood may be con-
siderably reduced if given on an intermittent basis.
Furthermore, in many tuberculosis programmes in
sub-Saharan Africa, the drug is not given strictly on
an mg/kg weight basis but rather according to weight
bands. Careful clinical studies need to be done in HIV-
positive patients to find out whether anti-tuberculosis
treatment regimens, given on an intermittent basis,
are as effective as those given on a daily basis. These
studies should be coupled with research on rifampicin
pharmacokinetics to determine whether rifampicin
concentrations are adequate in HIV-positive patients
receiving intermittent treatment and whether rifam-
picin concentrations are adequate across the weight
distribution within the different weight bands.
One of the reasons for the higher death rate in
HIV-positive patients with smear-negative pulmo-
nary tuberculosis may be the weaker regimen (often
standard treatment) given to such patients. The opti-
mum treatment regimen for smear-negative patients
needs to be determined.
29.11.9
Adjunctive Treatments
Adjunctive therapy may be needed to reduce early
deaths. Several reasons for early deaths in HIV-infected
patients with tuberculosis have been delineated-late
presentation of patients with severe and extensive
tuberculosis; life-threatening HIV-related complica-
tions such as severe anaemia and bacteraemia; and
the occurrence of a Herxheimer-type reaction due to
the rapid killing of tubercle bacilli by anti-tuberculosis
drugs (Anglaret et al. 1999). Due to resource implica-
tions, the identification of bacteraemia in many Afri-
can hospitals would be difficult, and ill patients might
A. Zumla and J. M. Grange
require an empirical course of antibiotics to treat
commonly occurring infections due to Streptococ-
cus pneumoniae and non-typhoidal Salmonella spp.
There appears to be some benefit of corticosteroids
in decreasing mortality from tuberculosis pericardial
effusion (Wragg and Strang 2000).
Although co-trimoxazole prophylaxis for the pre-
vention of opportunistic infections was shown in Cote
d'Ivoire to significantly reduce the death rate (Anglaret
et al.1999; Wragg and Strang 2000; Wiktor et al.1999),
it is unknown whether this effect will be replicated
elsewhere in Africa (Maynart et al. 2001). There is a
wide variation of prevalence of opportunistic patho-
gens and patterns of antibiotic susceptibilities of these
pathogens and the effectiveness of co-trimoxazole
prophylaxis requires validating through other con-
trolled trials. Despite provisional recommendations by
UNAIDS (2000b) that co-trimoxazole should be given
to all people in Africa living with AIDS (by definition
this includes HIV-positive patients with tuberculosis),
it would be prudent to gather more evidence about the
effectiveness, feasibility, optimum timing and cost-
effectiveness of this intervention. There are currently
three placebo-controlled clinical trials in Central
Africa that may well answer some of these questions.
Even if co-trimoxazole is effective and its use is
feasible, its widespread use by AIDS patients might
have serious adverse consequences, such as increased
drug resistance This could compromise the treatment
of acute respiratory infections in children, in whom
co-trimoxazole is first-line therapy, as well as malaria
in countries that use sulphadoxine-pyrimethamine as
their first-line treatment. Other antibiotics, such as
oral quinolones, which have good antibacterial effects
against non-typhoidal Salmonella species, may play
a role in chemoprophylaxis. Non-antibiotic interven-
tions such as multivitamins and mineral supplements
such as zinc and selenium may improve cell-mediated
immunity in HIV-positive patients (Mocchegiani and
Muzzioli 2000; Baum et al. 2000), and their efficacy
should be determined in placebo-controlled trials.
29.12
Issues Relating to Control of Tuberculosis
in HIV-Endemic Areas
29.12.1
Quality of Health Care
Regular clinical care and treatment for HIV-related
complications in patients on anti-tuberculosis treat-
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
ment may be far from adequate in resource-con-
strained health facilities in sub-Saharan Africa. The
adequacy of clinical care needs to be formally docu-
mented. If such deficiencies are identified, research
into health systems and quality-of-care issues will
need to be undertaken, with a particular focus on
staff morale and motivation. The capacity and quality
of health care provided by the local health service and
the health status of the staff (clinical officers, clini-
cians, nurses, technical personnel) available to care
for patients influence the outcome of the treatment of
tuberculosis. The health staff in many African coun-
tries has the same HIV-seroprevalence rates as the
general adult population, and in some urban areas
this approaches or exceeds 30%. High absentee rates
from work due to illness or attending funerals, and
high death rates of the health care staff due to AIDS
threaten the capacity of many developing countries
to deliver good and effective health care. Tuberculosis
programmes are no exception to this serious threat.
29.12.2
The WHO DOTS Strategy
The best way of controlling tuberculosis in high HIV-
prevalent countries in sub-Saharan Africa appears
to be the WHO DOTS strategy. This is a six-point
strategy incorporating-
- Government commitment to tuberculosis control
- A regular supply of high quality drugs, free at the
point of delivery to the patient
- Microscopy services for diagnosis
- Supervision of therapy
- Audit of the efficacy of the control programme
- Training of staff
Countries need to be assisted in implementing
this strategy, which must be adaptable to the chal-
lenges posed by the HIV epidemic if credibility is to
be maintained, and operational research may be very
useful in this setting. A recent study has been con-
ducted in Botswana (Kenyon et al. 1999), a country
with a good DOTS programme achieving high cure
rates and very low levels of drug resistance. Botswana
is, however, a country that has seen escalating HIV-
infection rates in the last 5-10 years and this has been
associated with escalating tuberculosis case rates. It is
clear from this study that DOTS alone may not be suf-
ficient to control tuberculosis in areas with epidemic
HIV infection, and additional strategies may be
needed for such control if reductions in the incidence
of tuberculosis are to be achieved. Tuberculosis pro-
475
grammes should integrate more with AIDS control
programmes, because it is now apparent that HIV!
AIDS control is essential for tuberculosis control.
29.12.3
BeG Vaccination and HIV Serostatus
There is a risk that vaccination with BCG, a living
attenuated vaccine, will cause infectious complica-
tions in HIV-positive persons (Talbot et al. 1997).
There is a small increase in the incidence of adverse
effects of BCG in the children of HIV-infected
women, but most such effects are mild (O'Brien et
al. 1995). In one study, complications occurred in 9 of
68 HIV-infected children 3-35 months after neonatal
BCG vaccination. Regional lymphadenopathy with or
without fistula formation occurred in seven and sys-
temic disease in two children (Besnard et al. 1993).
Accordingly, the safety of BCG vaccination in regions
with a high incidence of HIV infection requires
consideration. WHO (1987) has advised that, while
persons known to be HIV positive should never be
given BCG, routine immunisation of infants should
nevertheless continue in areas with a high incidence
of tuberculosis and HIV infection. In the UK it is
recommended infants born to mothers known or
suspected to be HIV positive that should not receive
BCG unless they are subsequently shown to be HIV
negative (World Health Organization 1987).
29.12.4
Governmental Responses to the Threat Posed
by HIV and Tuberculosis
Some countries have introduced information, education
and communication campaigns for the general public
and front-line health workers about the need for early
submission of sputum samples from patients with a
chronic cough. There is, however, little evidence that this
has had any effect in reducing diagnostic delays. Algo-
rithms are in place for the diagnosis of smear-positive
pulmonary tuberculosis based on chest symptoms, lack
of response to antibiotics, negative sputum smears, and
chest radiography. Moves are afoot to decentralise the
initial phase of treatment to peripheral health centres
and the community, with the use of family-based mem-
bers, community-care groups or shopkeepers to admin-
ister directly observed therapy. Many patients have dif-
ficulties attending health centres for directly observed
therapy on a daily basis. Therefore some programmes
based on rifampicin-containing regimens have moved
476
from daily treatment in the initial phase to intermittent
treatment, either three times or twice a week.
The WHO DOTS strategy is principally aimed at
detecting and treating patients with smear-positive
pulmonary tuberculosis patients. The growing burden
of smear-negative tuberculosis is now recognised and
several programmes now address such patients.
29.12.5
National Tuberculosis Control Programmes
The overall objective of tuberculosis control is to
reduce mortality, morbidity, and transmission of the
disease until it no longer poses a threat to public health.
The strategy appears simple in principle but is difficult
in practice. Standardised combination chemotherapy
is provided to, at least, all sputum smear-positive tuber-
culosis patients. This treatment cures the disease and
prevents future transmission of infection within the
community. Targets for tuberculosis control include
curing 85% of detected new smear-positive tuberculo-
sis cases and detecting 70% of the existing cases. The
achievement of a high cure rate is the highest prior-
ity because tuberculosis programmes with high cure
rates are thought to attract a large number of existing
cases within their catchment areas. Rates of treatment
completion are lower in DOTS programmes in high
HIV-prevalent countries in sub-Saharan Africa. This is
not because the delivery of DOTS is worse, but prin-
cipally because of high case fatality rates due to HIV
coinfection (Sonnenberg et al. 2001).
29.13
Conclusions
The emergence of the HIV and tuberculosis pandemics
over the last 2 decades have had a devastating effect on
tuberculosis world-wide, especially in poor developing
countries. In 2001, around 10% of all cases of tubercu-
losis will be HIV-related, rising to over 20% in Africa.
In addition, 30% of the expected 3 million or more
AIDS-related deaths in 2001 were due to tuberculosis.
It is expected that the problem will likewise become
extensive in Asia over the next decade. Many African
countries report that their health care facilities are
overwhelmed by the additional burden of HIV-related
tuberculosis and urgent action is therefore required to
stem this epidemic before it becomes totally unman-
ageable with devastating human and economic conse-
quences that will be felt throughout the world.
A. Zumla and J. M. Grange
References
Abrams EJ, Weedon J, Bertolli J, Bornschlegel K, Cervia J,
Mendez H et al (2001) Aging cohort of perinatally Human
Immunodeficiency Virus-infected children in New York
City. New York City Pediatric Surveillance of Diseases
Consortium. Pediatr Infect Dis J 20:511-517
Ackah AN, Coulibaly D, Digbeu H et al (1995) Response
to treatment, mortality, and CD4 lymphocyte counts in
HIV-infected persons with tuberculosis in Abidjan, Cote
d'Ivoire. Lancet 345:607-610
Ahmed Y, Mwaba P, Chintu C, Grange JM, Ustianowski A,
Zumla A (1999) A study of maternal mortality at the Uni-
versity Teaching hospital, Lusaka, Zambia: the emergence
of tuberculosis as a major non-obstetric cause of maternal
death. Int J Tuberc Lung Dis 3:8:675-680
Aisu T, Raviglione MC, van Praag E et al (1995) Preventive
chemotherapy for HIV-associated tuberculosis in Uganda:
an operational assessment at a voluntary counselling and
testing centre. AIDS 9:267-273
Anglaret X, Chene G, Attia A, Lafont S, Combe P, Manlan K,
N'Dri-Yoman T, Salamon R (1999) Early chemoprophyalxis
with trimethoprim-sulphamethoxazole for HIV-l-infected
adults in Abidjan, Cote d'Ivoire: a randomised trial. Lancet
353:1463-1468
Arthur G, Nduba VN, Kariuki SM, Kimari J, Bhatt SM, Gilks
CF (2001) Trends in bloodstream infections among
Human Immunodeficiency Virus-infected adults admitted
to a hospital in Nairobi, Kenya, during the last decade. Clin
Infect Dis 33:248-256
Awil PO, Bowlin SJ, Daniel TM (1997) Radiology of pulmonary
tuberculosis and Human Immunodeficiency Virus infec-
tion in Gulu, Uganda. Eur Respir J 10:615-618
Badri M, Ehrlich R, Wood R, Pulerwitz T, Maartens G (2001)
Association between tuberculosis and HIV disease pro-
gression in a high tuberculosis prevalence area. Int J Tuberc
Lung Dis 5:225-232
Batungwanayo J, Taelman H, Dhote R et al (1992) Pulmonary
tuberculosis in Kigali, Rwanda: impact of Human immu-
nodeficiency virus infection on clinical and radiographic
presentation. Ann Rev Respir Dis 146:53-56
Baum MK, Miguez-Burbano MJ, Campa A, Shor-Posner G
(2000) Selenium and interleukins in persons infected with
Human Immunodeficiency Virus type 1. J Infect Dis 182
[Suppll):S69-S73
Bellamy R, Ruwende C, Corrah T, McAdam KP, Whittle H, Hill
AVS (1998) Variations in the NRAMPlgene and suscepti-
bility to tuberculosis in West Africans. N Engl J Med 338:
10:640-644
Bern C (1996) the value of naked eye examination of biopsied
lymph nodes in the diagnosis of tuberculous lymphadeni-
tis. Trop Doct 26:10-13
Bern C (1997) Human immuno-deficiency virus positive
tuberculous lymphadenitis in central Africa: clinical pre-
sentation of 157 cases. Int J Tuberc Lung Dis 3:215-219
Bern C, Patil P, Elliott A et al (1993) The value of wide-needle
aspiration in the diagnosis of tuberculous lymphadenitis
in Africa. AIDS 7:1221-1225
Besnard M, Sauvion S,Offredo C et al (1993) Blanche S Bacil-
lus Calmette-Guerin infection after vaccination of Human
Immunodeficiency Virus-infected children. Pediatr Infect
Dis 12:993-997
Bocchino M, Sanduzzi A, Bariffi F (2000) Mycobacterium tuber-
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
culosis and HIV co-infection in the lung: synergis immune
dysregulation leading to disease progression. Monaldi Arch
Chest Dis 55:381-388
Borun M, Sajduda A, Pawlowska I, McFadden JJ, Dziadek J
(2001) Detection of Mycobacterium tuberculosis in clinical
samples using insertion sequences ISII0 and IS990. Tuber-
culosis (Edinb) 81:271-278
Brasil Ministerio da Saude (1999) Coinfeccao TB/HIV/AIDS.
Bol Epidemiol AIDS 6
Bruchfeld J, Aderaye G, Palme IB, Bjorvatn B, Kallenuis G,
Lindquist L (2000) Sputum concentration improves diag-
nosis of tuberculosis in a setting with a high prevalence of
HIV. Trans R Soc Trop Med Hyg 94:677-680
Bucher HC, Griffith LE, Guyatt GH, Sudre P, Naef M, Sendi P,
Battegay M (1999) Isoniazid prophylaxis for tuberculosis in
HIV infection: a meta-analysis of randomized controlled
trials. AIDS 13:501-507
Burman WJ, Jones BE (2001) Treatment of HIV-related tuber-
culosis in the era of effective antiretroviral therapy. Am J
Resp Crit Care Med 164:7-12
Cantwell MF, Binkin NJ (1997) Impact of HIV on tuberculosis
in sub-Saharan Africa: a regional perspective. Int J Tuberc
Lung Dis 1:205-214
Cauthen GM, Dooley SW, Onorato 1M et al (1996) Transmission
of Mycobacterium tuberculosis from tuberculosis patients
with HIV infection or AIDS. Am J Epidemioll44:69-77
Centers for Disease Control (1993) The 1993 revised classifica-
tion system for HIV infection and expanded surveillance
case definition for AIDS among adolescents and adults.
Morbid Mortal Wkly Rep 41:1-19
Centers for Disease Control and Prevention (1998) Pre-
vention and treatment of tuberculosis among patients
infected with Human Immunodeficiency Virus: principles
of therapy and revised recommendations. MMWR Morb
Mortal Wkly Rep 47:1-58
Centers for Disease Control and Prevention (2000) Updated
guidelines for the use of rifabutin or rifampin for the treat-
ment and prevention of tuberculosis among HIV-infected
patients taking protease inhibitors or non-nucleoside
reverse transcriptase inhibitors. MMWR Morb Mortal
Wkly Rep 49:185-189
Chaisson RE, Schecter GF, Theuer CP, Rutherford GW, Echen-
berg DF, Hopewell PC (1987) Tuberculosis in patients with
acquired immunodeficiency syndrome. Clinical features,
response to therapy, and survival. Am Rev Respir Dis 136:
570-574
Chaisson RE, Clermont HC, Holt EA, Cantave M, Johnson MP,
Atkinson J, Davis H, Boulos R, Quinn TC, Halsey NA (1996)
Six-month supervised intermittent tuberculosis therapy in
Haitian patients with and without HIV infection. Am J
Respir Crit Care Med 154:1034-1038
China Tuberculosis Control Collaboration (1996) Results of
directly observed short-course chemotherapy in 112,842
Chinese patients with smear-positive tuberculosis. Lancet
347:358-362
Chintu C, Zumla A (1997) Pediatric tuberculosis and the HIV
Epidemic. In: Zumla A, Johnson M, Miller RF (eds) AIDS
and respiratory medicine. Chapman and Hall, London, pp
153-163
Chintu C, Luo C, Bhat G, Raviglione M, du Pont H, Zumla A
(1993) Cutaneous hypersensitivity reactions to thioacet-
azone in the treatment of tuberculosis in Zambian children
infected with HIV. Arch Dis Child 68:665-668
477
Chintu C, Luo C, Bhat G et al (1995) Impact of the Human
Immunodeficiency Virus type-Ion common paediatric
illnesses in Zambia. J Trop Paediatr 41:348-353
Chintu C, Mudenda V, Lucas S, Nunn A, Lishimpi K, Maswahu
D, Kasolo F, Mwaba P, Bhat G, Terunuma H, Zumla A (2002)
Lung diseases at necropsy in African children dying of
respiratory illnesses. Lancet 360:985-990
Chretien J (1990) Tuberculosis and HIV. The cursed duet. Bull
Int Union Tuberc Lung Dis 65:25-28
Clerici M, Shearer GM (1994) TheThl-Th2 hypothesis of HIV
infection: new insights. Immunol Today 15:575-581
Coker R (1998) Lessons from New York's tuberculosis epi-
demic. Br Med J 317:616
Coker R, Miller R (1997) HIV associated tuberculosis. Br Med
J 314:1847
Colebunders R, Bastian I (2000) A review of the diagnosis and
treatment of smear-negative pulmonary tuberculosis. Int J
Tuberc Lung Dis 4:97-107
Coovadia HM (1991) Tuberculosis in children. In: Coovadia
HM, Banatar SR (eds) A century of tuberculosis - a South
African perspective. Oxford University Press, Cape Town,
pp 91-105
Cruciani M, Malena M, Bosco 0, Gatti G, Serpelloni G (2001)
The impact of Human Immunodeficiency Virus type 1 on
infectiousness of tuberculosis: a meta-analysis. Clin Infect
Dis 33:1922-1930
Dean GL, Edwards SG, Ives NJ, Matthews G, Fox EF, Navaratne
L, Fisher M, Taylor GP, Miller R, Taylor CB, de Ruiter A, Poz-
niak AL (2002) Treatment of tuberculosis in HIV-infected
persons in the era of highly active antiretroviral therapy.
AIDS 16:75-83
Di Lonardo M, Isola NC,Ambroggi M, Rybko A, Poggi S (1995)
Mycobacteria in HIV-infected patients in Buenos Aires.
Tubercle Lung Dis 76:185-189
Diagbouga S, Fumoux F, Ledru E, Sanou PT, Barro D, Marchal
G (1998) Lack of direct correlation between CD4+ lym-
phocyte counts and induration sizes of the tuberculin skin
test in Human Immunodeficiency Virus type 1 seropositive
patients. Int J Tuberc Lung Dis 2:317-323
Donald PR (2000) Childhood tuberculosis. Curr Opin Pulmon
Med 6:187-192
Donald PR (2002) Childhood tuberculosis - out of control?
Curr Opin Pulm Med 8:178-182
Donald PR, Fourie B, Grange JM (eds) (1995) Tuberculosis in
childhood. Van Schaik, Pretoria, SA
Elliott AM, Foster SD (1996) Thiacetazone: time to call a halt?
Tubercle Lung Dis 77:27-29
Elliott AM, Luo N, Tembo G, Halwiindi B, Steenbergen G,
Machiels L, Pobee J, Nunn P, Hayes RJ, McAdam KP (1990)
Impact of HIV on tuberculosis in Zambia: a cross-sectional
study. Br Med J 301:412-415
Elliott AM, Hayes RJ, Halwiindi B et al (1993) The impact of
HIV on infectiousness of pulmonary tuberculosis: a com-
munity study in Zambia. AIDS 7:981-987
Enarson DA; Rieder HL; Arnadottir T (2000) Management of
tuberculosis: a guide for low income countries, 5th edn.
Pmi, Frankfurt
Espinal MA, Perez EN, Baez J, Henriquez L, Fernandez K, Lopez
M, Olivo P, Reingold AL (2000) Infectiousness of Mycobac-
terium tuberculosis in HIV-l-infected patients with tuber-
culosis: a prospective study. Lancet 355:275-280
Festenstein F, Grange JM (1993) Tuberculosis and the acquired
immune deficiency syndrome. J Appl Bacteriol 71:3036
478
Fishman JE, Saraf-Lavi E, Narita M, Hollender ES, Ramsinghani
R,Ashkin D (2000) Pulmonary tuberculosis in AIDS patients:
transient chest radiographic worsening after initiation of
antiretroviral therapy. AJR Am J Roengenol 174:43-49
Fitzgerald DW, Desvarieux M, Severe P et al (2000) Effect of
post-treatment isoniazid on prevention of recurrent tuber-
culosis in HIV-1-infected individuals: a randomised trial.
Lancet 356:1470-1474
Foster SD, Godfrey-Faussett P, Porter J (1997) Modelling the eco-
nomic benefits of tuberculosis preventive therapy for people
with HIV: the example of Zambia. AIDS 11:919-925
Friedman LN, Williams MT, Singh TP, Frieden TR (1996)
Tuberculosis, AIDS, and death among substance abusers
on welfare in New Yo'rk City. N Engl J Med 334:828-833
Fylkesnes K, Musonda RM, Kasumba K, Ndhlovu Z, Mluanda
F, Kaetano L, Chipaila CC (1997) The HIV epidemic in
Zambia: socio-demographic prevalence patterns and
indications of trends among childbearing women. AIDS
11:339-345
Gilks CF, Brindle RJ, Otieno LS et al (1990) Extrapulmonary
and disseminated tuberculosis in HIV-l-seropositive
patients presenting to the acute medical services in Nai-
robi. AIDS 4:981-985
Godfrey-Faussett P, Githui W, Batchelor B et al (1994) Recur-
rence of HIV-related tuberculosis in an endemic area may
be due to relapse or reinfection. Tubercle 75:199-202
Goletti D, Weissman D, Jackson RW, Graham MH, Vlahov D,
Klein RS, Munsiff S, Otorna L, Cauda R, Fauci AC (1996)
Effect of Mycobacterium tuberculosis on HIV replication:
role of immune activation. J ImmunoI157:1271-1276
Goyal M, Shaw RJ, Bannerjee DK, Coker RJ, Robertson BD,
Young DB (1997) Rapid detection of multidrug resistant
tubeculosis. Eur Respir J 10:1120-1124
Halsey NA, Coberly JS, Desormeaux Jet al (1998) Randomised
trial of isoniazid versus rifampicin and pyrazinamide for
prevention of tuberculosis in HIV-l infection. Lancet 351:
786-792
Hargreaves NJ, Kadzakumanja 0, Phiri S, Nyangulu DS, Salani-
poni FM, Harries AD, Squire SB (2001) What causes smear-
negative pulmonary tuberculosis in Malawi, an area of high
HIV seroprevalence? Int J Tuberc Lung Dis 5:113-122
Harries AD (1998) Tuberculosis in HIV-infected persons with
special emphasis on sub-Saharan Africa. J Infect 37:205-209
Harries AD, Maher D (1996) TB/HIV a clinical manual (WHO/
TB/96.200). World Health Organisation, Geneva
Harries AD, Parry C, Nyongonya Mbewe L, Graham SM, Daley
HM et al (1997) The pattern of tuberculosis in Queen
Elizabeth Central Hospital, Blantyre, Malawi: 1986-1995.
Int J Tuberc Lung Dis 1:346-351
Harries AD, Nyangulu DS, Kang'ombe C et al (1998) Treatment
outcome of an unselected cohort of tuberculosis patients in
relation to Human ImmunodeficiencyVirus status in Zomba
Hospital, Malawi. Trans R Soc Trop Med Hyg 92:343-347
Harries AD, Hargreaves NJ, Kaluwa 0, Salaniponi FM (2001a)
Preventing antiretroviral anarchy in sub-saharan Africa.
Lancet 358:410-414
Harries AD, Hargreaves NJ, Kemp J, Jindani A, Enarson DA,
Maher D, Salaniponi FM (2001b) Deaths from tuberculosis
in sub-saharan African countries with a high prevalence of
HIV-I. Lancet 357:1519-1523
Hudson CP, Wood R, Maartens G (2000) Diagnosing HIV-asso-
ciated tuberculosis: reducing costs and diagnostic delay. Int
J Tuberc Lung Dis 4:240-245
A. Zumla and J. M. Grange
Huebner RE, Castro KG (1995) The changing face of tubercu-
losis. Annu Rev Med 46:47-55
Ikeogu MO, Wolf B, Mathe S (1997) Pulmonary manifestations
in HIV seropositivity and malnutrition in Zimbabwe. Arch
Dis Child 76:2:124-128
Johnson JL, Okwera A, Hom DL, Mayanja H, Kityo CM,
Nsubuga P, Nakibali JG, Loughlin AM, Yun H, Mugyenyi
PN, Vernon A, Mugerwa RD, Ellner n, Whalen CC (200l)
Duration of efficacy of treatment of latent tuberculosis
infection in HIV-infected adults. AIDS 15:2137-2147
Joint Tuberculosis Committee of the British Thoracic Society
(1998) Chemotherapy and management of tuberculosis in
the United Kingdom: recommendations. Thorax 53:536-548
Jones BE, Otaya M, Antoniskis D et al (1994) A prospective
evaluation of antituberculosis therapy in patients with
Human Immunodeficiency Virus infection. Am J Respir
Crit Care Med 150:1499-1502
Kenyon TA, Mwasekaga MJ, Huebner R, Rumisha D, Binkin
N, Maganu E (1999) Low levels of drug resistance amidst
rapidly increasing tuberculosis and Human Immunodefi-
ciency Virus co-epidemics in Botswana. Int J Tuberc Lung
Dis 3:4-11
Klausner JD, Ryder RW, Baende E et al (1993) Mycobacterium
tuberculosis in household contacts of Human Immuno-
deficiency Virus type I-seropositive patients with active
pulmonary tuberculosis in Kinshasa, Zaire. J Infect Dis
168:106-111
Kox LF, Jansen HM, Kuijper S, Kolk AH (1997) Multiplex PCR
assayfor immediate identification of the infecting species
in patients withmycobacterial disease. J Clin Microbiol 35:
1492-1498
Lalavani A, Pathan AA, Durkan H, Wilkinson KA, Whelan A
et al (2001) Enhanced contact tracing and spatial tracking
of Mycobacterium tuberculosis infection by enumeration of
antigen-specific T cells. Lancet 357:2017-2021
Lindbrathen A, Gaustad P, Hovig B, Tonjum T (1997) Direct
detection of Mycobacterium tuberculosis complex in clini-
cal samples from patients in Norway by ligase chain reac-
tion. J Clin Microbiol 35:3248-3253
Lucas S, Nelson AM (1994) Pathogenesis of tuberculosis in
Human Immunodeficiency Virus-infected people. In:
Bloom BR (ed) Tuberculosis: pathogenesis, protection and
control. ASM Press, Washington, pp 503-513
Lucas SB, de Cock KM, Hounnou A et al (1994) Slim disease
in Africa: the contribution of tubeculosis. Br Med J 308:
1531-1533
Lucas SB, Peacock CS, Hounnou A, Brattegard K, Koffi K,
Honde M et al (1996a) Disease in children infected with
HIV in Abidjan, Cote-d'Ivoire. Br Med J 312:335-338
Lucas SB, Honnou A, Koffi K, Beaumel A, Andoh J, DeCock
KM (1996b) Pathology of paediatric Human Immunodefi-
ciency Virus infections in Cote d'Ivoire. East Afr Med J 73
[SuppI5):S7-S8
Luo C, Chintu C, Bhat G, Raviglione M, Diwan V, DuPont HL,
Zumla A (1994) Human immunodeficiency virus type-l
infection in Zambian children with tuberculosis: changing
seroprevalence and evaluation of a thiacetazone-free regi-
men. Tuberc Lung Dis 75:111-115
Luu C, Chintu C, Bhat G, Raviglione M, Diwan V, Dupont HL,
Zumla A (1994) Human immunodeficiency virus type-l
infection in Zambian children with tuberculosis: changing
seroprevalence and evaluation of a thioacetazone-free regi-
men. Tuberc Lung Dis 75:110-115
Tuberculosis and Co-infection with the Human Immunodeficiency Virus
Malkin JE, Prazuck T, Simonnet F et al (1997) Tuberculosis
and Human Immunodeficiency Virus infection in West
Burkina Faso: clinical presentation and clinical evolution.
Int J Tuberc Lung Dis 1:68-74
Mallory KF, Churchyard GJ, Kleinschmidt I et al (2000) The
impact of HIV infection on recurrence of tuberculosis
in South African gold miners. Int J Tuberc Lung Dis 4:
455-462
Maynart M, Lievre L, Sow PS, Kony S, Gueye NF, Bassene E,
Metro A et al (2001) Primary prevention with co-trimoxa-
zole for HIV-1 infected adults: results of the pilot study in
Dakar, Senegal. J Acquir Immune Def Syndr 1:26
Mocchegiani E, Muzzioli M (2000) Therapeutic application of
zinc in Human Immunodeficiency Virus against opportu-
nistic infections. J Nutr 30 [5S Suppl]:1424S-1431S
Moro ML, Errante I, Infuso A, Sodano L, Gori A, Orcese CA,
Salamina G, D'Amico C, Besozzi G, Caggese L (2000)
Effectiveness of infection control measures in controlling
a nosocomial outbreak of multidrug-resistant tuberculo-
sis among HlV patients in Italy. Int J Tuberc Lung Dis 4:
61-68
Msamanga GI, Fawzi WW (1997) The double burden of HlV
infection and tuberculosis in sub-saharan Africa. N Engl J
Med 337:849-851
Mukadi YD, Maher D, Harries A (2001) Tuberculosis case
fatality rates in high HlV prevalence populations in sub-
Saharan Africa. AIDS 15:143-152
Murray J (1998) Tuberculosis and HIV infection: a global
perspective. Respiration 65:335-342
Murray cr, Styblo K, Rouillon A (1990) Tuberculosis in devel-
oping countries: burden, intervention and cost. Bull Int
Union Tuberc Lung Dis 65:6-24
Murray cr, DeJonghe E, Chum HJ, Nyangulu DS, Salomao
A, Styblo K (1991) Cost effectiveness of chemotherapy
for pulmonary tuberculosis in three sub-Saharan African
countries. Lancet 338:1305-1308
Murray J, Sonnenberg P, Shearer SC, Godfrey-Faussett P (1999)
Human immunodeficiency virus and the outcome of treat-
ment for new and recurrent pulmonary tuberculosis in
African patients. Am J Respir Crit Care Med 159:733-740
Murray J, Sonnenberg P, Shearer SC, Godfrey- Faussett P (1999)
Human immunodeficiency virus and the outcome of treat-
ment for new and recurrent pulmonary tuberculosis in
African patients. Am J Respir Crit Care Med 159:733-740
Mwaba P, Mwansa J, Chintu C, Pobee J, Scarborough M, Ports-
mouth S, Zumla A(2001) Clinical presentation, natural his-
tory and cumulative death rates in 230 adults with primary
cryptococcal meningitis in Zambian patients treated under
local conditions. Postgrad Med J 77:769-773
Mwinga A, Hosp M, Godfrey-Faussett P et al (1998) 1\vice
weekly tuberculosis preventive threrapy in HIV infection
in Zambia. AIDS 12:2447-2457
Nambuya A, Sewankambo N, Mugerwa J, Goodgame R, Lucas S
(1988) Tuberculous lymphadenitis associated with Human
Immunodeficiency Virus infection in Uganda. J Clin Pathol
41:93-96
Nunn P, Kibuga D, Gathua S et al (1991) Cutaneous hypersen-
sitivity reactions due to thiacetazone in HIV-l seropositive
patients treated for tuberculosis. Lancet 337:627-630
Nunn P, Mungai M, Nyamwaya J, Gicheha C, Brindle RJ, Dunn
DT, Githui W, Were JO, McAdam KP (1994) The effect of
Human Immunodeficiency Virus type-Ion the infectious-
ness of tuberculosis. Tuberc Lung Dis 75:25-32
479
O'Brien KL, Ruff AJ, Louis MA, Desormeaux J, Joseph DJ,
McBrien M, Coberly J, Boulos R, Halsey NA (1995) Bacillus
Calmette-Guerin complications in children born to HlV-1-
infected women with a review of the literature. Pediatrics
95:414-418
Pablos-Mendez A, Raviglione MC, Laszlo A, Binkin N, Rieder
HL, Bustreo F, Cohn DL, Lambregts-van Weezenbeek CS,
Kim SJ, Chaulet P, Nunn P (1998) Global surveillance for
antituberculosis-drug resistance, 1994-1997. World Health
Organization-International Union against Tuberculosis and
Lung Disease Working Group on Anti-Tuberculosis Drug
Resistance Surveillance. N Engl J Med 338:1644-1649
Pape JW, Jean SS, Ho JL, HafnerA,Johnson WD Jr (1993) Effect
of isoniazid prophylaxis on incidence of activetuberculo-
sis on disease progression of HIV infection. Lancet 342:
268-272
Parry CM, Kamoto 0, Harries AD, Wirima JJ, Nyirenda CM,
Nyangulu DS, Hart CA (1995) The use of sputum induc-
tion for establishing a diagnosis in patients with suspected
tuberculosis in Malawi. Tuberc Lung Dis 76:72-76
Peloquin CA, Nitta AT, Burman WJ et al (1996) Low antitu-
berculosis drug concentrations in patients with AIDS. Ann
Pharmacother 30:919-925
Perriens JH, St Louis ME, Mukudi MB et al (1995) Pulmonary
tuberculosis in HlV-infected patients in Zaire: a controlled
trial of treatment for either 6 or 12 months. N Engl J Med
32:779-784
Pillay K, Colvin M, Williams R, Coovadia HM (2001) Impact of
HIV-I infection in South Africa. Arch Dis Child 85:505-551
Pithie AD, Chicksen B (1992) Fine-needle extrathoracic Iym-
phnode aspiration in HIV-associated sputum-negative
tuberculosis. Lancet 340:1504-1505
Pozniac AL, Latif AS, Neill P et al (1992) Pulmonary Kaposi's
sarcoma in africa. Thorax 47:770-773
Quigley MA, Mwinga A, Hosp M, Lisse I, Fuchs D, Porter JD,
Godfrey-Faussett P (2001) Long-term effect of preventive
therapy for tuberculosis in a cohort of HIV-infected Zam-
bian adults. AIDS 15:215-222
Report (1992) Subcommittee of the Joint Tuberculosis Com-
mittee of the British Thoracic Society. Guidelines on the
management of tuberculosis and HlV infections in the
United Kingdom. Br Med J 304:1231-1233
Report (2001) Global tuberculosis control (WHO/CDS/
2001.287). World Health Organization, Geneva
Rose AM, Watson JM, Graham C, Nunn AJ, Drobniewski F,
Ormerod LP, Darbyshire JH, Leese J (2001) Public Health
Laboratory Service/British Thoracic Society/Department
of Health Collaborative Group.Tuberculosis at the end of
the 20th century in England and Wales: results of a national
survey in 1998. Thorax 56:173-179
Saks AM, Posner R (1992) Tuberculosis in HIV positive
patients in South Africa; a comparative radiological study
with HlV negative patients. Clin RadioI46:387-390
Sanduzzi A, Fraziano M, Mariani F (2001) Monocytes/
macrophages in HIV infection and tuberculosis. J Bioi
Regul Homeost Agents 15:294-298
Schaaf HS, Geldenduys A, Gie RP, Cotton MF (1998) Culture-
positive tuberculosis in Human Immunodeficiency Virus
type I-infected children. Pediatr Infect Dis J 17: 599-604
Schaefer R, Goldberg R, Sierra M, Glatt AE (1989) Frequency
of Mycobacterium tuberculosis bacteremia in patients with
tuberculosis in an area endemic for AIDS. Am Rev Respir
Dis 140:1611-1613
480
Schauf V, Rom WN, Smith KA,Sampaio EP, Meyn PA, Tramon-
tana JM, Cohn ZA, Kaplan G (1993) Cytokine geneactiva-
tion and modified responsiveness to interleukin-2 in the
blood oftuberculosis patients. J Infect Dis 168:1056-1059
Scientific Statement (1998) Treatment regimens in HIV-
infected tuberculosis patients. An official statement of
the International Union Against Tuberculosis and Lung
Disease. Int J Tuberc Lung Dis 2:175-178
Small PM, Shafer RW, Hopewell PC et al (1993) Exogenous
reinfection with multidrug-resistant Mycobacterium tuber-
culosis in patients with advanced HIV infection. N Engl J
Med 328:1137-1144
Sonnenberg P, Murray J, Glynn JR, Shearer S, Kambashi B,
Godfrey-Faussett P (2001) HIV-l and recurrence, relapse,
and reinfection of tuberculosis after cure: a cohort study in
South African mine workers. Lancet 358:1687-1693
Soriano EJ, Mallolas J, Gatell JM et al (1988) Characteristics of
tuberculosis in HIV-infected patients: a case-control study.
AIDS 2:429-432
Starke JR (1993) Childhood tuberculosis. A diagnostic
dilemma. Chest 104:393-404
Starke JR, Taylor-Watts RT (1989) Tuberculosis in the paediat-
ric population in Houston, Texas. Pediatrics 84:28-35
Taha TE, Graham SM, Kumwenda NI et al (2000) Morbidity
among Human Immunodeficiency Virus infected and
uninfected African children. Pediatricsl06:1-8
Talbot EA, Perkins MD, Silva SFM, Frothingham R (1997) Dis-
seminated bacille Calmette-Guerin disease after vaccina-
tion: case report and review. Clin Infect Dis 24:1139-1146
Tshibwabwa-Tumba E, Mwinga A, Pobee JOM and Zumla A
(1997) Radiological features of pulmonary tuberculosis in
963 HIV-infected adults in three Central African Hospitals.
Clin RadioI52:837-841
UNAIDS (1998) AIDS epidemic update, Dec 1998. World
Health Organization, Geneva
UNAIDS (2000a) AIDS epidemic update, Dec 2000. UNAIDS,
World Health Organisation, Geneva
UNAIDS (2000b) Use of co-trimoxazole prophylaxis in adults
and children living with HIV/AIDS in Africa. Recom-
mendations and operational issues. Apr 2000. url: http:
I Iwww.unaids.org/publications/documents/care/general!
recommendations-eng.pdf
Vetter KM, Djomand G, Zadi F, Diaby L, Brattegaard K, Timite
M,Andoh J,Adou JA,de Cock KM (1996) Clinical spectrum
of Human Immunodeficiency Virus disease in children in a
West African city. Pediatr Infect Dis J 15:438-452
Waddell RD, von Reyn CF, Baboo KS, Mwinga A, Chintu C,
Zumla A (1999) The effects of BCG immunization and
Human Immunodeficiency Virus infection on dual skin test
reactions to purified protein derivative and Mycobacterium
avium sensitin among adults in Zambia. Int J Tuberc Lung
Dis 3:255-260
Wallis RS, Johnson JL (2001) Adult tuberculosis in the 21st
century: pathogenesis, clinical features and management.
Curr Opin Pulmon Med 7:124-133
Watson JM, Meredith SK, Whitmore-Overton E, Bannister B,
Darbyshire JH (1993) Tuberculosis and HIV: estimates of
the overlap in England and Wales. Thorax 48:199-203
A. Zumla and J. M. Grange
Whalen C, Horsburgh CR, Hom D, Lahart C, Simberkoff M,
Ellner J (1995) Accelerated course of Human Immunodefi-
ciency Virus infection after tuberculosis. Am J Respir Crit
Care Med 151:129-134
Whalen CC, Johnson JL, Okwera A et al (1997) A trial of
three regimens to prevent tuberculosis in Ugandan
adults infected with the Human Immunodeficiency Virus.
Uganda-Case Western Reserve University Research Col-
laboration. N Engl J Med 337:801-808
Wiktor SZ, Sassan-Morokro M, Grant AD, Abouya L, Karon
JM, Maurice C et al (1999) Efficacy of trimethprim-sul-
phamethoxazole prophylaxis to decrease morbidity and
mortality in HIV-l infected patients with tuberculosis
in Abidjan, Cote d'lvoire: a randomized controlled trial.
Lancet 353:1469-1475
Wilkinson D, Davies GR, Connolly C (1996) Directly observed
therapy for tuberculosis in rural South Africa, 1991
through 1994. Am J Public Health 86: 1094-1097
Wilkinson D, Gilks CF (1998) Increasing frequency of tuber-
culosis among staff in a South African district hospital:
impact of the HIV epidemic on the supply side of health
care. Trans R Soc Trop Med Hyg 92:500-502
Wilkinson D, Squire SB, Garner P (1998) Effect of preventive
treatment for tuberculosis in adults infected with HIV:
systematic review of randomised placebo controlled trials.
BMJ 317:625-629
World Health Organization (1987) HIV and routine childhood
immunization. Wkly Epidemiol Rec 62:297-299
World Health Organization (1994) TB - a global emergency
(WHO/TB/1994.177). WHO, Geneva
World Health Organization (1996) Tuberculosis in the Era of
HIY. A deadly partnership. WHO, Geneva
World Health Organisation (2001) Global tuberculosis control.
WHO report (WHO/CDS/TBI2001.287). WHO, Geneva
Wragg A, Strang JI (2000) Tuberculous pericarditis and HIV
infection. Heart 84:127-128
Yates MD, Pozniak A, Grange JM (1993) Isolation of mycobac-
teria from patients seropositive for the Human Immuno-
deficiency Virus (HIV) in South East England: 1984-1992.
Thorax 48:990-995
Zar HJ, Hanslo D, Tannenbaum E, Klein M,Argent A, Eley B et
al (2001) Aetiology and outcome of pneumonia in Human
Immunodeficiency Virus-infected children hospitalized in
south Africa. Acta Paediatr 90:119-125
Zhang Y, Nakata K, Weiden M, RomWN (1995) Mycobacterium
tuberculosis enhances humanimmunodeficiency virus-l
replication by transcriptional activation at the long termi-
nal repeat. J Clin Invest 95:2324-2328
Zumla A, James DG (1999) Granulomatous Infections - an
overview. In: James DG, Zumla A (eds) Granulomatous
disorders. Cambridge University Press, Cambridge, pp
103-121
Zumla A, Chintu C (1995) Childhood tuberculosis and infec-
tion with the Human immunodeficiency virus. J R Coli
Phys 29:2:92-95
Zumla A, Rook GAW, Mwaba P, Lucas S (1999) Tuberculosis.
In: James DG, Zumla A (eds) Granulomatous disorders.
Cambridge University Press, Cambridge, pp 132-160
30 Spinal Tuberculosis
M. MONIR MADKOUR, M. WASEF AL SEBAI, KHALAF R. AL MOUTAERY

CONTENTS Hippocrates later described spinal tuberculosis

and classified the causes of spinal diseases as: (a) the
30.1 Introduction 481 existence of "tubercles" in the lung, (b) spinal injury,
30.2 Epidemiology 482
30.3 Pathogenesis 482 (c) age and extreme fatigue of the spine, (d) painful
30.4 Clinical Features 483 conditions (Marketos and Skiadas 1999). Hippocrates
30.5 Cranio-cervical Junction Tuberculosis 483 clearly associated the pathogenesis of spinal tubercu-
30.6 Spinal Tuberculosis with Neurological losis to its existence in the lungs as "tubercles:'
Deficits 484 Galen, "glorious Galen;' "Clarissimus Galenus" or
30.7 Clinical Features in Children 485
30.8 Imaging Features 485 Galen of Pergamon, lived during the 2nd century A.D.
30.8.1 Plain Radiography 485 At the age of 20 years, he went to Alexandria, Egypt to
30.8.2 Conventional Tomography 486 enrich his medical knowledge. In Egypt he had the
30.8.3 Myelography 486 chance to examine anatomy closely and learned about
30.8.4 Scintigraphy 486
human dissections. Alexandria was the most impor-
30.8.5 CT Scans 486
30.8.6 MRI Imaging 486 tant medical center in the old world and was famous
30.9 Diagnosis 487 for anatomical studies among many other medical spe-
30.9.1 Case Illustration 487 cialties. Galen learned the neuroanatomy of the spine
30.10. Spinal Tuberculosis Versus Spinal in Alexandria (van Staden 1992). Galen endorsed the
Brucellosis 489
Hippocratic division of the disease and the association
30.11 Anti-Tuberculous Chemotherapy
for Spinal Tuberculosis 490 between spinal and pulmonary tuberculosis. He also
References 490 noted that vertebral tuberculosis above the diaphragm
occurs at a young age with dramatic development and
it is difficult to cure. Below the diaphragm, vertebral
tuberculosis was often associated with abscess forma-
tion on both sides of the lumbar spine, in the psoases
30.1 and in the groin (Marketos and Skiadas 1999). In 1779,
Introduction spinal tuberculosis became known as Pott's Caries or
disease. Sir Percival Pou, a London surgeon born in
Features of spinal tuberculosis has been noted, 1714, described spinal tuberculosis with subsequent
reported and described by the Ancient Egyptians development of Pott's paralysis or paraplegia.
5000 years ago. They noted in their paintings and Spinal deformities, kyphosis and paraplegia de-
writings its occurrence among ordinary people and scribed since ancient times remains the main concern
among the nobles (Fig. 1.1, Chap. 1 Tuberculosis in today when dealing with spinal tuberculosis. Spinal
Ancient Egypt). surgery developed in Hong Kong in the 1950s in
regard to establishing spinal stability and correcting
deformity. Other aspects of spinal tuberculosis, such
M. M. MADKOUR, MD, DM, FRCP
as the epidemiology, pathogenesis, clinical and imag-
Consultant, Department of Medicine, Riyadh Armed Forces ing features as well as the problem of spinal growth
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia in children post-operatively, the late consequences of
M. W. AL SEBAI spinal canal narrowing and neurological deficits will
Consultant Spinal Surgeon, Riyadh Armed Forces Hospital, be discussed in this chapter. The following chapter
P.O. Box 7897, Riyadh 11159, Saudi Arabia
K. R. AL MOUTAERY, MD, FRCS (Ed), FACS
is on spinal surgery, where our imaging features of
Head of Neurosurgery, Riyadh Armed Forces Hospital, spinal tuberculosis will be presented together with
P.O. Box 7897, Riyadh 11159, Saudi Arabia the surgical management and follow-up.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
482
30.2
Epidemiology
Skeletal tuberculosis is one of the most common
extrapulmonary manifestations of the disease, both
in developed and developing countries (Pertuiset et al.
1999).The incidence ofskeletal tuberculosis is rising in
recent years with the resurgence of tuberculosis due to
the human immunodeficiency virus (HIV) epidemic.
In Zambia, Jellis (2002) reported a fourfold increase
in the incidence of osteoarticular tuberculosis, 5 years
after the start of HIV epidemic in 1980s. This author
reported 188 cases of osteoarticular tuberculosis
between 1991 and 1995 and found that 17% of children
and 60% of adults were HIV positive Oellis 1996). It is
estimated that 2 million or more patients worldwide
have active spinal tuberculosis (Rajasekaran et al.
1998). Skeletal tuberculosis accounts for 1-5% of all
tuberculosis cases and 15% of extrapulmonary tuber-
culosis and spinal involvement accounts for approxi-
mately 50% of all skeletal tuberculosis.
In a recent report from Denmark, Houshian and
colleagues (2000) studied the epidemiology of bone
and joint tuberculosis during the period 1993-1997.
They found 95 cases of bone and joint tuberculosis,
accounting for 4% of all tuberculosis cases and 15%
of extrapulmonary cases and the spine was affected
in 50%. These authors noted that immigrant popula-
tions were more affected than Danes and the diagno-
sis was often delayed for several months or years.
In New York, the incidence of spinal tuberculosis
was reported to be rapidly increasing in recent years
(Rezai et al. 1995). The Medical Research Council
(MRC) in England and Wales surveyed the number
of tuberculosis notifications for 6 months in 1983,
and found active tuberculosis in 3002 patients.
Respiratory tuberculosis alone was found in 2032
(68%),745 (25%) had non-respiratory disease alone
and 225 (7%) had both respiratory and non-respira-
tory tuberculosis. Bone and joint tuberculosis was
reported in 150 patients (5% of all patients and
15% of non-pulmonary tuberculosis). The spine
was the most common skeletal site and was found
in 60 patients (40%). This survey noted that the age
was over 55 years in 50% in the white population;
while in the Indian subcontinent migrants the age
was less than 35 years in 60% of the cases. In our
own series of 2484 patients with tuberculosis seen
between 1983 and 2000 inclusive, the spine, bone
and joints were affected in 292 patients (11.76%).
Spinal tuberculosis with and without joint involve-
ment was encountered in 212 patients (72.6%) (see
Table 30.1).
M. M. Madkour et aI.
30.3
Pathogenesis
Tuberculous spondylitis occurs as a result of hema-
togenous spread of M. tuberculosis bacilli to the
vertebrae (Tuli 2002; Jellis 1995). M. bovis accounts
for 5-10% of tuberculous spondylitis in develop-
ing countries (O'Reilly and Daborn 1995). In Latin
America, M. bovis accounts for 2% of pulmonary
and 8% of extra pulmonary tuberculosis (Cosivi et
al. 1998). Non-tuberculous mycobacterial organisms
account for 1-4% of tuberculosis cases (Sanders and
Horowitz 1995; Shembekar and Babhulkar 2002).
Rarely, non-tuberculous mycobacterial organism can
also lead to spondylitis, particularly among patients
co-infected with HIV. M. Xenopi is most commonly a
pulmonary pathogen in immunosuppressed patients
or those with underlying lung disease. Only four
cases of spine infection with this organism have been
reported (Danesh-Clough et al. 2000; Govender et al.
2000; Leibert et al. 1996; Weiner et al. 1998). Hema-
togenous spread of the bacilli from a distant source,
commonly pulmonary, via epiphyseal vessels to the
epiphyseal part of the vertebral body adjacent to the
superior or inferior end plates (Doub and Badgley
1932; Boachie-Adjei and Squillante 1996; Tuli 2002).
Branches of metaphyseal anastomozing arteries
pass superiorly and inferiorly, joining the upper and
lower anastomozing arteries together and crossing
the intervertebral disc to link up with the metaphy-
seal systems above and below. Therefore, spread of
infection through these communicating arteries to
vertebrae above and below usually occur. The cen-
tral part of the vertebral body may be the site of
initial localization of infection via nutrient arteries
and branches of the spinal arteries. This may be the
case when a single vertebral body is initially affected.
Subsequently, spread to other vertebrae may occur
via subligamentous spread. Rarely, it may start in
the posterior element of the vertebra (Weaver and
Lifeso 1984).
Inflammatory cellular infiltrates with granuloma,
necrosis and caseation reactions may occur. Granu-
lomata may enlarge and spread to form an inflam-
matory tissue mass with abscess formation to involve
the entire vertebral body, the disc and adjacent verte-
brae. The infection may spread beneath the anterior
or posterior longitudinal ligaments to affect other
vertebrae. Softening, demineralization and destruc-
tion by infection may lead to vertebral body wedging
or total collapse (Boachie-Adjei and Squillante 1996).
Abscess formation may progress and increase in size
as the infection progress (Fig.31.10 a, b, c - Chap. 31).
Spinal Tuberculosis
The accumulation of this abscess may become tense
and under pressure and may extend to areas with
least mechanical resistance or may penetrate through
a neighboring organ (Bailey et al. 1972). Adhesions,
fibrosis and swelling of the surrounding anatomical
structures usually follow. Rarely, collapse of the ver-
tebral body may not occur and the infection may be
contained in one or two vertebrae where calcification
and sequestrum formation may take place {Fig. 31.9
a, b, 31.27c - Chap. 3l}. In the cervical spine, inflam-
matory swelling and abscess formation may appear
behind the perivertebral fascia and form a retropha-
ryngeal mass that may cause dysphasia {Fig. 31.13,
31.23, 31.28 - Chap.31). The abscess content may
extend to the sternomastoid muscle and form a neck
cold abscess. It may track down to the mediasti-
num and communicate via fistula to the esophagus,
trachea, aorta or pleural cavity (Bailey et al. 1972).
Tuberculous spondylitis of the thoracic spine may be
accidentally depicted by chest radiography. Abscess
formation may appear as a well-defined posterior
mediastinal mass {Fig. 31.8 - Chap.3l}.
As the disease progresses with vertebral body col-
lapse, inflammatory granulomatous tissue and abscess
mass may bulge posteriorly into the spinal canal.
Mechanical compression of the cord, compromising
its blood supply through pressure or by inducing vas-
culitis may occur, leading to paraplegia {Fig. 31.2,31.5,
31.6,31.12,31.13,31.15,31.16,31.17,31.20 - Chap.3l}
(Hsu et al.1988; Tuli et al.1967; Martin 1971). Posterior
abscess formation can be noted clinically as a soft-
tissue fluctuating swelling close to the spine. Rupture
of this abscess through the skin with discharging sinus
may occur (Fig. 31.11e - Chap. 31).
The abscess contents may rarely track down and
follow intercostal nerve course pointing to the tho-
racic wall with cold abscess or discharging sinus
formations. Lumbar spine tuberculous abscess
commonly penetrates through the sheath of the
psoas muscle and gravitates down to the iliac fossa
(Boachie-Adjei and Squillante 1996) {Fig. 31.4, 31.6e
- Chap.3l}. It may progress further beneath the
inguinal ligament to the thigh or to the iliac crest
anteriorly or posteriorly with abscess swelling (Tuli
et al. 1967). Anti-tuberculous drugs with or with-
out surgical intervention will arrest this destruc-
tive inflammatory process. Healing takes places in
patients responding to treatment with resulting
fusion between adjacent infected vertebrae. Narrow-
ing of the spinal canal and deformities may occur
with possible long-term neurological deficits later on
(Martin 1970,1971; Rajasekaran and Shanmugasun-
daram 1987).
483
30.4
Clinical Features
Spinal tuberculosis has no clinical features that are
specific to the disease. It may simulate other spinal
diseases including brucellosis, pyogenic infections,
rheumatic diseases or spinal malignancy {Madkour
and Sharif 200l}. High index of suspicion by the
treating clinician is essential in order to avoid delay
in diagnosis with development of neurological defi-
cits. Proper identification of risk factors, particularly
among high-risk groups, is important. The disease
is commonly misdiagnosed initially for months or
years, both in developing and developed countries.
In developing countries, spinal tuberculosis
affects younger age groups, including children and
infants. In developed countries, spinal tuberculosis
is rare and mostly affects the elderly. However, with
the recent resurgence of tuberculosis due to the HIV
epidemic, spinal tuberculosis is on the rise among the
younger age groups. Male predominance is reported
by many authors in developing as well as developed
countries. The most serious complication of spinal
tuberculosis is the development of neurological
deficits and disabilities. These may occur in 10-46%
of patients with spinal tuberculosis and more com-
monly among those with dorsal spine region involve-
ment (Griffith 1979; Hodgson and Stock 1960; Moon
et al. 1996).
30.5
Cranio-cervical Junction Tuberculosis
Clinical features ranges from early, non-specific,
insidious symptoms to severe neurological com-
plications, and death may occur due to atlanto-
axial instability or cervico medullary compression
{Fig. 31.17 a, b, c, d, e, f, g, 31.27 - Chap. 3l} (Kanaan
et al. 1999). Night fever, sweating, loss of appetite,
weight loss and weakness are commonly present at
the onset of the disease {Table 30.l}. Backache may
be the sole presenting clinical feature in the absence
of constitutional symptoms. The onset of the disease
is usually gradual and it may take several months or
years for presentation and to establish the diagnosis.
The onset may rarely be acute, particularly among
those co-infected with HIV (Sigh et al. 1998). Other
clinical features of active tuberculosis of lung paren-
chyma, pleura or lymph nodes may also be present.
These may provide clues to the cause and nature of
the spinal disease. Active or old pulmonary and pleu-
484 M. M. Madkour et al.

Table 30.1. Spinal tuberculosis: clinical and investigative features

AI-Othman Rasit Leibert Our own

et al. (2001) et al. (2001) et al. (1996) series
n=69 n=53 n=26 n=212

Age (mean) years 52.8 40.2 45 34.3

Sex (male %) 53 70 70 47
Backache (%) 84 94 100 92
Fever (%) 32 57 64
Kyphosis (%) 17 22 33.5
Paresis (%) 28 44 4 36
Pulmonary tuberculosis (%) 7 18 7 26
Positive imaging (%) 80 47 94
Paraspinal abscess (%) 80 45 68
Positive histology (%) 57 44 (317)42 69
Positive culture (%) 70 (717)100 53
Had surgery (%) 46 57 11 43

ral tuberculosis may be found in 5-65% of patients 30.6

with spinal tuberculosis (Fig. 31.8d - Chap. 31) (AI-
Othman et al. 2001). Pleural effusion may mask an
underlying spinal involvement on plain chest radi-
ography. Accidental discovery of spinal tuberculosis
may be found during investigations of symptoms
not directly related to the spine. Dysphagia, dyspnea,
hoarseness of voice, neck pain and swelling may be
the initial presenting feature of retropharyngeal
abscess (Fig. 31.28 - Chap. 31) due to cervical spine
tuberculosis (Tirri et al. 2000; Pollard and EI-Beheiry
1999; AI-Soub 1996; Ventura et al. 1996).
It may also present as a cold abscess in the neck
with or without discharging sinus formation. Pos-
terior mediastinal shadow (Fig. 31.8 - Chap. 31)
depicted by chest radiography may be an accidental
finding in patients with lung parenchymal or pleural
tuberculosis (Fig. 31.8 - Chap. 31).
Localized back pain at the site of spinal involve-
ment aggravated on active movement is a common
feature. Pain may be referred along the spinal
nerve and may be misdiagnosed as neuralgia, sci-
atica (Fig. 31.9 - Chap. 31) or abdominal conditions
(Nussbaum et al. 1985; Humphries et al.1986). Local-
ized spinal tenderness is usually elicited by direct
percussion of the affected spine. Kyphosis or severe
kyphotic deformities are commonly noted on clinical
examination.
The frequency of clinical symptoms, signs and
other investigative features of spinal tuberculosis can
be seen in Table 30.1.
Cold abscess with or without discharging sinus
may also be found in the chest wall, paraspinal area
(Fig. 31.11 e - Chap. 31), abdomen, groin or the thigh
(Fig. 191 - Chap. 19).
Spinal Tuberculosis with Neurological
Deficits
Spinal tuberculosis may cause neurological deficits
either early due to active disease, which is the most
common cause, or rarely as late onset after treatment
(Luk and Krisha 1996; Bilsel et al. 2000; Rajeswari et
al. 1997).
Active spinal tuberculosis may be associated with
neurological deficit in 27% and up to 63% of patients
(Jain et al. 1999a,b; Srivastav and Sanghavi 2000). It is
more commonly reported when craniocervical junc-
tion or CI-C2 spine are involved (Fig. 31.17, 31.27, 31.28
- Chap. 31). Bhojray and colleagues (2001) reported 25
patients seen over a 12-year period with craniocervi-
cal junction tuberculosis. Conservative treatment of 16
patients and surgery on nine patients achieved good
results with no serious complications. Other authors
reported similar association of high incidence of neu-
rological deficits and craniocervical junction tubercu-
losis (Edward et al. 2000; Jain et al.1999; Allali et al. 2000;
Vaigya et al.1995; Ventura et al.1996; Parry et al.1999).
Neurological deficits are commonly manifested as
paraparesis or paraplegia (Fig. 31.2, 31.3, 31.5, 31.6,
31.7, 31.12, 31.13, 31.14 - Chap.31) with variable
degrees of severity. Spinal tuberculosis affects adults
as well as children with similar features of neurologi-
cal deficit manifestations (Bilsel et al. 2000; Moon et
al. 1996). Spinal tuberculosis of the cervical region
may rarely cause monoplegia, hemiplegia or quad-
riplegia (Dharnrni et aI. 2001; Lrnejjeti et aI. 2000;
Vaidya et al.1995; Hsu et al. 1984).
Neurological deficits due to active spinal tubercu-
losis is usually caused by compression of the cord by
Spinal Tuberculosis
granulomatous tissue mass and abscess or rarely as a
result of vasculitis and vascular compression. Sublux-
ation of the atlantoaxial joint leading to compression
of the medulla oblongata and upper spinal cord is
rarely reported (Allali et al. 2000).
Encroachment on the spinal canal by tuberculosis
abscess and granulomata was assessed by Jain et al.
(1999) in 15 patients with spinal tuberculosis without
neurological deficits. These authors indicated that an
encroachment of up to 76% of the spinal canal by
tuberculous tissue is compatible with undisturbed
neural status.
Paraparesis or paraplegia may develop many
years after treatment with anti-tuberculous chemo-
therapy. Late onset paraplegia may occur as a result
of reactivation of the disease, severe form of unstable
kyphotic deformities, bony bridge compression or
spinal stenosis above healed tuberculous kyphosis
(Luk and Krishna 1996; Rajeswari et al. 1997; Bilsel
et al. 2000).
Painful radiculopathy of the arm with muscle weak-
ness in the hand with a claw deformity may be the ini-
tial presenting feature. Difficulty in the diagnosis may
delay the management. Gopalakrishnan and Krishna
(2002) reported a patient with cervicothoracic junc-
tion spinal tuberculosis with radiculopathy.
30.7
Clinical Features in Children
Spinal tuberculosis in children is almost exclusively
reported from developing countries, particularly
India, Africa, Hong Kong and Korea. In children, the
disease is more severe, extensive and involves more
vertebrae than in adults. It causes more spinal defor-
mities, kyphosis and neurological deficits (Fig. 31.5,
31.17, 31.19, 31.28 - Chap. 31). The most frequent
clinical manifestations of spinal tuberculosis in chil-
dren include back pain and deformities, weakness of
the legs with difficulty in walking, painful torticollis
or dysphagia (Rajasekaran et al. 1998; Rajasekaran
2001; Akhadder et al. 1999; Lukhele 1996; Beekarun
et al. 1995; Sayi and Mlay 1995).
Neurological symptoms and signs of paraparesis
or paraplegia are common among children with
spinal tuberculosis. It occurs in 25-63% of children
with spinal tuberculosis. The diagnosis may be ini-
tially difficult but medical treatment alone or in com-
bination with debridement surgery are commonly
associated with neurological recovery. Kyphosis
and deformities are common in children with spinal
485
tuberculosis. Its progression may continue after med-
ical and surgical treatment, particularly with anterior
approach surgery (Rajasekaran 2001; Schulitz et al.
1997). However, many authors have different views
on the best approach, either anterior or posterior, of
operative debridement of the spine and the outcome
of kyphosis (Upadhyay et al. 1996).
30.8
Imaging Features
30.8.1
Plain Radiography
Initial early changes at the metaphyseal area anteri-
orly, adjacent to superior and inferior end plates, are
difficult to depict on plan radiography (Sharif et al.
1989; Weaver and Lifeso 1984) (Fig. 31.1 - Chap. 31).
Later, decalcification, lytic lesions may appear and
are best seen on lateral view (Harisinghani et al.
2000; McGuinness 2000) (Fig. 31.1a, b - Chap. 31).
The anterior boundary of the vertebral body will
show diminution of the bone intensity. As the infec-
tion progresses, the overlying cortex and the disc
destruction occur with loss of the vertebral body
height and narrowing of the disc space (Weaver
and Lifeso 1984; Harisinghani et al. 2000). Spread of
infection through the intermetaphyseal communicat-
ing vessels and beneath both longitudinal ligaments,
leads to involvement of contiguous or rarely to dis-
tant (skip lesions) vertebral involvement (Fig. 31.7,
31.9,31.13,31.20 - Chap. 31).
As the anterior part of the vertebral bodies are the
common site of infection, wedging with kyphosis is
common (Weaver and Lifeso 1984). Rarely when the
initial infection affects the central part (equatorial
region) of the vertebral body, decalcification (lytic
lesions) may be seen on lateral view (Fig. 31.14 -
Chap. 31). Paravertebral granulomatous tissue mass
with abscess formation is usually present at the time
of patient's presentation. In the cervical spine, it will
appear as a retropharyngeal mass (Fig. 31.17a and
31.23a - Chap. 31). In the thoracic spine, the abscess
lies between the spine and the pleural reflections of
the lungs and will appear as a mediastinal fusiform
soft-tissue opacity (Fig.31.8a, 31.17b - Chap.31).
Lung parenchymal shadowing due to active tuber-
culosis may also be seen on chest radiography. In
the abdomen, paraspinal abscesses are more difficult
to be seen on plain radiography. The absence of the
psoas muscles soft-tissue shadow is a helpful clue to
486
its presence and calcification may be noted (Fig. 31.4a
- Chap. 31). Posterior element involvement (pedicles,
arch, spinous process) is usually secondary to verte-
bral body infection but rarely may occur in isolation
(Fig. 31.11, 31.14 - Chap. 31). Subsequent instability
and vertebral displacement may occur (Fig. 31.1 b,
31.13, 31.14 - Chap.31). Lytic lesions or localized
decalcification in the posterior element due to tuber-
culosis may be confused with neoplastic disease.
30.8.2
Conventional Tomography
Conventional tomography is still used in some
developing countries where computed tomography
(CT) and magnetic resonance imaging (MRI) scan-
ning are not available (AI Arabi et al. 1992). Loss of
contrast between various tissues is a disadvantage,
but eliminating the images of overlying structures is
its main advantage. It is a useful imaging modality
in assessing occipitocervical junction tuberculosis
(Fig. 31.16, 31.17 - Chap. 31).
30.8.3
Myelography
Extradural granulomatous mass and abscess may only
be depicted by myelography in poor developing coun-
tries where CT and MRI facilities are not available (AI
Arabi et al. 1992) (Fig. 31.2, 31.12, 31.15 - Chap. 31).
30.8.4
Scintigraphy
Scintigraphy is a very useful imaging modality in
determining skip spinal involvement or detecting
other infected sites that are not symptomatic or sus-
pected. For details, please see the chapter entitled,
Scintigraphy
30.8.5
CTScans
CT or its more advanced versions will show all the
changes that are depicted on plain radiography (Sharif
et al.1995a, 1989; Lin-Greenberg and Cholankeril1990;
AI Arabi et al.1992; Jain et al.1993; McGuinness 2000).
In addition, CT shows the surrounding soft-tissues,
paraspinal abscesses and the spinal canal. The extent
M. M. Madkour et al.
of the vertebral damage, disc destruction and soft-
tissue calcification are better seen by CT scan than by
plain radiography. The main disadvantage of CT scan
is that only the chosen field is visualized, not the whole
area affected. CT guided abscess drainage and tissue
biopsy are mandatory and needed during investiga-
tion and imaging assessment of spinal tuberculosis.
Early vertebral body lesions may appear as focal low-
density, homogenous areas with regular boundaries
and are the characteristic findings on CT scan.
As the infection progresses, intraosseous abscesses
with bone fragments and pus will be seen in the ver-
tebral body as depicted by CT scan (Fig. 31.1c, 31.2d
& e, 31.4, 31.6, 31.10 - Chap. 31). Destruction of the
vertebral body and intervertebral disc, development of
paravertebral abscess and posterior element involve-
ment can be seen as the disease advances (Fig. 318d & f,
31.13 - Chap. 31). Encroachment of the spinal canal by
granulomatous tissue and abscess, extradural compres-
sion and displacement ofthe cord can be assessed by CT
scan (Fig. 31.13d, 31.15, 31.17, 31.18, 31.27 - Chap.31).
In late stage, both osteolytic and sclerotic areas may be
present in the vertebral body, particularly after initiat-
ing treatment (Fig. 31.15 - Chap. 31). Severe kyphosis
and dislocation may be difficult to image axially by CT
scan (Fig. 31.22, 31.24, 31.26d - Chap. 31).
30.8.6
MRllmaging
The MRI is the diagnostic imaging of choice for
spinal tuberculosis. It has the advantage of detect-
ing skip spinal lesions, anterior and posterior liga-
ment involvement and extradural extension of the
disease (Sharif et al. 1990, 1995; McGuinness 2000).
Decreased signal intensity of the affected vertebrae,
loss of disc height and para-vertebral swelling are
typically depicted by T 1 weighted images (Fig. 31.5,
31.7,31.9,31.16, 31.17f, 31.27 d & e - Chap. 31). T z-
weighted images often show increased signal inten-
sity of the vertebrae affected and the surrounding
inflammatory soft-tissue around it. Post-gadolinium
contrast images will show the rim enhancement of the
surrounding tissues, the anterior and posterior liga-
ments, the dura mater and will also show the degree
of cord compression (AI-Mulhim et al. 1995).
Para-vertebral abscesses will be depicted as iso-
tonic or low-intensity signals on T 1 weighted images.
High signal intensity of the para-vertebral abscesses
will be depicted on T z weighted images. Soft-tissue
calcifications are difficult to recognize by MRI and
better seen on CT scans.
Spinal Tuberculosis 487

30.9 mild diabetes mellitus that was well controlled by diet

Diagnosis and glibenclamide 5 mg daily. He had a myelogram
and CT scan of the lumbar spine while in Qatar that
Spinal tuberculosis may initially be difficult to diag- revealed a soft tissue density in the spinal canal as
nose or differentiate from other causes of spinal disor- well as paravertebral region at L4-5 levels. Isotope
ders such as infections, malignancies or rheumatic dis- bone scan showed increased uptake at L5. The patient
eases. The scarcity of distinctive clinical features may declined surgery and went to private doctors and
cause delay in diagnosis with subsequent increased received non-steroidal anti-inflammatory drugs.
morbidity. Imaging modalities play an important On arrival at our hospital the patient looked ill,
investigative role in providing clues to the diagnosis with fever, severe low back pain that was radiating
when used in conjunction with clinical, microbiologi- to the right lower limb, and difficulty in walking. His
cal and histopathological findings. Granulomatous temperature was 39.8°C but other vital signs were
infections of the spine are common in developing normal. There was no lymphadenopathy, hepatic or
countries, particularly those due to tuberculosis and splenic enlargement, and other systems were normal.
brucellosis and rarely due to fungal infections or sar- Neurological examination showed loss of lumbar
coidosis (Madkour and Sharif 2001). lordosis, paraspinal muscle spasm and tenderness in
In developing countries where both tuberculosis the L4-5 region. Straight leg raising was restricted to
and brucellosis are endemic, spinal involvement by 30° bilaterally. Knee and ankle tendon reflexes were
either or both infections can be difficult to diag- normal, and both plantars were flexor. There were no
nose. In our part of the world, when both diseases sensory deficits.
are endemic, all patients presenting to our medical The hemogram was normal, apart form a raised
facilities with back pain are routinely screened for ESR (35 mm/h). Brucella agglutinins showed Agg:
brucellosis (Madkour 1989; Madkour and Sharif 2560, IgG (2ME): 640, Coomb's: negative. Blood cul-
2001; Madkour et al. 1985, 1987, 1988; Sharif et al. tures were negative.
1989, 1990, 1995). Samples for histopathological and Plain radiographs of the spine showed erosion
microbiological examinations obtained by CT guided at the posterior inferior aspect of the body of L5
tissue biopsies and pus aspirations of spinal lesions (Fig. 30.1). Isotope bone scan showed increased
and paraspinal abscess are essential for the confir- uptake at L4 and L5 regions. CT scan of the spine
mation of the diagnosis. The detection of active lung showed erosion and destructive infective lesion of L5.
parenchymal tuberculosis, lymph nodes enlarge- MRI of the lumbar spine showed diffuse loss of signal
ment, cold abscess or discharging sinuses provides from the bodies of L4 and L5 on the short TR view.
additional sources of samples for histopathological Long TR view showed extensive soft tissue component
and microbiological evidences of the diagnosis. in the posterior aspect of L4 and L5 (Fig. 30.1). Soft
Other laboratory data, such as the hemogram, sedi- tissue (granulation) in the spinal canal was noted. The
mentation rate (ESR), biochemical parameters and disc space was within normal limits. CT scan-guided
tuberculin skin test do not provide confirmatory needle biopsies of the vertebral body and soft tissues
evidences of the diagnosis. were carried out. They showed chronic inflammatory
cell infiltrates but no granulomata. The patient was
initially treated with doxycycline 100 mg twice daily
30.9.1 and netilmicin 150 mg 8-hourly intravenously with
Case Illustration regular measurements of plasma levels.
The patient did not respond to treatment and
A 42-year-old man was referred to our hospital in remained febrile. Anterior spinal decompression and
January 1988. He had had brucellosis diagnosed bone grafting of the L4-5 region was carried out. At
in Qatar in May 1987. Brucella melitensis organ- operation an abscess, originating in the L4-5 region
isms were grown from the blood and his brucella and extending down to the iliac fossa, was noted.
agglutinins were high. He used to drink goat and About 50 ml of pus was drained from the abscess.
camel's milk. He was treated with a combination Extensive granulation tissue was seen in the L4-5
of streptomycin and tetracycline for 2 weeks, and region, pressing anteriorly on the dura. A smear of
tetracycline alone for a further 4 weeks. Although the drained pus showed tubercle bacilli.
he responded initially to this treatment, the patient Netilmicin and doxycycline were replaced with
remained symptomatic with persistent low back pain, rifampicin 600 mg, isoniazid 300 mg and ethambutol
fever, chills, sweating and weight loss. The patient had 1.2 g daily. The patient responded well to anti-tuber-
 488 M. M. Madkour et al.

b e

Fig. 30.1. A 42-year-old man with brucellar and tuberculous spondylitis.

Patient's brucella titer on admission was 1:20, 840. B. melitensis was iso-
lated from the blood. a Frontal radiograph showing early formation of a
lateral osteophyte in the right side (arrow). b Lateral radiograph. There is
a minimal reduction of the disc at L4/L5. Minimal cortical loss is noted in
the posterior aspect of the inferior end plate ofL4 (arrow). c Anterior scinti-
gram showing diffuse increased uptake in the bodies of L4 and L5. The disc
is still seen as a photopenic area. d Patchy bone destruction in the bodies
of L4 and L5, and large extradural mass compromising the spinal canal. e
Mid-sagittal cut on magnetic resonance imaging (short TRITE) showing
loss of signal in L4 and L5. The disc space does not appear to be reduced
but there is a large extradural component that was high signal compared
to cerebrospinal fluid (arrows). Acid-fast bacilli were isolated from around
c 50 ml of fluid drained from a paraspinal abscess at operation
Spinal Tuberculosis 489

culous chemotherapy and the temperature settled. The
back pain and sciatica improved. Culture of the pus did
not grow the tubercle organisms. He is being followed
up in the orthopedic as well as brucellosis clinics.
This patient had spondylitis probably due both
to concomitant brucella and tuberculous infections.
While brucellosis was treated, his symptoms per-
sisted. It is only at operation that tubercle bacilli were
positively identified by direct smear. Cultures from
the pus were negative.
30,10,
Spinal Tuberculosis Versus Spinal
Brucellosis
Several studies were carried out by our group,
at Riyadh Military Hospital, on the clinical and
imaging features of spinal brucellosis as well as a
comparison with spinal tuberculosis. Both of these
granulomatous diseases of the spine are common in
Saudi Arabia. They share many clinical and imag-
ing features and co-infections, although rare, often
lead to difficulties in diagnosis (Madkour 1989;
Madkour and Sharif 2001; Madkour et al. 1985, 1987,
1988; Sharif et al. 1989,1990,1995). We encountered
occasional cases of spinal tuberculosis in children
but have never seen it due to brucellosis in this age
group. The onset of symptoms of spinal tuberculosis
is commonly gradual, but in brucellosis spinal symp-
toms are usually acute. Tables 30.2-30.4 summarize
the clinical and imaging features of 17 patients with
spinal brucellosis and 15 patients with spinal tuber-
culosis that we prospectively evaluated.
Pyogenic spinal infections usually have a more
acute presentation with symptoms of sepsis and sep-
ticemia. No cardiosis and actinomycosis may present
with more insidious onset and course (Sharif et al.
1990). Metastatic disease of the spine may be dif-

Table 30.2. Patient population, age and sex distribution, anatomic sites of lesions and invasive procedures performed. Note: num-
bers in parentheses are percentages

Type of Sex distribution Age distribution Number of Site of lesions Patients who
infection (years) affected sides underwent

Male Female Range Average Cervical Thoracic Lumbar Surgery Biopsy

Brucellar 12 5 43-70 61 19" 3(16) 3(16) 13(68) 6

spondylitis
(n=17)
Tuberculosis 7 8 30-66 51 15 0 11(73) 4(27) 13 5b
spondylitis
(n=15)

"Two patients had double lesions in two different sites

b Two patients underwent both percutaneous biopsy and surgical decompression

Table 30.3. Plain radiographic findings. Note: none of the patients had involvement of posterior elements. Numbers in parentheses
are percentages

Type of Number Bone destruction Bone sclerosis Disk involvement Paravertebral Associated
infection of affected of single vertebra of single vertebra soft tissue spinal deformity
vertebrae mass (per site) (per site)

Focal Diffuse Focal Diffuse Disk space Disk gas

collapse

Brucellar 38 20 1(2.6) 18" 0 16 5 0 2(10.5)

spondylitis
(n=17)
Tuberculous 34 4 29(82) 0 11 14 0 lIb 9'
spondylitis
(n=15)

" Anterior osteophytes (parrot's beak) were seen in 14 vertebrae

b Abscesses not identified on radiographs included an upper thoracic lesion (T4 to T5) and three lumbar lesions.
'No associated spinal deformity was detected in six patients (three thoracic and three lumbar lesions)
490
Table 30.4. High-resolution findings. Note: numbers in parentheses are numbers of sites
Type of Number Bone destruction Bone sclerosis of single vertebra
infection of affected of single vertebra
vertebra
Focal Diffuse Focal Diffuse Abscess Granulation Loss of
Brucellar 34(16) 26 8 17 0 0
spondylitis
(n=14)a
Tuberculous 32(24) 2 32 0 11 14'
spondylitis
(n=14)b
a Ninepatients underwent contrast-enhanced computed tomography
b Allpatients underwent contrast-enhanced computed tomography
'Three patients with thoracic lesions had vertebral abscesses communicating with coexisting pleural collections; one of these
abscesses was extending into the erector spinae
ficult to differentiate without a tissue biopsy. Other
diseases that may mimic spinal tuberculosis include
fungal infections, hydatid disease and syphilis.
30.11
Anti-Tuberculous Chemotherapy
for Spinal Tuberculosis
Chemotherapy for spinal tuberculosis remains the
cornerstone of treatment and should be started early,
promptly and for adequate period to avoid the occur-
rence ofneurological complications (Boachie-Adjei and
Squillante 1996; Jain 2002; Shembekar and Babhulkar
2002; Tuli 2002). Prompt chemotherapy can reverse
neurological deficits and minimize the potential dis-
ability resulting from spinal tuberculosis. Neurological
deficits in association with active spinal tuberculosis
can be reversed and potential disabilities can be mini-
mized successfully with combination of chemotherapy
and surgical decompression (Govender et al. 2001a-c;
Bailey et al. 1972; Chahal and Jyothi 1990; Garst 1992;
Hodgson and Stock 1960; Lifeso et al.1985; Lifeso 1987;
Martin 1970; Tuli et al. 1967). A combination of che-
motherapy and surgical decompression is associated
with successful outcome in most patients with spinal
tuberculosis complicated with neurological deficits
(Medical Research Council 1978). Surgery for patients
without neurological deficits may only be required
in about 6%, while among those with neurological
deficits, it may be required in up to 60% (Upadhyay
et al. 1993). The overall surgical rate of spinal tuber-
culosis with and without neurological deficit was
reported as 49.5% by Lifeso and colleagues from
M. M. Madkour et al.
Paraspinal soft tissue abnormality
(per site)
Bone frag- Disk Epidural
tissue muscle fat ments or gas Extension
planes calcifications
6 11 8 9
0 5 0 14
Riyadh, Saudi Arabia (Lifeso et al. 1985; Lifeso 1987).
The chemotherapy success rate for osteoarticular TB
is estimated as greater than 90% (Barnes and Barrows
1993; Satoskar et al.1999). The outcome of chemother-
apy, however, is guarded by some problems such as the
chronicity, development of drug resistance and toxic-
ity, poor patient compliance and the presence of other
diseases that compromise patients immunity such as
HIV, diabetes mellitus and alcoholism (Shembekar
and Babhulkar 2002). The most commonly used regi-
men of chemotherapy is the initial use of first line
drugs (rifampicin and isoniazid with pyrazinamide
or streptomycin) for 2 months in a daily dosing. This
is followed by a minimum of 4- to 8-month period
using two drugs (rifampicin and isoniazid) in a daily
dosing. Other regimen using two or three times a week
given as directly observed therapy (DOT) can also be
used (see chapter on DOT). For patients with drug
resistant tuberculosis, please refer to Chap. 46 (multi-
drug resistant TB). Non-tuberculous mycobacterial
infections are usually resistant to first and second
line drugs. Amikacin, fluoroquinolone, rifabutin and
clarithromycin are effective agents for the treatment
of these infections (Barnes and Barrows 1993; Mandell
and Petri 1987; Shembekar and Babhulkar 2002).
References
Akhaddar A et al (1999) Craniocervical junction tuberculosis
in children. Rev Rhum Engl Ed 66:739-742
Al Arabi KM, al Sebai MW et al (1992) Evaluation of radiological
investigations in spinal dtuberculosis. Int Orthop 16:165-167
Allali F et al (2000) Atlantoaxial tuberculosis: three cases. Joint
Bone Spine 67:481-484
Spinal Tuberculosis
Al-Mulhim FA, Ibrahim EM et al (1995) Magnetic resonance
imaging of tuberculous spondylitis. Spine 20:2287-2292
AI-Othman A et al (2001) Tuberculous spondylitis: analysis of
69 cases from Saudi Arabia. Spine 26:E565-E570
Al-Soub H (1996) Retropharyngeal abscess associated with tuber-
culosis of the cervical spine. Tuberc Lung Dis 77: 563-565
Bailey Hi, Gabriel M, Hodgson AR et al (1972) Tuberculosis of
the spine in children. Operative findings and results in one
hundred consecutive patients treated by removal of the lesion
and anterior grafting. J Bone Joint Surg 54A:1633-1657
Barnes PF, Barrows SA (1993) Tuberculosis in the 1990's. Ann
Intern Med 119:400-410
Beekarun DD et al (1995) Atypical spinal tuberculosis in chil-
dren. J Pediatr Orthop 15:148-151
Bhojraj et al (2001) Tuberculosis of the craniocervical junc-
tion. J Bone Joint Surg Br 83:222-225
Bilsel N et al (2000) Late onset paraplegia. Spinal Cord 38:
669-674
Boachie-Adjei 0, Squillante RG (1996) Tuberculosis of the
spine. Orthop Clin North Am 27:95-103
Chahal AS, Jyothi SP (1990) The radical treatment of tubercu-
losis of the spine. Int Orthop 4:93-99
Cosivi 0, Grange JM et al (1998) Zoonotic tuberculosis duets
Mycobacterium bovis in developing countries. Emerg Infect
Dis 4:59-70
Danesh-Clough T et al (2000) Mycobacterium xenopi infection
of the spine: a case report and literature review. Spine 25:
626-628
Dhammi IK, Singh S, Jain AK (2001) Hemiplegic/monoplegic
presentation of cervical spine (C l-C2) tuberculosis. Eur
Spine J 10:540-544
Doub HP, Badgley CE (1932) The roentgen signs of tuberculo-
sis of the vertebral body. Am J RoentegenoI27:827-837
Edwards RJ et al (2000) Management of tuberculomas of the
craniocervical junction. Br J Neurosurg 14:600
Garst RJ (1992) Tuberculosis ofthe spine: a review of236 oper-
ated cases in an underdeveloped region from 1954 to 1964.
J Spinal Disord 5:286-300
Gopalakrishnan D, Krishna KN (2002) Cervico-thoracic
junction spinal tuberculosis presenting as radiculopathy.
Neurol India 50:93-94
Govender S et al (2000) Spinal tuberculosis in HIV positive
and negative patients: immunological response and clinical
outcome. Int Orthop 24:163-166
Govender S et al (2001a) Anterior spinal decompression in
HIV positive patients with tuberculosis. A prospective
study. J Bone Joint Surg Br 83:864-867
Govender S et al (2001b) Long-term follow-up assessment of
vascularized rib pedicle graft for tuberculosis Kyphosis. J
Pediatr Orthop 21:281-284
Govender S et al (2001c) Tuberculosis of the cervicodorsal
junction. J Pediat Orthop 21:285-287
Griffths DLL (1979) The treatment of spinal tuberculosis. In:
McKibbin B (ed) Recent advances in orthopaedics. Churchill
Livingstone, Edinburgh, pp 1-17
Harisinghani MG et al (2000) Tuberculosis from head to toe.
Radiographics 20:449-470
Hodgson AR, Stock FE (1960) Anterior fusion for the treat-
ment of tuberculosis of the spine: the operative findings
and results of treatment in the first one hundred cases. J
Bone Joint Surg 41A:295-310
Houshian S et al (2000) Bone and joint tuberculosis in Denmark:
increase due to immigration. Acta Orthop Scand 7:312-315
491
Hsu LCS, Cheng CL et al (1984) Tuberculosis of the lower cer-
vical spine (C2 to C7). J Bone Joint Surg 66B:I-5
Hsu LCS, Cheng CL et al (1988) Potts paraplegia of late onset.
J Bone Joint Surg 70B:534-538
Humphries MJ et al (1986) Spinal tuberculosis presenting with
abdominal symptoms - a report of two cases. Tubercle 67:
303-307
Jain AK (2002) Treatment of tuberculosis of the spine with neu-
rologic complications. Clin Orthop Relat Res 398:75-84
Jain AK,Aggarwal A, Mehrotra G (1999a) Correlation of canal
encroachment with neurological deficit in tuberculosis of
the spine. Int Orthop 23:85-86
Jain AK, Kumar S, Tuli SM (1999b) Tuberculosis of the spine
(CI-D4). Spinal Cord 37:362-369
Jain R, Sawhney S, Berry M (1993) Computer tomography of
vertebral tuberculosis: patterns of bone destruction. Clin
RadioI47:196-199
Jellis JE (1995) Bacterial infections: bone and joint tubeculosis.
Bailliere's Clin Rheumatol 9:151-159
Jellis JE (1996) Orthopaedic infection associated with HIV
disease. Surgery 12:175-177
Jellis JE (2002) Human immunodeficiency virus and osteoar-
ticular tuberculosis. Clin Orthop Relat Res 398:27-31
Kanaan IV et al (1999) Craniocervical junction tuberculosis: a
rare but dangerous disease. Surg Neurol 51:21-25
Leibert E et al (1996) Spinal tuberculosis in patients with
human immunodeficiency virus infection: clinical presen-
tation, therapy and outcome. Tuberc Lung Dis 77:329-334
Lifeso RM (1987) Atlantoaxial instability in adults. J Bone Joint
Surg 69B:183-187
Lifeso RM, Weaver P, Harder EH (1985) Tuberculous spondyli-
tis in adults. J Bone Joint Surg 67A:1405-1413
Lin-Greenberg A, Cholankeril J (1990) Vertebral arch destruc-
tion in tuberculosis. J Comput Assist Tomogr 14:300-302
Lmejjati M et al (2000) Tumor-like tuberculosis of the sacrum.
Joint Bone Spine 67:468-470
Luk KD, Krishna M (1996) Spinal stenosis above a healed
tuberculous Kyphosis. A case report. Spine 21:1098-1101
Lukhele M (1996) Tuberculosis of the cervical spine. S Afr
Med J; 86:663-556
Madkour et al (1987) Radiological features of Brucella spondy-
litis. American Roentgen Ray Society, 87th annual meeting,
paper no 1,154,155
Madkour MM (1989) Occupational related infectious arthri-
tis. In: Balint GP, Buchanan WW (eds) Baillihes clinical
rheumatology: occupational rheumatic diseases, vol 3.
Saunders, London, pp 157-192
Madkour M, Sharif H (2001) Bone and joint imaging. In:
Madkour M (ed) Madkour's brucellosis, 2nd edn. Springer,
Berlin Heidelberg New York, pp 90-132
Madkour M et al (1985) Brucellosis in Saudi Arabia. Saudi
Med J 6:324-332
Madkour M et al (1988) Osteoarticular brucellosis: results of
bone Scintigraphy in 140 patients. AJR 150:1101-1105
Mandell GL, Petri WA Jr (1987) Drugs used in the chemother-
apy of tuberculosis, M avium complex disease and leprosy.
In: Hardmen JG, Gilman AG, Limbird LE (eds) Goodman
and Gillman's: The pharmocological basis of therapeutics,
9th edn. McGraw-Hill, New York, pp 1155-1174
Marketos SG, Skiadas PK (1999) Galen: a pioneer of spine
research. Spine 24:2358-2362
Martin NS (1970) Tuberculosis of the spine: J Bone Joint Surg
52B:613-628
492
Martin NS (1971) Paraplegia. A report on 120 cases. J Bone
Joint Surg 53B:596-608
McGuinness F (2000) Tuberculous spondylitis. In: McGuinness
F (ed) Clinical imaging of non-pulmonary tuberculosis.
Springer, Berlin Heidelberg New York, pp 43-80
Medical Research Council (1978) Working part on tuberculo-
sis of the spine: a controlled trial of anterior spinal fusion
and debridement in the surgical management of tubercu-
losis of the spine in patients on standard chemotherapy: a
study in two centers in South Africa. Tubercle 59:79-105
Moon MS, HA KY, Sun DH et al (1996) Pott's paraplegia. Clin
Orthop 323:122-128
Nussbaum ES et al (1985) Spinal tuberculosis: a diagnostic and
management challenge. J Neurosurg 83:243-247
O'Reilly LM, Daborn CJ (1995) The epidemiology of Myco-
bacterium bovis infections in animals and man. Tubercle
Lung Dis 16:1-46
Okuyama Y et al (1996) Tuberculous spondylitis (Pott's
disease) with bilateral pleural effusion. Intern Med 35:
883-885
Parry 0, Bhebhe E, Levy LF (1999) Non-traumatic paraplegia
in a Zimbabwean population - a retrospective survey. Cent
Afr J Med 45:114-119
Pertuiset E et al (1999) Spinal tuberculosis in adults. A study
of 103 cases in a developed country 1980-1994. Medicine
(Baltimore) 78:309-320
Pollard BA, El-Beheiry H (1999) Pott's disease with unstable
cervical spine, retropharyngeal cold abscess and progres-
sive airway obstruction. Can JAnaesth 46:772-775
Rajasekaran S (2001) The natural history of post-tuberculous
Kyphosis in children. Radiological signs which predict late
increase in deformity. J Bone Joint Surg Br 83:954-962
Rajasekaran S, Shanmugasundaram TK (1987) Prediction of
the gibbus deformity in tuberculosis of the spine. J Bone
Joint Surg 69A:503-509
Rajasekaran S et al (1998) Tuberculous lesions of the lumbo-
sacral region. A IS-year follow-up of patients treated by
ambulant chemotherapy. Spine 23:1163-1167
Rajeswari R et al (1997) Late onset paraplegia - a sequelae to
Pott's disease. A report on imaging, prevention and man-
agement. Int J Tuberc Lung Dis 1:468-473
Rasit AJ et al (2001) The pattern of spinal tuberculosis in Sar-
awak General Hospital. Med J Malaysia 56:143-140
Rezai AR et al (1995) Modern management of spinal tubercu-
losis. Neurosurgery 36:87-97
Sanders WE Jr, Horowitz EA (1995) Other mycobacterium spe-
cies. In: Mandell GL, Benett JE,Dolin R (eds) Principles and
practice of infectious disease, 4th edn. Wiley, New York, pp
2264-2272
M. M. Madkour et al.
Satoskar RS,Bhandarkar SD,Ainapure SS (1999) Chemotherapy
of tuberculosis. Pharmacology and pharmacotherapeutics,
16th edn. Popular Prakhashan, Mumbai, India, pp 713-729
Sayi EN, Mlay SM (1995) Tuberculosis of the spine in children
at Muhimbili Medical Center - Dar es Salaam. East Afr
Med J 72:46-48
Schulitz KP et al (1997) Growth changes of solidly fused
Kyphotic bloc after surgery for tuberculosis. Comparison
offour procedures. Spine 22:1150-1155
Sharif HS et al (1989) Brucellar and tuberculous spondylitis:
Comparative-imaging features. Radiology 171:419-425
Sharif HS et al (1990) Granulomatous spinal infection: MR
imaging, Radiology 177:101-107
Sharif HS et al (1995a) Role of MRI and CT imaging in the
management of tuberculous spondylitis. Radiol Clin North
Am 33:787-804
Sharif HS et al (1995b) Role of MRI in the management of
spinal infections. AJR Am J RoentgenoI158:1333-1345
Shembekar A, Babhulkar S (2002) Chemotherapy of osteoar-
ticular tuberculosis. Clin Orthop Relat Res 398:20-26
Singh Bet al (1998) Isolated cervical tuberculosis in patients with
HIV infection. Otolaryngol Head Neck Surg 118:766-770
Srivastava S, Sanghavi NG (2000) Non-traumatic parapare-
sis: Aetiological, clinical and radiological profile. J Assoc
Physic India 48:988-990
Tirri R et al (2000) Pott's disease of the lower cervical spine in
a diabetic patient. Monaldi Arch Chest Dis 55:205-207
Tuli SM (2002) General principles of osteoarticular tuberculo-
sis. Clin Orthop Relat Res 398:11-19
Tuli SM, Srivastava TP, Varma BP, Sinha GP (1967) Tuberculo-
sis of the spine. Acta Orthop Scand 38:445-458
Upadhyay SS, Sell P, Saji MJ et al (1993) Seventeen year pro-
spective study of surgical management of spinal tubercu-
losis in children. Spine 18:1704-1711
Upadhyay SS et al (1996) The effect of age on the change in
deformity after anterior debridement surgery for tubercu-
losis of the spine. Spine 21:2356-2362
Vaidya MK, Shah GV, Bharucha KE (1995) Pregnancy and its
outcome in quadriplegia due to Pott's spine. Int J Gynaecol
Obstet 49:319-321
Van Staden H (1992) The discovery of the body: human dis-
section and its cultural contexts in ancient Greece. Yale J
Bioi Med 65:223-241
Ventura G et al (1996) Pott's Disease of the cervico-occipital
junction in an AIDS patient. Tuberc Lung Dis 77:188-190
Weaver P, Lifeso R (1984) The radiological diagnosis of tuber-
culosis of the adult spine. Skeletal RadioI12:178-186
Weiner BK et al (1998) Mycobacterium avium intracellulare:
vertebral osteomyelitis. JSpinal Disord 11 :89-91
31 Surgical Management of Spinal Tuberculosis
M. WASEF AL SEBAI, M. MONIR MADKOUR, K. R. AL MOUTAERY

CONTENTS 31.1
History of Tuberculosis Spinal Surgery
31.1 History of Tuberculosis Spinal Surgery 493
31.2 Pathogenesis, Clinical and Laboratory
Features 494
Surgical intervention in the treatment of spinal
31.3 Imaging Features 494 tuberculosis (TB) antedated the discovery of the anti-
31.3.1 Plain Radiography 494 biotics. Although some of these innovative surgical
31.3.2 Conventional Tomography 514 procedures were successful, the outcome was poor
31.3.3 Myelography 514 due to the fact that surgery was not supplemented
31.3.4 CT Features 514
31.3.5 MRI Features 514
with the yet undiscovered anti-microbial agents.
31.3.6 Scintigraphy 518 Surgical treatment of TB spondylitis was first
31.4 Treatment ofTB Spine 518 described by Sir Percival Pott in 1779 who performed
31.5 Antibiotics With or Without Surgery autopsies on patients with TB spondylitis, which was
(MRC Trials) 518 later named after him, i.e.,"Pott's disease" (Pott 1936).
31.6 Operative Treatment 519
31.7 Indications for Surgical Intervention 519 Laminectomy was the most popular surgical
31.8 Approaches and Surgical Techniques 522 intervention for the management of TB spondylitis
31.8.1 Needle Biopsy till Seddon in 1934-1935 abandoned it as it caused
(Under Fluoroscopic or CT Guidance) 522 further spinal instability.
31.8.2 Drainage of Abscesses 522
Posterior spinal fusion was first described by two
31.8.3 Debridement of Tuberculous Lesion 522
31.8.4 Spinal Fusion 523
separate authors in 1911 (Albee 1911; Hibbs 1911).
31.8.5 Instrumentation 523 The idea of posterior spinal fusion was abstracted
31.9 Decompressive Techniques and Treatment from the fact that the natural sequence of events
of Paraplegia 524 of TB joint healing was associated with ankylosis.
31.10 Surgical Treatment of Kyphosis 524
These two authors described their procedure using
31.11 Cervical TB 530
31.12 Summary 530 this observation and aimed at speeding recovery.
References 532 However, such a surgical technique did not arrest
the progression of kyphosis or alter the outcome
of paraplegia. Posterior decompression through
costotransversectomy was successful in surgical
management of paraplegia (Menard 1894). How-
ever, most of these patients developed secondary
infections.
Anterior spinal surgical technique was described
by Ito et al. in 1934 as a "New radical operation for
Pott's disease:' This approach certainly provided
a wider field of exposure, which allowed better
M. W. AL SEBAl, FRCS, FACS
Consultant Spinal Surgeon, Riyadh Armed Forces Hospital,
debridement and fusion.
Honorary Assistant Professor, King Saud University, P.O. Box After the discovery and availability of antibiot-
7897, Riyadh 11159, Saudi Arabia ics in 1945, antibiotics and anterior spinal surgical
M. M. MADKOUR, MD, DM, FRCP technique in combination made a major impact on
Consultant, Department of Medicine, Riyadh Armed Forces the prognostic outcome of TB spine. Subsequently,
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia
K. R. AL MOUTAERY, MD, FRCS (Ed), FACS
several trials on treating TB spine using anti-TB anti-
Head of Neurosurgery, Riyadh Armed Forces Hospital, biotics with or without surgery have been reported in
P.O. Box 7897, Riyadh 11159, Saudi Arabia the world literature.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
494
One of these major trials was carried out in
1956 by Hodgson and Stock in Hong Kong. These
authors adopted the anterior spinal approach with
radical excision and strut graft fusion to prevent
both kyphosis as well as late-onset paraplegia. They
were successful in their trial and managed to reduce
post-operation hospitalization period remarkably.
Similar large-scale prospective trials on treatment
of TB spondylitis were sponsored by the British
Medical Research Council (BMRC) working party.
Several reports from BMRC were published between
1973 and 1999 (BMRC 1973a,b, 1974, 1974, 1976,
1978a,b, 1982, 1985, 1986, 1989, 1998, 1999). Some
of these reports will be discussed later in the text of
this chapter.
31.2
Pathogenesis, Clinical
and Laboratory Features
See spondylitis and neuro-TB chapter.
31.3
Imaging Features
I. Plain radiography
II. Conventional tomography
III. Myelography
IV. Computed tomography (CT)
V. Magnetic resonance imaging (MRI)
VI. Scintigraphy
31.3.1
Plain Radiography
Early vertebral body metaphyseal TB osteomyelitis
may initially appear as normal, using plain radi-
ography. Later, as tuberculoma lesions expand and
enlarge, the vertebral body becomes decalcified and
edematous can be depicted by plain radiography,
computed tomography (CT) and magnetic resonance
imaging (MRI). Decalcification of the lower or upper
anterior margin of the vertebral body may appear
as diminution of the intensity of the "White-stripe"
boundary (Fig.31.1). These early changes can be
demonstrated better on lateral view than anterior-
posterior images. As the infection progresses it
M. W. Al Sebai et aI.
spreads to the cortex and the intervertebral disc,
leading to its destruction.
As the vertebral body softened, bulging of the disc
occurs - "Intervertebral herniation of disc". This will
lead to decrease in vertebral-body height and slight
narrowing (in contrast to pyogenic infection) of the
disc space.
Spread of infection via the intermetaphyseal
communicating vessels may lead to infection of
vertebral bodies above and/or below the original
focus of infection (Fig. 31.2). As the anterior part of
the vertebral bodies is primarily affected, associated
with intervertebral disc sequestration and hernia-
tion, kyphosis and wedging of the vertebral bodies
follow (Fig. 31.2). Unequal involvement of vertebral
bodies and muscle spasm secondary to unilateral
psoas abscess may produce scoliosis in few cases.
(Figs. 31.3-31.5). A distant skip lesion high up in the
thoracic or cervical spine may occur in about 4%
of patients (Figs. 31.6, 31.7). Sharif and colleagues
(1993) from our hospital reported skip lesions at
different levels. Paravertebral abscess may be pres-
ent at the time of presentation of the patient. In the
thorax, the abscess masses can be readily seen on
plain radiography of the mediastinum. They appear
as more dense and fusiform in shape in contrast to
the lucent lung tissue (Figs. 31.2, 31.8). Paraspinal TB
abscesses are difficult to visualize if they occur at the
dorso-lumbar junction. In the lumbar region, para-
spinal abscesses can be detected when they obliterate
the retroperitoneal tissue planes or if they contain
calcification (Fig. 31.9). Rarely, sclerotic lesions of
the vertebral bodies, as part of healing, may occur in
patients with active disease process (Fig. 31.10).
Posterior-element infection with bone destruc-
tion may involve one or both of the vertebral
pedicles, the whole of the neural arch or the spinous
process (Fig. 31.11).
Neural arch tuberculous infection associated with
paraspinal abscess was noted in seven of our patients
(Fig. 31.11). Tuberculous infection may occur in
single vertebral body and may lead to vertebral col-
lapse with plain radiography appearances similar to
those of metastatic carcinoma (Figs.31.12, 31.13).
Such presentations were noted in five adult patients
in our series (a feature that is less common in adults
than children). Subluxation secondary to spinal TB
was reported in the cervical spine, particularly the
upper cervical spine; similar findings were seen
in a few patients with thoracolumbar involvement
(Figs. 31.1,31.14). Advanced cases with affection of
long segment of the spine are currently rarely seen
(Fig. 31.15).
 Surgical Management of Spinal Tuberculosis 495

a b

c d
Fig. 31.1. Early tuberculous spondylitis in a 17-year-old male who presented with back pain. a Lateral x-ray shows early lesion
at lower inferior border of L[ (arrow). b There is destruction of the lower part of L[ and the upper part of Lz. L1 body is pos-
teriorly displaced with facet joint subluxation 6 weeks later (arrow). c Computed tomography reveals minimal degrees of bone
destruction, small soft-tissue shadow and bifacetal subluxation (arrows). d Lateral x-ray film taken 14 months after a two-stage
operation shows sound fusion and maintained reduction of subluxation
496 M. W. Al Sebai et al.

Fig. 31.2. This is a 26-year-old female who presented with back pain and incomplete paraplegia. a, b Anteroposterior and lateral
radiographs show classical Pott's disease of the lower half of T9 and the upper part of TIO. Note: the paravertebral soft-tissue
shadow on AP view (arrowheads) and the degree of kyphosis on the lateral view (40°). c Myelography shows a total spinal block-
age of the dye at TIO (arrow). d, e Post contrast computed tomography scan shows osteolytic lesion of T9, 10 and 11. Note: the
cord compression by the epidural abscess (arrowheads) with dye appearance at T9 and TIO levels
 Surgical Management of Spinal Tuberculosis 497

a b

c
Fig. 31.3. This is a 28-year-old male who presented with back pain and progressive paraplegia. a Plain radiography demonstrat-
ing bilateral paravertebral abscess (arrowheads) and scoliosis of 19° because of unequal involvement of vertebrae. b Lateral
view shows severe destruction of TI2, partial affection of lower part of TIl, loss of the disc space in between TIl and TI2
(arrow) and localized kyphosis of 29°, c, d Anteroposterior and lateral radiography after two-stage operation shows correction
of deformity on both planes
 498 M. W. Al Sebai et al.

a c

Fig. 31.4. This is a 26-year-old female who presented with back

pain, loin pain and abdominal swelling. a Anteroposterior
radiography shows minimal destruction of L2 and L3 bodies
and scoliosis mainly because of muscle spasm. Note absence of
psoas shadow on the same side (arrow). b Computed tomogra-
phy shows huge unilateral abscess (arrowhead). c Post-opera-
b tive x-ray shows correction of deformity

a b[>
Fig.31.5. This is a 13-year-old girl who presented with back pain, deformity and progres-
sive paraplegia. She was referred as a case of congenital kyphoscoliosis with neurological
complication. a, b Plain radiographic examination demonstrates kyphoscoliosis with L" looks
like a hemivertebra (black arrowhead). Note that the pedicle and posterior elements are not
seen on one side (white arrowhead) with obliteration of the disc space between T12 and L,.
 Surgical Management of Spinal Tuberculosis 499

c d

e f

g h
c, d, e Magnetic resonance imaging (MRI) study indicates diminished signal in the body of L[ on n (arrow), and increased
signal intensity on T2 (arrowhead). On post-gadolinium n, MRI, there is enhancement of the remaining destroyed body, there
is extensive anterior spinal abscess (arrow) and cord compression by epidural abscess (arrowhead). fAxial scan demonstrates
affection of posterior and anterior columns of L[ (arrowheads). g, h Post-operative x-ray films following two-stage operation
shows correction of deformity and fusion
 500 M. W. Al Sebai et al.

c
Fig. 31.6. This is a 25-year-old lady who presented with abdominal pain and fever. Investigations revealed incomplete intes- l>
tinal obstruction and destructive lesion in the lower thoracic spine causing lower extremities weakness. Lumbar spine was
noticed to be minimally affected. She was operated on by two-stage operation for the lower thoracic lesion 2 weeks after
start of antituberculous treatment. Patient continued to have constitutional symptoms and lymphopenia. Lumbar spine lesion
continued to deteriorate and necessitated radical debridement after extension of the fixation and fusion of the lumbar spine.
a Lateral plain radiograph shows severe destruction of TIO with localized kyphosis (arrowhead) as well as early lesion at
L3 (arrowhead) with intact disc spaces between L2> 3 and L3> 4' b Non-contrast computed tomography (CT) demonstrating
Surgical Management of Spinal Tuberculosis 501

severe destruction of TID, 11 including costovertebral junction and the pedicle on left side (arrowhead). There is a big ante-
rior and posterior abscess. c Non-contrast CT of the lumbar spine shows marginal anterior osteolytic lesion (arrowhead)
with anterior abscess. d Lateral view x-ray, 4 months following two-stage operation of the thoracic spine and correction of
its deformity. Note the complete collapse of L3 with localized kyphosis (arrowhead). e Non-contrast CT demonstrates frag-
mentary lesion of L3 (arrowhead) with bilateral psoas abscess. Note the left-sided lesion at Ls (open arrow). f Post-operative
plain radiograph, 18 months following extension of fixation and fusion to the lumbar spine and radical debridement and
fusion between L2 and L4
502 M. W Al Sebai et al.

d
Fig. 31.7. This is a 60-year-old lady who presented with back pain, girdle pain and paraparesis with signs of upper motor neuron t>
disease. a Lateral radiograph demonstrates collapse of T6 with obliteration of the disc space between T5 and T6 vertebrae
(arrow). b,c T1 and T2 weighted, sagittal scans demonstrate destruction of T6 with intraosseous abscess at T5. Note: the unusual
high intensity signals in the bodies of T5 and T6 on T1 scali (thick arrow). There is also diffused lesion behind the cord with
increased signal intensity on both T1 and T2 imaging indicating fat (lipomatosis) (thin arrow). d, e T1 and T2 weighted, axial
 Surgical Management of Spinal Tuberculosis 503

scans demonstrate the cord is compressed between the anterior destructive lesion (arrow) and the posterior lipomatosis (arrow-
head). f Lateral view of the lumbosacral spine in the same patient, shows destructive lesion at L4, 5 and the sacrum (arrow). g
Sagittal T1 weighted magnetic resonance imaging. Tuberculous spondylitis of L4,5 and the sacrum with diffuse anterior abscess
(arrow). h Lateral plain radiograph of the thoracic spine following radical debridement and fusion between T4 and T6
 504 M. W. Al Sebai et al.

a L- ~ ~ ...
c

b d [>
Surgical Management of Spinal Tuberculosis 505

Fig. 31.8. This is a 21-year-old male admitted by the thoracic surgeon because of tuberculous mediastinal abscess. Evaluation
of the spine preoperatively did not show major spinal affection. While the patient was walking 5 days following chest drain-
age, he fell down paraplegic. a Plain chest radiography shows huge mediastinal mass (arrows). b, c Plain radiographs of the
thoracic spine show haziness of vertebral bodies with reduction of the disc spaces (open arrows) with no gross collapse or
instability. d Computed tomography reveals the huge mediastinal shadow (A) surrounding the spine without destruction.
Note: also bilateral pleural effusion (arrows). e Five days following drainage anterior-posterior radiograph shows translation
of vertebrae at the level of T6, 7 (arrowhead). Myelography shows total blockage at the T8 level. f Post-contrast computed
tomography demonstrates translation of vertebrae with shearing at the pedicles (arrowhead) and cord transection. g. Ante-
rior-posterior radiography, 4 years following surgery shows fusion at translation site
 506 M. W. AI Sebai et al.

a b

Fig. 31.9. This is a 60-year-old male, presented with severe low

back pain, left sciatica and inability to walk. a, b Plain anterior
posterior and lateral radiographs show severe destruction of
L4 , less severe destruction of L3, localized kyphosis and sur-
rounding calcification. Note: the destruction of facet joints at
L3, L4 level (arrow). c, d Tl and T2 weighted studies reveal
loss of the Lr L4 disc, (thin arrow) with extensive destruction
of L4 vertebrae (thick arrow). e, f Post-gadolinium sagittal and
axial scans demonstrate total destruction of L4 body, enhanc-
ing L3 and Ls (arrowhead) and low-intensity signals represents
caseating material (arrow). Note: also on axial cut, the nerve
root is surrounded by granulation tissue (curved arrow) and
severe thecal compression (arrowhead). g Post-operative lat-
eral radiograph shows anterior fusion and posterior fixation.
c Note: the correction of kyphosis [>
 Surgical Management of Spinal Tuberculosis 507

f g
 508 M. W. Al Sebai et al.

a b

Fig. 31.10. a, b Plain radiograph in a 45-year-old lady shows L2-L 3 disc space narrowing, unilateral and anterior osteophytes
(arrowheads) with destruction of L3 upper and plate. Serological tests for Brucellosis were negative. c Post-contrast computed
tomography demonstrates anterior-marginal bone destruction (arrowhead in the upper cut) and multi-loculated abscess with
surrounding enhancement (arrowhead in the lower cut). d Lateral plain radiograph shows anterior debridement and fusion
3 months post-operative
 Surgical Management of Spinal Tuberculosis 509

a c

b ~-""'d

Fig.31.11. a Lateral radiograph of a 40-year-old lady who presented with back pain, shows no obvious abnormality. b Ante-
rior-posterior radiograph reveals destruction of the lower half of the spinous process of L4 (arrows). c Non-contrast computed
tomography shows destruction of the spinous process (arrow) with surrounding paraspinal and subcutaneous abscess (A). d
Lateral radiograph 2 years after debridement shows no further changes
 510 M. W. Al Sebai et a!.

a -""=~::-'_""';~""';'----"_. C

Fig. 31.12. This is a 60-year-old lady who presented with back pain and complete
paraplegia diagnosed as a metastatic disease of the spine a Plain radiographic exami-
nation demonstrates vertebra plana at T5 (arrow). b The conventional myelogram
demonstrates the vertebral-body lesion (arrow) with compression of the spinal canal
contents (open arrow). c Post-contrast computed tomography shows fragmentation
lesion with small soft-tissue shadow (arrow). Note: also cord compression (arrow-
head) d Lateral radiograph after combined operation of posterior decompression,
d fusion and fixation followed by anterior radical debridement and fusion

Fig. 31.13. This is a 30-year-old male who presented with neck pain, back pain and inability to walk because of weakness of the lower
limbs and painful left hip joint. a, b Plain radiograph shows destructive lesion C6, C7 and T1 and kyphotic deformity with huge retro-
pharyngeal abscess (long arrow). There is subluxation C6, C7 on lateral radiograph (thick arrow) and cervicothoracic scoliosis because
of unilateral destruction of C7 as shown on anterior-posterior radiograph (arrow). c, d plain radiographs of the thoracic spine shows
another lesion affecting mainly T8 (vertebra plana) (arrow). e Computed tomographs of the cervicothoracic region revealed affection
of C7 mainly, including the right pedicle and transverse process (arrowhead). fThoracic computed tomographs show destruction of the
anterior and posterior elements ofT8 (arrow) with cord compression by an epidural abscess (arrowhead). g Computed tomographs of
the left hip show soft-tissue shadow (arrow) and decreased bone density (open arrow). h Lateral radiograph of the cervical spine fol-
lowing combined operation shows correction of deformity. i Thoracic spine lateral radiography after combined operation demonstrates
correction of deformity, posterior fixation and anterior fusion
 Surgical Management of Spinal Tuberculosis 511

d e

g
 512 M. W. Al Sebai et al.

d e
Fig.31.14. This is a 45-year-old lady who presented with back pain, inability to walk and incomplete paraplegia. a Lateral
radiograph shows a lesion involving two vertebra. The disc space is obliterated and the upper vertebra is dipped into the lower
one. b Diagram shows the bifacetal dislocation Tll-Tl2. c Non-contrast computed tomography shows TIl body inside lower
part of TI2 in the upper scan and bare facet sign in the lower scan because of dislocation (arrows). d Sagittal reformat of the
computed tomography show the facet dislocation posteriorly (arrow). e Radiography 10 months after combined operation shows
fusion and reduction of dislocation
 Surgical Management of Spinal Tuberculosis 513

d
Fig. 31.15. a Radiography of a 40-year-old lady with chronic back pain and deformity of six years duration. Severe destruction
of the lumbar spine with kyphosis and shortness of the trunk. b Anterior-posterior radiograph demonstrates the disorganiza-
tion of the facet joints and calcification. c The conventional myelogram demonstrates the thecal indentation. d Post-contrast
computed tomography show the destruction of the vertebrae TI2 and all lumbar vertebrae. e Lateral radiography following
debridement and fibular graft fusion between TIl and the sacrum
514
31.3.2
Conventional Tomography
Conventional tomography remains as an important
imaging modality particularly in countries where
TB is common in poor underdeveloped parts of the
world with lack of CT and MRI facilities. It also has
the advantage of the absence of overshadowing of
overlying structures.
However, the loss of densities between various tis-
sues limits its value (Fig. 31.16). It is of considerable
value in the occipito-cervical junction as it shows the
extent of bone destruction and subluxation between
the first and second cervical vertebrae (Fig. 31.17).
Conventional tomography is also useful in depict-
ing TB infection at the cervico-thoracic junction and
the dorsal spine (Figs. 31.16,31.17).
31.3.3
Myelography
Water-soluble contrast myelography is still used in
poor underdeveloped countries with endemic TB.
However, it has been superseded by CT, CT-myelog-
raphy or MRI in rich industrialized countries. Its
value is limited to certain groups of patients with
posterior subligamentous abscesses or extradural
tuberculous granulomatous masses protruding in
the spinal canal that are not depicted by plain radi-
ography or conventional tomography. An extradural
space-occupying lesion change appears as displace-
ment of the contrast column or even total spinal
blockage (Fig. 31.2).
31.3.4
CT Features
Axial CT scanning is ideal for showing the extent of
bone involvement in TB but is less efficient in soft
tissue mass lesions, particularly extradural granu-
lomatous encroaching on the spinal canal. Another
disadvantage of CT is the difficulty in the detection of
skip distant spinal lesions that may occur and are not
included in the site of imaging. Some recent advances
in the development of soft and hardware may help in
soft tissue involvement but are not as cost effective.
Early changes characteristically appear as rarefac-
tion, homogenous focal low-density lesions with
regular boundaries. Clarification appears as the
disease advances with subsequent abscess formation
(Fig. 31.9).
M. W. Al Sebai et al.
Destruction of the cortical bone and intervertebral
disc occur as the disease advances and these can be
clearly depicted by axial and sagittal CT (Fig. 31.2).
Loss of clarity in the fat planes surrounding the ver-
tebra is an early feature that indicates paravertebral
extension of the infection.
In advanced tuberculous vertebral osteomyelitis,
loss of normal bone architecture, extensive frag-
mentation, kyphosis, multiloculated paravertebral
abscesses and subligamentous spread of infection
are characteristic findings (Fig. 31.10). Spinal canal
encroachment by extradural inflammatory tissue
granulomas and abscess formation may enhance
with contrast media if pus is present (Fig. 31.4).
The four distinctive patterns of vertebral body
destruction have been described by Jain et al. in
1993 as fragmentary, osteolytic, subperiosteal and
localized sclerotic. The fragmentary type is the most
common and is described by most authors and noted
in 47% of patients (Fig. 31.18).
CT-guided needle biopsy and aspiration is an
important tool for diagnosing TB of the spine. CT
with intrathecal injection of a small volume of low
concentration metrizamide has been used to evaluate
extradural encroachment of the spinal cord (Bront et
al. 1983; McGraham and Dublin 1985). However, it is
an invasive procedure and carries the risk of further
spread of infection (AI Arabi et al.1992).
31.3.5
MRI Features
MRI imaging modality is the investigation of choice
for spinal infection. It has good tissue differentiation,
bone marrow visualization, as well as excellent depic-
tion of the spinal canal, its content and the paraver-
tebral soft tissue. It does not expose the patient to
ionizing radiation. However, it has a poor definition
of bone and calcified tissue than CT modality.
At the time of presentation, MRI may show verte-
bral end-plate destruction, loss of disc space height
and paraspinal abscess. However, early changes of
tuberculous spondylitis may only appear in a single
vertebral body, which may simulate neoplastic dis-
ease. A Tl-weighted image may show an area of a
homogeneous low signal of intensity at the infected
area (commonly the anterior inferior part of the
vertebral end-plate). The T2-weighted images may
show an increased signal in the infected area of the
vertebral body. Occasionally, posterior areas of the
vertebral body may show similar changes. As the
disease advances, a vertebral abscess may appear as a
 Surgical Management of Spinal Tuberculosis 515

a b

c d

Fig. 31.16. a Conventional tomography in a young man presented with

progressive paraplegia. There is destruction of the upper thoracic ver-
tebrae with kyphosis. b, c Pre and post-gadolinium II scans of II to TS
tuberculous spondylitis. There is obliteration of the disc spaces and big
anterior collection (arrow). Note: the extensive epidural abscess press-
ing the cord. (Small arrow in Tl and arrowhead in enhanced Tl). d T2
weighted imaging show high signal intensity in the bodies, in the disc
spaces and the abscesses. Note: prevertebral abscess (arrow) and epidu-
ral abscess (arrowhead). e Conventional tomography done 6 weeks after
e debridement and fusion shows correction of kyphosis
 516 M. W. Al Sebai et al.

a b

c d

g
e Fig. 31.17. This is an 8-year-old girl who presented with painful torticollis, inability to support
her head and paraparesis. a Lateral radiograph shows invagination of the odontoid process (arrowhead) and retropharyngeal collec-
tion, (arrow), and subluxation Cl>~' b Anterior-posterior radiograph demonstrates upper mediastinal shadow (arrow) and decreased
intervertebral disc spaces at the cervicothoracic region. c Non-contrast computed tomographs of the craniocervicaljunction reveal
destructive lesion of the occipital condyles, CI and C2 (thin arrows). There is distortion of the anatomy of the region with invagination
of the odontoid process (thick arrows). d Non-contrast computed tomographs of the cervicothoracic junction reveal destruction of the
vertebrae with fragmentation and cord compression (arrow). e Conventional tomography was useful in this case to show the severe
kyphosis in the upper thoracic spine and subluxation ofTl-T2. (curved arrow). Note: also invagination of odontoid process (arrow). fA
sagittal Tl and T2 scans show anterior medulla compression by the tip of odontoid process (arrowhead) and prevertebral abscess with
increased signal intensity in both Tl and T2 because of proteinaceous material. There is destruction of the upper thoracic vertebrae
with severe cord compression (arrow). g Post-operative radiography show reduction of craniocervical as well as CI, C2 subluxation.
This was treated by traction and brace. The thoracic spine demonstrates the correction of kyphosis following combined operation
 Surgical Management of Spinal Tuberculosis 517

a b

Fig. 31.18. This is a 39-year-old lady who was

transferred from another hospital 3 months
following laminectomy. She had back pain
and inability to walk because of neuro-
logical deterioration following laminectomy.
a, b Plain radiographs show destructive
lesion affecting T9, 10 and 11 with extensive
destruction of TlO (arrow) and kyphosis.
There is absence of posterior elements of
TlO and II because oflaminectomy (arrow-
heads). c Post-contrast computed tomo-
graphs show extensive destruction of TlO
with fragmentation (small arrow) and cord
compression (arrow) in spite oflaminectomy
(arrowhead). d Post-operative lateral radio-
graph shows correction of deformity and
c fusion following two-stage operation
518
low-intensity signal with rim enhancement by gado-
linium contrast (Fig. 31.9).
Early changes of the disc in Tl-images, do not
show signal changes or may show blurring of the
disc-space margins. In T2-irnages, high-signal
intensity in the disc is uncommon, unlike early disc
involvement in pyogenic infection.
Subligamentous spread, anterior or posterior, is
depicted better on Sagittal Tl-imaging (Fig. 31.5) as
isointense with other structures.
In Tl-images, high intensity signal beneath the
anterior longitudinal ligament may appear with
margins enhancement by gadolinium.
Multiloculated large paravertebral abscesses may
show isointensity or low intensity signals on Tl-
weighted images. T2-weighted images will show a high
signal as well as the size and extent of the abscesses.
31.3.6
Scintigraphy
Please refer to Chap. 30.
31.4
Treatment of T8 Spine
The goals of treatment are to eradicate the infection
and to treat neurological deficit and spinal deformity.
The best method of treatment should produce rapid
fusion, prevent late recurrence after clinical healing
and prevent the development of progressive and
severe deformity.
The traditional treatment of Pott's disease prior
to the discovery of effective chemotherapy was pro-
longed immobilization, utilizing prolonged bed rest
and/or body casts (Dobson 1951). Although many
of the patients did remarkably well, mortality was in
the range of 20% and 20-30% of the patients suffered
recurrence of infection (Dobson 1951).
The clinical availability of streptomycin in 1944,
para-amino salicylic acid in 1950 and isoniazid in
1951 ushered in a new area in the management
of Pott's disease. In 1962, Konstam and Blesovsky
reported the management results of 207 patients
with Pott's disease who were treated with chemo-
therapy and allowed to walk without bracing (Kon-
stam and Konstam 1958). Only 27 patients under-
went operations for abscess drainage and 86%
made complete recoveries. One hundred patients
had to be withdrawn from the study because of
M. W. Al Sebai et al.
poor drug compliance and only 35 patients were
monitored for longer than 2 years. This was the
first study to show that chemotherapy without
long-term immobilization would be used to treat
Pott's disease successfully.
31.5
Antibiotics With or Without Surgery
(MRC Trials>
The multinational prospective study on the efficiency
of conservative chemotherapy and surgical treat-
ment coordinated by the Medical Research Council
(MRC) has just completed a final IS-year follow up
report. They conducted a series of randomized clini-
cal trials of the following treatment modalities: (1)
chemotherapy with immobilization via either strict
bed rest or body cast, (2) outpatient chemotherapy
with mobilization, (3) chemotherapy and debride-
ment of obviously infected bone without fusion, (4)
radical operation of anterior resection and debride-
ment with autologous bone-strut grafting (Medical
Research Council Working Party 1973a, b, 1974a, b,
1976, 1978a, b, 1982, 1985, 1986, 1989). The first two
modalities of non-operative treatment regimens
resulted in a favorable outcome (return to normal
activity without pain or neurological deficit) in 85%
and 86%, respectively, with radiographic evidence
of fusion in 36% and 67%, respectively. Operative
debridement resulted in no better outcome, although
patients who were debrided tended to have an ear-
lier resolution of abscesses. Overall, the patients
treated without radical debridement experienced
an increase in kyphosis of II°,30% had an increase
in kyphosis of II° to 30° and 10% had an increase
of up to 50°. Although it appears from these studies
that conservative treatment is as effective as surgical
intervention for earlier and milder diseases, there are
still reservations on the effect of such treatment for
more severe diseases (Luk 1999). The optimal dura-
tion of anti-tuberculous chemotherapy required for
complete recovery is still debated. The duration cur-
rently recommended by most experts is 12 months
(Moon 1997; Pertuiset 1999). Shorter durations of
6-9 months have been advocated in adults. These
trials failed to resolve this issue because of meth-
odological inadequacies regarding sample size and
statistical analysis (Pertuiset 1999).
Surgical Management of Spinal Tuberculosis 519

31.6 3. Unresponsiveness to medical therapy as mani-

Operative Treatment fested by development or progression of neuro-
logical deficits, spinal deformity, intractable pain
Active surgical intervention of the diseased area in and progression of disease
the care of Pott's disease was begun even before the 4. Non-compliance with medications
introduction of specific anti-tuberculous chemo- 5. Non-diagnostic biopsy
therapy, even though most of the authors considered
chemotherapy as the mainstay treatment for TB. Still, Tuli described a middle-path regimen of selective
lesions could be safely treated without surgery to be operative treatment (Tuli 1975). He considered sur-
defined (Boachie-Adjei and Squillante 1996). gery for cases with neurological deficits that failed to
improve during the initial trial of chemotherapy. All
other patients received chemotherapy with surgery
reserved for posterior lesions, persistent active infec-
31.7 tion, instability, doubtful diagnosis or recurrence
Indications for Surgical Intervention of neurological deficit. Lifeso, from Saudi Arabia,
modified surgical indications (Lifeso et a1.1985). The
Hodgson and Stock proposed that the operation, in authors indicated the need for immediate anterior
combination with chemotherapy, be done as early as decompression and fusion for complete paralysis,
possible after the diagnosis for the following reasons profound neurological deficits due to cervical or
(Hodgson and Stock 1956): upper thoracic lesion and gross destruction of the
1. Diseased material may be obtained for a definitive cervical spine or any severe kyphosis associated
diagnosis with active disease. They recommended that patients
2. The patient's general condition improves dra- who have slight or no neurological deficit and slight
matically immediately after the evacuation of the kyphosis can be safely treated with medical therapy
abscess alone, and close observation should then be the rule.
3. In the thoracic spine, especially in children, early In our own series we separated the cases into three
decompression of the abscess is necessary, as the groups: (1) Early cases with no or minimal bony affec-
abscess tends to extend up and down the spine tion are treated conservatively. Drugs to be continued
rapidly for 12 months, with four drugs in the first 2 months
4. Bone sequestra, sequestrated intervertebral discs, and two drugs for the remaining 10 months. (2) Cases
caseous material and avascular bone may be with bony destruction and involvement of disc spaces
removed and placement of an anterior strut graft are treated surgically by anterior radical debridement
under compression will lead to early fusion and fusion, except in cases with posterior arch affec-
5. Late recurrence is uncommon after radical exci- tion, which needs posterior surgery. (3) Cases with
sions of the diseased focus and solid fusion have affection of three or more vertebrae, affection of poste-
been achieved rior and anterior columns and progressive deformity
6. Increasing deformity may be prevented while the patient is on medical treatment and bed rest.
7. Paraplegia may be prevented These three categories are considered as radiological
8. In the presence of established paraplegia, anterior signs of instability of the spine. For these cases, both
decompression and bone grafting lead to rapid anterior radical debridement and fusion and posterior
recovery fixation and fusion are needed (Figs. 31.19,31.20) (Al
9. The anterior approach gives accurate diagnosis Sebai et al. 2001). In our patients this has resulted in
and exposure for dealing with penetration of an improvement in the correction of the deformity as
organs, especially of the lung well as encouraging neurological recovery and allow-
ing early mobilization.
Rezai et al. considered the criteria for surgical
management, in combination with chemotherapy, of
PoU's disease as (Rezai et al. 1995):
1. Neurological deficit including acute neurological
deterioration
2. Spinal instability with more than 50% vertebral
body collapse or destruction and spinal deformity
of more than 5°
 520 M. W. Al Sebai et al.

a
c

Fig.31.19. a Lateral radiography of 3-year-old girl who presented with

chronic back pain and incomplete paraplegia, shows destruction of the
vertebrae no to L2• There is kyphosis of 86°. b Post-operative lateral
radiograph following radical debridement and anterior fusion between
no and L3. It shows correction of kyphosis to 55°. c Post-operative lateral
radiograph following second-stage of posterior fusion and instrumenta-
tion reveals further correction of kyphosis to 26°. (With permission from
b International Orthopaedics 25,2001)

l>
Fig. 31.20. This is a 56-year-old lady who complained of back pain and chest wall pain. Chest radiograph and lateral view of
the thoracic spine revealed no gross abnormality. She had increasing pain 4 months later with inability to walk because of
incomplete paraplegia. a, b Lateral plain radiographs at this stage showed compression of T7 and lytic lesions affecting L[ and
L2 with preservation of disc spaces (arrows). c Bone scan revealed multiple foci of increased uptake interpreted as most likely
metastatic disease of the spine. d Post-intravenous contrast computed tomographs show destructive lesion of T7 extending to the
pedicle and costovertebral junction with severe cord compression, which is pushed to the right side (arrowheads). Note absence
of soft-tissue shadow around the vertebrae (arrow). e Computed tomographs of the lumbar spine show destruction of L} and L2
with fragmentation (arrowhead) and surrounding enhancing soft-tissue shadow (curved arrow). f Post-gadolinium T1 images
show destruction of T7 and L1 vertebrae. (arrows) Note enhancing epidural lesion compressing the cord at T7 and to a lesser
degree at L}. g, h Lateral radiographs of the thoracic and lumbar spine, 3 weeks following start of anti-tuberculous treatment
and bed rest. There is progressive collapse of the vertebrae and obliteration of the disc space between L1 and L2 • (arrows). i, j
Lateral and anterior-posterior radiographs following combined operation showing fusion and fixation
Surgical Management of Spinal Tuberculosis 521

e
522
31.8
Approaches and Surgical Techniques
31.8.1
Needle Biopsy
(Under Fluoroscopic or CT Guidance)
This technique is usually used for the thoracic and
lumbar spine disease. The approach depends on the
anatomic region and part of the vertebra involved.
Spinal needle is used for aspiration to get cytologi-
cal diagnosis while Tru-cut biopsy could be used for
histopathological diagnosis. Definite cytological
diagnosis was obtained in up to 88.5% of cases (Kang
et al. 1999). Biopsy of the cervical spine poses spe-
cial problems because of the proximity to many vital
organs. With CT, the exact relationship of the skeletal
lesion to the adjacent structures can be established,
allowing a safe route to be selected (Kattapuram and
Rosenthal 1987). Gupta et al. reported the use of ultra-
sound guidance for needle biopsy of lytic lesions of
the cervical spine in four cases without complications
(Gupta et al. 1993).
M. W. Al Sebai et al.
31.8.2
Drainage of Abscesses
Aspiration or surgical drainage was carried out for
patients with a large cold abscess because it was
thought that evacuation of the abscess improves the
patient's general condition and rapid progression of
the abscess along the spine was prevented. This has
been shown to be ineffective and surgical drainage
of a cold abscess alone is no longer recommended
(Moon et al. 1987, 1996). However, in certain cases
with huge psoas abscesses, with or without minimal
spinal affection, presenting with abdominal mass
causing discomfort to the patient, the patient may
get quick relief after drainage (Fig. 31.21).
31.8.3
Debridement of Tuberculous Lesion
To eradicate the tuberculous lesion, anterior debride-
ment was done first by Ito et al. in 1934. Kondo and
Yamada wrote the end results of focal debridement
Fig.31.21. Non-contrast computed tomo-
graphs of a 26-year-old male who presented
with back pain, abdominal pain and systemic
manifestations. There are huge bilateral multi-
loculated abscesses (A) extending from the
upper abdomen down to the pelvis
Surgical Management of Spinal Tuberculosis 523

in spinal TB and its indications in 1937 (Kondo and and fusion can, however, be associated with fall-off
Yamada 1937). Hodgson and Stock popularized the in the post-operative kyphos, frequently because
anterior spinal surgery and described the anterior of slipping of the graft, graft fracture, protrusion,
radical surgery (Hong Kong operation) (Hodgson absorption or non-union and, in part, because of the
and Stock 1956). continued growth in the posterior elements in the
Focal debridement can effectively remove the dis- growing child (Bailey et al. 1972). In severe cases, the
eased tissue and can evacuate the abscess; however, it combined approach of anterior and posterior spinal
does not prevent the progression of kyphosis, espe- fusion gives the best results.
cially if compared with the radical operation in both Comparison of four procedures of fusion revealed
adults and children (Upadhyay et al. 1993, 1994). that the combined fusion and anterior debridement
Simple debridement gives no long-term advantage guaranteed an equal growth of the anterior and
over ambulant chemotherapy alone and, therefore, posterior heights in children (Schulitz et al. 1997).
is no longer accepted as a preferred method of treat- Long-term follow-up for children who had anterior
ment (Medical Research Council Working Party and posterior fusion without instrumentation for
1982,1985). extensive disease and kyphosis, revealed solid fusion
In the Hong Kong operation, the spine is and improvement of the kyphotic deformity (Altman
approached anteriorly so that the affected area may et al.I996). However, the use of posterior instrumen-
be dealt with more directly. The sequestrated bone tation in the face of active disease helps in providing
and caseous material must be debrided back to bleed- early fusion, prevention and correction of kyphosis
ing bone above and below and back to the posterior as well as earlyambulation (Boachie-Adjei and Squil-
longitudinal ligament. The decompression should lante 1996; Moon et al.1995, 1996).
go back to the dura in cases of neurological deficit
when spinal decompression is necessary. The angular
deformity is corrected by insertion of a strut graft. A 31.8.5
slightly oversized graft is put under compression by Instrumentation
springing open the kyphosis during insertion. The
choice of graft material is based on considerationsThe role of posterior instrumentation and fusion in
of graft incorporation and structural support. The the treatment of spinal TB has only recently been
reported (Korkusuz et al.1997; Moon et al.1995, 1996;
strut graft should be strong yet osteogenic in nature,
Rezai et al.1995). In infectious diseases, instrumenta-
tri-cortical or bi-cortical iliac crest grafts are ideal,
tion introduces a foreign body that acts as a focus, the
but frequently the area to be grafted is too large for
the iliac grafts to be sufficient. Longer struts can be
infection is notoriously resistant to antibiotic therapy
obtained from the fibula or ribs. These should be and usually requires removal of the instrument (Gris-
supplemented with iliac bone because the fibula is tina and Costerton 1985; Gristina et al. 1985). How-
strong but mostly cortical bone (non-osteogenic). ever, both clinical and microbiological results suggest
Bradford and Daher described the use of vascular- that posterior instrumentation is not associated with
persistence or recurrence of spinal tuberculous infec-
ized rib grafts for stabilization of kyphosis (Bradford
and Daher 1986). Good results were also obtained bytion and is useful to provide immediate stability and
the use of allograft in 47 children by Govender andprotect against the development of a kyphotic defor-
Parbhoo (1999). mity (Oga et al. 1993). In adults with deformity and
paraplegia, Moon et al. found the combined two-stage
operation to be most successful (Moon et al. 1995,
31.8.4 1996). In their cases, anterior surgery was preceded
Spinal Fusion by posterior instrumental stabilization surgery.
Additional posterior spinal fixation after anterior
Posterior spinal fusion was used to hasten recovery decompression and fusion was reported by others
since 1911 (Albee 1911; Hibbs 1911). This procedure (Abramovitz et al. 1986; Graziano and Sidhu 1993;
did not prevent progressive kyphosis or development Jeanneret and Magerl 1994). Despite the general
of paraplegia. Debridement followed by anterior acceptance of the Hong Kong operation for the
fusion offers the advantage of debridement and a treatment of spinal TB, Guven and co-workers have
result in the diminution of the kyphotic deformity used posterior instrumentation and fusion without
(Hodgson and Stock 1956, 1960; Medical Research anterior debridement in ten neurologically intact
Council Working Party 1974). Anterior debridement patients. They achieved clinical and radiological
524
evidence of fusion in all cases (Guven et al. 1994).
Chemotherapy was instituted 2 weeks pre-opera-
tively and continued for a mean period of 11 months.
Lee et al. also used transpedicular instrumentation
as an adjunct in the treatment of thoracolumbar and
lumber spine with early stage bone destruction (Lee
et al. 1999). Combined posterior decompression and
internal fixation was reported by Rath et al., avoiding
the risks of anterior approaches for the elderly and
debilitated patient (Rath et al. 1996). Jeanneret and
Magerl used percutaneous debridement and external
spinal fixation as an alternative procedure to conser-
vative or more invasive operative treatment modali-
ties in the following conditions: (a) painful lesion of
the spine with minimal bone loss, not amenable to
efficient orthotic stabilization, (b) when emergency
decompression of the spine is mandatory and ante-
rior decompression is not possible emergently, (c)
osteomyelitis of the spine at L51S1, and (d) in the
presence of infective wounds, making internal pos-
terior stabilization unsuitable (Jeanneret and Magerl
1994). They had good results, as far as healing of
infection, neurological outcome and kyphotic defor-
mitywith shorter period of bed rest. A small number
of recent reports deal with anterior spinal cord
decompression, block bone grafting and anterior
internal fixation in cases of vertebral osteomyelitis
(Kostuik 1983; Redfern et al.1988; Yilmaz et al.1999).
Yilmaz et al. reported 22 patients who had TB of the
spine with moderate to severe localized kyphosis and
16 patients who had more than two involved levels;
all had stabilization with anterior instrumentation
(Yilmaz et a1.1999). They obtained an average correc-
tion of kyphosis of 64% and there was no recurreJ;lce
of the disease. However, anterior instrumentation
should not be used to correct kyphotic deformity
when the posterior column is affected.
31.9
Decompressive Techniques
and Treatment of Paraplegia
The first case of tuberculous spondylitis treated suc-
cessfully by laminectomy was in 1882. In the early
part of the 20th century it has become a common
procedure for patients with Pott's paraplegia. Seddon
condemned the procedure because it removes the
integrity of the posterior arch and may lead to insta-
bility and further neurological damage (Figs. 31.18,
31.22) (Seddon 1934/1935). Laminectomy is consid-
ered contraindicated in the usual form of TB (Fel-
M. W. Al Sebai et al.
lander 1975; Hodgson et al. 1964; Kemp et al. 1974).
The only indication for laminectomy in the treatment
of spinal TB is atypical disease involving the neural
arch (Fellander 1975; Kemp et al. 1974; Rahman et al.
1987; Rand and Smith 1989).
Decompression in the thoracic spine may be per-
formed through a transthoracic approach, through a
costotransversectomy or by an extrapleural approach.
Transthoracic approach is more successful than costo-
transversectomy (Kirkaldy-Willis and Thomas 1965).
The extrapleural approach has the theoretic benefit
of avoiding the tuberculous empyema. However, no
studies have demonstrated any actual advantage of an
extrapleural approach over a standard thoracotomy.
Decompression in the lumbar spine may be done
by retroperitoneal approach. A left-sided approach is
preferred because the arterial structures are easier to
deal with than the venous counterpart. In approach-
ing the lumbosacral junction, the common and exter-
nal iliac vessels are mobilized.
Cervical spine infection has a high incidence of
cord compression, more than 40%. Hsu and Leong
reported excellent results from using the Hong Kong
procedure via anterior approach in conjunction with
medical treatment (Hsu and Leong 1984).
31.10
Surgical Treatment of Kyphosis
Kyphosis is one of the two major complications of
spinal TB. It has been common in patients treated
with chemotherapy alone (Moon et al. 1995). Almost
3% of cases of TB of the spine develop a severe
kyphotic deformity. The patients at risk are those
who develop the disease under the age of 10 years,
and who had involvement of three or more vertebral
bodies (Tuli 1995). A severe kyphosis is more than
a cosmetic disfigurement because nearly all such
patients develop cardiopulmonary dysfunction,
painful impingement between the ribs and pelvis
and compression of the spinal cord with paraplegia
at an average of 10 years after the onset of the dis-
ease (Smith et al. 1996; Tuli 1995). Severe kyphosis is
not only difficult to treat, but also dangerous, with a
high complication rate. There are two clinical types of
kyphosis, mobile and rigid (Moon et al.1995). Mobile
kyphosis could be treated by traction, posterior
fusion, anterior radical surgery or two-stage opera-
tion (Moon 1997). Skull traction was found effective
in correction of non-rigid kyphosis in cervical spine
(Fig.31.23) (Moon et al. 1987, 1996). A two-stage
 Surgical Management of Spinal Tuberculosis 525

Fig.31.22. This is a 50-year-old lady, referred from another hospital because of

persistent back pain and inability to walk 6 months following laminectomy and dis-
traction rod fixation for lumbar tuberculous spondylitis. a Lateral radiograph show
extensive destruction of L4 , lumbar kyphosis and dislodged rods. b, c Functional
flexion and extension radiographs reveal gross instability. d Non-contrast computed
tomographs demonstrate destruction of the body of L4 (arrow). They show evidence
of previous laminectomy (arrowheads) and dislodgement of the rods (small arrow-
heads). e, f Plain radiographs, 2 years following staged operation show fusion and
correction of deformity f
 526 M. W. Al Sebai et al.

a b

Fig.31.23. a Lateral radiograph of a young man that shows

destructive lesion of C6-7 with obliteration of the disc in
between and kyphotic deformity. Note subluxation of the facet
joints C6-7 (long arrow) and large prevertebral soft-tissue
shadow (short arrows). b Lateral radiography following skull
traction demonstrates reduction of subluxation (arrows) and
correction of deformity. c Post-operative plain radiography
c shows fusion and correction of deformity

operation could be done either with the posterior
stage done first or as the second stage. Anteri6lr
radical debridement followed by posterior resection
and instrumentation was recommended by Yau et
al. (1974) (Fig. 31.19). Combined two-stage opera-
tion, where posterior instrumentation is done first,
achieves good correction of kyphosis, provided the
deformity is not fixed and severe has been reported
by others (Fig. 31.24) (Moon 1991; Moon et al. 1995,
1996). The rigid deformity could be corrected by pos-
terior closing wedge osteotomy, two-stage operation
or multi-stage operation (Fig.31.25) (Guven et al.
1994; Wu et al. 1996). Because of technical demands
and higher risk of neurological injury, correction of
a severe kyphosis without neurological deficit should
not be done for cosmetic reasons alone. In a paralytic
case, partial correction can be attempted together
with decompressive surgery (Fig. 31.26).
 Surgical Management of Spinal Tuberculosis 527

a b

Fig.31.24. This is a 25-year-old lady who

has had kyphosis since early childhood
following a chronic illness. She presented
with recent onset back pain and weakness
of lower limbs. a Lateral plain radiography
shows kyphosis of 75° between T8 and L4
and destruction of the vertebrae T9 to L3•
Note the lordosis of thoracic spine above
the kyphos (arrow). b Anterior-posterior
view demonstrates the crowding of ribs,
loss of disc spaces between no and L4 and
increased density of the bone. c Non-con-
trast computed tomography demonstrates
areas of destruction (arrowheads) because
of reactivation of tuberculous spondylitis.
d Lateral radiography following combined
operation shows correction of kyphosis to
30° and fusion with fibular graft between
c T8 and L4
 528 M. W. Al Sebai et al.

Fig. 31.25. This young lady presented with back pain and deformity. She had
kyphosis and cauda equina syndrome since childhood. a Lateral radiogra-
phy shows kyphosis of 72° between TIO and L2 • There are fused, deformed
Til, 12 and L[ vertebrae at the apex (arrow). b Anterior-posterior plain
radiograph demonstrates previous laminectomy between TI2 and L2 and
old myodil with evidence of arachnoiditis (arrow). c Lateral radiography
2 years following staged operation shows fusion and correction of defor-
c mity to 22°
 Surgical Management of Spinal Tuberculosis 529

a c

Fig. 31.26. This is a 35-year-old male who started to have back pain and progressive
paraplegia. He had kyphosis since early childhood. a, b Lateral and anterior-poste-
rior plain radiographs show kyphosis 92° between T8 and Lz and fusion of T9 to L1
with no soft-tissue shadow (arrow). c The conventional myelogram demonstrates
stretching of the cord over the kyphos without blockage of the dye (arrow). d Post-
contrast computed tomographs at the apex of the kyphos revealed indentation of
the thecal sac (arrowheads). Note absence of signs of reactivation or surrounding
soft-tissue shadow. e, £lateral and anterior-posterior plain radiographs, 3 years fol-
lowing two-stage operation. Note: fusion both anteriorly and posterolaterally as well
as partial correction of kyphosis f
 530 M. W. Al Sebai et al.

31.11 bifacetal dislocation for the first time in a IS-year-old

Cervical T8
TB of the cervical spine is uncommon. Its incidence,
according to the few reports in the literature, varies
from 3% to 5% (Dobson 1951; Hsu and Leong 1984;
Martin 1970). Because of its low incidence, it has not
been included in the MRC trials (Medical Research
Council Working Party 1973a, b, 1974, 1976, 1978a, b,
1982,1985,1986,1989). Lifeso recognized three stages
of CI-2 tubercular infection and recommended sur-
gical treatment for all of them (Lifeso 1987). We
treated four cases with upper cervical spine TB; two
of them had fractured odontoid process following
car accidents. We assume that the infection precipi-
tated the fracture (Fig. 31.27). The other two cases
had subluxation with destruction of the vertebrae.
Three of these four cases needed posterior fixation
and fusion. The fourth case was treated by trac-
tion and chemotherapy, while surgery was directed
to the cervico-thoracic lesion in the same patient,
which was responsible for the neurological deficit
(Fig.31.17). Spinal instability following TB of the
spine has been described. Three patients with vary-
ing degrees of subluxation following cervical TB were
reported from Malaysia (Arumagasamy et al. 1977).
Although subluxation of cervical spine TB has been
rarely reported, bifacetal dislocation has not been
described before 1991. We reported cervical spine
girl with cervical tuberculous spondylitis (Fig. 31.28).
(AI Arabi and Al Sebai 1991).
TB of the cervico-thoracic spine junction can be
challenging both in visualization on plain radiogra-
phy as well as in surgical approach. In our own series,
we noted diagnostic delay resulting in neurological
deficit in 16 of our patients. The surgical approach of
C7 TB was better performed from the lower neck in
our series (Fig. 31.13). In T1 spinal TB, our approach
was periscapular one (Fig. 31.16). However, Horner
Syndrome developed post-operatively in one patient
who had periscapular approach, which gradually
recovered over the subsequent 6 months.
31.12
Summary
In recent years, since the discovery of antibiotics,
treatment of spinal TB has been revolutionized.
Several trials aimed at finding the best methods of
treatment have been discussed. We also highlighted
the use of antibiotics alone or in conjunction with
surgical intervention in the management. Continu-
ous debate among authors on the indications of sur-
gery has been discussed. In our own experiences, we
have noted the occurrence of unusual sites of spinal

Fig.31.27. a Lateral radiograph of the cervical spine in a 62-year-old man following a car accident shows fracture odontoid
process with posterior displacement. Note increased prevertebral shadow (arrow). b Following traction, lateral radiograph dem-
onstrates reduction of displacement and the retropharyngeal soft-tissue shadow (arrow). c Non-contrast computer tomographs
demonstrate partial destruction of the occipital condyles and anterior arch of C1(thin arrow). Note fragments of bone that have
been extruded into an anterior abscess (thick arrow). d, e T1 magnetic resonance imagine (MRI) indicates diminished signal
of odontoid process and Cl while T2 MRI shows increased signal intensity with retropharyngeal abscess (arrowheads). f The
post-gadolinium scan outlines the retropharyngeal, tuberculous abscess (arrow). Note degenerative lesion at C5-6. g Lateral
radiography following aspiration of prevertebral abscess and posterior fixation of Cl-2
 Surgical Management of Spinal Tuberculosis 531

f
 532
Fig. 31.28. a Lateral radiography of the cervical spine in a 15-year-old girl shows bifacetal dislocation C3-4 (arrow). There is
extensive anterior, retropharyngeal tuberculous abscess displacing the airway forward (small arrows). b Post-operative lateral
radiograph taken 4 months following two-stage operation demonstrates reduction of dislocation, fusion and fixation
involvement among our patients, such as posterior
element, occipito-cervical and cervico-thoracic junc-
tions. We also highlighted the diagnostic and surgical
difficulties that we encountered during the manage-
ment of these unusual cases. Surgical treatment of
paraplegia and kyphosis were also discussed.
References
Abramovitz IN, Batson RA, Yablon JS (1986) Vertebral osteo-
myelitis: the surgical management of neurologic complica-
tions. Spine 11:418-420
AI Arabi KM, Al Sebai MW (1991) Bifacetal dislocation fol-
lowing tuberculosis of the cervical spine. Tubercle 72:
294-298
AI Arabi KM, AI Sebai MW, AI Chakaki M (1992) Evaluation
of radiological investigations in spinal tuberculosis. Int
Orthop 16:165-167
Albee FH (1911) Transplantation of a portion of the tibia into
the spine for Pott's disease: a preliminary report. JAMA 57:
885-886
AI Sebai MW et al (2001) Operative treatment of progressive
deformity in spinal tuberculosis. Int Orthop 25:322-325
Altman GT, Altman DT, Frankovitch KF (1996) Anterior and
posterior fusion for children with tuberculosis of the spine.
Clin Orthop Relat Res 325:225-231
Arumagasamy N, Pyn CC, Keng KK (1977) Tuberculous
abscesses of the cervical spine with quadriparesis. Surg
Neurol 8:35-40
Bailey HL et al (1972) Tuberculosis of the spine in children. J
Bone Joint Surg 54A:1633-1657
Boachie-Adjei 0, Squillante RG (1996) Tuberculosis of the
spine. Orthop Clin North Am 27:95-103
M. W. Al Sebai et al.
Bradford DS, Daher XH (1986) Vascularized rib grafts for sta-
bilization of Kyphosis. J Bone Joint Surg 68B:357-361
Bront ZM, Burke VD, Jeffrey RB (1983) CT in the evaluation of
spine infection. Spine 8:358-364
Dobson J (1951) Tuberculosis of the spine. Analysis of the
results of conservative treatment and of the factors influ-
encing the prognosis. J Bone Joint Surg 33B:517-531
Fellander M (1975) Paraplegia in spondylitis: results of opera-
tive treatment. Paraplegia 13:75-88
Govender S, Parbhoo AH (1999) Support of the anterior
column with allografts in tuberculosis of the spine. J Bone
Joint Surg 81:106-109
Graziano GP, Sidhu KS (1993) Salvage reconstruction in acute
and late sequelae from pyogenic thoracolumbar infection.
J Spinal Discord 6:199-207
Gristina AG, Costerton JW (1985) Bacterial adherence and
the glycocalyx and their role in musculoskeletal infection.
Orthop Clin North Am 15:517-535
Gristina AG et al (1985) Bacterial adherence and the patho-
genesis of osteomyelitis. Science 228:99-103
Gupta RK et al (1993) Ultrasound-guided needle biopsy oflytic
lesions of the cervical spine. J Clin Ultrasound 21:194-197
Guven 0 et al (1994a) A single stage posterior approach and
rigid fixation for preventing Kyphosis in the treatment of
spinal tuberculosis. Spine 19:1039-1043
Guven 0, Yalcin S, Karahan M (1994b) Eggshell procedure in
correction of neglected cases of Pott's Kyphosis. Proceed-
ings of the 5th biannual conference of European Spinal
Deformities Society. Birmingham, England. Springer,
Berlin Heidelberg New York, pp 84-85
Hibbs RA (1911) An operation for progressive spinal deformi-
ties. NY State Med J 93:1013-1016
Hodgson AR, Stock FE (1956) Anterior spinal fusion: a pre-
liminary communication on the radical treatment of Pott's
disease and Pott's paraplegia. Br J Surg 44:266-275
Hodgson AR, Stock FE (1960) Anterior spinal fusion for the treat-
ment of tuberculosis of the spine. J Bone Joint Surg 42A:295
Surgical Management of Spinal Tuberculosis
Hodgson AR et al (1964) A clinical study of 100 consecutive
cases of Pott's paraplegia. Clin Orthop 36: 128-150
Hsu LCS, Leong JCY (1984) Tuberculosis of the lower cervical
spine (C2 t C7). J Bone Joint Surg 66B:I-5
Ito H, Tsuchiya J, Asami G (1934) A new radical operation for
Pott's disease: report of then cases. J Bone Joint Surg 16:
499-515
Jain R, Sawhney S, Berry M (1993) Computed tomography of
vertebral tuberculosis: patterns of bone destruction. Clin
RadioI47:196-199
Jeanneret B, Magerl F (1994) Treatment of osteomyelitis of the
spine using percutaneous suction/irrigation and percuta-
neous external spinal fixation. J Spinal Disord 7:185-205
Kang Met al (1999) CT guided fine needle aspiration biopsy
of spinal lesions. Acta RadioI40:474-478
Kattapuram SV, Rosenthal DI (1987) Percutaneous biopsy of
the cervical spine using CT guideance. AJR 149:539-541
Kemp HBS, Jackson JW, Shaw NC (1974) Laminectomy in
paraplegia due to infective spondylitis. Br J Surg 61:66-72
Kirkaldy-Willis WH, Thomas TG (1965) Anterior approaches
in the diagnosis and treatment of infections of the verte-
bral bodies. J Bone Joint Surg 47A:87-110
Kondo E, Yamada K (1937) End results of focal debridement
in bone and joint tuberculosis and its indications. J Bone
Joint Surg 39A:27
Konstam PG, Blesovsky A (1962) The ambulant treatment of
spinal tuberculosis. Br J Surg 50:26-38
Konstam PG, Konstam ST (1958) Spinal tuberculosis in south-
ern Nigeria: with special reference to ambulant treatment
of thoraco-lumbar disease. J Bone Joint Surg 40B:26-32
Korkusuz F, Islam C, Korkusuz Z (1997) Prevention of
postoperative late Kyphosis in Pott's disease by anterior
decompression and intervertebral grafting. World J Surg
21:524-528
Kostuik JP (1983) Anterior spinal cord decompression of
lesions of the thoracic and lumbar spine, techniques, new
methods of internal fixation. Spine 8:512-531
Lee TC et al (1999) Transpedicular instrumentation as an
adjunct in the treatment of thoracolumbar and lumbar
spine tuberculosis with early stage bone destruction. J
Neurosurg 91 [SuppI2]:163-169
Lifeso R (1987) Atlanto-axial tuberculosis in adults. J Bone
Joint Surg 69(B):183-187
Lifeso RM, Weaver P, Harder EH (1985) Tuberculosis spondy-
litis in adults. J Bone Joint Surg 67A:1405-1413
Luk KD (1999) Tuberculosis of the spine in the new millen-
nium. Eur Spine J 8:338-345
Martin NS (1970) Tuberculosis of the spine. A study of the
results of treatment during the last twenty-five years. J
Bone Joint Surg 52B:613-628
McGraham JP, Dublin AB (1985) Evaluation of spinal infection
by plain radiographs, computed tomography, intrathecal
metrizamide and CT guided biopsy. Diagn Imag Clin Med
54:11-20
Medical Research Council Working Party on Tuberculosis of
the Spine, First Report (1973a) A controlled trial of ambu-
lant out-patient treatment and in-patient rest in bed in the
management of tuberculosis of the spine in young Korean
patients on standard chemotherapy. A study in Masan,
Korea. J Bone Joint Surg 55B:678-697
Medical Research Council Working Party on Tuberculosis
of the Spine, Second Report (1 973b) A controlled trial of
plaster of Paris jackets in the management of ambulant
533
outpatient treatment of tuberculosis of the spine in chil-
dren on standard chemotherapy: a study in Pusan, Korea.
Tubercle 54:261-282
Medical Research Council Working Party on Tuberculosis
of the Spine, Third Report (1974a) A controlled trial of
debridement and ambulatory treatment in the manage-
ment of tuberculosis of the spine in patients on standard
chemotherapy: a study in Bulawayo, Rhodesia. J Trop Med
Hyg 77:72-92
Medical Research Council Working Party on Tuberculosis of the
Spine, Fourth Report (1974b) A controlled trial of anterior
spinal fusion and debridement in the surgical management
of tuberculosis of the spine in patients on standard chemo-
therapy: a study in Hong Kong. Br J Surg 61:853-866
Medical Research Council Working Party on Tuberculosis of
the Spine, Fifth Report (1976) A five-year assessment of
controlled trials in in-patient and out-patient treatment
and plaster of Paris jackets for tuberculosis of the spine in
children on standard chemotherapy. Studies in Masan and
Pusan, Korea. J Bone Joint Surg 58B:399-411
Medical Research Council Working Party on Tuberculosis of
the Spine, Sixth Report (1978a) Five year assessments of
controlled trials of ambulatory treatment, debridement
and anterior spinal fusion in the management of tubercu-
losis of the spine: studies in Vulawayo (Rhodesia) and in
Hong Kong. J Bone Surg 60B:163-177
Medical Research Council Working Party on Tuberculosis of
the Spine, Seventh Report (1978b) A controlled trial of
anterior spinal fusion and debridement in the surgical
management of tuberculosis of the spine in patients on
standard chemotherapy: a study in two centers in South
Africa. Tubercle 59:79-105
Medical Research Council Working Party on Tuberculosis of
the Spine, Eighth Report (1982) A 10 year assessment of
a controlled trial comparing debridement and anterior
spinal fusion in the management of tuberculosis of the
spine in patients on standard chemotherapy in Hong Kong.
J Bone Joint Surg 64B:393-398
Medical Research Council Working Party on Tuberculosis of
the Spine, Ninth Report (1985) A 10 year assessment of
controlled trials of inpatient and outpatient treatment and
of plaster of Paris jackets for tuberculosis of the spine in
children on standard chemotherapy: studies in Masan and
Pusan, Korea. J Bone Joint Surg 6B:I03-110
Medical Research Council Working Party on Tuberculosis
of the Spine, Tenth Report (1986) A controlled trial of
six month and nine month regimens of chemotherapy in
patients undergoing radical surgery for tuberculosis of the
spine in Hong Kong. Tubercle 67:243-259
Medical Research Council Working Party on Tuberculosis of
the Spine, Eleventh Report (1989) A comparison of 6 or 9
month course regime of chemotherapy in patients receiv-
ing ambulatory treatment or undergoing radical surgery for
tuberculosis of the spine. Indian J Tuberc [Suppl] 36:1-21
Medical Research Council Working Party on Tuberculosis of
the Spine. Thirteenth report (1998) A IS-year assessment
of controlled trials of the management of tuberculosis of
the spine in Korea and Hong Kong. J Bone Joint Surg 80B:
456-462
Medical Research Council Working Party on Tuberculosis of
the Spine. Fourteenth Report (1999) Five-year assessment
of controlled trials of short-course chemotherapy regi-
mens of 6, 9 or 18 months duration for spinal tuberculosis
534
in patients ambulatory from the start or undergoing radi-
cal surgery. Int Orthop 23:73-81
Menard V (1894) Causes de paraplegie dans Ie mal de Pott.
Son traitment Chirurgical Par l'ouverture directe de goyer
tuberculeuse des vertebras. Rev Orthop 5:47-54
Moon MS (1991) Combined posterior instrumentation and
anterior interbody fusion for active tuberculosis Kyphosis
of the thoraxolumbar spine. Curr Orthop 5: 177-179
Moon MS (1997) Tuberculosis of the spine. Controversies and
a new challenge. Spine 22:1791-1797
Moon MS et al (1987) Conservative treatment of tuberculosis
of the thoracic and lumbar spine in adults and children. Int
Orthop 11:315-322
Moon MS et al (1995) Posterior instrumentation and anterior
interbody fusion for tuberculosis Kyphosis of dorsal and
lumbar spines. Spine 20:1910-1916
Moon MS et al (1996) Pott's paraplegia - 67 cases. Clin Orthop
Relat Res 323:122-128
Oga M et al (1993) Evaluation of the risk of instrumentation as
a foreign body in spinal tuberculosis. Clinical and biologic
study. Spine 18:1890-1894
Pertuiset E (1999) Medical therapy of bone and joint tubercu-
losis in 1998. Rev Rhumat 66:152-157
Pott P (1936) Remarks on that kind of palsy of the lower limb,
which is frequently found to accompany a curvature of the
Spine, and is supposed to be caused by it. Williams and
Wilkins, Baltimore MD, pp 271-323 (Medical classics, vol I)
Rahman N, Al Arabi KM, Khan FA (1987) Atypical forms of
spinal tuberculosis. Acta Neuro Chir 88:26-33
Rand C, Smith MA (1989) Anterior spinal tuberculosis:
paraplegia following laminectomy. Ann R Coli Surg Engl
71:105-109
Rath SA et al (1996) Neurosurgical management of thoracic
and lumbar vertebral osteomyelitis and discitis in adults:
a review of 43 consecutive surgically treated patients. Neu-
rosurgery 38:926-933
M. W. Al Sebai et al.
Redfern RM et al (1988) Stabilization of the infected spine. J
Neurol Neurosurg Psychiatry 51:803-897
Rezai AR et al (1995) Modern management of spinal tubercu-
losis. Neurosurgery 36:87-97
Schulitz KP et al (1997) Growth changes of solidly fused
Kyphotic block after surgery for tuberculosis. Comparison
of four procedures. Spine 22:1150-1155
Seddon HJ (1934/1935) Pott's Paraplegia: prognosis and treat-
ment. Br J Surg 22:769-799
Sharif HS et al (1993) Role of CT and MR imaging in the man-
agement of tuberculous spondylitis. Radiol Clin North Am
33:787-804
Smith IE et al (1996) Kyphosis secondary to tuberculosis
osteomyelitis as a cause of ventilatory failure. Chest 110:
1105-1110
Tuli SM (1975) Results of treatment of spinal tuberculosis by
"Middle-path" Regime. J Bone Joint Surg 57B:13-23
Tuli SM (1995) Severe Kyphotic deformity in tuberculosis of
the spine. Int Orthop 19:327-331
Upadhyay SS et al (1993) 17-year prospective study of surgi-
cal management of spinal tuberculosis in children. Hong
Kong operation compared with debridement surgery for
short and long term outcome of deformity. Spine 18:
1704-1711
Upadhyay SS et al (1994) Surgical management of spinal
tuberculosis in adults. Hong Kong operation compared
with debridement surgery for short and long term outcome
of deformity. Clin Orthop Relat Res 302:173-182
Wu SS et al (1996) Management of rigid post-traumatic
Kyphosis. Spine 21:2260-2267
Yau AMC et al (1974) Tuberculous Kyphosis: correction with
spinal osteotomy, halopelvic distraction, and anterior and
posterior fusion. J Bone Joint Surg 56A:1419-1434
Yilmaz C et al (1999) Anterior instrumentation for the
treatment of spinal tuberculosis. J Bone Joint Surg 81A:
1261-1267
32 Tuberculosis of the Central Nervous System
M. ZUHEIR AL-KAWI

CONTENTS by Hippocrates (460-377 B.C.). Aristotle (384-322 B.C.)

observed that persons associated with a person affected
32.1 History 535 by"phthisis" may contract the disease, thereby alluding
32.2 Epidemiology 536
32.3 Microbiology 536 to the possibility of its contagious nature.
32.4 Pathogenesis 536 Avicenna (Ibn Sina, 980-1037) described the
32.5 Pathology 537 cerebral involvement by hot or cold swelling (inflam-
32.6 Clinical 538 mation). The latter is thought to represent chronic
32.6.1 Tuberculoma 538
meningitis (most likely tuberculous) that frequently
32.6.2 Tuberculous Spondylitis 538
32.6.3 Tuberculous Meningitis 539 led to death as he noted.
32.7 Complications 539 With the industrial revolution in Europe, a mas-
32.7.1 Hydrocephalus 539 sive population shift toward inner cities resulted
32.7.2 Vascular Complications 540 in crowding and created conditions that favored
32.7.3 Hyponatremia 540
the spread of infection. Consequently, TB became a
32.7.4 Cranial Neuropathies 540
32.7.5 Seizures 541 major cause of death between the late 17th and early
32.7.6 Tuberculous Encephalopathy 541 20th centuries. Schoenlein (1793-1864) coined the
32.7.7 Adhesive Arachnoiditis 541 term tuberculosis to highlight the gross pathological
32.7.8 Myelopathy 541 appearance of small lumps caused by granulomata
32.8 Diagnosis 541
and the word "phthisis" fell into disfavor. Laennec
32.9 Management 543
32.10 Prognosis 543
(1781-1826), who was credited with the invention of
References 544 the stethoscope, described in detail the auscultative
findings in pulmonary TB.
Following on the research done by Villemin
(1827-1892) to show that TB was transmissible, Koch
32.1 (1843-1910) described the necessary postulates for the
History proof of a contagious nature of any illness. The condi-
tions he described were widely accepted as classical
Tuberculosis (TB) has affected humans since antiq- teachings and were as follows: finding the pathogen
uity. References to what is believed to be tuberculous in every lesion in the body, being able to culture the
infection were made in some writings from ancient pathogen outside the patient's body and reproducing
Egypt and Babylon. Typical spine deformity of Pott's the disease by inoculation into animals. By applying the
disease appear in numerous drawings of hunchbacks aforementioned postulates to TB he was able to estab-
on the walls of ancient Egyptian tombs. By morbid lish the infectious nature of the Bacillus tuberculosis.
anatomy, spinal TB can be traced back to about The glory of his discovery, however, was tarnished by
2000 B.C., as was shown by psoas abscess found in the tuberculin blunder. He touted a prepared glycerin
a mummy. extract of the bacillus as a secret cure for TB. In fact, the
Microscopically, Zimmerman demonstrated the extract injected in large quantities to patients with TB
tuberculous bacillus in a mummy of a child (Cave 1939; did cause many deaths. By the turn of the last century,
Zimmerman 1979). Symptoms of TB were recognized Osler noted that in 1911, "in a population of one mil-
lion, seventeen hundred persons died from TB" (Osler
1921). At the Pasteur Institute, Calmette and Guerin
M. z. AL-KAWI, MD, FACP
Senior Consultant Neurologist & Deputy Chairman, Depart- (1921) produced a live vaccine prepared by successive
ment of Neurosciences, King Faisal Specialist Hospital and subculturing of a strain of mycobacterium bovis. This
Research Center, P.O. Box 3354, Riyadh 11211, Saudi Arabia was what came to be known later as BeG vaccine.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
536
Until the middle of the last century, treatment
remained generally supportive-aimed at boosting
the innate defenses of the body to overcome the
infection. It was not until 1943that specific treatment
against the causative organism was started. Chemo-
therapy began in earnest with the advent of strepto-
mycin followed by PAS in 1946 and then isoniazid
(INH) in 1951; the latter heralded the era of modern
effective anti-tuberculous chemotherapy.
32.2
Epidemiology
It is estimated that one billion people are infected
with M. tuberculosis worldwide. Active TB claims
nearly 8 million new victims each year (Barnes
and Barrows 1993) Among them, 15% will develop
extrapulmonary infection and 6% of that is menin-
geal (Kochi 1991). Therefore, we can estimate that
there are approximately 70,000 new victims of TB
meningitis per annum. Most initial infections with
M. tuberculosis remain clinically silent. The lifetime
risk of developing a clinical case of TB after an ini-
tial silent infection has been estimated at 10%. TB of
the central nervous system (CNS) is the most seri-
ous form of extrapulmonary TB, and about 10% of
immuno-competent patients who develop clinical TB
manifest CNS involvement (Udani et al. 1971). TB of
the CNS has a higher incidence in children with no
particular gender preponderance.
Prior to the acquired immune deficiency syn-
drome (AIDS) epidemic, good steps were made in the
prevention and treatment of TB and its CNS compli-
cations. In a review of a large number of autopsies
done in Germany from 1955 to 1969, there was a
statistically significant decrease in frequency of TB
meningitis to about 1/7 when compared with the data
from 1924 to 1938. This difference was attributed to
the BCG vaccination and to the effective anti-tuber-
culous medications (Weigel 1976).
In spite of the advances made in the fight against
infectious diseases, the overall prevalence of TB is
again on the increase worldwide. The AIDS epidemic
created circumstances favoring the spread of TB both
in the developing and developed countries alike.
Extrapulmonary TB is considered an AIDS-defin-
ing condition. The likelihood of contracting clinical
TB is much higher in the AIDS population (Selwyn
et al' 1992). The proportion of AIDS patients who
develop TB is estimated at 5-9% (Sanchez-Portocar-
rero et al. 1999). The importance of this observation
M. Z. Al-Kawi
is highlighted by the fact that TB in a human immu-
nodeficiency virus (HIV)-infected patient may pres-
ent as an overwhelming systemic disease (Gachot et
al. 1990). CNS TB may be the initial presentation of
AIDS and has a high mortality among such affected
individuals. Tuberculous meningitis is the most fre-
quent meningeal infection in the HIV patients living
in areas with high prevalence of TB (Berenguer et
al. 1992). Infection with HIV may increase the risk
of developing TB meningitis five-fold from 2% of
patients to 10% (Tagliati et al.I994).
Another growing at-risk population is the medi-
cally compromised patients with systemic diseases
or organ failure who are kept alive with modern
management techniques.
32.3
Microbiology
The predominant organism in human infections
is M. tuberculosis. The organism M. tuberculosis is
considered a gram-positive bacterium, though it is
difficult to stain by the standard method. Its growth
is very slow in culture, which clinically correlates with
the slow infection it causes in vivo. Similarly, a long
treatment course with anti-tuberculous medications
is necessary to eradicate infection and to insure non-
resurgence. The genome of M. tuberculosis has been
sequenced and it includes approximately 4000 genes
in a circular chromosome consisting of over 4 million
base pairs (Cole et al. 1998).
32.4
Pathogenesis
The microorganism reaches the CNS in the course of
dissemination following a primary infection. Entry
into the host is most frequently airborne. The fung
tissue is infected first along with a regional lymph
node. Several factors determine the clinical picture
and outcome of CNS TB. These include age, nutri-
tional status, load and virulence of infecting organ-
ism, immune deficiency and prior immunization
with BCG. The density of bacilli in a unit volume of
inspired air is dependent on severity of disease in the
infecting person and the effectiveness of air circula-
tion in the environment. The longer the exposure to
contaminated air, the more likely the infection is to
take place. Acid-fast bacilli do not produce toxins,
therefore, they cause no initial tissue reaction or
Tuberculosis of the Central Nervous System
inflammation if the host has not been sensitized
to tuberculoproteins (non-immune host). Primary
hematogenous dissemination usually occurs in
childhood and is clinically silent. Proliferation of
M. tuberculosis favors foci with high blood flow and
oxygenation. Vertebral column and lung apices are
common sites. Positive tuberculin test signals estab-
lishment of cell-mediated immunity that commonly
appears within 2 months. Granulomas may lead to
healing-albeit incomplete-of the infected foci or lead
to tissue destruction. Persons with impaired cell-
mediated immunity due to HIV infection or other
diseases are more likely to become infected with M.
tuberculosis after exposure than persons with normal
immunity. Intact cell-mediated immunity decreases
the chances of dissemination. Lymphocytopenia of
(CD4+) T-lymphocytes was reported in TB patients
who were seronegative for HIV (Kony et al. 2000)
Cell mediated immunity plays an important role
in both the host defense against TB and the tissue
destruction that characteristically occurs with the
infection. Tuberculoproteins activate T-lympho-
cytes, which produce cytokines that, in turn, activate
macrophages. Macrophages not only perform a bac-
tericidal role, but precipitate tissue destruction as
well (Crowle et al. 1983). The production of the free
radical nitric oxide is central in this process as it is
a vasodilator, inflammatory mediator and cytotoxic
(Anggard 1994).
Tumor necrosis factor alpha (TNF a) appears to
playa significant role in the host defense against TB
and in the production of systemic symptoms such as
weight loss and fever (Takashima et al. 1990). Recent
experience with infliximab, a TNF-a-neutralizing
agent developed for the treatment of rheumatoid
arthritis and Crohn's disease, caused dissemination
of latent tuberculous infection. This highlighted the
role ofTNF-a in the host defense against tuberculous
infection (Keane et al. 2001).
The meninges are infected by organisms carried in
the bloodstream. Primary hematogenous dissemina-
tion occurs early after primary infection, while rup-
ture of a subpial or subependymal granuloma (Rich
focus) may occur any time later in life.
Recently the serum and cerebrospinal fluid (CSF)
levels of vascular endothelial growth factor were
found to be higher in tuberculous meningitis and
they decreased in parallel with the clinical improve-
ment (Matsuyama et al. 2001).
The complexity of factors and varied method-
ologies are responsible, in part, for the variability
of features of CNS infection reported from different
centers. In a series from Riyadh; meningitis consti-
537
tuted 28% of cases, mass lesions 36% and neurologi-
cal sequelae of spinal TB 36% (Bahemuka et al.1988),
while in a pediatric population from London; men-
ingitis constituted 60%, tuberculomas in 13% and
mixed meningitis/tuberculoma in 26%. Vaccination
with BCG was shown to protect against tuberculous
meningitis (Awasthi and Moin 1999) and to modify
the outcome whereby death and severe sequelae are
avoided (Farinha et al. 2000).
32.5
Pathology
Tuberculomas are granulomatous masses of variable
sizes. They develop as the body's immune system
attempts to contain a focus of M. tuberculosis. Macro-
phages induced by T-lymphocytes engulf the bacilli
and form giant cells. Caseous material may appear at
the center and typically contains a few bacilli. Gliosis
and lymphocytic infiltration surround such foci. A
variable amount of edema can be demonstrated by
imaging studies. Tuberculomas are most frequently
seen intraparenchymal but flat granulomas may grow
on the meninges and have been called tuberculomas
en-plaque (Fig. 32.1a).
Tuberculous abscess of the brain is uncommon.
It represents failure of the immune mechanism
described above and consequently contains a high
count of bacilli. It is suspected on imaging studies
when an intensely enhancing thin wall surrounds
a content that displays imaging characteristics of
liquefaction.
In tuberculous meningitis, the infection develops
at a relatively rapid pace inside the subarachnoid
spaces as opposed to the slower rate of progression
in other organs. The basilar inflammation produces
thick gelatinous exudates, which entrap the cranial
nerves as they pass through the cisterns. The third,
fourth, sixth and eighth cranial nerves are involved
most frequently. Similarly, inflammation may induce
tuberculous endarteritis. Penetrating branches of
the anterior, middle, and posterior cerebral arteries
are most commonly involved. Those include ante-
rior choroidal, medial and lateral lenticulo-striate,
thalamogeniculate and thalamoperforating arteries.
Major arteries can also be thrombosed, causing mas-
sive infarctions. Granulomas surrounding blood ves-
sels at the base of the brain may give a characteristic
appearance on contrast enhanced computed tomog-
raphy (CT) or magnetic resonance imaging (MRI)
that is grape-cluster like.
538
Pathological features were noted to be different
in the HIV patients with tuberculous meningitis
since they show reduced and atypical inflammatory
response and extensive vasculopathy. The radiological
correlate of this finding was the relatively less menin-
geal enhancement and the absence of communicating
hydrocephalus on CT scan (Katrak et al. 2000).
32.6
Clinical
The CNS may be directly compromised by inflamma-
tion, mass lesion or infarction or it may be affected by
compression from collapsing supportive structures
(i.e., spine). In the first instance, the patient may
develop meningitis, parenchymal mass lesion (tuber-
culoma or abscess) or a combination thereof. In the
latter instance, spinal cord compression may occur as
a consequence of vertebral destructive infection and
epidural cold abscess (Pott's disease of the spine).
In order of severity, the combination of menin-
geal-parenchymal type ranks the highest, followed
by meningitis. The majority of patients present in a
sick state within a few weeks of infection. Cerebral
or medullary tuberculomas and spinal TB are rather
chronic smoldering infections with significant late
complications.
32.6.1
Tuberculoma
Tuberculomas are slow-growing inflammatory mass
lesions which may be associated with variable peri-
focal edema. Lesions are mostly intraparenchymal.
Any part of the CNS may be involved but it is more
common in the cerebral hemispheres, a distribution
that is probably proportional to the parenchymal
volume.
Tuberculoma may rarely be encountered in the
spinal cord (intramedullary). Spinal cord tubercu-
loma may be the only significant manifestation of
TB. Spinal or radicular pain and progressive weak-
ness are the usual presenting symptoms. Correct
diagnosis is possible in the presence of TB elsewhere.
Intramedullary neoplasm is usually suspected before
the correct diagnosis is reached (Kocen and Parsons
1970). Imaging studies show a fusiform swelling of
the cord along with a central area of iso- or slight
hyperintensity on T1WI, surrounded by edema that
is typically hyper-intense on T2WI.
M. Z. AI-Kawi
Intrasellar tuberculomas are rare and constituted
less than 2% of all intrasellar masses in some series.
They have female predominance and swelling of the
pituitary stalk is a useful sign (Sharma et al. 2000).
32.6.2
Tuberculous Spondylitis
Vertebrae are the most commonly involved part of
the skeletal system. Approximately half of tubercu-
lous infections of the skeletal system are localized
to the spine. During the initial hematogenous dis-
semination of M. tuberculosis, small foci of infection
settle in the skeletal system to be reactivated at a later
stage. The infection may as well spread from an adja-
cent infected lymph node. Any part of the vertebral
column may be affected, but over half of the patients
develop involvement of the thoracic spine, especially
the lower part (Alothman et al. 2001). Back pain usu-
ally precedes neurological manifestations. Tubercu-
lous spondylitis frequently presents as progressive
paraplegia. Early disease recognition may be ham-
pered by late presentation in the endemic area or by
unfamiliarity with the early symptoms in developed
parts of the world. This discrepancy is responsible
in part for the variability in the reported frequency
of neurological complications in tuberculous spon-
dylitis. By the time the patient presents for medical
help a few vertebrae have already been damaged.
The vertebral body adjacent to the disk space forms
the initial lesion. Infection of the intervertebral disk
soon follows. Typically, two or more adjacent verte-
brae are destroyed with formation of granulomas
or cold abscess in the surrounding paraspinal soft
tissues. The dura is a strong barrier to the spread of
infection into the CNS or development of meningitis.
The cord is compressed by the epidural abscess. Pain
is an early symptom caused by local inflammation.
Root irritation gives a radicular distribution of pain
around the trunk. Spinal pain may be dull and local-
ized over the involved vertebrae. It is protracted with
fluctuations in severity. Movement induces painful
spasms, and to prevent them patients may avoid
movements and become stiff. Neurological mani-
festations range from flaccid paraparesis in cases of
cauda equina compression to spastic quadriplegia
in cervical lesions. Pyramidal tract signs usually
precede sensory and sphincter involvement. Rarely
a sinus tract may drain to the skin.
The paraspinal abscess may be noted clinically
in the cervical or retropharyngeal region or can be
demonstrated on spinal radiograph. Systemic symp-
Tuberculosis of the Central Nervous System
toms are similar to tuberculous infection in other
parts of the body. These include low-grade fever,
anorexia, weight loss, anemia, lassitude and night
sweats. The long duration of symptoms, the develop-
ment of typical gibbus, thoracic spine involvement
and elevated erythrocyte sedimentation rate are sug-
gestive of Pott's disease as opposed to spinal epidural
infection or neoplastic conditions (Alothman et al.
2001; Buranapanitkit et al. 2001).
The diagnosis is suggested by imaging, including
plain radiographs of the spine, CT and MRI. The
diagnosis is confirmed by obtaining a tissue sample
by fine needle aspiration.
The treatment is essentially medical. Surgical inter-
vention can be avoided in many cases (Alothman et al.
2001). We noted good results with proper antitubercu-
lous treatment even in the face of neurological involve-
ment. In the presence of significant spondylitis and
compression, spinal decompression may be required.
Removal of pus or granulation tissue through a lami-
nectomy is sufficient when spondylitis is mild (Curling
et al.1990).
Epidural tuberculous infection is not unique to
spinal TB. Disseminated infection may, on rare occa-
sions, affect cranial epidural space and result in cere-
bral symptoms.
Illustrative case: a 24-year-old lady presented with
focal seizures on the left side with secondary gen-
eralization. She had a past history of fever and night
sweats. Chest films were suggestive of miliary appear-
ance. CT of the head showed an enhancing lesion in
the right frontal epidural space (Fig. 32.1a). The
material recovered from the lesion after trephine
procedure was caseous with a lot of acid-fast bacilli
consistent with tuberculous abscess. The patient
was treated with triple anti-tuberculous medica-
tions. She completed a 6-month course of therapy
and remained seizure free after anti-epileptic drugs
were discontinued.
32.6.3
Tuberculous Meningitis
Fever, headache, lethargy and altered mentation are
the most frequent presenting symptoms. Anorexia,
night sweats and stiff neck may occur earlier. It is
not uncommon for symptoms to be present for an
average of 6 weeks before the person seeks medi-
cal attention. Approximately 50% of patients have
meningeal signs early, while focal signs are seen in
up to 20% (Berenguer et al. 1992). Past history of
539
pulmonary TB is not an absolute necessity since less
than 50% of adults with tuberculous meningitis give
such a history.
The spread of infection to the subarachnoid
spaces results in diffuse meningitis that may rapidly
become complicated by vasculitis, cranial nerve pal-
sies, encephalitis or myelitis. Cranial neuropathies
may appear at any time during the course of infection
but are usually late. Optic, oculomotor and acoustic
are most common.
Some types of movement disorder may appear
during tuberculous meningitis. Tremor is the most
common, while chorea is mainly seen in young chil-
dren. It is believed that deep infarcts are responsible
for such manifestations (Alarcon et al. 2000).
32.7
Complications
32.7.1
Hydrocephalus
Hydrocephalus is a common complication of tuber-
culous meningitis. It may occur early or late in the
course of illness. Worsening headache and deteriorat-
ing level of consciousness are the main symptoms.
Adhesive ependymitis may sequester a part of the
ventricular system (usually temporal) and cause a
localized hydrocephalus with mass effect that can
lead to herniation. Imaging studies are necessary for
any sudden change in the condition of the patient
and hydrocephalus can be easily recognized. Diver-
sion of CSF flow is necessary and usually done by
ventriculoperitoneal shunting.
Illustrative case: a 37-year-old man presented with
headache, vomiting and low-grade fever for 2 weeks.
His examination showed lethargy and nuchal rigid-
ity. Brain CT showed enhancement at the basal cis-
terns but no mass lesions or hydrocephalus. Lumbar
puncture yielded a slightly xanthochromic fluid with
protein of 1200 mg/l, sugar 1.8 mmolll and 329 WBC/
mm 3• Few acid-fast bacilli were seen on smear. He
improved after treatment with rifampin, isoniazid,
pyridoxine and pyrazinamide. He was discharged
home after 2 weeks and was compliant with treat-
ment. He returned to the emergency room 4 weeks
after discharge, complaining of severe headache and
vomiting. He was afebrile but pupils showed right
pupil measuring 5 mm and left pupil 3 mm. Fundo-
scopic exam showed papilledema. On CT examina-
 540
a b
Fig. 32.1a, b. a Tuberculoma en-plaque in the frontal region. T1 WI with enhancement, sagittal section b CT showing right tem-
poral horn hydrocephalus. Sequestration of the temporal horn occured in the course of TB meningitis
tion there was a large hypodense mass lesion in the
right temporal lobe with density compatible with CSF
with transependymal edema in the temporal white
matter (Fig. 32.1b).
A ventriculoperitoneal shunt was placed in the
right temporal horn and fluid under high pressure
was released. The patient improved following sur-
gery and completed 9 months of anti-tuberculous
treatment.
32.7.2
Vascular Complications
Arteries that supply the brain emerge in the basal
cisterns. The same area that bears the brunt of
inflammation in tuberculous meningitis. The walls
of vessels immersed in the thick gelatinous exudate
are affected by inflammation and arteritis may lead
to vascular occlusion or thrombosis. Less commonly,
but more seriously, intraparenchymal hemorrhage
may occur.
32.7.3
Hyponatremia
Hyponatremia is common in tuberculous meningitis.
It is attributed to inappropriate anti-diuretic hor-
mone secretion (Singh et al. 1994). Seizure threshold
M. Z. Al-Kawi
may be lowered by hyponatremia and occurrence
of seizures may be enhanced by low serum sodium.
Extremely low levels. especially when the concentra-
tion decreases rapidly, may lead to encephalopathy.
Attention to electrolyte balance is imperative during
treatment.
32.7.4
Cranial Neuropathies
Optic: Obstruction of CSF pathways occurring in
tuberculous meningitis may result in increased
intracranial pressure and papilledema. Ultimately,
damage to the optic nerve occurs by ischemia. Direct
involvement of the optic nerve or the chiasm may
be due to arachnoiditis and chronic granulomatous
inflammation (Bruetsch 1948).
Oculomotor Palsies: similarly, oculomotor nerve may
be involved by microvascular thrombosis or by raised
intracranial pressure. Ocular movements and pupil-
lary reaction may be impaired. Ocular signs may
develop independently of the level of consciousness.
Deafness: Sensorineural hearing loss is a common
complication of meningitis in general and tuber-
culous meningitis in particular. The use of ototoxic
drugs may contribute to this morbidity. It is unlikely
to find evidence of labyrinthitis ossificans on CT of
Tuberculosis of the Central Nervous System
temporal bones in people suffering of hearing loss as
a sequel of tuberculous meningitis.
32.7.5
Seizures
Seizures are common presenting symptoms in cere-
bral tuberculomas. They occur later in tuberculous
meningitis and may be associated with a worse prog-
nosis. Rarely it may be the presenting manifestation
of an extraparenchymal lesion such as an epidural
TB abscess (Fig. 32.1a). They need to be treated by
a suitable anti-epileptic drug, keeping in mind the
interaction with anti-tuberculous chemotherapy and
the side effects on liver functions.
32.7.6
Tuberculous Encephalopathy
This complication was considered an allergic type-IV
hypersensitivity occurring in the CNS as a reaction
to tuberculous infection. It presents with an enceph-
alopathic picture of drowsiness rapidly progressing
to coma and death within weeks. There is mild CSF
inflammatory changes and diffuse brain edema.
Pathologically there is demyelination and perivas-
cular inflammation (Dastur and Udani 1966). It is of
interest that pulmonary TB without direct CNS infec-
tion was reported to be associated with neuromyeli-
tis optica syndrome. This was believed to represent
another remote immune-mediated complication as
well (Silber et al. 1990).
32.7.7
Adhesive Arachnoiditis
The thick inflammatory exudate at the base of the
brain and around the cord may cause adhesions that
compromise blood circulation to adjacent structures,
resulting in loculation or block of CSF flow or caus-
ing entrapment of nerve roots. Symptoms of cranial
nerve dysfunction, myelopathy or radiculopathy may
occur.
32.7.8
Myelopathy
Ischemia due to arteritis in the cord vasculature is
probably the major cause of myelopathy. The result-
541
ing disability is therefore frequently permanent. The
development of intramedullary granuloma is less
common and may have a better outcome with treat-
ment. Extramedullary mass effect in spinal tubercu-
lous meningitis (Fig 3.1) or a loculated CSF pocket
may cause pressure leading to myelopathy.
32.8
Diagnosis
As with many other diagnoses, a proper degree of
clinical suspicion is necessary. This can be easily
achieved in areas of the world with high prevalence
of TE. In the developed countries, however, a similar
high index of suspicion should be adopted when the
patient belongs to a population group susceptible to
HIV or already seropositive for HIV. Other indica-
tors for tuberculous infection include history of TB
in a family member or contact with a person known
to have open pulmonary TE. A positive Mantoux
intradermal skin tuberculin test may be helpful but
has inadequate sensitivity and specificity. It may be
positive in a high proportion of adult population
in endemic regions. However, severe infection may
be associated with skin anergy and negative test in
the presence of disseminated tuberculous infection.
The diagnosis is usually based on four sources of
evidence: the clinical manifestations, cerebrospinal
fluid examination in combination with CT scan
of the head, and evidence of TB elsewhere in the
patient's systems. Evidence for TB elsewhere is help-
ful but its absence does not rule out the diagnosis.
It should be looked for on chest film. The chance of
finding a lesion on a routine chest radiograph has
been reported with wide variability in the literature.
Hilar lymphadenopathy, primary complex or miliary
TB may be strongly suggestive but the latter may be
missed if not particularly sought. Chest CT is more
sensitive and should be performed in suspected cases.
Considering the pathogenesis of CNS TB, the same
hematogenous spread of infection should deliver the
organism in the vascular choroid layer of the eye. The
presence of choroid tubercle, however, is infrequently
reported. This may reflect the difficulty in examining
the peripheral regions of the fundus in a sick patient.
When present, a choroidal tubercle is characteristic
and appears as a small slightly raised rounded pale
yellowish lesion with indistinct margins.
Imaging of tuberculous meningitis usually shows
significant enhancement in the basal cisterns. CT
or MRI scan usually reveals basilar enhancement
542
assumed to correlate with the intensity of the inflam-
mation and the thickness of the exudates (Fig. 2.1).
The parenchyma shows evidence of deep cerebral
infarctions commonly in the distribution of the
thalamoperforating and lenticulostriate arteries (Fig.
2.3). Stenosis or occlusion of the intracranial portion
of the internal carotid artery and proximal segments
of its branches, the anterior cerebral and middle
arteries, may be identified on cerebral angiogram
(Fig. 2.6). It is not unusual to find ischemic areas
in addition to parenchymal inflammatory lesions
(tuberculomas).
Examination of CSF obtained by lumbar puncture
is the definitive test for diagnosis of tuberculous men-
ingitis. When the risk of herniation during lumbar
puncture is a concern in patients with increased
intracranial pressure, puncture may be performed
with a small gauge needle to retrieve a small amount
of CSF for essential diagnostic tests after the patient is
prepared by the administration of intravenous man-
nitol 20% (0.25-0.5 g/kg) over 30 min. The findings
of increased opening pressure, lymphocytic pleocy-
tosis (usually in the range of 10-500 cells/mm\ high
protein content and low glucose levels are typical but
not universal. Early in the course of infection, there
might be polymorphonuclear predominance.
CSF glucose is usually low in tuberculous men-
ingitis. Mean glucose values range from 22-38 mg
per 100 ml, but 12-15% of patients may have normal
glucose. Glucose should always be examined simul-
taneously in the blood and CSF. Failure to check the
blood level may be a source of error when diabetes
masks hypoglycorrhachia or iatrogenic hyperglyce-
mia occurs in patients receiving dextrose intrave-
nously. Glucose in the CSF falls below 40% of that in
the serum. The low glucose may not be only because
of consumption by organisms or white blood cells
as had been considered in the past, but the inflam-
matory reaction may cause a disturbance in the CSF
glucose transport system.
Protein levels are extremely variable and may be
influenced by the intensity of inflammatory reaction,
but more importantly by the presence of block due to
adhesive arachnoiditis. When protein levels exceed
1 gm/l00 ml, xanthochromia appears and the fluid
may form a pellicle on the top or may clot altogether.
Identifying acid-fast bacilli in the smear is an irre-
futable evidence of tuberculous meningitis. Examin-
ing the pellicle stained for acid-fast bacilli may prove
to be a richer diagnostic source than the fluid itself.
Alternatively, the last tube of lumbar puncture speci-
men is the best tube to recover acid-fast bacilli on a
smear. Positive smears in adults range from 10% to
M. Z. Al-Kawi
40% of cases of tuberculous meningitis. The sensitiv-
ity of the smear can be augmented by centrifugation
of a larger specimen volume and by intensive exami-
nation. It was shown to be directly related to the dili-
gence of the technologist performing the test and to
the time spent on the microscope.
Culture of M. tuberculosis from a CSF specimen
is the gold standard for the diagnosis of tuberculous
meningitis. The proportion of positive examinations,
however, is not high. Growth of acid-fast bacilli in
cultures of CSF requires 4-6 weeks and is positive
in approximately 75% of cases of tuberculous men-
ingitis. Its value is therefore confirmatory. Similarly,
IgG antibody against M. tuberculosis in spinal fluid
has been looked for by enzyme-linked immunosor-
bent assay (Kalish et al. 1983) but it is currently less
often used. The time it takes for intrathecal synthesis
of antibody renders the test insensitive in the early
stages when the diagnosis is most needed. Recent
advances in polymerase chain reaction (PCR) tech-
nology methods have shown a faster diagnostic yield.
It has a reported sensitivity of 50%, and a 10% false-
positive rate (Lin et al. 1995). However, PCR confir-
mation of tuberculous meningitis faces difficulties
because of inconsistent standards (Macher and
Goosby 1995). Currently, the use of PCR to diagnose
tuberculous meningitis from CSF specimen cannot
be a routine practice because of its unreliability in
detecting and confirming correctly the presence of
M. tuberculosis.
Diagnosis in patients with HIV infection is often
delayed. Factors suggestive of tuberculous meningi-
tis include endemicity of TB in the community, the
presence of pulmonary TB, and hypoglycorrhachia
(Theuer et al. 1990).
Tuberculomas manifest as enhancing mass
lesions, the usual triad of lesions which show no
vascular blush if cerebral angiography is performed
and are associated with signs that are less than would
normally be expected for the size and location of the
lesion. Biopsy is frequently necessary but a surgical
procedure on the CNS can be avoided (see manage-
ment of tuberculoma below). Since CNS TB is usually
secondary to tuberculous infection elsewhere, search
for an accessible tissue for biopsy and culture (Le.,
cervical or mediastinal lymph node) should be con-
sidered. Stereotactic brain biopsy is often the only
certain method for diagnosing mass lesions in HIV-
infected patients.
Spinal TB recognition on imaging studies is dis-
cussed in the chapter on imaging in this book. Fine
needle aspiration biopsy is a quick and reliable way
for obtaining pathological specimen for microscopic
Tuberculosis of the Central Nervous System
examination and bacteriological confirmation of
spinal TB. It can be done under radiological guid-
ance (CT or fluoroscopy) with little morbidity and a
high diagnostic yield, thereby avoiding more invasive
diagnostic procedures. Material obtained is also help-
ful for culture and sensitivity. This allows immediate
initiation of appropriate treatment, and decreases
hospitalization time (Francis et al. 1999).
32.9
Management
The treatment of CNS TB is more problematic than
treatment of pulmonary TB. To be effective, the drugs
have to cross the blood-brain or the blood-CSF bar-
riers in adequate concentrations.
Treatment of tuberculous meningitis should
be initiated with a regimen of at least three drugs.
Selection of drugs takes into account their abil-
ity to achieve an effective concentration across the
blood-brain or the blood-CSF barriers. Compli-
ance needs to be assured and that may necessitate
supervised intake in some instances. First line drugs
include isoniazid, pyrazinamide and rifampin. Eth-
ambutol or streptomycin may be added in the first
2 months, especially if drug resistance to M. tuber-
culosis is suggested by knowledge about local inci-
dence or prior treatment (Raviglione and O'Brien
1997). In a study of 150 CSF cultures from Egypt,
10% were resistant to isoniazid, 7% to ethambutol
and 3% to rifampin while none showed multi-drug
resistance (Girgis et al. 1998). Duration of treatment
may extend from 9 months to 2 years depending on
the patient's response. Blood tests may show a mild
to moderate rise in liver enzymes during the early
phase of therapy. This should not by itself constitute
an indication for interruption of treatment, unless a
serious liver dysfunction is suggested by a significant
persistent rise in the enzymes or the development
of jaundice. Peripheral neuropathy may develop in
patients treated with INH for TB, particularly when
pyridoxine is not supplemented.
Second-line drugs include: ciprofioxacin, amika-
cin, kanamycin, cycloserine, rifabutin and clofazi-
mine.
The use of corticosteroids is controversial but
is believed to improve outcomes in patients with
tuberculous meningitis who are HIV negative (Kent
et al. 1993). Indications for the use of glucocorticoids
have included severe disease, increased intracranial
pressure, mass effect from associated tuberculomas,
543
deterioration following the institution of antituber-
culous therapy, and spinal block. The rationale for
use of corticosteroids is to attempt to reduce the
inflammatory response through its potential effects
on cytokines, chemokines, and matrix metallopro-
teinases, and on inhibiting cell recruitment and the
synthesis of prostaglandins and leukotriene (Coyle
1999; Dooley et al. 1997).
When tuberculoma is suspected and the patient's
condition allows a watchful waiting, then a therapeu-
tic trial may spare the patient an invasive diagnostic
intervention (i.e., CNS biopsy). Our approach is to
start the patient on triple or quadruple anti-tubercu-
lous first line medications and to avoid the use of cor-
ticosteroids during the trial period. Corticosteroids
may confuse the results of a therapeutic trial by its
nonspecific anti-inflammatory effects or by its abil-
ity to make lymphoma deposits vanish. After 3 weeks
of triple therapy, imaging should be repeated. If the
mass lesion increases a biopsy is indicated. If the
lesion decreases or remains the same the trial should
be continued for an additional 3 weeks. The average
time for early visible decrease in the size of a tubercu-
loma responding to the therapeutic trial is 6-8 weeks.
Follow-up imaging in 6 weeks helps in deciding to
have a biopsy if no improvement is noted. Treatment
is continued if improvement is noted.
If the diagnosis is secured early, corticosteroids
may be used to ameliorate edema and reduce intra-
cranial pressure. In rare occasions, paradoxical
enlargement of lesions occurs during the first weeks
of therapy. Corticosteroids may be used to suppress
such a "paradoxical response" although their effec-
tiveness is not proven.
Initial therapy for most patients with HIV and
TB in the CNS should consist of a four-drug regi-
men because of the possibility of drug-resistant TB.
Rifampin-based or rifabutin-based drugs should be
administered. Compliance is necessary in the treat-
ment of TB in any host, but more so in patients with
HIV infection. Special attention should be paid to
interactions between anti-tuberculous medications
and anti-retroviral drugs.
32.10
Prognosis
Prognosis in tuberculous meningitis correlates with
the duration of illness before hospitalization, clinical
stage on admission and the presence or absence of
hydrocephalus.
544
The British Medical Research Council suggested
classifying the severity of meningitis into three
stages: in stage I are patients who are fully conscious,
rational, and do not have focal neurological signs.
Stage II patients are confused or have focal neuro-
logical signs such as hemiparesis or cranial nerve
palsy. Stage III patients are stuporous or comatose.
This classification may have some management and
prognostic implications. The higher the stage, the
worse the prognosis.
In a study from India, multivariate analysis identi-
fied focal weakness, low score on Glasgow coma scale
and abnormalities in somatosensory evoked poten-
tial as predictors of poor outcome (Misra et al. 2000).
Pediatric patients may have significant sequelae
especially if younger than 20 months or show evi-
dence of infarction or hydrocephalus on imaging
(Wallace et al. 1991).
In patients infected with mv, illness lasting more
than 2 weeks before admission and a CD4+ count
of less than 200 per cubic millimeter predict poor
prognosis (Berenguer et al. 1992). Depressed levels
of consciousness and hemiplegia were associated
with poor prognosis (Katrak et al. 2000). The pres-
ence of tuberculoma does not necessarily preclude a
good long-term prognosis in properly treated HIV-
infected patients (Malaskyand Reichman 1992).
Tuberculous spondylitis has a better prognosis
and effective treatment for a compliant patient can
lead to resolution of a significant part of neurological
manifestations.
References
Alarcon F, Duenas G, Cevallos N et al (2000) Movement dis-
orders in 30 patients with tuberculous meningitis. Mov
Disord 15:561-569
Alothman A, Memish ZA, Awada A et al (2001) Tuberculous
spondylitis: analysis of 69 cases from Saudi Arabia Spine
26:565-570
Anggard E (1994) Nitric oxide, mediator, murderer and medi-
cine. Lancet 343:1199-1206
Awasthi S, Moin S (1999) Effectiveness of BCG vaccination
against tuberculous meningitis. Indian Pediatr 36:455-460
Bahemuka M, Babiker MA, Wright SG et al (1988) The pattern
of infection of the nervous system in Riyadh: a review of
121 cases. Q J Med 68:517-524
Barnes PF, Barrows SA (1993) Tuberculosis in the 1990 s. Ann
Intern Med 119:400-410
Berenguer J, Moreno S, Laguna F et al (1992) Tuberculous
meningitis in patients infected with the human immuno-
deficiency virus. N Engl J Med 326:668
Bruetsch WL (1948) Etiology of optochiasmal arachnoiditis.
Arch Neurol Psychiatry 59:215
M. Z. Al-Kawi
Buranapanitkit B, Lim A, Kiriratnikom T (2001) Clinical
manifestation of tuberculous and pyogenic spine infec-
tion. J Med Assoc Thai 11:1522-1526
Cave AJE (1939) The evidence for the incidence of tuberculosis
in ancient Egypt. Br J Tuberc 30: 142
Cole ST, Brosch R, Parkhill J et al (1998) Deciphering the biol-
ogy of Mycobacterium Tuberculosis from the complete
genome sequence. Nature 393:537
Coyle PK (1999) Glucocorticoids in central nervous system
bacterial infections. Arch Neurol 56:796-801
Crowle AJ, Douvas GS, May MH (1983) The cellular and
molecular nature of human tuberculoimmunity. Bull Int
Union Tuberc 58:72-80
Curling aD, Gower DJ, McWhorter JM (1990) Changing
concepts in spinal epidural abscess: a report of 29 cases.
Neurosurgery 27:185
Dastur DK, Udani PM (1966) Pathology and pathogenesis of
tuberculous encephalopathy. Acta Neuropathol (Bed) 6:
311-326
Dooley DP, Carpenter JL, Rademachen S (1997) Adjunctive
corticosteroid therapy for tuberculosis: a critical reap-
praisal of the literature. Clin Infect Dis 25:872-887
Farinha NJ, Razali KA, Holzel H, Morgan G, Novelli VM (2000)
Tuberculosis of the central nervous system in children: a
20-year survey. J Infect 41 :61-68
Francis 1M, Das DK, Luthra UK et al (1999) Value of radio-
logically guided fine needle aspiration cytology (FNAC) in
the diagnosis of spinal tuberculosis: a study of 29 cases.
Cytopathology 10:390-401
Gachot B, Wolff M, Clair B et al (1990) Severe tuberculosis in
patients with human immunodeficiency virus infection.
Intensive Care Med 16:491-493
Girgis NI, Farid SZ, Mansour MM et al (1998) Tuberculous
meningitis, Abbassia Fever Hospital ... 1976-1996. Am J
Trop Hyg 58:28-34
Kalish SB, Radin RC, Levitz D et al (1983) The enzyme linked
immunosorbent assay method for IgG antibody to purified
protein derivative in cerebrospinal fluid of patirnts with
tuberculous meningitis. Ann Intern Med 99:630-663
Katrak SM, Shembalkar PK, Bijwe SR et al (2000) The clinical,
radiological and pathological profile of tuberculous men-
ingitis in patients with and without human immunodefi-
ciency virus infection. J Neurol Sci 181:118-126
Keane T, Gershon S, Wise RP et al (2001) Tuberculosis associ-
ated with infliximab, a tumor necrosis factor alpha-neutral-
izing agent. N Engl J Med 345:1098-1104
Kent SJ, Crowe SM, Yung A et al (1993) Tuberculous meningi-
tis: a 30-year review. Clin Infect Dis 17:987
Kocen RS, Parsons M (1970) Neurological complications of
tuberculosis: some unusual manifestations. Q J Med 39:17
Kony SJ, Hane AA, Larouze B (2000) Tuberculosis-associated
severe CD4+ T-lymphocytopenia in HIV-seronegative
patients from Dakar. J Infect 41: 167-171
Lin JJ, Harn HJ, Hsu YD, Tsao WL et al (1995) Rapid diagnosis
of tuberculous meningitis by polymerase chain reaction
assay of cerebrospinal fluid. J NeuroI242:147-152
Macher A, Goosby E (1995) PCR and the misdiagnosis of
active tuberculosis. N Engl J Med 332:128-129
Malasky C, Reichman LB (1992) Long-term follow-up of tuber-
culoma of the brain in an AIDS patient. Chest 101:278-279
Matsuyama W, Hashiguchi T, Umehara F et al (2001) Expres-
sion of vascular endothelial growth factor in tuberculous
meningitis. J Neurol Sci 186:75-79
Tuberculosis of the Central Nervous System
Misra UK, Kalita J, Roy AK et al (2000) Role of clinical, radio-
logical, and neurophysiological changes in predicting the
outcome of tuberculous meningitis: a multivariate analysis.
J Neurol Neurosurg Psychiatry 68:300-303
Osler W (1921) The evolution of modern medicine. Yale Uni-
versity Press, New Haven, CT
Raviglione MC, O'Brien RJ (1997) Tuberculosis. In: Fauci AS
et al (eds) Harrison's principles of internal medicine, 14th
edn. McGraw-Hill, New York, pp 1004-1014
Sanchez-Portocarrero J, Perez-Cecilia E, Romero-Vivas J (1999).
Infection of the central nervous system by Mycobacterium
tuberculosis in patients infected with human immunode-
ficiency virus (the new neurotuberculosis). Infection 27:313
Selwyn PA, Sckell BM,Alcabes P et al (1992) High risk of active
tuberculosis in HIV-infected drug users with cuaneous
anergy. JAMA 268:504
Sharma MC, Arora R, Mahapatra AK et al (2000) Intrasellar
tuberculoma - an enigmatic pituitary infection: a series of
18 cases. Clin Neurol Neurosurg 102:72-77
Silber MH, Willcox PA, Bowen RM (1990) Neuromyelitis optica
(Devic's syndrome) and pulmonary tuberculosis. Neurol-
ogy 40:934-938
Singh BS, Patwari AK, Deb M (1994) Serum sodium and
545
osmolal changes in tuberculous meningitis. Indian Pediatr
31:1345-1350
Tagliati M, Godbold J, Hassett J et al (1994) Neuromuscular
disorders in HIV infection: cross-sectional cohort analysis
of 250 patients. Neurology 44:A367
Takashima T, Ueta C, Tsuyuguchi I et al (1990) Production of
tumor necrosis factor alpha by monocytes from patients
with pulmonary tuberculosis. Infect Immun 58:3286-3292
Theuer CP, Hopewell PC, Elias D et al (1990) Human immuno-
deficiency virus infection in tuberculosis patients. J Infect
Dis 162:8
Udani PM, Parekh UC, Dastur DK (1971) Neurological and
related syndromes in CNS tuberculosis. Clinical features
and pathogenesis. J Neurol Sci 14:341
Wallace RC, Burton EM, Barrett FF et al (1991) Intracranial
tuberculosis in children: CT appearance and clinical out-
come. Pediatr RadioI21:241-246
Weigel B (1976) Changes of the tuberculosis of the central
nervous system since the introduction of prophylactic
BCG vaccination and tuberculostatic therapy. Zentralbl
Allg PathoI120:21-33
Zimmerman MR (1979) Pulmonary and osseous tuberculosis
. in an Egyptian mummy. Bull NY Acad Med 55:604-608
33 Imaging of Brain and Spinal Cord Tuberculosis
FRANCIS MCGUINESS

CONTENTS 33.1
Brain and Spinal Cord Tuberculosis
33.1 Brain and Spinal Cord Tuberculosis 547
33.2 Histopathology of Intracranial Tuberculosis 548
33.3 Tuberculous Meningitis (TBM) 548 Events during the past decade have dramatically
33.4 Imaging Characteristics of TBM 550 changed the nature and magnitude of the problem of
33.5 Tuberculomas of TBM 554 tuberculosis. Much of what many physicians learned
33.6 Parenchymal Cerebral Tuberculosis 555 in training about this disease is no longer true. In many
33.6.1 Parenchymal Tuberculomas 555 respects tuberculosis has become a new entity. (Snider
33.6.2 Imaging Characteristics of Tuberculomas 557
33.7 Paradoxical Response of Tuberculomas and Roper 1992)
to Treatment 560 While taking into account the coinfection of
33.7.1 Late Stage Appearances of Tuberculomas 560 AIDS and tuberculosis, the authors, Drs Snider and
33.8 Tuberculous Brain Abscesses 561 Roper, were also stressing the major global increase
33.9 Atypical Tuberculous Masses 561 in tuberculosis infection that occurred in the 1980s.
33.10 Miliary Cerebral Tuberculomas 562
33.11 Tuberculosis of the Calvarium This epidemic led to the World Health Organisation
with CNS Lesions 563 (WHO) declaring a worldwide emergency in April
33.12 Tuberculous Otitis Media and Tuberculosis 1993 and describing the spread of tuberculosis as a
of the Temporal Bone 564 neglected health crisis.
33.12.1 Imaging 565 In its most recent report in 2000, WHO estimates
33.13 Differential Diagnosis of TBM
and Parenchymal Tuberculosis 565 that 1.9 billion persons are infected with the tubercle
33.14 Tuberculous Radiculomyelopathy bacillus and have the potential to develop postpri-
and Myelitic Tuberculomas 570 mary tuberculosis (TB). That is one third of the
33.15 Imaging Methods in Spinal world's population (Pio and Chaulet 1998).
Neurotuberculosis 571 Central nervous system tuberculosis can affect any
33.15.1 Plain Radiography 571
33.15.2 Water-Soluble Contrast Myelography of the elements of the neural tissues as well as their
(WSCM) 572 investing meninges. Jinkins estimates that between 2%
33.15.3 CT Combined with Water-Soluble and 5% of patients with active tuberculosis elsewhere
Myelography 572 in the body and up to 10% of those with the AIDS
33.15.4 Magnetic Resonance Imaging 573 coinfection develop central nervous system lesions
33.16 Differential Diagnosis 575
33.16.1 Spinal Arachnoiditis 575 (Jinkins 1991; Kanamalla et al. 2000).
33.16.2 Infective Leptomeningitis 576 The infection shows a predilection for the
33.16.3 Spinal Meningitis in Neoplastic Conditions 576 younger age groups, many patients being under the
33.16.3.1 Meningeal Carcinomatosis 576 age of 5 years and the majority below the age of 30
33.16.4 Fungal Diseases 579 (McGuinness 2000; Jamieson 1995; Schoeman et al.
33.16.5 Spirochetal Disease 580
33.16.6 Other Granulomatous Diseases 581 1988; Cremin and Jamieson 1996). However, in the
33.16.7 Parasitic Diseases 582 West there is an increase in the incidence of the dis-
33.16.7.1 Schistosoma 582 ease in those over 60 years of age.
33.16.8 Neurocysticercosis 582 The structure and reproductive properties of
33.16.8 Demyelinating Disease 582 Mycobacterium tuberculosis are such that tubercu-
33.16.8.1 Multiple Sclerosis 582
References 583 lous infections develop insidiously. Mycobacterium
tuberculosis var. hominis is the common causal agent
of infection, although Mycobacterium tuberculosis
F. MCGUINESS var. bovis also causes a number of extrapulmonary
Apt. 169, AI Haurin El Grande, 2912 Malaga, Spain infections.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
548
Both bacilli have a protective layer of phospho-
lipids, phosphoglycolipids, and waxes, and are not
detected by Gram stain. Ziehl-Neelsen stain and fluo-
rescent staining methods reveal the organism, but in
only a few cases is the bacillus found in the initial
examination of cerebrospinal fluid (CSF).
The slow reproductive time of 48 hours makes the
organism difficult to culture, and means that clinical
presentation and biochemical assessment are para-
mount to making the initial diagnosis of tuberculous
meningitis (TBM), tuberculous meningomyelitis
(TBSM), and tuberculoma (Jinkins 1991; Kanamalla
et al. 2000; McGuinness 2000; Jamieson 1995; Cremin
and Jamieson 1996; Shah 2000).
33.2
Histopathology of Intracranial
Tuberculosis
The manifestations of intracranial tuberculosis are
diverse and are reflected in the imaging. In interpret-
ing the images, an understanding of the underlying
histopathology of the disease is helpful. Detailed
descriptions of these changes are to be found in the
literature (McGuinness 2000; Shah 2000; Dastur et al.
1995; Dastur 1972; Kirkpatrick 1991; Reid and Fallon
1992). A brief account of the major pathologic changes
in central nervous system tuberculosis follows.
Central nervous system tuberculosis is almost
invariably the result of hematogenous spread of
infection from a focus elsewhere in the body. In the
majority of cases, the primary source of the disease is
a focus of pulmonary or pulmonary-lymphatic tuber-
culosis. In infants and children, a primary pulmonary
focus leading to local lymph node involvement or the
development of progressive postprimary tuberculo-
sis is the common cause of spread of the infection
to the nervous system (McGuinness 2000; Jamieson
1995; Cremin and Jamieson 1996; Shah 2000). In
older patients (from adolescence onwards), reactiva-
tion of pulmonary disease, or miliary tuberculosis
is a common cause of hematogenous dissemination
to the central nervous system. Dissemination from
tuberculous foci in the gastrointestinal or genitouri-
nary systems, or from foci in bones and joints, the
paranasal sinuses, or the middle ear are less common
occurrences (Kanamalla et al. 2000; McGuinness
2000; Jamieson 1995; Cremin and Jamieson 1996;
Shah 2000; Dastur et al.1995).
Breakthrough of the infection from osteogenic
tuberculous foci in the spine or the cranial vault is
F. McGuiness
less frequent, due to the protective nature of the dura
mater. The dura mater, comprising two layers - the
outer layer fused to the periosteum and the inner
layer bound to it closely - forms a formidable bar-
rier to the spread of infection across its boundary
(McGuinness 2000; Shah 2000).
Although the initial response of the body to tuber-
culous infection is a nonspecific inflammatory one, a
cell-mediated immune response develops within 2 to
3 weeks. Further foci of tuberculous infection elicit
the development of granulomata. Granulomatous
tuberculous foci comprise central clumps of epithe-
liod cells and macrophages containing intracellular
tubercle bacilli. The core is surrounded by fibroblasts
and mononuclear inflammatory cells and often con-
tains Langerhans giant cells (Kanamalla et al. 2000;
McGuinness 2000; Cremin and Jamieson 1996; Shah
2000; Dastur et al.1995). These foci were described by
Rich in 1933 and bear his name (Rich and McCordock
1933; Rich 1944).
There are two theories concerning the develop-
ment of tuberculous lesions in the leptomeninges,
the cerebrum, and myelum. The common theory is
that Rich foci, resulting from hematologic spread,
develop in the leptomeninges and rupture into the
subarachnoid space. The cerebrospinal fluid has min-
imal protection against infection and the space and
leptomeninges become widely infected (Kanamalla
et al. 2000; McGuinness 2000; Dastur et al. 1995;
Dastur 1972). A second theory is that the hematog-
enous spread to the capillary and venule walls in the
leptomeninges and similar vessels penetrating the
cerebrum and myelum, causes inflammatory break-
down of the vessel walls and of the blood-brain bar-
rier. This allows spread of the infection both super-
ficially over the leptomeningeal surface and deeper
into the brain or spinal cord substance. This seems
the likely cause of the development of the meningo-
cerebritis characteristic of tuberculous meningitis,
and the associated tuberculomas show a preference
for development at the cortico-white matter inter-
face, where the peripheral blood vessels are at their
narrowest (Kanamalla et al. 2000; McGuinness 2000;
Cremin and Jamieson 1996; Shah 2000; Dastur et al.
1995; Dastur 1972).
33.3
Tuberculous Meningitis (T8M)
TBM develops when a microscopic granuloma, a Rich
focus, situated beneath the pia mater, in the cerebral
Imaging of Brain and Spinal Cord Tuberculosis
cortex, or in the wall of a cerebral venule, ruptures
into the subarachnoid space. Untreated, TBM is a
fatal disease and unless diagnosed and treated early
leads to calamitous neurologic defects. TBM is still
a common disease. This is especially so in the 24
high-burden countries named in the WHO report
on worldwide tuberculosis (Pio and Chaulet 1998),
but it also remains a problem in the economically
rich states of the world.
Wallace quotes approximately 4,000 cases of TBM
per year, reported to the Centers for Disease Control
of the United States (Wallace et al. 1991; Ogawa et al.
1987). Both Wallace and Hooijboer (Hooijboer et al.
1996) emphasize that although commonly occurring
in immigrant groups in the United States and Europe,
meningitis surprises the physician by expressing
itself among native-born children as well. TBM is
an insidious disease and a high clinical awareness
is essential to its diagnosis. Most published series
emphasize the fact that the common presentation
to the physician is after the initial stages, when the
illness is well established (Jinkins 1991; McGuinness
2000; Jamieson 1995; Schoeman et al. 1988; Cremin
and Jamieson 1996; Wallace et al. 1991; Hooijboer et
al. 1996; Kioumehr et al. 1994; De Castro et al 1995;
Jinkins et al.1995; Cremin 1995; Leiguarda et al.1988;
Waeker and Connor 1990).
The presentation is characterized by malaise, las-
situde, low grade fever and an intermittent headache.
Unlike the acute pyogenic and viral meningitides,
high fever, papilloedema, and focal cerebral symp-
toms are unusual in the early stages. In this phase of
the infection the clinical findings are similar to the
other chronic meningitides caused by fungus, para-
sites, meningeal carcinomatosis, Lyme disease, and
neurosarcoidosis (McGuinness 2000; Shah 2000).
The severity of the disease is classified according to
the Medical Research Council of the United Kingdom
staging system of 1948. Stage I disease is character-
ized by nonspecific symptoms in a conscious patient,
including headache, vomiting, irritability, or lethargy
with no paresis and a good general condition. Stage
II disease shows drowsiness, photophobia, meningeal
irritation, and focal neurologic signs with possible
seizures. Stage III disease presents with profound
changes in the sensorium, major neurologic signs,
and coma (McGuinness 2000; Wallace et al. 1991;
Waeker and Connor 1990; Medical Research Council
1948; Lincoln et al. 1960).
CT scanning is advised before lumbar puncture.
Examination of the cerebrospinal fluid after the first
few days, demonstrates a lymphocytic pleocytosis,
increased protein level, and diminished glucose.
549
Bacteriologic or histologic support for the diagno-
sis is sought by the purified protein derivative skin test
(PPD), which is particularly helpful in children pro-
ducing a high percentage of positives. Examination
of sputum, gastric washings, and CSF by direct smear
and culture should be carried out in all suspected
cases. CSF examination with acid-fast stains is, in
most reported series, positive in only a small propor-
tion of cases, usually less than 10% (McGuinness 2000;
Jamieson 1995; Cremin and Jamieson 1996; Wallace
et al. 1991; Hooijboer et al. 1996; Kioumehr et al. 1994;
De Castro et al1995; Jinkins et al. 1995; Leiguarda et
al. 1988). Waeker (Waeker and Connor 1990) reports
only one of 30 children having a positive smear, and
37% with positive CSF cultures. Shah (Shah 2000)
suggests a higher proportion of 90% of cases being
CSF-positive to acid-fast staining methods. Search for
the bacillus should also be made in the urine and in
cultures taken from any ear infection (McGuinness
2000; Wallace et al. 1991; Naranbhal et al. 1989). His-
tologic support or positive bacteriologic cultures may
derive from samples taken from other regions, remote
from the nervous system. Chest radiography is only
positive for visible evidence of tuberculosis in 40-50%
of cases, and a negative chest radiograph in no way
precludes the diagnosis of TBM (McGuinness 2000;
Jamieson 1995; Cremin and Jamieson 1996; Hooi-
jboer et al. 1996; De Castro et al 1995; Jinkins et al.
1995; Leiguarda et al.1988; Waeker and Connor 1990;
Ozates et al. 2000). Control examinations of family
members are of great importance and often reveal
active tuberculosis in parents or siblings (McGuinness
2000; Ogawa et al.I987).
The most effective radiologic investigation in cases
of TBM is a contrast-enhanced CT scan (CECT), and
where this investigation is available it should be car-
ried out immediately on presentation of the patient.
After the initial compromise of the leptomen-
inges by Rich foci, cell-mediated immunity leads
to the development of a glutinous exudate. In the
initial stages the exudate is largely confined to the
basal subarachnoid areas, but rapidly involves the
basal cisterns, particularly the interpeduncular
and suprasellar spaces (McGuinness 2000; Dastur
et al. 1995; Dastur 1972; Kirkpatrick 1991; Reid
and Fallon 1992). Spread of the exudate along the
periarterial spaces of the major arteries and their
branches follows, with encasement of the common
carotid, anterior, middle, and posterior cerebral
arteries. This facilitates spread of the exudate to the
ambient system, prepontine cistern, and the supra-
tentorial subarachnoid space, including Sylvius'
fissure. The exudate is an inflammatory response to
550
the presence of both intra-and extracellular bacilli.
The cell-mediated response from activated T cells
leads to an outpouring of lymphocytes, plasma
cells, and macrophages, which reorganize into
countless microscopic tubercles. In the later stages
of the disease, these tuberculous granulomas are
surrounded by fibroblasts, and caseous necrosis
occurs in the central core (Jinkins 1991; McGuin-
ness 2000; Jamieson 1995; Schoeman et al. 1988;
Cremin and Jamieson 1996; Shah 2000; Dastur et al.
1995; Dastur 1972; Wallace et al. 1991; De Castro et
a11995; Jinkins et al. 1995).
33.4
Imaging Characteristics of T8M
The common triad of imaging changes in TBM on
contrast-enhanced CT and gadolinium-enhanced
MRI are basal meningeal enhancement, hydroceph-
alus and parenchymal supratentorial infarctions.
These changes are the direct result of the underly-
ing histopathologic processes. Enhancement is due
to the breakdown of the blood-brain barrier in the
leptomeninges and the intense basal inflammatory
exudate, hydrocephalus due to the blockage of the
normal CSF pathways, and infarctions are the result
of spasm or closure of basal cerebral vessels.
Basal contrast enhancement in the early stages is
confined to the leptomeninges, although in chronic
cases the pachymenix may be extensively affected.
The enhancement follows the pathway of the basal
vessels and the perivascular or Virchow-Robin spaces.
As the inflammatory response to the infection devel-
ops, exudates are seen in the basal leptomeninges and
basal cisterns, especially in the ambient, sylvian, pon-
tine, and suprasellar areas. Widespread enhancement
of exudates is the common pattern, although rarely
focal enhancement has been reported involving
Sylvius' fissure (Shah 2000; Klingensmith and Datu
1978) and the left parietal leptomeninges (Elkeslassy
et al. 1997). Linked with the other two elements of the
triad this enhancement is highly suggestive of TBM
but is by no means pathognomonic. Other condi-
tions - bacillary or viral meningitis, fungal menin-
gitis, other granulomatous diseases, inflammatory
response to the rupture of a paracytic or dermoid
cyst, and meningeal metastatic disease - may dem-
onstrate basal meningeal enhancement, but usually
of a less intense nature and a narrower distribution
(Jinkins 1991; McGuinness 2000; Kioumehr et al.
1995).
F. McGuiness
Noncontrast CT studies (NCCT) are of little value
in the detection of meningeal disease. The beam
hardening artefacts due to the overlying cranium
and the bones of the posterior fossa make it difficult
to image the meninges. In the few cases where cal-
cification develops at an early stage in TBM, NCCT
has an advantage. The presence of basal exudates is
suspected when the subarachnoid space is obliter-
ated by isoattenuating tissue. With CECT these can
be readily demonstrated. The usual pattern is of
long continuous segments of enhanced leptomen-
inges, often with associated pial enhancement of the
perivascular intergyral spaces. In advanced cases,
evidence of cerebritis may be seen, signaled by areas
of hypoattenuation in the underlying cerebral cortex.
The most apparent leptomeningeal pathologic
change is evidenced by the predilection of post-con-
trast enhancement for the basal cisterns, particularly
the suprasellar region, the prepontine region, and
ambient system (Kanamalla et al. 2000; McGuinness
2000; Jamieson 1995; Cremin and Jamieson 1996;
Shah 2000; Dastur et al. 1995; Dastur 1972; Wallace
et al. 1991; Kioumehr et al.1994; De Castro et a11995;
Jinkins et al.1995; Waeker and Connor 1990; Ozates et
al. 2000; Kioumehr et al.1995). (Fig. 33.1a, b)
Magnetic resonance imaging (MRI),is the gold stan-
dard for the imaging of the brain and leptomeninges.
As in the case of CT imaging, precontrast studies are
often of little value in acquiring information on men-
ingeal infectious disease (Kanamalla et al. 2000; Shah
2000; Kioumehr et al. 1995; Meltzer et al. 1996). Post
intravenous gadolinium enhanced images possess an
inherent high contrast resolution and improved tissue
differentiation. These qualities linked with multipla-
nar imaging make it the imaging modality of choice
in intracranial infectious disease. It is unfortunate that
MRI examination is not universally available in most
of the areas of the world with a high burden of tuber-
culosis. In some of these areas a heightened clinical
awareness and often limited laboratory facilities are
the only aids available to the physician. Recent MRI
applications suggest that fluid attenuation inversion
recovery (FLAIR) acquisitions are helpful in cases
where meningeal disease is suspected (Kanamalla et
al. 2000; Singer et al.1998).
Gadolinium-enhanced Tl-weighted MRI images
readily reveal the extent of leptomeningeal enhance-
ment in TBM and the distribution of the lesions. They
will also bring to light the concomitant changes of
subcortical cerebritis, hydrocephalus, brain infarc-
tions, ependymitis, and associated tuberculomas
(Kanamalla et al. 2000; McGuinness 2000; Shah 2000;
Jinkins et al.1995; Gupta et al.1994). The second com-
 Imaging of Brain and Spinal Cord Tuberculosis 551

a b
Fig. 33.1 a Axial Tl-weighted, postgadolinium scan, demonstrating intense enhancement of the basal meninges in a case of tuber-
culous meningitis. The white arrow points to border-zone disease adjacent to the fourth ventricle. b A similar scan at a lower level.
There is enhancement of the surface of the brain stem and of exudates in the surrounding cisterns (arrow). On both scans there is
enhancement of the meninges over the convexities, and a number of small parenchymal granulomas. There is hydrocephalus

ponent of the imaging triad, hydrocephalus, is almost sary, although in TBM many stage I and stage II cases
universally present at the time of patient presentation. will require ventricular drainage early in the course of
It commonly develops in tandem with the glutinous treatment. (Fig. 33.2a, b).
basal exudates which cause disruption of CSF circula- Early development of hydrocephalus in TBM
tion and reabsorption. The hydrocephalus is charac- is emphasized by most authors (Jinkins 1991;
teristically of high pressure, and the periventricular Kanamalla et al. 2000; McGuinness 2000; Jamieson
white matter is suffused to a greater or lesser extent 1995; Schoeman et al. 1988; Cremin and Jamieson
with hydrostatic edema. Until recently the reabsorp- 1996; Kioumehr et al. 1994; De Castro et al 1995;
tion of CSF was thought to occur solely by the activity Jinkins et al.1995; Ozates et al. 2000; Gupta et al.1994;
of the arachnoid granulations overlying the cerebral Bonafe et al. 1985; Dastur et al. 1970; Artopulous et
convexities. Hydrocephalus in TBM was thought to be al. 1984; Chang et al. 1990). Wallace (Wallace et al.
secondary to obstruction of the flow of CSF around 1991) and Waecker (Waeker and Connor 1990) also
the midbrain and in the basal subarachnoid space note its early development and find it a more reliable
by leptomeningeal exudates, which also blocked the diagnostic pointer in TBM than the presence of basal
arachnoid granulations. Recent work by Greitz et al., exudates. Of Wallace's nine cases, ventriculomegaly
however, suggests a new concept of CSF absorption was present in seven, while it was present in all of
(Greitz et al. 1997). Using flow-sensitive MRI and Waecker's 30 cases. In Ozates' study of 289 TBM
radionuclear cisternography, they propose that the patients (Ozates et al. 2000), 204 developed hydro-
main site of absorption of CSF is through the walls of cephalus in the course of the disease, and it was pres-
cerebral arterioles, capillaries, and venules. In TBM, ent in 163 patients on presentation. In general, the
compromise of these vessels, by spasm and arteritis, severity of hydrocephalus corresponds to the sever-
resulting from encasement by exudate in the sub- ity of the infection, it is more pronounced in children
arachnoid space, would be an important factor in the than in adults, and serial CT studies show that it often
early development of communicating hydrocephalus. worsens during the first few weeks despite adequate
Rarely, in TBM the hydrocephalus is of the obstructing chemotherapy.
type, resulting from the obstruction of the aqueduct of Arteritis, vessel encasement by exudate, spasm, and
Sylvius or the foramina of Luschka and of Magendie by vessel occlusion lead to compromise of the arterial flow
subarachnoid granulomas or glutinous subarachnoid in the vessels of the circle of Willis and their branches.
exudates. It is important to recognize this complica- So leading to the third feature of the clinical triad, cere-
tion, as in these cases early shunt placement is neces- bral infarction. Cerebral cortical, white matter, and tha-
 552
a
Fig. 33.2a, b. Sagittal T1 and proton-density scans. There is a grape-like cluster of tuberculous granulomas obstructing the outlet
of the fourth ventricle, resulting in hydrocephalus (arrow)
lamic infarction is a major cause of permanent brain
damage, occurring despite adequate antituberculous
chemotherapy, and may be taken as a good indicator of
an eventual poor clinical outcome.
Angiographic and postmortem studies have dem-
onstrated occlusions ranging from closure ofthe supra-
sellar segment of the carotid artery to thrombosis of its
major branches, leading to large territorial infarctions
(McGuinness 2000; Dastur et al. 1995; Dastur 1972;
Leiguarda et al. 1988; Gupta et al. 1994). The typical
angiographic changes of irregularities in caliber ofves-
sels branching from the circle ofWillis, occlusion of the
medium-size arteries at the base of the brain, and early
venous drainage were described by Lehrer (Lehrer
H 1966), and when combined with a hydrocephalic
pattern of vessel distribution, were called by him, the
angiographic triad of TBM.
However, the smaller perforating branches are
more commonly affected than the larger arteries, and
occlusion of the medial-striate and thalamo-striate
perforating arteries, leads to a characteristic picture
of infarction in the thalamus and caudate nucleus.
As in many of the manifestations of TBM, infarc-
tion is more common in patients under 5 years of age
(Kanamalla et al. 2000; McGuinness 2000; Jamieson
1995; De Castro et al 1995; Leiguarda et al. 1988).
This underscores the notion that the branches of the
arteries of the circle of Willis become involved in the'
inflammatory process as they pass through the basal
exudates. In the case of children, the small caliber of
the vessels makes them more vulnerable and suscep-
tible to thrombotic occlusion than the larger caliber
vessels of adults. Postmortem studies have shown
F. McGuiness
that venules as well as arteries are involved in this
pathologic process. Large vessel infarction does how-
ever occur, even in adults, and may be demonstrated
by digital subtraction angiography (DSA) (Fig. 33.3)
or magnetic resonance arteriography (MRA). Large
vessel infarction in the territories of the anterior or
middle cerebral arteries leads to extensive cerebral
cortical infarction. These areas subsequently develop
encephalomalacia and segmental areas of overlying
cortical atrophy. At a later stage, areas of calcification,
best seen on CT examination, have been described in
these areas (McGuinness 2000; Ogawa et al. 1987; De
Castro et al1995; Jinkins et al. 1995).
Although there are no reported cases of clinical
deterioration after DSA, the examination is today
deemed unnecessary in cases of TBM (McGuin-
ness 2000; Rochas-Echeverri et al. 1996). Impending
infarction is suggested by a shaggy appearance of the
perivascular exudates on CFCT, and the results of
infarction are clearly demonstrated on both CT and
MRI examinations (Ogawa et al. 1987). Gupta has
demonstrated the presence of large and small vessel
arteritis by MRA studies, a noninvasive method of
examination (Gupta et al.1994).
At post mortem, the incidence of infarction is
around 40% (Dastur et al. 1995; Dastur 1972; Dastur
et al. 1970). Published studies of investigation by CT
examination suggest a discovery rate of around 28-
38%. MRI studies are more sensitive, having a discov-
ery rate of around 50% and demonstrate that many of
these lesions are hemorrhagic, which may lead to cavi-
tation (McGuinness 2000; Jamieson 1995; Schoeman et
al. 1988; Cremin and Jamieson 1996; Shah 2000; Wal-
 Imaging of Brain and Spinal Cord Tuberculosis 553

lace et al.1991; Jinkins et al.1995; Leiguarda et al.1988;
Gupta et al.1994; Rochas-Echeverri et al.1996).
Schoeman (Rochas-Echeverri et al.1996),described
a group of 27 children investigated with MRI, with
much higher rates of infarction, 20 of whom presented
with stage II or III disease, had midbrain or basal gan-
glia lesions. Ten presented with brain stem, parahip-
pocampal gyri, or hypothalamic changes. By contrast,
Ozates in a review of 289 TBM cases - 214 children
and 75 adults - examined by CT, reports a lower overall
incidence of infarctions of 13%. Infarctions are seen on
CT scans as areas of hypoattenuation. On Tl-weighted
gadolinium-enhanced MRI (Tl Gd) the areas of infarc-
tion stand out as hypodense lesions (Fig. 33.4a, b). On
T2 MRI, acquisitions infarctions show a high signal
and are difficult to differentiate from areas both of
leptomeningitis and border-zone disease.
Border-zone changes are a form of cerebritis that
develops beneath the inflamed leptomeninges, and
these are frequently seen in the superficial cerebral
cortex, the cerebral peduncles abutting onto the
interpeduncular fossa, and the midbrain and pons
adjacent to areas of cisternal enhancement. They are
usually present in advanced disease and indicate a
poor prognosis (McGuinness 2000; Jamieson 1995;
Cremin and Jamieson 1996). (See Fig. 33.5a, b.)

Fig. 33.3a,b. Avertebral artery angiogram with much wall

irregularity in the basilar artery and the posterior fossa
a branches, resulting from tuberculous arteritis (arrows)

a b
Fig. 33.4a, b. Extensive, bilateral, thalamic infarction and hydrocephalus on coronal MRI (arrows). Images courtesy of Dr J. Lotz,
Riyadh Military Hospital, Saudi Arabia
 554
Infarctions are commonly supratentorial, and
brain stem infarction only occurs in 2-3% of cases,
mostly in infants and children. The clinical syn-
dromes resulting from TBM cerebral and brain stem
infarctions encompass a wide variety of patterns,
ranging from focal or generalized seizures, menta-
tion and gaze disturbances, monoplegias, hemiple-
gias, and nuclear cranial nerve lesions. Although rare,
the arteritis of TBM can lead to aneurysm formation
with fatal consequences, and a small number of such
cases have been described (McGuinness 2000; Jinkins
et al. 1995; Leiguarda et al. 1988; Gupta et al. 1994;
Cross et al. 1995). One such case, reported by Cross
et al., responded to endovascular occlusion treatment
of the aneurysm, after presenting with life-threaten-
ing epistaxis (Cross et al.1995).
a The fifth and seventh nerves are sometimes affected
b
Fig.33.5a, b. Sagittal and axial, T2-weighted images
demonstrate extensive basal exudate and border-zone
disease. This encroaches on the anterior margin of the
right cerebellar hemisphere (arrows)
F. McGuiness
33.5
Tuberculomas of T8M
The wide distribution of microscopic granulomas in
the subarachnoid space, leads to the development of
tuberculomas. These tend to be small in size and are
generally seen related to the supratentorialleptomen-
inges. Parenchymal tuberculomas complicating TBM
tend to be found at the cortico-white matter border,
although deeper masses may be found. In adults,
most are supratentorial, while in children, tuberculo-
mas of the cerebellum and brain stem are commoner.
Initially solid, tuberculomas mature to develop cen-
tral caseating necrosis. Often developing in clusters
they display the radiologic properties of granulomas.
When solid they are isodense or hypodense to brain
tissue on CT and MRI scans carried out without
contrast medium. Postcontrast studies demonstrate
homogeneous enhancement in solid lesions and an
intense ring enhancement in tuberculomas undergo-
ing central caseating necrosis. An area of edema is
often present and in TBM the lesions are usually no
greater than 0.5 cm in diameter (Fig. 33.6).
Basal inflammatory exudate affects the cranial
nerves as well as the vessels that course through it. Cra-
nial nerve deficits are a common complication of stage
II and III TBM. The optic nerve or chiasma is open to
compression due to suprasellar leptomeningitis. The
long, basal, intracranial courses of the third, fourth,
and sixth nerves expose them to exudate damage.
Fig.33.6. MRI of an enhancing, grape-like cluster of tuber-
culomas (arrow). Note also the ependymitis in the temporal
horn of the right lateral ventricle (arrow head). This was in a
terminal case of tuberculous meningitis
 Imaging of Brain and Spinal Cord Tuberculosis
as they leave the midbrain. Gupta has demonstrated
enhancement of the inflamed nerve sheaths on T1 Gd
MRI (McGuinness 2000; Jinkins et al. 1995; Gupta et
al. 1994) (Fig. 33.7a, b). Permanent damage results as
healing converts the basal exudates into dense, fibrotic
material, with the pattern of deficit varying from
patient to patient (McGuinness 2000; Jinkins et al.
1995; Gupta et al.1994; Artopulous et al.I984).
Long-standing TBM may in some instances lead to
infection of the dura mater leading to thickening and
pachymeningitis. Common sites include the floor of
a
b
Fig. 33.7. a Postgadolinium T1 image demonstrating enhance-
ment of a left-sided tuberculous infection of the middle ear
(arrow). There is infection of the temporal lobe (arrow head).
b A coronal image confirms the middle ear infection (arrow)
and the associated temporal-lobe tuberculoma. Central areas
oflow signal caseation are seen (arrow head), as is a surround-
ing zone of edema (curved arrows)
555
the middle cranial fossa, the dura overlying the cere-
bral convexities, and the tentorium. Calcification may
be present. Contrast studies reveal intense homoge-
neous enhancement of the affected areas (Shah 2000;
Ng et al.1996; Praharaj et al.I997).
Isolated areas of calcification of the dura mater
and cases showing punctate gyral calcification are
reported (Kanamalla et al. 2000; McGuinness 2000;
Shah 2000; Wallace et al. 1991; De Castro et aI1995).
Calcification may be an early feature or may occur
years later, after successful treatment of the infection.
33.6
Parenchymal Cerebral Tuberculosis
Parenchymal cerebral tubercular disease occurs spo-
radically, and as a complication of TBM. It may be
subdivided into
1. Parenchymal tuberculomas
2. Tuberculous abscess
3. Tuberculomas en Plaque
4. Miliary tuberculomas
33.6.1
Parenchymal Tuberculomas
In cases ofTBM,the incidence ofcerebral tuberculoma
varies. Atlas suggests a low incidence of 10% (Atlas
1991), but Jinkins found higher rates in Saudi Arabia:
in a study of 80 patients with intracranial tubercu-
losis, 11 % had compound meningeal/parenchymal
lesions and 89% had parenchymal tuberculomas. Of
all mass lesions in his practice, Jinkins found that
10-15% were tuberculomas (Jinkins 1991).
In developing countries, sporadic parenchymal
tuberculomas make up a large proportion of intra-
cranial space-occupying lesions. In patients in India,
Gupta suggests that between 10% and 40% of all
intracranial mass lesions are tuberculomas (Gupta
et al. 1988; Gupta et al. 1990). Salgado points out an
increasing incidence of tuberculomas in developed
countries (Salgado et al.I989). Cremin and Jamieson
(Schoeman et al. 1988; Cremin and Jamieson 1996)
recorded parenchymal tuberculoma formation in
10% of their cases in children. Sporadic cerebral
tuberculous masses had a different clinical presen-
tation from lesions in TBM, which were multiple in
15-20% ofTBM patients.
Tuberculomas may be single or multiple, and when
multiple, may demonstrate differing degrees of devel-
 556
opment at the same time. Most tuberculomas occur
at the cortico-white matter junction, suggesting an
origin by hematologic spread (Fig. 33.8a, b). A small
number develop by direct spread from Rich lesions in
the pia mater, in which case the overlying meninges
enhance on CECT or T1 Gd MRI. The origin, in a few
cases, is by spread from a venous sinus (McGuinness
2000). In the less common forms of tuberculomas of
the basal ganglia, brain stem, and cerebellum, dif-
ferentiation from primary or secondary neoplasms
is a problem, especially in adults (Fig. 33.9). Beneath
the tentorium, both single and multiple lesions are
described, but account for only between 1% and 5%
of tuberculomas (Jinkins 1991; Kanamalla et al. 2000;
a but any age group may be affected (Jinkins 1991;
b developments in MRI, have defined groups of appear-
Fig. 33.8a, b. Postcontrast computer scans of a young patient.
As well as hydrocephalus, the scans demonstrate multiple
enhancing, subcortical tuberculomas in the frontal and pari-
etal regions, and at a higher level beneath the convexity
F. McGuiness
Fig. 33.9. Sagittal postcontrast MRI demonstrates an isolated
tuberculoma of the brain stem. It is impossible on this image
to differentiate the appearance from that of a primary brain
stem tumor
McGuinness 2000; Schoeman et al. 1988; Cremin and
Jamieson 1996; De Castro et al 1995; Jinkins et al.
1995; Gupta et al. 1988; Salgado et al. 1989; Shen et
al. 1990; Draouat et al. 1987; Vengsarkar et al. 1986;
Van Dyke 1988; ,Jinkins et al.1987; Al Deeb et al.1992;
Gupta et al.1991; Rajshekhar and Chandy 1997).
All age groups are affected. In developing coun-
tries, tuberculomas show a predilection for children,
Kanamalla et al. 2000; McGuinness 2000; Atlas 1991;
Gupta et al. 1988; Salgado et al. 1989; Vengsarkar et al.
1986; Van Dyke 1988; Jinkins et al. 1987; Abugali et al.
1994). In developed countries, the elderly, as well as
patients with a susceptibility, or those suffering from
diabetics, alcoholicism, drug-abuse, or AIDS, are likely
to be affected, but exceptions occur and even infants
may develop tuberculomas (Vallejo et al. 1994). The
onset of symptoms is insidious and gradual eleva-
tion of intracranial pressure, focal epilepsy, and focal
neurologic deficits are the common presenting signs.
Fever may be present. Clinically, differentiation from
tumor, fungus, or paracytic disease is difficult, and
this is also a problem radiologically. Many radiolo-
gists have expressed the opinion that the radiologic
diagnosis of tuberculoma is always tentative. Recent
ances in which the probability of the diagnosis of
parenchymal tuberculoma is extremely high, espe-
cially in areas where tuberculosis is endemic. CT and
MRI studies carried out without contrast agents are
Imaging of Brain and Spinal Cord Tuberculosis
Fig. 33.10. Precontrast CT demonstrates an isodense right
frontal, subcortical, tuberculous mass (arrow). There is a sur-
rounding zone of edema and a pressure effect on the lateral
ventricle (arrow head)
likely to overlook both isodense parenchymal tuber-
culomas and TBM (Fig. 33.10). In cases of suspected
tuberculoma, a careful search for an active focus of
infection in other systems should be made. For this,
sputum testing, gastric washings, CSF study, and
chest radiography (including lateral views in children
(Cremin and Jamieson 1996), should be carried out.
Abdominal ultrasound or CT will demonstrate peri-
toneal tuberculosis if present, and in women, pelvic
disease can be the source. Urine bacteriology is also
necessary. However in some patients, if biopsy of the
tumor is thought inadvisable, the response to antitu-
berculous therapy will be the only way to make the
diagnosis (Jinkins 1991; McGuinness 2000; Schoeman
et al. 1988; Cremin and Jamieson 1996; Gupta et al.
1988; Gupta et al. 1990; Gupta et al. 1991).
33.6.2
Imaging Characteristics ofTuberculomas
In the early phase, tuberculomas are solid granu-
lomas and are slightly hypodense or isodense to
normal brain tissue on NECT. After contrast these
solid lesions, which are characteristically round, or
occasionally oval or lobulated, enhance homoge-
neously on CECT. When first discovered, tubercu-
lomas are usually about 2 cm in diameter and may
be single or multiple. There is accompanying low
density white matter edema, which may be intense
557
in these early lesions. The resulting mass effect
may displace the ventricular system and give rise
to hydrocephalus if foramina are compromised.
This is more likely if the lesion lies in the mesen-
cephalon or in the brain stem. Supratentorial sites
are commoner.
As central caseation and encapsulation develops,
the pattern of enhancement on CECT becomes het-
erogenous centrally and the outer capsule enhances
as a ring. A cerebral edema surrounding these
caseating lesions is usually less intense than is the
case in the solid granulomas. On angiography and
dynamic CT, they are centrally avascular. Tuber-
culomas discovered at this stage are often up to
5 cm in diameter but in endemic regions have been
demonstrated up to 8 cm in size, and these may
show a laminated pattern, resulting from alternat-
ing phases of granuloma formation and caseation
(Gupta et al. 1988, Gupta et al. 1990). At the next
stage, central necrosis of the tuberculoma develops.
The fluid or semifluid center of the mass does not
enhance on CECT (Fig. 33.11a-c). Characteristi-
cally, a dense ring of enhancing tissue surrounds
a hypodense core. This ring comprises glial tissue
and the compressed collagen capsule of the tuber-
culoma. Jinkins suggests that from the very start,
new fragile blood vessels develop in this zone, and
the passage of contrast through leaky walls allows
the blood-brain barrier to be breached and the
granuloma to enhance, either totally, or at a later
stage peripherally (Jinkins et al.1987).
Enhancement is a rough index of activity of the
lesion, and in the early stages of the disease may
increase in intensity, however with inception of
treatment it eventually wanes. Involution of the
mass during treatment is a slow process that may
take months or years, although some tuberculomas
exhibit a reduction in size in as few as 4-6 weeks after
commencement of treatment. Other cases may take
months or even years to resolve. Any increase in size
during treatment raises the question of drug resis-
tance, misdiagnosis, or paradoxical enlargement.
Some small lesions may resolve without therapy
(Vengsarkar et al. 1986), but in countries where cys-
ticercosis is also endemic, the true diagnosis in these
cases is often in doubt (Gupta et al.1990).
In children, Cremin and Jaimeson describe grape-
like clusters of tuberculomas that may be associated
with solid lesions elsewhere (Schoeman et al. 1988;
Cremin and Jamieson 1996).
In ring enhancing lesions the thickness of the
ring may vary around the circumference. Usually it
is continuous but when broken is difficult to differ-
558
Fig. 33.11. a Precontrast CT reveals an isodense, right occipital mass (arrows). b After contrast injection, there is strong ring
enhancement around a caseating or necrotic center (arrow head). c Postcontrast CT demonstrates ring enhancement of solid
tuberculomas in the right posterior cerebral white matter (arrows)
entiate from metastatic disease (De Castro et al. 1995,
Vengsarkar et al. 1986). Multiple lesions often show
different stages of development and difference in size.
Large, single lesions are difficult to differentiate from
neoplasms and lymphoma, especially when situated
deep in the midbrain, the brain stem, or cerebellum.
DSA demonstrating avascularity is considered as
helpful in these cases (Jinkins et al.1987).
The MRI appearances of tuberculomas are
described in the literature, but with wide variations
in appearance. Many of these descriptions were made
before the availability of contrast studies. Later stud-
ies are more consistent. That of Jinkins et al. is the
standard classification (Jinkins et al. 1995).
1. The noncaseating granuloma is Tl hypointense
to brain tissue and T2 hyperintense, enhancing
homogeneously with postgadolinium contrast
study. (Fig. 33.12a-d)
2. A solid caseating granuloma is hypointense to
isointense on Tl images and isointense to hypoin-
tense on T2, with a hypointense rim, depending on
the degree of capsule development. With contrast
examination a strong rim enhancement is seen.
Both these types of lesion are surrounded by a zone
of edema, which is hypointense on Tl and hyper-
intense on T2, remaining unchanged in appearance
on postgadolinium examination. (Fig. 33.13)
3. When central necrosis or liquefaction occurs,
the central signal is Tl hypointense. On T2, the
necrotic material gives a fairly homogeneous
hyperintense signal, and the surrounding capsule
appears hypointense. Tl-weighted postgadolin-
ium MRI study reveals intense ring enhancement
of the lesion.
F. McGuiness
In those cases of long-standing laminated lesions
as yet no postcontrast studies have been published,
so the enhancing characteristics are not yet known.
These lesions show alternating bands of iso- and
hypointense tissues on Ti, and of hypo- and hyper-
intense tissue on T2 (Gupta et al. 1990). The reason
for shortening of the T2 signal in some tuberculomas
is not clear, but may be the result of the presence of
paramagnetic free radicals in enclosed macrophages,
distributed inhomogeneously throughout the lesion
(Gupta et al. 1991). There is a general agreement that
the imaging appearances of intracranial tuberculo-
mas are nonspecific and that the ability of tuber-
culous lesions to mimic other diseases of pyogenic,
fungal, neoplastic, and paracytic origin leads to a
wide range both of appearances and of differen-
tial diagnoses (Jinkins 1991; Kanamalla et al. 2000;
McGuinness 2000; Schoeman et al.1988; Cremin and
Jamieson 1996; De Castro et al 1995; Jinkins et al.
1995; Gupta et al. 1988; Jinkins et al. 1987; Al Deeb et
al.1992; Bhargava and Tandon 1980).
Wechman suggested that one particular CT pat-
tern was specific to tuberculoma and was supported
by van Dyke (Wechman 1979; van Dyke 1988). The
pattern which they named the "target sign" com-
prises a round or oval lesion, isodense or slightly
hyperdense on NECT, containing a small central
nidus of calcification. With examination by CECT,
a peripheral enhancing ring was seen as well as the
central calcific lesion (Fig. 33.14a, b). But neither
author had access to MRI study. Van Dyke noted the
appearance in 10 of 30 patients with tuberculomas,
all of whom were black South Africans. The appear-
ance has only rarely been described in other parts
Imaging of Brain and Spinal Cord Tuberculosis
Fig. 33.13. An axial nonenhanced,
T2-weighted MRI of an occipital
tuberculoma of mixed signals (arrow).
There is extensive surrounding
edema (arrow heads)
559
Fig. 33.12. a Axial, postgadolinium
MRI. A group of right-sided
thalamic tuberculomas. These
are solid and the right lateral
ventricle is effaced (arrow). Note
also the extensive enhancement
of the thickened meninges over
the left convexity, in this young
patient with tuberculous menin-
gitis. b Coronal, postgadolinium
MRI. There is an irregular, ovoid
tuberculoma in the right side of
the brain stem (arrows). c A sagit-
tal, postgadolinium study of a
patient with three tuberculomas.
All enhance but show some central
relative hypodensity. The lesions
are seen in the parasagittal, cer-
ebellar, and brain-stem locations
(arrows). d An axial, postgado-
linium MRI scan reveals a single
enhancing tuberculoma high in
the brain stem
Fig. 33.14a, b. A case of tuberculoma. a A lesion with density slightly higher
than the surrounding brain tissue. Note the central calcification, as well as the
mass effect and the dilatation of the posterior horn of the lateral ventricle. b
After injection of contrast medium, showing intense enhancement of the rim
of the lesion, and the typical appearance of the "target sign"
560
of the world, namely North Africa, India, and Saudi
Arabia (Draouat et al. 1987; Vengsarkar et al. 1986;
Abduljabbar 1991). Recent studies by Bargallo et al.
in Spain, augmented by MRI, have shown the sign
present in cases of AIDS, complicated by intracranial
toxoplasmosis in one case and lymphoma in another.
In a third case it was seen in an elderly woman with
a pyogenic brain abscess (Bargallo et al. 1996). In
van Dykes' cases, the sign may be due to a specific
response of a local population to tuberculosis. Other
factors must also be considered. The presence of
central calcification suggests a long-term lesion, so
delay in presentation at a hospital may be one factor.
However, peripheral enhancement indicates an active
lesion. The target sign is therefore a suggestive but
nonspecific indicator of a tuberculoma.
33.7
Paradoxical Response of Tuberculomas
to Treatment
Since first described by Lees, McLeod, and Marshall
in 1980 (64), a number of cases have been cited where
tuberculomas have developed or increased in size
during apparently adequate antituberculosis therapy
of nonresistant tuberculous organisms. None of these
patients were of Caucasian origin, but included Indi-
ans, Chinese, Vietnamese, and one North American
black (Teoh et al. 1987).
The reasons for this paradoxical phenomenon have
not yet been discovered. Some patients have associated
tuberculous lymph node disease, and the lymph nodes
also increase in size. This last feature is well recognized
during antituberculous therapy, a possible explana-
tion being a local hypersensitivity and inflammatory
response, resulting from the release of tuberculopro-
tein from dead or dying mycobacteria (Campbell and
Dyson 1971). However, while infected lymph nodes
contain large numbers of mycobacteria, this is not the
case in parenchymal cerebral lesions, where the num-
bers of mycobacteria are small.
The relapse time from the apparently successful
treatment of the tuberculoma and the onset of new
focal signs,convulsions, or raised intracranial pressure,
due to paradoxical expansion, varies between 1 month
and 18 months. Patients have come from a wide range
of age groups, from infants to the elderly (Teoh et aI.
1987). The phenomenon occurs both in parenchymal
tuberculoma and in tuberculoma-complicating TBM
and may throw the diagnosis into doubt. In some
F. McGuiness
cases, biopsy of the lesion may be necessary to obtain
histologic confirmation of the diagnosis. Brain biopsy
or surgery are not without danger in tuberculosis
and ideally should be reserved for those cases where
intervention is required due to the effects of raised
intracranial pressure or where malignancy, fungal,
or paracytic disease cannot otherwise be eliminated
from the diagnosis (Vengsarkar et al. 1986; Bouchama
et al.1991).
33.7.1
Late Stage Appearances ofTuberculomas
The behavior of parenchymal tuberculomas during
antituberculous therapy is related to the size of the
lesion at the time of initial diagnosis. Small, single
or multiple lesions of less than 1 em in diameter
usually disappear completely during therapy, often
within the first few months and almost invariably
within 1 year. This is the norm in children (Schoe-
man et al. 1988; Cremin and Jamieson 1996). This is
the case if the lesion is a solid granuloma, without
central caseation or calcification. Larger nonsolid
or calcifying lesions above 2 em in diameter usually
take between 2 and 3 years to resolve, especially if
the lesion is both large and lobulated, as in those
described by Bhargava (Bhargava and Tandon
1980). Jinkins (Jinkins et al. 1987) in his study
of 80 patients with parenchymal tuberculomas,
described 57 patients with isolated granulomatous
lesions and followed them up with CECT until
therapy could be safely stopped. He took failure to
enhance on CT study as an indication of healing.
Many of these cases required between 18 months
and 2 years therapy before resolution. Fourteen of
57 eventually showed a normal CT. Eight showed
a residual, nonenhancing, calcified lesion, and 12
displayed calcification with associated overlying
cerebral atrophy. Seventeen patients showed focal
cerebral atrophy alone. One patient continued to
demonstrate an enhancing lesion after 2 years of
chemotherapy. Five patients were lost to follow-
up. This is a higher percentage of calcified lesions
than in some other studies. Wilkinson (Bouchama
et al. 1991) and Reed (Wilkinson et al. 1971) in two
separate series stated that 38% of TBM patients
showed late stage calcification, but that only 1%
and 6% of parenchymal tuberculoma developed
this phenomenon in their respective studies. This
is in agreement with van Dyke, only one of whose
30 cases developed significant late stage calcification
(Fig. 33.15a, b) (van Dyke 1988).
 Imaging of Brain and Spinal Cord Tuberculosis 561

a b

Fig. 33.15a, b. Pre- and postcontrast computer scans after craniotomy. a There is extensive cortical and white matter calcifica-
tion and some cortical atrophy (arrows). b Postcontrast scanning shows a number of areas of enhancement within the lesion
(arrows). These areas represent continued activity in this tuberculoma some months after commencement of antituberculous
chemotherapy

33.8 from pyogenic abscesses. They show similar MRI

Tuberculous Brain Abscesses
Although uncommon, true tuberculous abscesses
develop from parenchymal tuberculous granulomas
or the spread of tuberculous foci in the meninges to
the brain substance in patients with TBM. Cremin
and Jaimeson have pointed out that two distinct types
of necrosis occur in tuberculomas. Microscopically,
those possessing a structure of fibrovascular ele-
ments, as demonstrated with reticulin stains, undergo
gummatous necrosis, where the inflammatory granu-
latomatous tissue undergoes necrosis. Cremin points
out that this type of central necrosis gives an MRI
signal on T2 studies that is isodense to brain tissue
and surrounded by a dense zone of edema. Tuber-
culomas of this type do not develop into abscesses
(Schoeman et al. 1988,Cremin and Jamieson 1996).
Those tuberculomas that do convert into abscesses
show a different structure on microscopy as they
have no reticulin elements, being composed entirely
of cellular elements. The necrosis of these epitheli-
oid cells, macrophages, and polymorphonucleocytes
passes through a phase of inspissation or caseation
to liquefaction and the development of a true abscess
(Cremin and Jamieson 1996).
These are usually isolated lesions, occasionally
occurring simultaneously with other solid tubercu-
lomas. Radiographically they are indistinguishable
characteristics, being oval or round in shape with
a thin, strongly enhancing wall that, on T1 Gd MRI
examination, is in marked contrast to the hypodense
liquid center (Fig. 33.16a-c).
A point of clinical importance is that, while solid
or caseating tuberculomas contain few bacteria, the
tuberculous abscess contains the mycobacterium in
large numbers. If such an abscess ruptures into the
ventricular system or the subarachnoid space, a dev-
astating ventriculitis or meningitis ensues. Drainage
of these lesions is therefore a dangerous procedure.
Seeding of daughter lesions along the needle track
has been described by De Castro (De Castro et al
1995) who points out that in the West, abscesses are
more likely to develop in the elderly, or in the immu-
nosuppressed.
33.9
Atypical Tuberculous Masses
Tuberculoma 'en plaque' lesions occur more com-
monly as a complication of TBM than as isolated
tuberculomas. En plaque lesions may also occur
when a tuberculoma abuts onto the meningeal sur-
face of the brain and a proliferation of granuloma-
tous tissue extends into the adjacent meninx. When
 562 F. McGuiness

a b

Fig. 33.16. a A group of right-frontal tuberculous abscesses in the subcortical

white matter, causing midline shift. Postgadolinium imaging demonstrates an
intense enhancement of the abscess wall (arrow), while the liquid contents
remain of low signal (arrow head). b Two deep, midline tuberculous abscesses
in the region of the tentorium (arrows). There is a smaller collection posteriorly
(arrow head). c Two right, frontal tuberculous abscesses, one of which has been
sucessfully drained (arrow). The second shows the characteristic appearance
of an enhancing thin wall surrounding hypodense fluid. (arrow head). There is
hydrocephalus. This was in a late stage case of tuberculous meningitis

present as isolated lesions, the clinical presentation
is similar to any other space-occupying lesion (Gee
et al. 1992).
A common site of such lesions is the tentorial
edge, where the resemblence to meningioma causes
difficulty in diagnosis (Fig. 33.17a). The lack of vas-
cularity on DSA is a helpful pointer. Other primary
and secondary tumors may be differentiated in the
same way, but stereotactic biopsy will be necessary
in some cases (Bouchama et al.1991; Rajshekhar and
Chandy 1997).
When tuberculous meningeal mass lesions are
confluent with the cerebral cortex, there is extensive
gyral enhancement on both CECT (Fig. 33.17b, c) and
TlGd images, the underlying cortex and white matter
show edema beneath the lesion. This edema will be
hypodense on NECT and Tl MRI, and hyperintense on
T2, showing no enhancement on Tl Gd acquisitions.
Other atypical tuberculomas are exceedingly
rare. Graveli (Graveli et al. 1998) describes a lesion
of the cavernous sinus, with the characteristics of
a meningioma. Shah et al. treated a patient with a
tuberculoma of the sphenoid sinus showing bone
destruction of the superior section of the clivus and a
soft tissue mass, and a second patient who presented
with a tuberculoma in the brain as a mass in the tectal
plate (Shah et al.1993; Shah et al.1994).
33.10
Miliary Cerebral Tuberculomas
Miliary tuberculosis is a multiorgan infection which
develops when the host defenses prove inadequate
in controlling the generalized spread of the disease.
It is characterized by Widespread 2-3-mm granulo-
mas, typically seen throughout the lung fields, and
histologically present in other organs. However, on
presentation only 50% of cases have an abnormal
 Imaging of Brain and Spinal Cord Tuberculosis 563

Fig. 33.17. a An enhancing en plaque tuberculoma at the tentorial edge (arrow). Difficult to differentiate from a meningioma. b, c
Pre- and postcontrast CT scans demonstrate a hyperdense lesion in the frontoparietal region. After intravenous contrast injection,
there is extensive enhancement of the adjacent meninges. This is associated with ring enhancement of a I-cm tuberculoma with a
central, hypodense nidus (arrows). Images courtesy of Dr Ng Chang Gung, Memorial Hospital, Kwei Shan, Tao Yuan, Taiwan

chest radiograph. At autopsy between 47% and 55% 33.11

of cases are described with cerebral involvement
(Slavin et ai. 1980).
In children, 68% of those with tuberculous men-
ingitis are reported to show concomitant miliary
lesions in the brain (Jinkins 1991; Gee et ai. 1992).
Gupta pointed out that cerebral miliary tuberculosis
may be more common than previously realized, and
found typical lesions in seven patients with pulmo-
nary miliary tuberculosis, who had no symptoms or
signs of central nervous system disease (Gupta et ai.
1997).
Small solid granulomatous lesions are isodense
on NEeT and isointense on T1 images. On T2, these
small lesions give a high signal. Larger caseating
lesions are revealed as low signal on T2. After con-
trast injection, small lesions enhance solidly whilst
caseating lesions show ring enhancement. Gupta
pointed out that the majority of lesions occur at the
gray/white matter interface. This suggests embolic
hematologic spread, and arrest at the site of nar-
rowing of cortical arterioles. A similar mechanism
influences the distribution of cerebral metastases
(Atlas 1991).
Tuberculosis of the Calvarium with CNS
Lesions
Tuberculosis of the calvarium is a rare condition
but in a small proportion of reported cases the
inner table is penetrated to give rise to a tubercu-
lous epidural abscess. This in turn has led to focal
tuberculosis in the underlying cerebrum. There are
three distinct types of lesion. The most common is a
single osteolytic lesion in the skull vault. This is usu-
ally a well-defined, punched-out defect with smooth
but occasionally irregular margins. A thin border of
sclerotic bone has been described (McGuinness 2000;
Dastur 1972). The inner table is first involved, and the
inner defect is often more extensive that the outer
(Le Roux et ai.1990). The central fragment of bone is
often sequestrated, a button sequestrum, and there is
painless overlying soft tissue swelling, while in some
cases a cutaneous sinus may develop. The bone defect
may cross the suture line. A less frequent appearance
is of a more diffuse lesion with ill-defined margins.
In both types of lesion there may be a surrounding
zone of osteoporosis. Some authors have noted that
 564
previous trauma had occurred at the site of the lesion
(Le Roux et al.1990). The second type of presentation
is common in India and Africa. Multiple bone defects
are seen, and these may be associated with cystic
bone lesions elsewhere in the long bones of the skel-
eton or with dactylitis (McGuinness 2000; Cremin
and Jamieson 1996; De Castro et al1995; Le Roux et
al.1990; Wessels et al.1998). In the third presentation,
known as "Pott's puffy tumor:' the scalp swelling is
the major component and the underlying button
sequestrum and calvarial osteomyelitis are only
revealed on skull radiography or CT (Fig. 33.l8a,b)
(McGuinness 2000; Le Roux et al. 1990).
a the socially deprived, and in Natal is an important
b for many years. The condition is usually painless.
Fig. 33.18. a Pott's puffy tumor. CT study of a soft tissue lesion
in the right frontal region demonstrates an extensive swelling,
isodense to brain tissue. There is thinning of the bone of the
underlying calvarium. b Postgadolinium MRI. This defines the
subcutaneous abscess as a low signal collection, surrounded
by a ring of enhancing tissue. The tuberculous osteitis of the
frontal bone and an underlying epidural inflammatory lesion
are demonstrated
F. McGuiness
In cases where underlying infection occurs within
the skull or cerebrum, a wide range of clinical pat-
terns have been described (Le Roux et al. 1990). The
common complication is seizure, but diabetes insipi-
dus, vision disturbance, 3rd and 6th nerve palsies,
mild hemiparesis, and proptosis have been described
(McGuinness 2000; Cremin and Jamieson 1996; De
Castro et al 1995; Le Roux et al. 1990; Patankar et al.
2000; Gupta et al.1989).
33.12
Tuberculous Otitis Media
and Tuberculosis of the Temporal Bone
In the pre-chemotherapy era, tuberculous otitis
media was common. In 1915, Turner commented that
50% of cases of otitis media in infants were tubercu-
lous, the figure falling to 2% in adolescents. In the
post-1950s period, it has become rare in the West,
and although sporadic cases occur in Caucasians,
most cases are found in immigrant communities
(McGuinness 2000; De Castro et al 1995; Buchanan
and Rainer 1988; Cavallin and Muren 2000).
In South Africa, the condition is common among
cause of facial palsy in children. In Singh's 43 patients,
17 developed lower motor neurone 7th nerve palsy
(Singh 1991).
In young African patients, 50% have pulmonary
tuberculosis, and otitis media may be part of a more
widespread central nervous system disease, while
hematogenous spread may have occurred from foci in
the liver, kidney, or spine. Often the infection in the ear
is bilateral. In the West, isolated unilateral otitis is more
common. Target groups with immune problems may
be infected. One of Lee and Drysdale's cases had been
treated with steroids for rheumatoid arthritis (Lee and
Drysdale 1993).
Although rare in the West, if not treated, it remains a
chronic, disagreeable condition that leads to deafness
in the affected ear. Complications, such as mastoiditis
with sinus formation, bony sequestration, and intra-
cranial infection can arise. Profuse otorrhea continues
On clinical examination, the drum is seen to be
perforated or partially destroyed. Visually the drum
and mucosa are pale, as are granulations when
present. The middle ear fills with caseous material,
which has the appearance of cottage cheese. Biopsy
of the granulations reveals caseating granulomas,
and immediate screening sometimes reveals AAFB.
Imaging of Brain and Spinal Cord Tuberculosis
It is essential to send material for culture as almost
invariably a misleading secondary pyogenic infection
will have occurred (McGuinness 2000; Buchanan and
Rainer 1988).
33.12.1
Imaging
Plain radiographs show clouding of the mastoid cells
or osteoporosis and bone destruction in the middle
ear. These nonspecific changes are confirmed by con-
ventional tomography. Axial imaging where available
is much superior. Standard middle ear protocols reveal
bone destruction and soft tissue masses. And destruc-
tion of the ossicles may be documented (Hoshino et
al. 1994). On MRI studies postcontrast TlGd images
demonstrate the extent of the inflammatory change.
Granulation tissue enhances and caseating tissue
remains isodense, or hypodense intracranial extension
of the disease is shown (Fig. 33.7a, b). The common
intracranial complications are tuberculomata, often
of the temporal lobe, meningitis, and neural deficits
in nearby cranial nerves (VI, VII, VIII, XI, and XII).
Grewal et al. describe a tuberculoma of the mastoid
bone (Grewal et al. 1995).
Differentiation from other chronic conditions
depends on culture and biopsy. Of infections, chronic
otomastoid cholesteatoma and fungal granulomatous
diseases may show similar appearances. Wegener's
granulomatosis, Langerhans cell histiocytosis, rhab-
domyosarcoma, and recurrent bacterial infections
may all produce similar appearances (Mumtaz et al.).
33.13
Differential Diagnosis of TBM
and Parenchymal Tuberculosis
It is clear from the radiologic literature that there
are no specific CT or MRI findings in either TBM or
parenchymal cerebral tuberculosis, and that the diag-
nosis rests on a combination of clinical assessment,
imaging appearance, and response to therapy. A wide
range of pathologic processes, including malignancy,
pyogenic and fungal infections, paracytic infestations,
and a range of other disconnected conditions, ranging
from trauma to vascular malformation and aneurysm,
can mimic either TBM or parenchymal tuberculosis
(Jinkins 1991; Kanamalla et al. 2000; McGuinness
2000; Cremin and Jamieson 1996; Shah 2000; Wallace
et al. 1991; De Castro et al 1995; Jinkins et al. 1995;
565
Rochas-Echeverri et al. 1996; Cross et al. 1995; Atlas
1991; Bargallo et al. 1996; Ng et al. 1996; Rajshekhar
and Chandy 1997; Graveli et al.1998; Shah et al.1993;
Le Roux et al. 1990).
Basal meningitis with infarctions is commonly
seen in Haemophilus injluenzae infection. Infarc-
tions are also common in aspergillosis and mucor-
mycosis, where direct spread from the paranasal
sinuses is usual. In these cases, infarctions follow
extension of the condition though vessel walls, and
are seen in the cortex and subcortex, but major vessel
infarctions also occur. In disseminated cerebral coc-
cidioidomycosis (DCC), intense basal meningeal
enhancement involving the sylvian systems is seen,
and is often more intense than that of TBM. However,
cases where basal enhancement is absent, even in the
presence of hydrocephalus, occur in both conditions
(McGuinness 2000; Schoeman et al. 1988; Dublin and
Phillips 1980). In DCC, the basal ganglia and white
matter lesions are more diffuse than in TBM, and
focal granulomas and basal ganglia infarctions are
rare. Ventriculitis is also an early finding in DCC, and
a late stage development in TBM (McGuinness 2000;
Dublin and Phillips 1980). Geographically the distri-
bution areas of DCC, histoplasmosis, blastomycosis,
and TB overlap, giving rise to problems of differen-
tiation (Ogawa et al.1987).
In cryptococcus neoformans infection, basal
meningeal enhancement, arteritis, and basal ganglia
infarctions are seen. But in cryptococcal infection the
organism can usually be isolated from the CSF and
recognized using rapid laboratory techniques.
In spirochetal disease, basal meningitis can develop
together with parenchymal granulomatous gummas.
In syphilitic infection these are often attached to the
meninges, and associated cerebral atrophy and basal
fibrous nonenhancing pachymeningitis is another
diagnostic clue, as well as laboratory findings. Tick-
borne Lyme disease is another spirochetal infection that
may affect the cerebrum. Granulomas in the basal gan-
glia can be present, but unlike those of tuberculosis are
diffuse, and more solid lesions lack ring enhancement.
Contrast enhancement of the CSF, on Gd MRI, has
been described in both syphilis and Lyme disease,
as well as in cryptococcal infection (Good and Jager
2000, Sakamoto et al. 1997). Abscess formation in
bacterial, mycobacterial, and fungal disease cannot
be differentiated radiologically. Bacteriologic or
histologic examination, as well as the response to
medical treatment, is the only means of differentiat-
ing among them.
Single or multiple malignant tumors may mimic
tuberculomas. The ring enhancement of tumors is
566
often of varying thickness, and the ring is sometimes
incomplete. The surrounding zone of edema tends to
be less intense than in tuberculoma. With DSA, tuber-
culomas are avascular (Jinkins 1991; McGuinness
2000; Rochas-Echeverri et al.1996; Gupta et al.1991).
Leptomeningitis carcinomatosis shows a patchy
enhancement of the basal meninges (McGuinness
2000; Kioumehr et al. 1995).
Primary lymphomas tend to be periventricular in
position, and may cross the midline, passing through
the corpus callosum, unlike supratentorial tuber-
culomas. Lymphomas have limited mass effect, are
structurally diffuse, and show little or no edema and
no ring enhancement (McGuinness 2000). Second-
ary lymphomas commonly affect the meninges, in
contrast to the usual subcortical position of multiple
tuberculomas. Single en plaque tuberculomas can
cause some confusion, but enhance intensely, and
have to be differentiated from meningiomas. They
are avascular structures on DSA and MRA.
Cysticercosis (CC) and TB have similar geographi-
cal distributions. CC lesions are distributed through-
out the body and the diagnosis is often made from
the biopsy of subcutaneous lesions.
In cerebral CC, the various radiologic appear-
ances of developing and degenerating cysts are
described in the literature (Rajshekar et al.1993; Jena
et al. 1988). The mural node of the living cyst can be
defined by MRI, but may be difficult to define on CT
studies. After death of the parasite, an intense local
reaction and ring enhancement on T1 Gd studies are
seen, with a central low signal area. Punctate calcifi-
cation is often present on CT, while on MRI the vari-
ous stages of development and death of the cyst are
seen more clearly. At all stages, however differentia-
tion is difficult and a trial of drugs may be necessary
to make the diagnosis (Rajshekar et al. 1993; Jena et
al. 1988; Rajshekhar and Chandy 1996). Rajshekhar
points out that in a series of 16 patients examined
by both CECT and T1 Gd MRI revealing a solitary CC
granuloma, no additional information was acquired
from the MRI studies. Cost plays an important role
in the management of the generally underprivileged
patients presenting with CC disease. He sees thin-slice
contrast-enhanced CT as the examination of choice
(Fig. 33.19a-d) (Rajshekhar and Chandy 1996).
Toxoplasmosis produces intracerebral pathology
that may be confused with tuberculoma. Toxoplas-
mosis is the most common infestation in AIDS,
where tuberculosis may coexist. Cystic solid or ring
enhancing lesions are described, often with a central
enhancing nodule (McGuinness 2000; Atlas 1991).
The distribution of the lesions in the subcortex
F. McGuiness
and deep white matter is similar in both condi-
tions. However, in toxoplasmosis, subependymal
lesions are common, the ring enhancement is usu-
ally thicker and more irregular than in tuberculoma
and hemorrhage into the lesion often occurs. Gupta
has described hemorrhage in some tuberculomas
analyzed by MRI and proton spectroscopy, but such
lesions are uncommon (Gupta et al. 1991). Antitoxo-
plasmic therapy often reduces the size of the lesions
in between 2 to 3 weeks, another way of differentiat-
ing them from tuberculomas.
Neurosarcoidosis occurs in 5% of sarcoidosis
cases and is usually a subacute condition, clinically
unlikely to be confused with TBM. However the men-
ingeal enhancement of sarcoidosis produces some
similarities to TMB. Subfrontal meningeal enhance-
ment in a patient who is not acutely ill is a common
manifestation, and the suprasellar cisterns, parasellar
spaces, and hypothalamic parenchyma are commonly
involved, which can result in diabetes insipidus. Neu-
rosarcoid changes are usually diffuse and show gyral
and subcortical enhancement, and extension along
the perivascular spaces. They are therefore likely to
be confused with tuberculous border-zone disease.
Parenchymal granulomas also occur. They can be
of wide distribution above and below the tentorium.
They are usually solid enhancing lesions and show
little surrounding edema (Fig. 33.20). Hydrocephalus
only occurs when periventricular infiltrating lesions
cause CSF obstruction. MRI examination is particu-
larly useful in demonstrating the spread of pathologic
granulomatous tissue along the vessels in the Vir-
chow-Robin spaces (Urback et al. 1997). When pres-
ent, destructive lesions of the calvarium are another
differential finding. Sarcoid lesions of the skull vault
offer a point of biopsy, as do sarcoid lesions in other
parts of the body.
Other granulomatous disorders are unlikely to be
confused with tuberculosis. Intracranial neurobrucel-
losis is a rare disorder even in areas where brucellosis
is endemic. The incidence varies from 4% to 10% in
reported series of cases of brucellosis (97). Men-
ingitis tends to be chronic and intermittent rather
than acute, although fever, neck stiffness, lethargy,
and impaired consciousness are described. As the
perineurium, as well as the vessel walls, are affected
cranial nerve lesions are common. Madkour notes
that sensorineural hearing loss may be the present-
ing sign, but the 6th, 7th, 2nd, and 5th cranial nerves
are also commonly affected. Meningeal enhancement
if present is mild and hydrocephalus uncommon.
However, deep white matter granulomas have been
seen (Fig. 33.21a, b). CT scans may be normal except
 Imaging of Brain and Spinal Cord Tuberculosis
a b
c apparent limits of the granuloma itself, in both cases
when mild hydrocephalus is present. MRI studies are
of help, because evidence of enhancement of cranial
nerves is more likely in brucellosis than in TBM
(Fig. 33.22a-d) and intracerebral lesions when pres-
ent are clearly imaged (Madkour 2001, Shakir 1986).
Other intracranial mass lesions that may be simi-
lar in appearance to tuberculoma include those of
tumefactive multiple sclerosis (TMS). In this condi-
tion, large, contrast-enhancing, white matter lesions,
surrounded by vasogenic edema are found. They are
usually low signal on Tl and high signal on T2, with
minimal ring enhancement on TlGd acquisitions.
The surrounding edema may be very extensive, and
567
---:z.;.o,-d
Fig. 33.19a-d. Coronal and sagittal contrast-enhanced MRI.
a· Coronal and b sagittal, contrast-enhanced MRI in two
patients with solitary cysticercus granulomas. Showing a
poorly circumscribed area of enhancement, lying beyond the
the clinical presentation will usually differentiate
TMS from single tuberculoma. The neurologic deficit
is generally more widespread, and there can be a his-
tory of previous acute spinal myelitis (Fig. 33.23a,b)
(Miller et al. 1989). Granulomatous masses in Can-
dida meningitis, a common complication of AIDS,
leukemia, and lymphoma, are generally extra-axial
and fail to enhance on Tl Gd images.
Cryptic angiomatous malformations presenting
as mass lesions, develop a layered appearance as a
result of local episodes of bleeding into the angioma.
The presence of methhemoglobin accounts for the
MRI appearance. In the rare cases of layered tuber-
 568 F. McGuiness

Fig. 33.20. Coronal Tl-weighted postgadolinium images. Multiple, enhancing, solid granulomatous lesions are seen in the cere-
brum and brain stem (arrows). The appearances are indistinguishable from multiple tuberculomas. The patient was a proven
case of multisystem sarcoidosis

a b

Fig. 33.21. a Postcontrast CT revealing periventricular enhancement (arrow) and edema (arrow heads). This was a case of cere-
bral brucellosis. b Diffuse, high signal, thalamic lesions in cerebral brucellosis
 Imaging of Brain and Spinal Cord Tuberculosis 569

a b

c d
Fig. 33.22a-d. A case of brucella polyneuritis in a child. a Contrast-enhanced Tl-weighted image. Swelling and contrast enhance-
ment of the prechiasmic portion of the right optic nerve (arrow) b Swelling of the left fifth cranial nerve and marked contrast
enhancement of the fifth cranial nerves and gasserian ganglion in Meckel's cave (arrows). c Marked contrast enhancement of
both seventh and eighth cranial nerves (arrows). d Coronal Tl-weighted image shows a well-defined, low signal, cystic mass
with peripheral contrast enhancement, anteromedial to the left seventh and eighth cranial nerves (arrow)

Fig. 33.23. a Sagittal, postgadolinium MRI demonstrates two cerebel-

lar tuberculomas (arrows). b A case of multiple sclerosis (MS). There
are similar lesions in the cerebrum but in MS they are slightly less
well defined (arrows); there is a poorly defined lesion in the cerebel-
a lum (arrow head)
570
culoma, MRI spectroscopy demonstrates that tuber-
culomas have low levels of iron and other metallic
elements (Gupta et al. 1991). Intracranial aneurysm
may be differentiated from tuberculoma by MRI or
DSA examinations. Where CT is the only available
modality, small aneurysms of the circle of Willis
have caused diagnostic problems (Gucuyener et al.
1993). Aneurysm is an uncommon complication of
tuberculous arteritis, but does occur and can lead to
intracranial hemorrhage, epistaxis, or bleeding from
the aural canal (Leiguarda et al. 1988, Cross et al.
1995). MRI will differentiate between tuberculomas
and aneurysms.
33.14
Tuberculous Radiculomyelopathy
and Myelitic Tuberculomas
The study of spinal cord tuberculosis has made great
strides since the advent of contrast-enhanced com-
puter tomography (CECT) and magnetic resonance
(MRI) studies. Before that conventional myelography
was limited, both in its scope, and by the nature of
the disease process (Chang et al. 1989).
Excessive glutinous exudates associated with
spinal meningitis impede the normal circulation of
CSF, making lumbar puncture and the introduction
of contrast agents difficult, and in the later stages
almost impossible. Postmortem studies show that
the exudates are very extensive, often filling the
subarachnoid space completely, and compressing the
cord and other spinal contents, with resulting cord
edema and ischemia. Cord and nerve root swelling,
as well as the development of granulomatous tissue,
rapidly produces spinal block (Dastur et al.1995).
In the late stages gliotic, collagen, and fibrotic
pathologic tissue forms, causing irreversible changes
in the theca and myelum.
If tuberculous meningitis (TBM) and tuberculous
spinal meningitis (TBSM) are present together, the
deficits in the peripheral nervous system may be
obscured by the low level of consciousness of the
patient. Acute onset of back pain, paraesthesia, muscu-
lar weakness and, sphincter dysfunction are common
features of TBSM. However, an insidious progressive
pattern also occurs, mimicking intraspinal tumor,
polyradiculopathy, or spinal demyelination. This type
has also been described as developing many years
after apparently resolved intracranial TBM (Chang et
al. 1989). In treated TBSM, inactive fibrotic and glial
tissue remain in the spinal canal, and combined with
F. McGuiness
cord cavitation result in permanent neurologic deficits
(McGuinness 2000).
In all cases, it is essential to investigate both the
brain and the spinal lesion, as concomitant intracra-
nial and intraspinal lesions are common (McGuinness
2000; Tandon 1978; Gupta et al.1997; Shen et al.1993).
In a small number of cases, CT or MRI at the level of a
spinal block will reveal tuberculous osteomyelitis of a
vertebral body or, rarely, an isolated tuberculous epi-
dural granuloma (McGuinness 2000).
Wadia and Dastur divided tuberculous radiculo-
myelopathy into two groups:
1. Primary, arising from a focus of tuberculosis out-
side the central nervous system (CNS).
2. Secondary, arising from intracranial TBM or from
spinal tuberculous osteomyelitis.
In a series of 70 cases, they found the primary type
to be more common (Wadia 1973; Wadia and Dastur
1969). The secondary type of TBSM is either the
result of downward spread ofTBM infection, through
the cerebrospinal fluid, or of the spreading outward
and upward from a focus of spinal osteomyelitis,
commonly in the lumbar or low thoracic regions.
More recently, Dastur has confirmed that 50% of
cases are of the primary type and over 30% follow the
pattern of spreading down from intracranial TBM
(Dastur et al. 1995). This is supported by the find-
ings of Gupta. Of 20 consecutive cases of intraspinal
tuberculosis, 75% were of the primary type and 25%
secondary to TBM. None of the cases had vertebral
tuberculous osteomyelitis (Gupta et al.1994).
Geographical and racial differences may affect the
mechanisms of spread, as Chang et aI., in a study of
13 cases, reported 85% associated with TBM, one case
resulting from tuberculous spinal osteomyelitis, and
only one primary case of an extramedullary tubercu-
loma arising low in the thoracic spine (Chang et al.
1989).As in TBM, the pressure effect of exudate in the
confines of the spinal canal, combined with arteritis,
gives rise to small vessel occlusion, with resulting
ischemic myelitis. Local hematologic spread results
in single or scattered intra-axial tuberculomas. The
development of granulomatous tissue in the exudates
produces thickening and deformity of the nerve root
theca which at a later stage is converted to fibrous
tissue. All elements of the theca and axial nervous
system can be involved in the infection.
Both leptomeningeal TB and tuberculous radicu-
lomyelopathy are descriptive terms in current use,
but as not all of the intraspinal elements need to be
simultaneously affected, the term spinal neurotuber-
culosis, used by Dastur is also applicable (Dastur et al.
1995) (Figs. 33.24, 33.25).
Imaging of Brain and Spinal Cord Tuberculosis
Fig.33.24. Sagittal Tl-weighted postgadolinium scan of the
thoracic spine. Markedly thickened spinal meninges and
thick enhancing exudates compress the thoracic spinal cord
and displace it anteriorly (arrows). A case of spinal tubercu-
lous meningitis
Fig. 33.25. A single tuberculous lesion expanding the cord at
the Th4 level. Ring enhancement with a central, lower signal
due to a tuberculoma
As in TBM, the spread of the infection along the
arachnoid-pia mater complex is rapid. Granulomas
develop on the leptomeningeal and cord surface,
leading to leptomeningeal thickening, and in the
areas of cord involvement, to fusiform expansion of
571
the cord. Pitting and excavation of the cord surface
often develops, involving the superficial neural tracts
in the area. Varying patterns of neurologic deficit
are present, because myelitis as well as nerve root
pathology underlies the sensory and motor fall-out.
The myelitic lesions are usually confined to short
lengths of the cord, with intervening normal seg-
ments, and some cases show a true transverse myeli-
tis. Cord edema involving considerable lengths of the
myelum is another pattern. Fusiform swelling of the
cord and eventually cavitation are described (Chang
et al.1989; McGuinness 2000; Gupta et al.1994; Kumar
et al. 1993).
Spinal block occurs commonly, and although it
may result from the development of granulomatous
tissue at any level in the spinal canal, its usual site
is at the level of the conus medullaris. As impedi-
ments to CSF flow develop, spinal puncture becomes
increasingly difficult and a dry tap may result. The
CSF becomes thick, at low pressure, and xantho-
chromic, due to a high protein content. Injecting of
intrathecal contrast agents is increasingly difficult
as granulomatous tissue advances and the flow of
contrast is impeded. In the absence of MRI facilities,
cervical myelography is sometimes necessary, but the
same problems of impeded flow can be encountered
in this area also. The clinical presentation of TBSM
is similar to a large number of conditions, ranging
from cord tumors to demyelinating disorders, and
polyneuropathies. Tuberculous myeloradiculopathy
should always be included in the differential diag-
nosis of spinal lesions, although rare, isolated TBSM
does occur from tuberculous spondylitis or epidural
granuloma, as well as that arising from TBM. Tuber-
culous disease in other body systems outside the ner-
vous system occurs in between 50% and 80% of cases
of TBSM (McGuinness 2000; Gupta et al. 1994), so the
absence of tuberculosis elsewhere in no way excludes
the possibility of tuberculous radiculomyelopathy.
33.15
Imaging Methods
in Spinal Neurotuberculosis
33.15.1
Plain Radiography
Plain radiography is of little direct help, except in
cases where TBSM is associated with tuberculous
spondylitis. However, as a method of exclusion of
other causes of spinal cord or nerve root pathology,
572
it can be rewarding. If arachnoiditis is due to the rem-
nants of oily contrast medium the cause is demon-
strated, as is the case in some patients with vertebral
neoplasm, metastasis, myeloma, or lymphoma.
33.15.2
Water-Soluble Contrast Myelography (WSCM)
CT and MRI are not available in the majority of hos-
pitals worldwide, and this is particularly the case in
many of the 24 countries with a high burden of tuber-
culosis (WHO. Global tuberculosis control. 2000). By
use of a basic radiography unit, myelography of the
spinal axial nervous system may be acheived, even
without a tilting table (McGuinnes 2000). Using
myelography, it is possible in the majority of cases to
define the lesion and the extent of the extramedullary
disease. The disadvantage is that the pia-arachnoid
and nerve root systems are well seen, but only the
surface of the myelum can be studied. Variations of
volume of the myelum are defined, but the presence
of myelitis, intramedullary granulation lesions, and
ischemic foci are not demonstrated. Also, it cannot be
ascertained whether the lesions of the subarachnoid
space and nerve roots are actively inflamed or rep-
resent scar tissue. As is the case in TBM, the changes
seen in TBSM are nonspecific. Bacterial and fungal
inflammatory diseases are mimicked, as are parasitic
disease, tumor, arteriovenous malformation, sarcoid-
osis, polyradiculopathies, and demyelating disease.
An attempt should be made to examine the full
length of the cord, because in TBSM, multiple lesions
are common, with areas of normal cord and theca
intervening. In cases of TBSM, a normal myelogram
is uncommon (Gupta et al. 1994). The usual patterns
of the disease are as follows:
1. Irregular filling of the subarachnoid space, due to
the presence of granulomatous exudates, granulo-
mas, and thickening of the dentate ligaments.
2. Thickening of the nerve roots of the cauda equina
and the paired nerve roots, particularly in the
lower thoracic and lumbar regions.
3. Partial or total extramedullary blockage at the
level of the conus medullaris or in the lower tho-
~ racic region. Blockage at higher levels, Th5, Thl-
C7, and C5, do occur, but are less common.
4. Long, vertical~ band-like filling defects, perhaps the
result of thickening of the anterior midline septum.
5. Variations in the dimensions of the myelum, due
to myelitis and cord edema. Most intramedullary
tuberculomas in meningitis are too small to cause
visible cord expansion.
F. McGuiness
6. Large mass lesions due to arachnoid or pial granu-
lomas, simulating extramedullary tumours. These,
in practice, are usually found posterior to the cord
and in the lower thoracic area.
7. Epidural spinal abscess, due to extension of tuber-
culous vertebral osteomyelitis or, rarely, isolated
epidural abscess without vertebral involvement.
8. Multiple filling defects in the contrast column,
either fine and Widespread or larger and coarser,
due to granulomatous lesions. Thecal granuloma-
tous tissue also gives rise to surface irregularities
and variations in the dimensions of the thecal sac
(Chang et al. 1989; McGuinness 2000; Gupta et al.
1994; Kumar et al. 1993).
On lumbar puncture in both TBM and TBSM in
the early stages, a polymorph-leucocytic response
may be found, suggesting a bacterial cause. Once
the immune system has been triggered, then the
picture changes to the characteristic findings of CSF
pleomorphism, raised protein, and low sugar levels.
In contrast to the appearances of spinal extramed-
ullary tumor, the margins of granulomatous tissue
are usually irregular and lack the smooth outline
of most tumors. Thickening of the elements of the
cauda equina is unlikely to be confused with the more
serpentine appearance of vascular malformations of
the spinal canal. Other granulomatous masses due to
syphilitic gummas, or neurosarcoid are confusing,
as well as those occurring in paracytic disease, such
as schistosomiasis. Similar appearances of coarse
filling defects may be the result of leptomeningeal
carcinomatosis or due to the lymphomas, but the
clinical findings outside the nervous system will
usually point to the correct diagnosis in these cases
(McGuinness 2000).
33.15.3
CT Combined with Water-Soluble Myelography
CT examination in spinal inflammatory disease has
the disadvantage that many images may be neces-
sary to pinpoint multilevel pathology. Sagittal recon-
structions are not of sufficient quality to define small
lesions within the myelum, and there is relatively
poor tissue differentiation between CSF, exudates,
and the myelum (McGuinnes 2000). Gross volume
changes of the myelum are recognized, and Chang
et al. describe a pear-shaped cross section of the cord,
in the lower thoracic region as being a common find-
ing in cases of TBSM (Chang et al. 1989). Improved
images are obtained by combining CT with contrast
Imaging of Brain and Spinal Cord Tuberculosis
myelography, although the disadvantage remains
that if multiple level lesions are present, they may lie
outside the area programed for examination. Small
intramedullary TB lesions are isodense or hypodense
on CT, and contrast examination adds little informa-
tion in such cases. However, extra-axial granulomas
may enhance, and if present epidural or paravertebral
abscess will be recognized by postcontrast studies.
33.15.4
Magnetic Resonance Imaging
In differentiating between the various tissues in the
spinal canal, and in its multiplanar image acquisi-
tion, MRI is excellent in defining the pathology of
spinal neurotuberculosis. This is especially so in
the area of the cauda equina and the lumbar nerve
roots (McGuinness 2000; Gupta et al. 1994; Kumar et
al. 1993; Ross 1987; Gero et al. 1991), although some
authors consider that water-soluble contrast CT has
advantages (Chang et a11998). However, there is no
doubt that in delineation of lesions within the cord,
gadolinium-enhanced Tl-weighted MRI studies are
superior to WSCM and to CT myelography.
Active tuberculous lesions appear isodense or
slightly hypodense on Tl images and either solid and
hyperintense, or hyperintense with central area of
low signal, on T2 imaging. After gadolinium-contrast
injection, there is intense enhancement of tuberculo-
mas, granulomatous tissue, and thecal inflammatory
lesions. On Tl Gd studies in the areas affected by the
disease, there is loss of differentiation between the
spinal cord, the CSF, and the spinal meninges.
These types oflesions are present either segmentally
or throughout the length of the spinal canal. There are
also variations in the caliber of the myelum, with thick-
ening due to cord edema, or narrowing following cord
compression. Areas of edema show a high signal on T2
images, but do not enhance after contrast injection.
Tuberculomas of the cord are commonly of increased
signal or isodense on T2. They are differentiated from
areas of edema by enhancement after gadolinium on
Tl acquisitions. There is no particular pattern of dis-
tribution of tuberculomas or extramedullary granulo-
mas, except that those lesions causing spinal block are
usually found in the lower thoracic or upper lumbar
regions. The MRI findings are not specific to the dis-
ease, and are to be seen in other infections, fungal dis-
eases, neoplasm, and paracytic infestations.
Isolated epidural granulomatous masses are
described by Gupta in the absence of both TBSM and
tuberculous spondylitis (Gupta et al. 1994). In MRI
573
studies of the cervical spine, areas of the brain stem
and medulla oblongata are invariably included in the
disease (Shen et al. 1993). Isodense tuberculomas, if
present will only be revealed on postcontrast studies.
In all cases of suspected spinal tuberculosis, both the
spine and the brain should be studied (McGuinness
2000; Gupta et al. 1997; Shen et al. 1993). Tubercu-
lous radiculopathy of the cauda equina is not well
demonstrated by noncontrast MRI. This is in contrast
to nonspecific lumbar arachnoiditis, where MRI has
been a useful tool (Ross 1987). In suspected lumbar
radiculopathy, Tl Gd studies are usually only carried
out in the sagittal plane. There is a case for including
a coronal, postgadolinium study in the protocol to
improve the definition of the cauda equina, the emerg-
ing nerve roots, and the associated pathologic changes
(McGuinness 2000).
WSCM is still a valuable method for imaging the
changes of both active tuberculous radiculopathy
and chronic adhesive tuberculous radiculopathy.
However, a normal WSCM does not exclude spinal
tuberculosis. Gupta describes five cases with normal
conventional myelography, where subsequent Tl Gd
studies revealed lesions in the myelum (Gupta et
al. 1994). Despite its limitations in failing to image
changes in the myelum, the superior definition of
long stretches of nerve roots makes WSCM an impor-
tant method of investigation, and one that is available
to all radiologists (McGuinness 2000).
Tl Gd MRI is important in the differentiation
of active tuberculous granulomatous tissue from
chronic fibrotic posttuberculous scarring, and in
separating edema in the myelum from tuberculoma.
Both fibrotic tissue and edema fail to enhance on
TlGd examination (Chang et al. 1989; McGuinness
2000; Gupta et al. 1994). Although uncommon in the
acute stage, syrinx formation and cavitation of cord
lesions is described (Chang et al. 1989; McGuinness
2000; Sanchez Pernaute et al. 1996). Cavitation is
more likely to develop in thoracic cord lesions than in
other areas (Fig. 33.26a-c). Loculation of CSF, result-
ing from local obliteration of the subarachnoid space,
is a common finding in both acute and suba<:ute pre-
sentations. Parenchymal lesions, depending on the
stage of development, appear isodertse or slightly
hypodense on unenhanced Tl images, and may not
be visualised. T2 images usually show homogeneous
low-signal appearances, but mixed-signal lesions
are also described (Wadia and Dastur 1969; lena et
al. 1991). In two cases of tuberculoma reported by
lena, noncontrast T2 images revealed low-signal
lesions (Fig. 33.27a, b). In Tl studies of intramedul-
lary tuberculoma, both solid enhancing and ring
 574 F. McGuiness

b c

Fig. 33.26. a A noncontrast study demonstrates expansion of the cord by an isodense lesion,
with an associated area of syringomyelia beneath it (arrows). b,c Sagittal and axial postcon-
trast scans show the ring enhancement of a tuberculoma (arrows). Images courtesy of Dr
a Pernaud, University Hospital, Maques de Valldecilla, Santander, Spain

Fig. 33.27a, b. An intramedullary, tuberculous lesion extending over three cervical segments (arrows), with a mixed signal on
Tl and a high signal on T2. Biopsy should be avoided in such cases
Imaging of Brain and Spinal Cord Tuberculosis 575

enhancing lesions are seen. The MRI changes may be 33.16

summarized as follows: Differential Diagnosis
Intramedullary
33.16.1
1. Cord edema or infarction
Spinal Arachnoiditis
2. Cord tuberculoma
3. Cord cavitation
Nerve root thickening and deformity confined to
Meningeal the lumbar region is characteristic of nonspecific
1. Fluid loculation spinal arachnoiditis (NSA). Once a common sequela
2. Obliteration of the subarachnoid space to the use of oily contrast media in myelography,
3. Meningeal enhancement (Fig. 33.28) it is now more likely to follow spinal surgery or
4. Adherent and thickened nerve roots spinal anesthesia. The thickened nerve roots are
5. Meningeal tuberculomas displaced posteriorly and may be adherent to the
6. Intradural extramedullary granulomatous dura, leading to the appearance of an ~mpty dural
mass lesions (Fig. 33.29) sac, although central clumping is also described.
Adherance leads to some degree of spinal block, and
contrast introduced above the level of the lesion on
WSCM produces a "dripping candle" appearance, as
contrast flows down through the thickened nerve
roots (Ross 1987). In TBSM, all elements of the
spinal canal contents are commonly involved in
the infection, so that the dura, spinal cord surface,
and nerve roots all show inflammatory changes, and
usually at a number of levels. In the acute phase, the
infection enhances on TlGd MRI. In the long stand-
ing fibrotic lesions, the tissues are nonenhancing as
is the case in NSA.
Enhancement of nerve roots occurs when the
blood-nerve root barrier breaks down and is
reported in subarachnoid hemorrhage, trauma,
ischemia, inflammatory radiculopathy, and demyela-
tion as well as axonal degeneration (Jinkins 1993). In
tuberculosis, the younger age groups are commonly
affected, while in NSA, cases tend to be between the
ages of 30 to 50 years. Hypertrophic interstitial poly-
Fig. 33.28. Sagittal Tl, postgadolinium scanning reveals local neuritis, a rare condition of the lumbar nerve roots
enhancement of the meninges and cauda equina (arrows) in in the young, is differentiated by the slow progress
a case of spinal tuberculous meningitis and lack of inflammatory change (McGuinness 2000;
Kumar et al.1993; Donovan Post et al.1990).

..
Fig. 33.29. Axial, midthoracic, postgadolinium scan. There
are two large paravertebral abscesses with ring enhancement
(arrows). The cord is displaced anteriorly by an epidural col-
lection (arrow head). There is tuberculous spondylitis in the
vertebral body and transverse process (curved arrow)
576
Fig.33.30. Sagittal scans of the thoracic spine demonstrate
irregular, low signal thickening of the meninges at a number
of levels (arrows). Bacteriologic study of the CSF confirmed
enterococcal meningitis
33.16.2
Infective Leptomeningitis
Isolated spinal meningitis is unusual, and the majority
ofspinal cases are secondary to spread from intracranial
meningeal infection. The most common agents are viral
or pyogenic (Fig. 33.30). The clinical presentation of the
meningitides, with fever, neck, back and radicular pain,
often linked with changes in mentation, sensory loss and
muscle weakness, leads to examination of the CSF. In
the acute stage of TBSM, there is a polymorphonuclear
cell reaction, with a slightly raised protein, suggestive
of pyogenic infection. The characteristic pleomorphism
and high protein levels of TBSM, develop only after the
triggering of the cell-mediated immune response. This
underlines the need for repeated CSF examinations.
Unfortunately the results of immediate bacteriologic
examinations are usually negative, although rapid cul-
ture methods are helpful, in centers where they are avail-
able (McGuinness 2000; Wadia 1973). In those cases
with coexistent or recent TBM, the onset ofspinal symp-
toms is virtually diagnostic of TBSM, and conventional
or CT myelography will demonstrate a supportive (but
nonspecific) picture (Chang et al. 1989; McGuinness
2000). TBSM is a disease of the young (WHO. Global
tuberculosis control 2000; Gupta et al. 1994; Gero et
al. 1991). In nongranulomatous infections, the mean
age is higher (Donovan Post et al. 1990). However, the
upsurge in tuberculous infection in older age groups
must also be considered.
MRI changes of dural thickening, linear enhance-
ment of the cord surface, and nodular enhancement
F. McGuiness
of the nerve roots of the cervical, thoracic, and lumbar
regions are common to pyogenic meningitis, dis-
seminated coccidioidomycosis and syphilis (Gero
et al. 1991; Sharif 1992; Phillips et al. 1990). These
diagnoses are rapidly confirmed by bacteriologic,
immunologic, or biochemical examination of the CSF
and serum.
Linear enhancement confined to the nerve roots
is described in aseptic meningitis and in immune-
mediated meningitis (Gero et al. 1991; Phillips et al.
1990). AIDS patients with secondary central ner-
vous system cytomegalo-virus infections develop
a polyradiculomyelitis. This demonstrates a strong
enhancement of the conus and cauda equina nerve
roots. These roots are thickened, but there is no irreg-
ularity or nerve root clumping. The CSF protein level
is raised, there is a polymorphonuclear cell reaction,
and serum CVM antibody titres are raised (Hansman
Whiteman et al.1994).
In some cases of spinal bacterial meningitis there
is no demonstrable abnormality on T1 Gd MRI
examination (Donovan Post et al.1990). In TBSM, the
imaging appearances are invariably advanced at the
time of presentation of the patient (Gero et al.I991).
In spinal infections, the clinical assessment of the
patient is paramount, while the imaging documents
the extent of the disease. (McGuinness 2000).
33.16.3
Spinal Meningitis in Neoplastic Conditions
33.16.3.1
Meningeal Carcinomatosis
TBSM and meningeal carcinomatosis share many
imaging features on WSCM, CT, and MRI (Chang
et al. 1989; McGuinness 2000; WHO. Global tuber-
culosis control 2000; Phillips et al. 1990; Krol et al.
1988). The neoplasms commonly metastasizing to
the spinal leptomeninges are primary carcinomas
of the breast and lung as well as lymphomas and
melanomas. In most instances the primary focus
will already have manifested itself in other body
systems (Fig. 33.31).
Multiple cytologic examination of the CSF is
necessary, as sensitivity is low. Although, in the case
of lymphoma, CSF lymphocytosis can be unusually
high (McAllister and O'leary 1987; Williams et al.
1990). Also mass lesions associated with lymphoma
are often epidural and a paravertebral element is
sometimes present (McGuinness 2000) (Fig. 33.32).
Bone marrow changes seen on MRI are also helpful
Imaging of Brain and Spinal Cord Tuberculosis
Fig.33.31. Sagittal scan of the thoracic spine
reveals multiple, low signal lesions (arrows).
These were foci of meningeal carcinomatosis
Fig. 33.32. Postgadolinium scan of the
thoracic spine with posterior meningeal
enhancement in the upper area and patchy
enhancement in the lower area (arrows). A
case of spinal meningeal lymphoma with
meningeal infiltration
577
in differentiating metastases from TBSM (Krol et al.
1988; McAllister and O'Leary 1987).
Meningeal and nerve root thickening, bunching,
nodular lesions, as well as surface changes in the
conus medullaris, are common to both conditions.
In intracranial tumors, spread to the spinal menin-
ges by "dropping down" will give rise to local areas
of similar imaging patterns, but less widespread
than in TBSM (Chang et al. 1989; McGuinness 2000;
Phillips et al. 1990; Hansman Whiteman et al. 1994;
Blews et al.1990). In intramedullary tumors the same
clinical pattern is seen as in tuberculoma of the
myelum. Tuberculomas are usually less than 5 mm
in diameter, usually do not expand the cord, in which
case they are not seen on WSCM. MRI defines tuber-
culomas, which are commonly associated with TBM
or TBSM (McGuinnes 2000; Gupta et al.1997; Shen et
al. 1993; Wadia and Dastur 1969; Gupta et al. 1994).
Isolated (Jinkins 1993; Sanchez Pernaute et al. 1996;
Jena et al. 1991; Donovan Post et al. 1990; Rhoton et
al. 1988 (Figs. 33.33-33.35) and multiple lesions are
described (McGuinness 2000; Gupta et al. 1997; Shen
et al.1993; Wadia and Dastur 1969; Gupta et al.1994)
(Fig. 33.36).
The common primary tumors of the myelum,
astrocytome, ependymomas and hemangioblastomas
generally show different MRI characteristics, and are
more likely to expand the cord, to contain cystic ele-
Fig. 33.33. Sagittal postgadolinium scan of the brain demon-
strating a cord tuberculoma at the cervico-medullary junction.
Ring enhancement with a hypodense center and expansion
of the cord
578 F. McGuiness

Fig. 33.34. A high signal tuberculoma, low in the thoracic cord.

There is a central necrotic nidus (arrow)

Fig. 33.35. An axial, Tl postgadolinium scan of the neck, with

fat supression, demonstrates a small intensely enhancing solid
tuberculoma of the cervical cord (arrow)

Fig. 33.36. Multiple

supra- and infratentorial
tuberculomas demon-
strate ring enhancement
(arrows). There is a
single tuberculoma of the
thoracic cord
Imaging of Brain and Spinal Cord Tuberculosis 579

ments, to be less well defined and to lack intense ring
enhancement (Jinkins 1991). Ependymomas enhance
patchily, due to the presence of cystic components
(Sanchez Pernaute et al. 1996; Slasky et al. 1987).
Multiple small secondary deposits in the myelum
present a diagnostic problem, but at this stage of the
disease manifestations elsewhere .and the presence of
bone secondaries defined by radiographs, isotope, or
MRI examination are usual (Fig. 33.37). Myelomatous
meningitis has similar appearances to TBSM, carci-
nomatosis, and some granulomatous conditions. The
characteristic vertebral changes and laboratory find-
ings differentiate the condition from TBSM (Quint et
al.1995).
may be acute or chronic. Extensive thickening of the
leptomeninges in the basal cisterns and over consider-
able contiguous areas of the cervical theca are seen.
In disseminated coccidioidmycosis (DCC) cervi-
cal meningeal thickening of such a degree as to cause
flattening of the cervical cord is sometimes present
(McGuinness 2000). The cervical spine lesions are
the result of extension from intracranial meningitis,
which shows features similar to TBM and TBSM,
including parenchymal lesions. The two conditions
are differentiated by culture of the organisms from
the CSF, by complement fixation antibody titre stud-
ies, and, if necessary meningeal biopsy at the time
of shunting for high pressure hydrocephalus. Blood
vessels are invaded in fungal diseases and infarctions
occur. Wrobel describes two cases of anterior spinal
artery occlusion in DCC with cervical meningeal
enhancement in TlGd MRI studies (Wrobel et al.
1992).
Histoplasmosis also disseminates in the CNS, and
although rare, the changes in the leptomeninges and
myelum are similar to those of TBSM (Desai et al.
1991).

Fig.33.37. Sagittal thoracolumbar scan, demonstrating mul-

tiple, high signal bone marrow metastases (arrows). There is
also irregular thickening of the meninges, due to further meta-
static deposits, in this case of carcinoma of the nasopharynx

Biopsy of TB lesions of the myelum should be

avoided and a trial of antituberculosis therapy is pref-
erable, because the histology of tuberculosis may be
unclear in a frozen section (McGuinness 2000).

33.16.4
Fungal Diseases
Fig. 33.38. Mixed signal in extensive mycetoma of the poste-
rior cervical soft tissues (arrows). The fungus has extended
Disseminated fungal infections produce images simi- into the spinal canal, both anteriorly and posteriorly, resulting
lar in appearance to TB (Fig. 33.38). These diseases in cord compression (arrow heads)
 580 F. McGuiness

In patients with AIDS, opportunistic infections population (McGuinness 2000,Villoria et al. 1992;
of the central nervous system reflect the diseases Whiteman 1997).
endemic in the geographical area. TBM and TBSM
tend to be early complications and affect between 2%
and 18% of patients. The higher levels of infection, 33.16.5
reported in Spain, appear related to the underlying Spirochetal Disease
trigger of the disease being drug abuse as opposed
to other causes (Villoria et al. 1992). Toxoplasmo- Tick-borne Lyme disease, due to Borrelia burgdor-
sis and cryptococcal infections usually appear Jeri, is widespread in Europe and in North America.
at a later stage of the disease, when the immune When the CNS is involved, lymphocytic pleomor-
response has reached a lower level. Although the phism and raised CSF protein are seen. The onset is
changes are widespread in the CNS, the inflamma- milder and slower than in TBM or TBSM, but both
tory response is diminished. The enhancement of radiculopathy and myelopathy occur. Although the
lesions seen on both CECT and T1 Gd MRI is less MRI appearances are not usually so dramatic as in
than in patients with a normal immune response. TB, enhancement of both the leptomeninges and the
This is also the case in the appearances of AIDS- CSF has been described (Fig. 33.39a-c). More often
related CNS tuberculosis, when compared with the enhancement is confined to the pial surface of the
imaging appearances of similar cases in the general cord (Demaerel et al. 1994).

Fig. 33.39. a Sagittal contrast-enhanced Tl-

weighted (500/30) spin echo image of the cervical
spine, shortly following injection. There is mild
contrast enhancement of the CSF in the pontine
and chiasmic cisterns, which gives a slightly
higher signal than that in the fourth ventricle.
Note clear differentiation between the avidly
enhancing meninges and the less intense CSF,
in this case of confirmed neuroborreliosis. b, c
Coronal images obtained 30 min after the spinal
images, showing markedly enhanced CSF in the
basal cisterns and around the spinal cord. Note
the normal signal of the CSF in the ventricles c
Imaging of Brain and Spinal Cord Tuberculosis 581

Fig. 33.40. Extensive meningeal thickening and

enhancement posteriorly (arrows). The result of
syphilitic pachymeningitis

In neurosyphilis, spinal gummas are exceedingly Expansion of the myelum on WSCM is docu-
rare, but have similar imaging characteristics to tuber- mented (Hitchon et al. 1984,Huang and Haq 1987;
culomas. The more common expression of syphilis Kelly et al. 1988). MRI has confirmed granulomas,
is hypertrophic leptomeningitis, or pachymeningitis, and these lesions may extend over a number of spinal
a chronic condition affecting the cervical segments. segments, and enhance on Tl Gd MRI studies (Nesbit
Enhancement is seen on TlGd MRI (Fig. 33.40). et al. 1989; Williams et al. 1990). CSF findings are of
Enhancement in the myelum, corresponding to pos- lymphocytic pleomorphism and raised protein, but
terior column pathology, has also been described, in the CSF sugar level is normal (McGuinness 2000). If
cases of tertiary syphilis (Gero et al. 1991; Tashiro et al.
sarcoid of the myelum is suspected, MRI imaging of
1987). Testing serum and CSF for levels of venereal dis- the brain will often reveal characteristic granulomas
ease research laboratory (VDRL) antigen are part of of chiasmal, hypophyseal, and paraventricular dis-
the protocol in all cases of CNS disease (McGuinness tribution (Greco and Steiner 1987; Nesbit et al. 1989;
2000). Williams et al. 1990; Hayes et al. 1987). CSF enhance-
ment in intracranial sarcoid has recently been
described (Good and Jager 2000).
33.16.6 Isotope imaging with Ga 67 reveals sarcoid
Other Granulomatous Diseases lesions in other organs, and hilar lymph nodes
are often enlarged on chest radiographs, or on
Neurosarcoidosis is a noncaseating granulomatous mediastinal CT. Biopsy of the conjunctiva, liver, or
condition with systemic manifestations. Patho- lymph nodes, or a transbronchial biopsy, enables a
logic change in the CNS is commonly intracranial histologic diagnosis in a high proportion of cases
and spinal neurosarcoidosis is a rare complication (McGuinness 200). Neurobrucellosis is commonly
(Hitchon et al. 1984; Huang and Haq 1987; Kelly et intracranial, and meningitis, cerebritis, and cranial
al. 1988; Greco and Steiner 1987). As it is essentially neuropathies occur. Although rarely involving the
curable with steroid therapy, it is important to distin- spinal elements of the CNS, brucella radiculopa-
guish parenchymal sarcoid from tuberculomata and thies are described. Swelling of the nerve roots
spinal tumors. Sarcoid granulomas expand the cord involved is present, and they enhance on postcon-
and infiltrate the meninges. If surgical intervention trast MRI studies. High titres of serum agglutinat-
is necessary with biopsy of the meninges, culture ing antibodies are present, as well as a history of
specimens for TB should be routine and cord biopsy exposure to infected animals or their products (Tali
avoided (Hitchon et al. 1984). et al. 1996).
582 F. McGuiness

33.16.7 defects. Intramedullary cysts are usually found in the

Parasitic Diseases
33.16.7.1
Schistosoma
Schistosoma mansoni (SM), S. haemotobium (SH),
and the Eastern form, S. japonicum (SJ), all cause
granulomatous lesions in the CNS (Scrimgeour and
Gajdusek 1985).
Sporadic cases are being frequently reported in
nonendemic areas, as both immigration and acces-
sibility by air travel increases. SM or SH ovae are
deposited in the brain, myelum, or meninges. SH ovae
are presumed to pass through the vertebral plexus of
veins, and in cases of hepatosplenic schistosomiasis,
SM passes through pulmonary arteriovenous shunts
to the brain. SJ is more likely to infest the brain than
the spinal cord (Murphy et aI. 1998; Silbergleit and
Silbergleit 1992; Dupuis et aI.1990).
Young males are the most commonly affected, and
although rare, the condition should be considered in
any case of myelitis, where the patient has recently
traveled in an endemic area. The clinical presentation
is similar to local tuberculous myelitis, with low back
pain, lower limb paraesthesia, and loss of sphinc-
ter function as the usual symptoms. The disease
is normally confined to the thoracic cord or conus
medullaris regions. Myelography and MRI demon-
strate local expansion of the cord, and with T1 Gd
imaging, both pial and intramedullary enhancement
are described (McGuinness 2000; Murphy et aI. 1998;
Silbergleit and Silbergleit 1992; Dupuis et al. 1990).
Lymphocytic pleomorphism and raised CSF protein
levels are found, but there is also an eosinophilia.
Serum and CSF antibodies against schistosomiasis
are present.
lower thoracic region, and the MRI characteristics
are of a multiseptate lesion similar in appearance to
a cystic neoplasm with enhancing margins. As cavita-
tion of the cord may be associated with tuberculomas
(McGuinness 2000; Sanchez Pernaute et al. 1996;
Slasky et al. 1987), they must be included in the dif-
ferential diagnosis. Laboratory testing for cysticerco-
sis-specific antibody is positive in a high proportion of
instances (Gupta et aI. 1994; Lotz et aI. 1988; Castillo
et aI. 1998).
33.16.8
Demyelinating Disease
33.16.8.1
Multiple Sclerosis
In all cases of suspected spinal myelitis or space
occupation, CT or MRI of the brain should be car-
ried out, as silent intracranial lesions may be present
(Fig. 33.41). Spinal lesions in multiple sclerosis (MS)
are often preceded by an episode of acute myelitis,
followed by recovery. Later signs of intracranial or
spinal MS recur. Dissemination in time is an impor-
tant feature of MS lesions (Hansman Whiteman et aI.
1994; Miller et aI. 1989; Posner et aI. 1983). The clini-
cal presentation is of great importance in MS, as the

33.16.8
Neurocysticercosis

Cysticercosis in the spine is rare. Lotz et aI., in a com-

bined study of pathology and imaging pointed out that
in the cranium the disease is more commonly found
in the subarachnoid spaces than in the parenchyma
(Lotz et al. 1988). Racemose cysts in the basal cistern
complex of the posterior fossa sometimes extend into
the cervical region as extramedullary lesions. Intra-
medullary lesions are very rare, but must be differen-
tiated from tuberculomas. WSCM and CT will show
expansion of the cord over a number of segments, or Fig.33.41. Enhancing lesions in the cord and brain stem
in the case of subarachnoid lesions, extensive filling (arrows). A case of multiple sclerosis
Imaging of Brain and Spinal Cord Tuberculosis
enhancing spinal lesions are difficult to differentiate,
using imaging methods, from tubrculomas, granulo-
mas, or tumors of the cord (McGuinness 2000).
References
Abduljabbar M (1991) Paradoxical response to chemotherapy
for intracranial tuberculoma:two case reports from Saudi
Arabia. J Trop Med Hyg 94:374-376
Abugali N, van de Kuyp PF, Annable W, Kumar M (1994) Con-
genital tuberculosis. Pediatr Infect Dis J 13:738-741
AI Deeb S, Yaqub BA, Sharif HS (1992) Neurotuberculosis: a
review. Clin Neurol Neurosurg 94:S30
Artopulous J, Chamelis Z, Christopulous S et al (1984) Sequen-
tial CT in TBM in infants and children. Comput Radiol 8:
271-277
Atlas SW (1991) Intracranial brain tumours.In: Atlas SW (ed)
MR imaging of the brain and spine. Raven press, New York,
pp 223-226
Atlas SW (1991) MRI of the brain and spine. In: Intracranial
infection. Raven Press, New York, pp 511-534
Bargallo J, Berenguer J, Garcia_Barrionuevo J et al (1996) The
"target sign": is it a specific sign of CNS tuberculoma. Neu-
roradiology 38:547-550
Bhargava S, Tandon PS (1980) Intracranial tuberculomas: a CT
study. Br J Radiol 55:935
Blews DE, Wang K, Kumar AJ, Robb PA et al (1990) Intradu-
ral spinal metastases in pediatric patients with primary
intracranial neoplasms: Gd-DTPA enhanced MR vs CT
myelography. J Comp Ass Tomogr 14:730-735
Bonafe A, Manelfe C, Gomez MC et al (1985) TBM: contribu-
tion of CT to its diagnosis and treatment. J Neuroradiol
12:302
Bouchama A, Zuheir al Kawa M, Kanaan I et al (199l) Brain
biopsy in tuberculoma: the risks and benefits. Neurosur-
gery 28:405-409
Buchanan G, Rainer EH (1988) Tuberculous mastoiditis. J
LaryngoIOtoll02:440-446
Campbell lA, Dyson AJ (1971) Lymphnode tuberculosis: a
comparison of various methods of treatment. Tubercle
58:171-179
Castillo M, Quencer RM,Donovan Post MJ (1998) MR of intra-
medullary spinal cysticercosis. AJNR 9:393©395
Cavallin L, Muren C (2000) CT findings in tuberculous oto-
mastioditis. Acta Radiologica 41 :49-51
Chambers ST, Hendrickse WA, Record C, Rudge P (1984) Par-
adoxical expansion of intracranial tuberculomas during
chemotherapy. Lancet 2:181-184
Chang K-H, Han M-H, Roh J-K et al (1990) Gd-DTPA
enhanced MR imaging in intracranial tuberculosis. Neu-
roradiology 32:19-25
Chang KH, Han MH, Choi YW (1989) Tuberculous arach-
noiditis of the spine: findings on myelography CT and MR
imaging. AJNR 10:1255-1262
Cremin BJ, Jamieson DH (1996) In: Childhood tuberculosis:
modern imaging and clinical concepts. Springer, Berlin
Heidelberg New York
Cremin BJ (1995) Tuberculosis: The resurgence of our
mostlethal infectious disease - a review. Pediatr Radiol
25:620-626
583
Cross DT, Moran cr, Brown AP et al (1995) Endovascular
treatment of episaxis in a patient with tuberculosis and
a giant petrous carotid aneurysm. Am J Neuroradiol 16:
1084-1086
Dastur D, Manghani DK, Udani PM (1995) Imaging of
tuberculosis and craniospinal tuberculosis: pathology
and pathogenic mechanisms. Radiol Clin North Am 33:
733-752
Dastur DK,Lalitha VS, Udani PM (1970) The brain and menin-
ges in TBM. Neurology (India) 18:86-100
Dastur DK, Manghani D, Udani PM (1995) Pathology and
pathological mechanisms in neurotuberculosis. Radiol
Clin North Am 33:733-752
Dastur DK (1972) Neurotuberculosis. In: Minckler J (ed)
Pathology of the nervous system, vol 3. McGraw-Hili, New
York, pp 2412-2422
Demaerel P, Wilms G, Van Lierde S et al (1994) Lyme dis-
ease in childhood presenting as primary leptomeningeal
enhancement without parenchymal findings on MR. AJNR
15:302-304
Desai SP, Bazan C, Hummell W, Jinkins JR (1991) Disseminated
CNS Histoplasmosis. AJNR 12:290-292
De Castro C, De Barros N, Campos Z et al (1995) CT scanning
of cranial tuberculosis. Radiol Clin North Am 33:771-769
Donovan Post MJ, Sze G, Quencer RM et al (1990) Gadolinium
enhanced MR in spinal infection. J Comp Ass Tomogr 14:
721-729
Draouat S, Abdenabi B, Ghanem M, Bourjat P (1987) Com-
puted tomography of cerebral tuberculoma. J Comput
Assist Tomogr 11:594-597
Dublin AB, Phillips HE (1980) CT of disseminated cerebral
coccidioidomycosis. Radiology 135:361-368
Dupuis MJ, Atrouni S, Dooms GC, Gonsette RE (1990) MR
imaging of schistosomal myelitis. AJNR 11 :782-783
Elkeslassy A, Weill A, Miaux Y (1997) Dilatation of deep
medullary veins in cortical venous occlusion due to focal
tuberculous meningitis. Neuroradiology 39:705-07
Gee GT, Bazan C, Jinkins JR (1992) Miliary tuberculosis involv-
ing the brain: MR findings. AJR 159:1075-1076
Gero B, Sze G, Sharif H (1991) MR imaging of intradural
inflammatory diseases of the spine. AJNR 12:1009-1019
Good CD, Jager HR (2000) Contrast enhancement of the
cerebrospinal fluid on MRI in two cases of spirochaetal
meningitis. Neuroradiology 42:448-450
Good CD, Jager HR (2000) Contrast enhancement of the
cerebrospinal fluid on MRI in two cases of spirochaetal
meningitis. Neuroradiology 42:448-450
Graveli AB, Redondo A, Salama Jet al (1998) Tuberculoma of
the cavernous sinus. Neurosurgery 42:179-182
Greco A, Steiner RE (1987) Magnetic resonance imaging in
neurosarcoidosis. Mag Reson Imaging 5:15-21
Greitz D, Greitz T, Hindmarsh T (1997) A new view on the
CSF'circulation with the potential for pharmacological treat-
ment of childhood hydrocephalus. Acta Paediatr 86:125-132
Grewal DS, Bhargava P, Mistry B, Gaikwad N (1995) Tubercu-
loma of the mastoid. J Laryngol OtoI109:232-235
Gucuyener K, Baykaner MK, Keskil IS (1993) Tuberculoma of
the suprasellar cistern: possible CT misinterpretation as
aneurysm. Pediatr RadioI23:153-154
Gupta PK, Kolluri VR, Chandramouli BA (1989) Calvarial
tuberculosis: a report of two cases. Neurosurgery 25:830
Gupta RK, Gupta S, Kumar S, Koli A et al (1994) MRI in intra-
spinal tuberculosis. Neuroradiology 36:39-43
584
Gupta RK, Gupta S, Singh D, Sharma et al (1994) MR imaging
and angiography in TBM. Neuroradiology 36:87-92
Gupta RK, Jena A, Sharma A, et al (1988) MR imaging of
intracranial tuberculomas. J Comput Assist Tomogr 12:
280-285
Gupta RK, Jena A, Singh AK (1990) Role of MR in the diagnosis
and management of intracranial tuberculomas. Clin Radiol
41:120-127
Gupta RK, Kohli A, Gaur V, Lal JH, Kishore J (1997) MRl of
the brain in patients with miliary pulmonary tuberculo-
sis, without symptoms or signs of central nervous system
involvement. Neuroradiology 39:699-704
Gupta RK, Kohli A, Gaur V, Lal JH, Kishore J (1997) MRI of
the brain in patients with milliary pulmonary tuberculo-
sis without symptoms or signs of central nervous system
involvement. Neuroradiology 39:699-704
Gupta RK, Pandey R, Khan EM et al (1991) Intracranial tuber-
culomas: MRI signal intensity. Correlation with histopa-
thology and localised proton spectroscopy. Magn Reson
Imaging 11:443-449
Hansman Whiteman ML, Dandapani BK, Sherbert RT, Dono-
van Post MJ (1994) MRI of AIDS-related polyradiculomy-
elitis. J Comp Ass Tomogr 18:7-11
Hayes WS, Sherman JL, Stern BJ, Citrin CM, Pulaski PD (1987)
MR and CT evaluation of intracranial sarcoidosis. AJR 149:
1043-1049
Hitchon PW, Ul Haque A, Olsen JJ et al (1984) Sarcoidosis
presenting as anintramedullary spinal cord mass. Neuro-
surgery 15:86-90
Hooijboer PJA, van des Vliet AM, Sinnige LGF (1996) Tuber-
culous meningitis in native Dutch children: report of four
cases. Pediatr Radiol 26:542-546
Hoshino T, Miyashita H, Asai Y (1994) Computed tomography
of the temporal bone in tuberculous otitis media. J Laryn-
goIOtoll08:702-705
Huang H, Haq N (1987) Spinal leptomeningeal sarcoidosis.
Neuroradiology 29:100
Jamieson DH (1995) Imaging of intracranial tuberculosis in
childhood. Pediatr Radiol 25: 165-170
Jena A, Banerji AK, Tripati RR et al (1991) Demonstration
of intramedullary tuberculomas by magnetic resonance
imaging: a report of two cases. Br J Radiol 64:555-557
Jena A, Sanchetee PC, Gupta RK (1988) Cysticercosis of the
brain shown by magnetic resonance. Clin Radiol 39:
542-546
Jinkins JR, Al Kawi MZ, Bashir R (1987) Dynamic computed
tomography of cerebral parenchymal tuberculomata. Neu-
roradiology 29:523-529
Jinkins JR, Gupta R, Chang KH, Rodriguez-Carbajal J (1995)
MRI of central nervous system tuberculosis. Radiol Clin
North Am 33:771-786
Jinkins JR (1991) Computed tomography of intracranial
tuberculosis. Neuroradiology 33:126-135
Jinkins JR (1993) MR of enhancing nerve roots in the unoper-
ated lumbosacral spine. AJNR 14:193-202
Kanamalla US, Ibarra RA, Jinkins JR (2000) Imaging of cranial
meningitis and ventriculitis. Neuroimaging Clin N Am 10:
2,309-331
Kelly RB, Mahoney PD, Cawley KM (1988) MR demonstra-
tion of spinal cord sarcoidosis:report of a case. AJNR 9:
197-199
Kioumehr F, Dadestanm MR, Rooholamini (1994) Central ner-
vous system tuberculosis: MRI. Neuroradiology 36:93-96
F. McGuiness
Kioumehr F, Dadestan MR, Feldman et al (1995) Post con-
trast MRI of the cranial meninges: leptomeningitis versus
pachymeningitis. J Comput Assist Tomogr 19:5,713-720
Kirkpatrick JB (1991) Neurologic infections due to bacteria,
fungi and parasites. In: Davis RL, Robertson DM (eds)
Textbook of neuropathology. Williams and Wilkins,
London, pp 719-803
Klingensmith WC, Datu J (1978) Tuberculous meningitis of
the sylvian fissure. Clin Nucl Med 3:315-317
Krol G, Sze G, Malkin M, Walker R (1988) MR of cranial and
spinal meningeal carcinomatosis:comparison with CT and
myelography. AJR 151:583-588
Kumar A, Montanera W, Willinsky R et al (1993) MR features
of tuberculous arachnoiditis. J Comp Ass Tomogr 17:
127-130
Lee PYC, Drysdale AJ (1993) Tuberculous otitis media: a dif-
ficult diagnosis. J Laryngol Otoll07:339-341
Lehrer H (1966) Angiographic triad of tuberculous menin-
gitis:a radiographic and clinico-pathological correlation.
Radiology 87:829-835
Leiguarda R, Berthier M, Starkstein S et al (1988). Ischaemicin-
farction in 25 children with tuberculous meningitis. Stroke
19:200-204
Le Roux PD, Griffin GE, Marsh HT, Winn HR (1990) Tubercu-
losis of the skull: case report and a review of the literature.
Neurosurgery 26:851-856
Lincoln EM, Sordillo SVR, Davies PA (1960) Tuberculous men-
ingitis in children. J Pediatr 57:807-823
Lincoln EM (1947) Tuberculous meningitis in children with
special reference to serous meningitis. Am Rev Tuberc 56:
75
Lotz J, Hewlett R,Alheit R, Bowen R (1988) Neurocysticercosis:
correlative pathomorphology and MR imaging. Neuroradi-
ology 30:35-41
Madkour M (2001) Neurobrucellosis. In: Brucellosis. Springer,
Berlin Heidelberg New York, pp 166-178
McAllister MD, O'leary DH (1987) CT myelography of sub-
arachnoid leukemic infiltration of the lumbar thecal sac
and lumbar nerve roots. AJNR 8:568-569
McGuinness FE (2000) Tuberculous radiculomyelopathy and
myelitic tuberculomas. In: Clinical imaging of non-pulmo-
nary tuberculosis. Springer, Berlin Heidelberg New York
McGuinness F (2000) Intracranial tuberculosis. In: Clinical
imaging of non-pulmonary tuberculosis. Springer, Berlin
Heidelberg New York, pp 27-42
Medical Research Council (1948) Streptomycin treatment of
tuberculosis. Lancet 1:582-596
Meltzer CC, Fukui MB, Kanal E, Smirniotopoulos JG (1996) MR
imaging of the meninges: normal anatomic features and
non-neoplastic disease. Radiology 201:297-08
Miller DH, Ormerod lEC, Rudge P et al (1989) The early risk
of MS following acute syndromes of the brain and spinal
cord. Ann Neuol 26:635-639
Miller DH, Ormerod IEC, Rudge P et al (1989) The early risk
of multiple sclerosis following isolated acute syndromes of
the brain and spinal cord. Ann Neurol 26:635-639
Mumtaz MA, Schwartz RH, Grundfast KM (1983) Tuberculo-
sis of the middle ear and mastoid. Pediatr Infect Dis J 2:
234-236
Murphy KJ, Brunberg JA, Quint DJ, Kazanjian PH (1998) Spinal
cord infection: myelitis and abscess formation. AJNR 19:
341-348
Naranbhal RC, Mathiassen W, Malan AF (1989) Congeni-
Imaging of Brain and Spinal Cord Tuberculosis
tal tuberculosis localised to the ear. Arch Dis Child 64:
738-740
Nesbit GM, Miller GM, Baker HL et al (1989) Spinal cord
sarcoidosis:a new finding at MR imaging with Gd-DTPA
enhancement. Radiology 173:839-843
Ng SH, Tang LM, Lui TN (1996) Tuberculoma en plaque. Neu-
roradiology 38:453-455
Ng SH, Tang LM, Lui TN et al (1996) Tuberculoma en Plaque.
Neuroradiology 38:453-455
Ogawa SK, Smith MA, Brennessel DJ, Lowy FD (1987) Tuber-
culous meningitis in an urban medical center. Medicine
66:317
Ozates M, Kemaloglu S, Gurkan F et al (2000) CT of the brain in
tuberculous meningitis: a review 289 patients. Acta Radiol
41 :2000-13-17
Patankar T, Varma R, Krishnan A et al (2000) Radiographic
findings in tuberculosis of the calvarium. Neuroradiology
42:518-521
Phillips ME, Ryals TJ, Kambhu SA, Yuh WTC (1990) Neoplastic
vs inflammatory meningeal enhancement with Gd-DTPA. J
Comp Ass Tomogr 14:536-541
Pio A, Chaulet P (1998) WHO, Tuberculosis handbook. WHO/
TB/98.259
Posner CM, Paty DW, Scheinberg L et al (1983) New diagnostic
criteria for MS: guidelines for research protocols. AnnNeu-
rol13:227-231
Praharaj SS, Sharma MC, Prasad K et al (1997) Unilateral
meningeal thickening: a rare presentation of tuberculous
meningitis. Clin Neurol Neurosurg 99:60-62
Quint DJ, Levy R, Krauss JC (1995) MR of myelomatous men-
ingitis. AJNR 16:1316-1317
Rajshekar V, Haran RP, Prakash GS, Chandy MJ (1993) Dif-
ferentiating solitary small cysticercus granulomas and
tuberculomas in patients with epilepsy. J Neurosurgery
78:402-407
Rajshekhar V, Chandy MJ (1997) Tuberculomas presenting
as isolated brain stem masses. Brit J Neurosurg II (2):
127-133
Rajshekhar V, Chandy M (1996) Comparative study of CT and
MRI in patients with seizures and a solitary cysticercus
granuloma. Neuroradiology 38:542-546
Reed MH, Ferguson CA (1978) The radiology of intracranial
tuberculosis in children. Can Assoc Radiol J 29:113
Reid H, Fallon RJ (1992) Bacterial infections. In: Adams JH,
Duchen WL (eds) Greenfields neuropathology. Edward
Arnold, London, pp 302-334
Rhoton EL, Ballinger WE, Quisling R, Sypert GW (1988) Intra-
medullary spinal tuberculoma. Neurosurgery 22:733-736
Rich AR, McCordock HA (1933) The pathogenesis of tubercu-
lous meningitis. Bull Johns Hopkins Hosp 52:5
Rich AR (1944) The pathogenesis of tuberculousis. Charles C
Thomas, Springfield
Rochas-Echeverri LA, Soto-Hernandez JL, Garza S et al (1996)
Predictive value of digital subtraction angiography in
patients with TBM. Neuroradiology 38:20-24
Ross JS, Masaryk TJ, Modic MT, Delamater R et al (1987) MR
imaging oflumbar arachnoiditis. AJR 149:1025-1032
Sakamoto S, Kitagaki H, Ishii K et al (1997) Gadolinium
enhancement of the cerebrospinal fluid in a patient with
meningeal fibrosis and cryptococcal infection. Neuroradi-
ology 39:504-505
Salgado P, Del Brutto OH, Talamas 0, et al (1989) Intracranial
tuberculoma: MR imaging. Neuroradiology 31:299-302
585
Sanchez Pernaute R, Berciano J, Robollo M, Pascual J (1996)
Intramedullary tuberculoma of the spinal cord with syrin-
gomyelia. Neuroradiology 38:S105-S106
Schoeman J, Hewlett R, Donald P (1988) MR of childhood
tuberculous meningitis. Neuroradiology 30:473-477
Scrimgeour EM, Gajdusek DC (1985) Involvement of the CNS
in schistosoma mansoni and S. Haematobium infection.
Brain 108:1023-1038
Shah GV, Desai SB, Malde HM, Naik G (1993) Tuberculosis of
the sphenoid sinus: CT findings. AJR 161:681-682
Shah GV, Sowan AM, Doctor MR, Shah PS (1994) MR findings
in a patient with a tuberculoma of the teetal plate. AJR
162:469-470
Shah GV (2000) Central nervous system tuberculosis: imaging
manifestations. Neuroimaging Clin N Am 10-2:355-374
Shakir RA (1986) Neurobrucellosis. Postgrad Med J 62:
1077-1079
Sharif H (1992) Role of MR imaging in the management of
spinal infections. AJR 158:1333-1345
Shen WC, Cheng TY, Lee SK (1990) Disseminated tuberculo-
mas of the spinal cord and brain demonstrated by MRI
with gadolinium-DTPA. Neuroradiology 35:213-215
Shen WC, Cheng TY, Lee SK (1993) Disseminated tuberculo-
mas of the spinal cord and brain demonstrated by MRI
with gadolinium-TPA. Neuroradiology 35:213-15
Silbergleit R, Silbergleit R (1992) Schistosomal granuloma of
the spinal cord: evaluation with MR imaging and intra-
operative sonography. AJR 158: 1351-1353
Singer MB, Atlas SW, Drayer BP (1998) Subarachnoid space
disease: diagnosis with fluid-attenuated inversion-recovery
MR imaging and comparison with gadolinium enhance
spin-echo MR imaging - blind reader study. Radiology
208:417-22
Singh B (1991) Role of surgery in tuberculous mastoiditis. J
LaryngoIOtoI105:907-915
Slasky BS, Bydder GM, Niendorf HP, Young JR (1987) MR
imaging with gadolinium DTPA in the differentiation of
tumour, syrinx and cyst of the spinal cord. J Comp Ass
Tomogr 11:845-850
Slavin RE, Walsh TJ, Pollack AD (1980) Late generalised tuber-
culosis:a pathological analysis of 100 cases in the pre-anti-
biotic and antibiotic eras. Medicine 59:352-366
Snider DE, Roper WL (1992) The new tuberculosis. N Engl J
Med 362:703-705
Tali ET, Keskin T, Oznur et al (1996) MRI of Brucella polyneu-
ritis in a child. Neuroradiology 38:S190-S192
Tandon PN (1978) Tuberculous meningitis (cranial and
spinal). In: Vinken PJ, Bruyn GW (eds) Infections of the
nervous system. Handbook of clinical neurology, vol 33.
Elsevier, Amsterdam pp 195-162
Tashiro K, Moriwaka F, Sudo K et al (1987) Syphilitic myelitis
with its MRI verification and succesful treatment. Jpn J
Psych NeuroI41:269-271
Teoh R, Humphries MJ, O'Mahony G (1987) Symptomatic
intracranial tuberculoma developing during the treatment
of tuberculosis: a report of 10 patients with a review of the
literature. QJM New Series 63 241:449-460
Urback H, KristofR, Zentner Jet al (1997) Sarcoidosis present-
ing as an inta-or extra-axial cranial mass: report of two
cases. Neuroradiology 39:516-519
Vallejo JG, Ong LT, Starke JR (1994) Clinical features, diagnosis
and treatment of tuberculosis in infants. Pediatrics 94:1-7
Van Dyke A (1988) CT of intracranial tuberculosis, with
586
special reference to the "target sign." Neuroradiology 30:
329-336
van Dyke A (1988) CT of intracranial tuberculosis, with specific
reference to the "target sign:' Neuroradiology 30:329-336
Vengsarkar US, Pisipaty RP, Parekh B, Panchal VG, Shetty MN
(1986) Intracranial tuberculoma and the CT scan. J Neu-
rosurg 64:568-574
Villoria MF, de la Torre J, Fortea F, Munoz L et al (1992)
Intracranial tuberculosis in AIDS: CT and MRI findings.
Neuroradiology 13:11-14
Wadia NH, Dastur DK (1969) Spinal meningitides with radicu-
lomyelopathy. Part 1: clinical and radiological features. J
Neurol Sci 8:239-260
Wadia NH (1973) Radiculomyelopathy associated with spinal
meningitides (arachnoiditis) with special reference to the
tuberculous variety. In: Spillane JD (ed) Tropical neurology.
Oxford University Press, London, pp 63-72
Waeker NJ Jr, Connor JD (1990) Central nervous systemtuber-
culosis in children: a review of 30 cases
Wallace RC, Burton EM, Barrett FF et al (1991) Intracranial
tuberculosis in children: CT appearances and clinical out-
come. Pediatr RadioI21:241-246
F. McGuiness
Wechman JM (1979) CT of intacranial tuberculomas. Clin
RadioI30:567-573
Wessels G, Hesseling PB, Beyers N (1998) Skeletal tuberculosis:
dactylitis and involvement of the skull. Pediatr Radiol 28:
234-236
Whiteman MLH (1997) Neuroimaging of central nervous
system tuberculosis in HIV-infected patients. Neuroimag
Clin N Am
WHO. Global tuberculosis control. WHO report 2000. Geneva,
Switzerland, WHO/CDS/TB/200.275
Wilkinson HA, Ferris AJ, Muggid AL (1971) Central nervous
system tuberculosis:a persistent disease. J Neurosurg 34:
15-22
Williams DW, Elster AD, Kramer SI (1990) Neurosarcoidosis:
gadolinium enhanced imaging. J Comp Assist Tomogr 14:
704-707
Williams MP, Olliff JFC, Rowley MR (1990) CT and MR in
parameningeal leukaemic masses. J Comp Ass Tomogr
14:736-742
Wrobel CJ, Meyer S, Johnson RH, Hesselink JR (1992) MR fid-
ings in acute and chronic coccidioidomycosis meningitis.
AJNR 13:1241-1245
34 Extraspinal Musculoskeletal Tuberculosis
M. MONIR MADKOUR, AIDA J. KUDWAH, MOHAMMED ABD EL BAGI

CONTENTS 34.1
Introduction
34.1 Introduction 587
34.1.1 Epidemiology 588
34.1.2 Pathogenesis 588 Tuberculous arthritis and osteomyelitis may present
34.1.3 Clinical Features 588 with indolent and often nonspecific clinical features
34.1.4 Diagnosis of Extraspinal Bone that constitute a challenge to the treating clinician
and Joint Tuberculosis 590 (CDC 1997; Cremin 1971; Jacobs et al. 1991; Rieder
34.1.5 Tuberculosis of the Hip Joint 590 et aI. 1990; Silva 1980).
34.1.6 Imaging Features of Tuberculosis of the Hip 591
34.1.7 Treatment 591 Such obstacles often lead to a delay in diagnosis
34.1.8 Tuberculosis of the Knee 591 of a disease which is curable, with subsequent joint
34.1.8.1 Case Illustration 591 destruction and disabilities. The duration of symp-
34.1.9 Imaging Features of Knee Tuberculosis 592 toms before diagnosis may be up to 20 years as
34.1.10 Tuberculosis of the Foot and Ankle 592
noted (Soler et al. 2001). Difficulties in early diagno-
34.1.11 Imaging Features of Foot
and Ankle Tuberculosis 593 sis are more noticeable in developing countries, for
34.2 Sacroiliac Joint Tuberculosis 593 example in Saudi Arabia and other countries (Ellis
34.3 Tuberculosis of the Joints et al. 1993; Negusse 1993), particularly in children
of the Upper Extremity 594 with more severe joint destruction and loss of func-
34.3.1 Treatment of Upper Limb Joint Tuberculosis 595
tion (Singh et al. 1992). This may be compounded
34.3.2 Sternoclavicular Joint Tuberculosis 596
34.3.3 Poncet's Disease 596 with partial, inadequate antituberculous treatment
34.3.4 Extraspinal Tuberculous Osteomyelitis 597 prior to presentation, which may reduce the effec-
34.3.5 Tuberculous Osteomyelitis of the Chest Wall 597 tiveness of clinicoradiological and microbiological
34.3.6 Tuberculous Tenosynovitis, Bursitis, diagnostic tests.
and Myositis 599
In coinfection of HIV and tuberculosis, where
34.3.7 Epidemiology 599
34.3.8 Pathogenesis 599 both conditions are of insidious onset and where
34.3.9 Clinical Features and Diagnosis 599 reactive arthritis or septic arthritis are also common
34.4 Multifocal Osteoarticular Tuberculosis 600 in HIV patients, the diagnosis of tuberculous arthri-
References 601 tis and osteomyelitis becomes difficult (Chretien
1990; Elliott 1990; Govinder et al. 2000; Jellis 1996;
Leibert et al. 1996; Ragni et al.1995; Soler et aI. 2001;
Veerken and Bermejo 1992; Wright et al.1996; Jellis
2002). Atypical mycobacterial infection of the mus-
culoskeletal system is increasingly reported. The
most frequent are Mycobacterium intracellulare, M.
fortuitum, M. kansasii, M. avium, and others (Ip and
Chow 1992; Williams and Riordan 1973; Cortez and
M. M. MAOKOUR, MD, DM, FRCP
Consultant, Department of Medicine, Riyadh Armed Forces
Pankey 1973; Chow et al. 1983,1987). See chapter on
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia nontuberculous mycobacteria.
M. ABO EL BAGI, MB BCh, DMRD, FSRRCSI The most common cause of extraspinal osteoar-
Department of Radiology, Riyadh Armed Forces Hospital, ticular tuberculosis is Mycobacterium tuberculosis
P.O. Box 7897, Riyadh 11159, Saudi Arabia
and to a lesser extent other related organisms such as
A. J. AL-KuoWAH, MD, DD, FRCP (Ed)
Consultant Dermatologist, Department of Dermatology,
M. africanum and M. bovis found in poor, developing
Riyadh Armed Forces Hospital, P.O. Box 7897, C-117, countries with high levels of raw milk consumption
Riyadh 11159, Saudi Arabia (Watts and Lifeso 1996; Kosin and Bishop 1991).

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
588
34.1.1
Epidemiology
Osteoarticular tuberculosis accounts for 1% to 5% of
all tuberculosis cases and for about 15% of extrapul-
monary disease (Weir and Thorton 1985). The spine is
involved in approximately 50%, joints in 30%, extraspi-
nal osteomyelitis in 19%, and tenosynovitis and bursi-
tis in 1% (Jaovisidha et al.1996; Martini et al.1986a,b,c;
Singh et al. 1992; Watts and Lifeso 1996).
Tuberculous arthritis most commonly affects
the weight-bearing joints, particularly the hips and
knees in equal frequencies, and less frequently the
ankles and feet. In the upper extremities the wrist
joint is the most frequently affected, followed by the
elbow and shoulder (Skoll and Hudson 1999). The
frequency of individual joint involvement varied
in different series, particularly from developing
countries where children and young adults are more
commonly affected. Singh and colleagues (1992)
reported a series of 104 children with osteoarticular
tuberculosis seen over a two-year period. The ages
ranged between 9 months and 18 years with male
predominance. The hip joints were affected in 14.9%,
the knees in 10.3%, and extraspinal osteomyelitis
occurred in 18.2%. The spine was affected in 43%.
Shah and colleagues (1992) reported 80 patients with
osteoarticular tuberculosis seen during hospitaliza-
tion over a 3-year period. About 60% of patients were
below the age of 40 years with female predominance
(57.5%). The spine was affected in only 21.2%, the
joints in 48.6%, and extraspinal tuberculous osteo-
myelitis occurred in 30%. Associated pulmonary or
renal tuberculosis was found in 36.25%. The knee
joints were affected in 15%, the hips in 11.2%, the
ankles in 6.2%, the elbows in 10%, the wrists in 3.7%
and the sacroiliac in 2.5%.
In Addis Ababa, Negusse (1993) reported 42 chil-
dren with osteoarticular tuberculosis attended the
Ethio-Swedish Children's Hospital over a period of
3 years. The hip joints were more affected than the
spine, in 22 patients and 15 patients, respectively.
Associated active pulmonary tuberculosis was found
in four patients.
34.1.2
Pathogenesis
Tuberculous arthritis may occur as a progression of
primary infection, particularly in children in devel-
oping countries (Resnick 1988). Localization of infec-
tion into the joint is commonly caused by hematog-
M. M. Madkour et ai.
enous spread of the bacilli from a primary focus such
as the lung. However, in more than 50% of patients
with tuberculous arthritis, no evidence appears in
imaging focused on the lung. Another possible mode
of transmission of infection is via lymphatic spread.
Joint involvement may also occur due to spread of
infection from epiphyseal bone eroding into the joint
space. And, another possible mode of transmission of
infection to the joint is direct inoculation (Davidson
and Horowitz 1979; Abdelwahat et al. 1998).
Granulomatous inflammatory tissue reaction takes
place in the joint synovium with granulation tissue,
caseation, necrosis, and joint effusion. Fibrin in the
synovial fluid may precipitate in the form of so-called
rice bodes, similar to those seen in rheumatoid arthritis
(Al-Qattan et al. 1994). A pannus of granulation tissue
may slowly progress and cause erosion and damage to
the articular cartilage and adjacent cancellous bone.
Slow cartilage and bone destruction in tuberculous
arthritis is due to the inability of M. tuberculosis bacilli
to produce proteolytic enzymes that are found in
pyogenic arthritis caused by Staphylococci and Strep-
tococci. However, as the disease progresses, complete
joint destruction may occur, similar to that caused by
pyogenic infections. In untreated joints, ankylosis and
sinus formation may result (Fig. 34.1).
34.1.3
Clinical Features
Tuberculous monoarthritis is the most common
presenting feature, but oligoarticular or polyarticular
involvement, on rare occasions, may occur (Valdazo
et al. 1990). The onset is usually insidious with mild
pain and swelling of the soft tissue surrounding the
affected joint. These features may be easily mistaken
as post-traumatic (Hunfeld et al. 1998). As the dis-
ease progresses, limitation of joint movement, cold
abscesses with draining fistulas, and destruction with
deformity of the joint may occur (Singh et al. 1992;
Dhillon et al. 2001a,b; Babhulkar et al. 2002). Clinical
features of tuberculous arthritis are commonly mild
and nonspecific, and many mimic other arthritic
diseases. Subsequently, the diagnosis is commonly
delayed (Kim et al. 1999). Constitutional symptoms,
such as fever, loss of appetite, and weight loss may be
lacking in patients with tuberculous arthritis.
Joint pain may be absent despite the presence of
swelling (Ruggieri et al. 1997; Hoffman et al. 2002).
In a series of 52 children with tuberculosis of the
knee, Hoffman and colleagues reported swelling in
all patients, pain in 65%, and draining sinuses in 10%.
 Extraspinal Musculoskeletal Tuberculosis 589

b e

Fig. 34.1a. Tuberculous discharging sinus below the left medial

aspect of the left knee. Note the discharge pus over the left
leg. b Note the swelling of the left knee, wasting of the thigh
and leg muscles, and shortening of the left leg. c Tuberculous
arthritis of the left knee. Frontal radiography demonstrates
advanced destructive changes. There are marked osseous
erosions mostly peripheral in location (open arrows), periar-
ticular osteoporosis, and diminished joint space (Phemister
triad). Complete obliteration with ankylosis of the joint space
medially has ensued. Note normal appearance of the right
knee. d Tc 99m bone isotope demonstrates high tracer uptake
in the left knee. e Lateral radiograph of lower lumbar spine
demonstrates evidence of healed brucella spondylitis. Note
marginal osteophytes (arrow heads), erosion of the anterior
aspect of the superior end plate of L4 vertebral body (open
c arrow) and preservation of the disc spaces (d)
590
An acute septic arthritis-like feature (warm,reddened,
and swollen joint) was noted in eight patients (15%).
Tuberculous arthritis may occur in patients with
systemic lupus erythematosus and mixed connective
tissue disease where systemic steroid are used for
treatment of these conditions (Stecher et al. 1992). In
such cases, delay in diagnosis is likely to be common.
The disease may simulate oligoarticular-onset
juvenile rheumatoid arthritis ORA). Al-Matar and
colleagues (2001) reported two Canadian children
who were initially diagnosed as JRA-a 25-year-old
Caucasian girl and a 6-year-old boy who had emi-
grated with his family from Somalia. Treatment with
systemic and intraarticular steroids were given with
no response. A considerable delay of the diagnosis
for 18 months was emphasized by the authors. They
related such a long delay to the rarity of the disease in
Canada and the similarity of the clinical and investi-
gative findings to those of JRA.
34.1.4
Diagnosis of Extraspinal Bone
and Joint Tuberculosis
Tuberculous arthritis is often nonarticular and
rarely affects more than one joint. Monoarthritis
should always be initially regarded as infectious till
proven otherwise. Tuberculous monoarthritis can
be diagnosed if the attending clinician maintains a
high level of vigilance and examines a detailed his-
tory and physical examinations for possible clues of
tuberculosis that may be found in other body organs
or systems. Among patients with monoarthritis, the
most common causes are infections, crystals, trauma,
spondyloarthropathy, and rheumatoid arthritis. The
differential diagnosis of tuberculous arthritis should
include pyogenic arthritis caused by bacteria or fungi.
In developing countries such as Saudi Arabia, brucel-
losis may be common and may simulate tuberculous
arthritis, although the onset is more acute (Madkour
2001). Sarcoidosis may be polyarticular, but is rarely
present as monoarthritis, which is intermittent and
persistent. Tuberculin skin test (PPD) may be posi-
tive in more than 90% of patients with osteoarticular
tuberculosis (Berney et al. 1972; Halsey et al. 1982;
Kerri and Martini 1985; Lee et al. 1995). The hemo-
gram shows normal white cell count and mildly
elevated ESR (Wang et al. 2000). Arthrocentesis and
synovial tissue biopsies are the most important diag-
nostic investigations in patients with osteoarticular
tuberculosis (Versefeld and Solomon 1982; Wollinsky
1994; Yao and Sartoris 1995; Suh et al. 1996; Sant and
M. M. Madkour et al.
Bajaj 1991). The synovial fluid may appear as clear or
turbid with a raised white blood cell count ranging
from 2,500 to 100,OOO/mm3 with 60% neutrophils.
Synovial fluid protein is usually raised and glucose
may be low in approximately 50% of patients, which
are nonspecific. Synovial fluid smear positivity may
range from 0% to 83% (Berney et al. 1972; Lee et al.
1995). Fluid culture may be positive in up to 50% of
patients (Ellis et al. 1993; Wang et al. 2000; Berney et
al. 1972). Synovial tissue biopsy may yield the bacilli
in 47% t090% of cases (Albers et al. 1984; Bush and
Schneider 1984; Berney et al. 1972; Halsey et al. 1982;
Lee et al. 1995). Histopathology of synovial tissue
biopsy may be positive in up to 90% of cases, and
when combined with microbiology, definitive diag-
nosis can be achieved in 95% (Gorbach et al. 1992).
Gen-Probe methods using PCR or its new ver-
sions, Amplified Mycobacterium Tuberculosis Direct
Test (AMTDT), are more sensitive than PCR with
sensitivity of 94.7% to 96% and specificity close to
100% (Gamboa et al. 1998; Wobeser 1996).
Imaging features of osteoarticular tuberculosis
are discussed in detail in the chapter "Imaging of
Bone and Joint Tuberculosis:'
34.1.5
Tuberculosis of the Hip Joint
The hip joint is one of the most common extraspi-
nal sites to be affected by tuberculosis. It is essential
that clinicians differentiate it from septic arthritis,
as both diseases are common in children and young
adults. Serious disabilities may occur if a diagnosis
is not made early. Mild hip pain with limping is the
most common early presenting clinical feature with
or without constitutional symptoms. Patients may
present late in the course of the illness with cold
abscesses, discharging sinuses, and hip deformities.
Hip pain tends to be worse at night and may lead to
sleep disturbances. In advanced stages, walking may
become difficult particularly with the occurrence of
pathological dislocation of the hip joint. Hip pain
and limitation of movements are readily elicited
with active or passive movements during physical
examination. The affected limb may be shorter than
the opposite one, with flexion, adduction, and inter-
nal rotation. Muscle wasting around the hip joint may
be noted. Clinical and imaging features of tubercu-
losis of the hip have been classified (see the chapter
"Imaging of Bone and Joint Tuberculosis") by several
authors to identify guidelines for diagnosis, assessing
the extent of joint damage, managing, and checking
Extraspinal Musculoskeletal Tuberculosis
for indications for total hip replacement (Shanmuga-
sundaram 1983; Babhulkar and Pande 2000).
34.1.6
Imaging Features ofTuberculosis of the Hip
We have devoted a separate chapter to imaging of
bone and joint tuberculosis from our own series;
please refer to the chapter "Imaging of Bone and
Joint Tuberculosis." For tuberculosis of the hip joint,
see also Figs. 16,24,25, and 26 in that chapter.
34.1.7
Treatment
The management of hip tuberculosis is a multidisci-
plinary task that involves clinicians, microbiologists,
histopathologists, occupational therapists, and physio-
therapists. Chemotherapy with duration of treatment
ranging from 18 to 24 months, for long-term therapy,
to short-term therapy of 6 months has been exten-
sively reported (American Thoracic Society 1986; Dutt
et al. 1986; Hannanchi et al. 1988; Tenth Report of the
Medical Research Council 1986; Dutt and Stead 1989;
Cohn et al. 1990; Combs et al. 1990).
Isoniazid and rifampicin with or without etham-
butol, streptomycin, and pyrazinamide are used in all
regimens.
In early phases of the disease, chemotherapy is
given with the application of bilateral traction to
ensure hip rest and supervised mobilization without
weight-bearing for 4 to 6 weeks. Lack of any favorable
response to this conservative approach to treatment
is an indication for surgery to prevent hip destruc-
tion. Synovectomy and joint debridement may be
useful but may not prevent painful fibrous ankylosis.
In advanced hip joint damage with restriction of
mobility and function, total hip replacement may
be necessary too (Kim et al. 2001; Yoon et al. 2001;
Babhulkar et al. 2002). Long-term follow-up after
total hip replacement for hip tuberculosis for up to
28 years was reported by Kim and colleagues (2001).
These authors reported the follow-up for up to 28
years of 60 patients with total hip replacement for
hip joint tuberculosis. Six patients had relapses of hip
tuberculosis and the prosthesis was removed from
three patients. The reasons for the relapses were either
inadequate drug dosage or duration, or poor compli-
ance of the patient. The authors also reported a 54-
year-old woman with left acetabular destruction and
calcifications. Tuberculosis was not diagnosed preop-
591
eratively. Tissue biopsies were sent for histopathology
and showed granulomas but these results were not
conveyed to the treating clinician. The patient was dis-
charged without chemotherapy. She returned 1 year
later with persisting discharging sinus. The diagnosis
was confirmed only by performing a PCR.
The final clinical outcome can be predicted by
the imaging features of hip tuberculosis at the time
of presentation. Normal and Perthes' type with pre-
served joint space tend to have a good prognosis.
Hips with atrophic, traveling acetabulum, protrusio
acetabuli, and mortar and pestle types will have poor
results (Babhulkar et al. 2002).
34.1.8
Tuberculosis of the Knee
Tuberculosis of the knee usually presents as chronic
monoarthritis with insidious onset of pain, swell-
ing, and limping, with or without the presence of
constitutional symptoms. There may be a history of
trauma to the knee. The duration of symptoms may
be up to 2 years before presentation. In about 15%
of patients the duration of symptoms may be 1week
with features of septic arthritis including fever, with
swollen, warm, and tender knee (Hoffman et al. 2002).
The knee may be held in flexion, and active or passive
movement during physical examination is associated
with severe pain. In advanced stages, knee defor-
mity with flexion, posterior and lateral dislocation,
and wasting of surrounding muscles may be noted
on examination. Discharging sinuses (Fig. 34.1a,b)
may be noted in about 26% of cases (Kerri and
Martini 1985; Friedman and Vishwa 1970). Because of
its gradual onset in most patients or even during acute
presentation the, diagnosis of tuberculous arthritis of
the knee is commonly delayed between 4 to 40 months
(Halsey et al. 1982; Newton et al. 1982; Kerri and
Martini 1985; Lee et al. 1995; Dhillon et al. 1998).
Evidence of pulmonary tuberculosis in patients with
knee tuberculosis maybe present in 10% to 47% ofpatients
(Friedman and Vishwa 1970; Chow and Yau 1980).
34.1.8.1
Case Illustration
A 45-year-old male shepherd living in a rural area in the
central province of Saudi Arabia presented with pain
and swelling of the left knee, limping, and discharging
sinus. The pain had started 12 years earlier, and he
related it to a minor trauma to the knee. He noticed
the intermittently discharging sinus over the left leg for
about 7-9 months before presentation to our hospital.
592
He denied any history of tuberculosis or constitutional
symptoms. He had a past history ofbrucellosis with low
back pain, fever, sweating, and chills 16 years ago, which
had been treated in a peripheral clinic and responded
well. He keeps sheep, goats, and camels in the back yard
of his house and regularly consumes raw milk.
The patient attended several hospitals for his left
knee and refused biopsies, arthroscopies, or knee
replacement. The patient looked well and had no
other systemic findings. The left knee was swollen with
flexion deformity of 15 degrees (Fig. 34.1a, b). There
was a discharging sinus below the medial aspect of
the left knee. The muscles of the thigh and leg were
wasted. The left leg was 3 cm shorter than the right.
Examination of the spine showed no deformities or
localized tenderness. Plain radiography of the knees
showed gross destruction of the left knee (Fig. 34.1c, d).
Bone scintigraphy showed increased uptake in the left
knee (Fig. 34.1e). The hemogram was normal and the
ESR was 48 mmlhr. His biochemical parameters were
normal. Brucella agglutinins were not raised; repeated
at one-month intervals, these remained at a low titer of
1:80. Blood culture for brucella was negative. Culture
of sinus discharge yielded Mycobacterium tuberculosis
after 4 weeks. Lumbar spine radiography showed a
healed old destructive lesion at the anterior superior
end plate with osteophyte formation. The intervertebral
disc space was preserved, and there was no vertebral
body collapse, osteoporosis, or deformities (Fig. 34.1e).
The patient was treated with antituberculous
chemotherapy for 1 year complemented with phys-
iotherapy and a high-heeled shoe. He refused total
knee replacement. The patient responded well to
treatment and the sinus healed and the ESR returned
back normal. The most likely cause of his old healed
spinal lesion at the fourth lumbar vertebra is bru-
cellosis. (The fourth lumbar vertebra is the most
common site of spinal brucella spondylitis. Infection
commonly starts at the anterior superior end plate,
and collapse of the vertebral body, unlike tubercu-
losis, is rare.) The intervertebral disc space was pre-
served and there were no deformities.
34.1.9
Imaging Features of Knee Tuberculosis
Full detailed imaging features of tuberculosis of the
knee using different modalities at various stages of the
disease can be seen in our chapter on "Imaging of Bone
and Joint Tuberculosis"; and also see Figs.4, 11,12, 14,
15,17,20,21,23, and 28 for knee tuberculosis.
Kerri and Martini (l985) described radiological
classification of 52 patients with tuberculosis of the
M. M. Madkour et al.
knee into four stages. Stage 1, with osteopenia and soft
tissue swelling. Stage 2, with erosions but normal joint
space. Stage 3, with narrow joint space. Stage 4, with
gross anatomic distortions. These classifications do
serve as predictive of the outcome of chemotherapy
alone or combined with surgery (Lee et al. 1995). Stage
1 and 2 had good outcomes, while in stage 3 and 4
with cartilage destruction had unfavorable outcomes.
Wilkinson (l969) graded the response to treatment
using the range of knee flexion and imaging features
as parameters. Results are considered excellent with
full range of movement and normal XR; good, with
flexion of more than 90 degrees and normal XR; fair,
with flexion between 35 and 90 degrees and narrow-
ing ofjoint space; poor, with worse flexion or ankylosis
and joint space narrowing.
Treatment of stage 1 and 2 knee tuberculosis using
chemotherapy alone provided good results (Gupta 1982;
Kerri and Martini 1985; Hoffman et al. 2002). Some
authors have used chemotherapy and surgical synovec-
tomy and debridement, but others found synovectomy
patients had more symptoms and stiffer knee postop-
eratively (Wilkinson 1969; Katayama et al. 1962; Gupta
1982). Early mobilization without the need for splinting
postoperatively with stage 1 and 2 was found to be as
effective as with splinting (Gupta 1982; Hoffmann et al.
2002). In stage 3 and 4 disease, however, flexion defor-
mity of the knee should be prevented by placing it in
splint. For adult patients with persistent pain or signifi-
cant knee joint destruction, arthrodesis or arthroplasty
may be necessary (Eskola et al.1988; Kim 1988).
34.1.10
Tuberculosis of the Foot and Ankle
Foot and ankle tuberculosis involve sites rarely involved
in patients with osteoarticular diseases. Large series are
reported from developing countries and only a few
reports of 1 to 3 cases from developed countries.
Dhillon and colleagues (2001, 2002) from India
reviewed the world literature and postulated that one-
fifth of the world population of tuberculosis patients
are in India. They also found that the reported inci-
dence of foot and ankle involvement varied from 3%
to 12%. Among the 224 patients with foot and ankle
tuberculosis, these authors reported that articular
involvement was more frequent than tarsal or small
bone osteomyelitis. Tuli (l997) from India reviewed
1,074 patients with skeletal tuberculosis seen over a
32-year period and found 154 patients with foot and
ankle disease (l4.3%). A more recent report by Dhil-
lon and colleagues from India (Dhillon and Nagi 2002)
described their own experience of 74 patients with
Extraspinal Musculoskeletal Tuberculosis
tuberculosis of the foot and ankle over an ll.5-year
period. There have been only a few case reports from
developed counties in recent years.
Dalldorf and colleagues (1994) from New York
reported a 34-year-old alcoholic man with right-
foot pain and swelling without any constitutional
symptoms. Delay in diagnosis was experienced by
these authors similar to many others. Eventually the
diagnosis was confirmed microbiologically from
open tissue and fluid materials. Nicklas et al. (1996)
from Ohio, USA, reported a 38-year-old female with
painless swelling of the second toe of the right foot.
Other systemic physical examinations were normal.
At operation, bone and soft tissue biopsies revealed
granulomatous inflammation and positive culture
of the bacilli after 6 weeks. Malhan et al. (2001) from
Yorkshire, UK, reported a 28-year-old man with a 2-
month history of pain and swelling of the right ankle
without constitutional symptoms. Open biopsy for
histopathology showed granuloma with negative
tissue culture. The diagnosis was confirmed by PCR.
Ruggieri and colleagues (1997) reported three
Italian children aged 14 months to 6 years with tuber-
culosis of the ankle. These authors experienced dif-
ficulties and delay in diagnosis. Ong and colleagues
(1998) from Singapore reported a 25-year-old woman
with left-foot pain without swelling. Six months later
during routine periodic health screening for foreign
domestic helpers, she had swelling and persistent pain
in the foot. Foot and chest radiography showed osteo-
porosis at the small joints and paravertebral abscess
with collapse of vertebral bodies of T7-T8 consistent
with Pott's disease.
The clinical features of foot and ankle tuberculosis
at the time of presentation are nonspecific and may
vary widely. Patients may present with painless swell-
ing of the ankle, foot, or toe without constitutional
symptoms. Pain may be the only presenting feature,
without swelling. Acute onset with pain, swelling, red-
ness, hotness, limping, and diminished motion may be
the presenting features. Constitutional symptoms may
be present in 45% of patients. The ankle may be held
in extension and muscular atrophy of the lower limb
may be noted. Sinuses may be noted in 23% of patients
at presentation. The duration of symptoms may range
from weeks to 4-5 years (Ruggieri et al. 1997; Dhillon
and Nagi 2002).
34.1.11
Imaging Features of Foot and Ankle Tuberculosis
Tuberculous lesions may be either osseous, articular, or
a combination of both as depicted by imaging modali-
593
ties. Imaging features are clearly described from our
own series, in the chapter "Imaging of Bone and Joint
Tuberculosis (please see Figs. 2, 3, 5, 31, 32, 36, and 37).
Chemotherapy alone is adequate for the treatment
offoot and ankle tuberculosis. Clinical symptoms may
disappear after 2-5 months, while imaging improve-
ments with consolidated cortex and disappearance of
periostitis may be achieved after 6 months. Surgical
intervention may occasionally be required for open
biopsy, debridement, or arthrodesis.
34.2
Sacroiliac Joint Tuberculosis
Tuberculosis of the sacroiliac joint is rare and
because the joint is deeply located with a limited
range of movement, signs of this disease are easily
overlooked (Soholt 1951). The main presenting
symptom is buttock pain on the affected side. Pain
may be poorly localized in the buttock and may be
mistaken for lumbar spine, hip, lower abdominal, or
pelvic diseases (Soholt 1951; Feldmann et al. 1981;
Goldberg and Kovarsky 1983). The onset of pain is
usually insidious over weeks or months.
It may be associated with limping and may cause
sleep disturbances. The pain may radiate to the lower
limb with a positive straight leg raising test (Pouchot
et al. 1988). Buttock abscess may rarely be present at
the time of presentation as we noted in three of our
patients (see Fig. 35.1, Fig. 35.22 and Fig. 34.2). Wast-
ing of the quadriceps muscle may be present. Dis-
charging buttock sinuses may be noted at the time of
presentation (Fig. 34.2).
Sacroiliac joint examination will reveal tenderness
on pressure and conventional stress tests will induce
pain at the affected side. Sacroiliac joint tuberculosis
is usually unilateral. Rarely, it may be associated with
evidences of osteoarticular tuberculosis in another
location such as the pubic symphysis (Fig. 35.22).
Constitutional symptoms may be lacking.
Chemotherapy alone for 12-18 months is usu-
ally sufficient. Symptoms improve gradually and
commonly subside after 4-6 months of treatment
(Pouchot et al. 1988). Bed rest and immobilization in
a cast may rarely be required. Surgery for drainage
of buttock abscess, and rarely for arthrodesis, may
be indicated if symptoms persist after 9 months of
treatment (Pouchot et al. 1988).
594 M. M. Madkour et al.

Table 34.1. Osteoarticular tuberculosis of the wrist, hand and

elbow literature review)

Authors (year) Sites No of Chemotherapy

cases ± surgery
Brashear et al (1975) Wrist 10 Surgery
Martini et al (1980) Elbow 29
Arafiles et al (1981) Elbow 11 Surgery
Benkeddache (1982) Hand & Wrist 27
Albers et al (1994) Wrist 101 Surgery
Bush et al (1984) Hand & Wrist 11 Surgery
Martini et al (1986) Elbow 42
Wrist 10
Other sites 22
of upper limb
Parkinson et al (1990) Elbow 5 Surgery
Fig. 34.2. Tuberculous sinus in the right buttock (white arrow) Al-Qattan et al (1994) Wrist 3 Surgery
in a 42-year-old patient with right sacroiliac joint tuberculosis. Vohra et al (1995) Elbow 10 Surgery
Visuthikosol et al Hand & Wrist 23 Surgery
(1996)
Chen et al (1997) Elbow 23 Surgery
34.3 Cedidi et al (1997) Hand & Wrist 1 Surgery
Skoll et al (1999) Wrist 7 Surgery
Tuberculosis of the Joints Elbow 4 Surgery
of the Upper Extremity Other site 1
Wang et al (2000) Hand & Wrist 11
Tuberculosis of the wrist and hand, elbow and shoul- Elbow 2
der occurs in approximately 10% of all cases of bone Other sites 3
Farnell et al (2001) Wrist 1 Surgery
and joint tuberculosis, and the wrist and hand are the
most frequently affected (Martini et al. 1986; Hunfeld Surgery includes: open biopsy, synovectomy & debridement,
et al. 1998; Skoll and Hudson 1999). drainage of cysts or abscesses, excision of sinuses, release of
carpal tunnel, carpectomy or arthrodesis
We reviewed the literature that was available to
us (see Table 34.1) to determine the frequency of
involvement of the hand and wrist, elbow and other
sites in the upper limb. We found 363 reported cases
of tuberculosis of the upper extremities-in the wrist
and hand in 206 (57%), the elbow in 130 (36%), and
other sites in 27 (7%).
Tuberculosis of the wrist and hand may present
with pain and swelling. It is usually monoarticular
and the onset is insidious with progressive loss of
motion (Chen et al. 1997; Skoll and Hudson 1999).
Discharging sinuses and symptoms of carpal tunnel
syndrome may be the presenting feature late in
Fig. 34.3. Tubercu-
the disease (Al-Qattan et al. 1994; Chen et al. 1997; losis osteomyelitis
Cedidi et al. 1998; Skoll and Hudson 1999). Consti- and arthritis of the
tutional symptoms may be lacking in less than 50% wrist. Plain radio-
of patients. Secondary bacterial infections after an graph demonstrates
evidence of marginal
injury to the hand may present with acute pyogenic
osseous erosions with
arthritis and mask the presence of tuberculous infec- soft tissue swelling
tion (Al-Qattan et al.1994). (white arrows). Focal
Soft tissue swelling, tenderness, and limitation of osteolytic lesion in
both flexion and extension motion of the wrist may be the distal radius due
detected on physical examination. As the disease pro- to tuberculous osteo-
myelitis (black arrow
gresses,bone and ligaments subluxation and deformities heads). There is also
may be noticed. Soft tissue swelling may form abscesses periarticular osteo-
and discharging sinuses at the time of presentation. porosis
Extraspinal Musculoskeletal Tuberculosis 595

The clinical features of elbow joint tuberculo- ultrasonography, CT, and MRI. They show soft tissue
sis are similarly insidious and monoarticular, and swelling with joint effusion, olecranon tip changes,
symptoms may last for months or years before pre- cubital fossa cystic mass, and evidence of early lower
sentation. As the disease progresses loss of motion humerus osseous involvement. In the shoulder joint,
and functional disability may occur. Acute onset of plain radiography may look normal and axial CT will
symptoms with local features of pyogenic arthritis depict the soft tissue swelling and erosion particu-
and constitutional symptoms may be the presenting larly in the subglenoid bone (Fig. 35.35).
features. Patients with elbow joint tuberculosis are
more likely to have enlarged supratrochlear lymph
nodes, which may caseate, and form an abscess or 34.3.1
draining sinus (Patel 2001). The soft tissue swelling Treatment of Upper Limb Joint Tuberculosis
and edema around the elbow can easily be felt at the
back of the elbow on both sides of the olecranon and Chemotherapy alone for 12-18 months, supple-
triceps tendon. The shoulder joint lesion usually mented with joint rest and physiotherapy, are suf-
starts at the glenoid {Fig. 35.35) or at the humeral ficient to cure early disease in most patients without
head. Granulation tissue formation, fibrosis, and the need of additional surgical intervention. (Ross-
mann and Mac Gregor 1995; Houston et al.1994; Bass
osseous destruction lead to pain, stiffness, and limi-
et a11994)
tation of the range of motion. The onset of the disease
is similarly insidious with nonspecific symptoms A volar wrist splint and regular occupational
that may last for months or years before presenta- therapy and physiotherapy are important supple-
tion. Symptoms and signs may be similar to other mentary therapies. The range of motion of the wrist
rheumatic diseases. As the disease advances, muscle will improve with physiotherapy in early stages of the
spasm and wasting, particularly of the deltoid and disease. Surgical synovectomy, debridement, drain-
supraspinatus, may occur. age of abscesses, excision of sinuses, release of carpal
The imaging features of tuberculosis of upper tunnel syndrome, carpectomy, or" even arthrodesis
limb joints are nonspecific. These features have beenfor advanced disease may be required (AI-Qattan et
classified by many authors into four stages describ- al. 1994; Skoll and Hudson 1999; Cedidi et al. 1998).
The prognosis of wrist joint tuberculosis is usually
ing the changes from early disease to late and to the
advanced, destructive inflammatory process. good. However, residual motion loss and functional
Features may range from normal appearances, disabilities, as well as relapse of sinuses, may occur.
soft tissue swelling, osteoporosis, bone erosions, focalTreatment of elbow joint tuberculosis follows the
same principle. Immobilization of the elbow in a plas-
cysts, joint space reduction, to advanced joint destruc-
ter of Paris back-slab with elbow at 90 degrees for as
tion. These classifications are helpful as a guideline in
predicting the response to chemotherapy and physio- long as 4 weeks, followed by active mobilization and
regular physiotherapy for up to 6 months. Surgery to
therapy alone or the possible need for surgical inter-
vention as well (Martini 1988). the elbow joint may be required for open drainage
Stage I: No bony lesions; localized osteoporosis of any abscess, synovectomy, and debridement (Skoll
Stage II: One or more erosions (or cavities) in the and Hudson 1999; Chen et al.1997). Complications of
bone; discrete diminution of the joint surgery were reported by many authors.
space. Chen and colleagues (1997) reported their expe-
Stage III: Involvement and destruction of the whole riences of 23 patients with elbow joint tuberculosis
joint without gross anatomic distortion. treated with chemotherapy and surgery, Fifteen had
Stage IV: Gross anatomic distortion. uneventful healing, four had injury and palsy of the
posterior interosseous nerve, three had persistent
Plain radiography is helpful in most patients but sinuses after excision, and two had developed sinuses
CT and MRI will depict the extent of lesions much after synovectomy and arthrodesis. The range of
better than plain radiography (see "Imaging of Bone movement was better in early stages of the disease
and Joint Tuberculosis"). Plain radiography of the (stage I and II) than in later stages. The management
wrist and hand (Fig. 34.3 plain radiograph of the outcome of elbow joint tuberculosis tends to be less
wrist with tuberculosis) clearly depicts soft tissue predictable and generally poorer than that of wrist
swelling, osteoporosis, bone erosions, and cysts. joint tuberculosis (Martini et al. 1980).
Various imaging modalities of the elbow joints can Tuberculosis of the shoulder joint usually responds
be seen in Figs. 35.33 and 35.34: plain radiographs, favorably to chemotherapy when it is initiated in the
596
early stages of the disease. The shoulder should be
immobilized in position-of-function in plaster of
spica or removable polythene brace for 3 months. The
optimum position is 80 degrees abduction, 30 degrees
forward flexion, and 30 degrees of internal rotation.
Fibrous ankylosis usually takes place but daily activity
can be performed using the scapulothoracic movement
for compensation. Persistent joint pain may be helped by
excision arthroplasty with debridement of the affected
synovium and bony parts. However, surgical arthrode-
sis of the gleno-humeral joint may be necessary if other
measures fail to control the persisting pain.
34.3.2
Sternoclavicular Joint Tuberculosis
Tuberculosis of the sternoclavicular joint is rare and
frequently misdiagnosed for pyogenic infection, other
rheumatic diseases,metastasis,or brucellosis (Madkour
2001). The infection usually starts at the medial end of
the clavicle (Fig. 35.30 and Fig. 36.7a-l). The disease is
usually unilateral but may rarely be bilateral (Dhillon
et al. 2001; Sipsas et al. 2001). The onset of symptoms
is usually insidious with pain and swelling of the joint
for 6 to 32 months. Mild to moderate restriction of
the motion of the shoulder may occur because of
pain. Constitutional symptoms may be present at the
time of presentation in approximately 50% of patients
(Dhillon et al. 2001). Painless swelling of the joint or
discharging sinus may be the presenting feature. Plain
radiography may not be helpful in depicting changes.
Computed tomography or MRI with contrast is ideal
for showing soft tissue swelling and erosion of the
clavicular head. An incidental retroclavicular lymph
node, paratracheal lymph adenopathy, subcutaneous
presternal abscess, destruction of D3 vertebral body
with paravertebral soft tissue component may also be
depicted, as we noted in one of our patients (Fig. 35.29a,
b, and Fig. 35.30).
Chemotherapy alone without surgery is sufficient
for the treatment of sternoclavicular joint tuberculo-
sis. However, Dhillon and colleagues in 2001 reported
nine patients with tuberculosis of the sternoclavicu-
lar joint and two required surgical debridement after
2-3 months of chemotherapy without response.
The symphysis pubis is a rare site of tuberculous
arthritis and may be accompanied by tuberculosis of
the sacroiliac joint as noted in one of the two patients
who presented to us with symphysis pubis tubercu-
losis (Fig.35.1 and Fig. 35.27). The first patient was a
50-year-old woman who presented to us with a I-year
history of progressive low back and pelvic pain; she had
M. M. Madkour et al.
no constitutional symptoms. She noticed progressive
difficulty walking due to the pain and was using non-
steroidal anti-inflammatory drugs to relieve that pain.
Frontal radiography of the pelvis depicted widening,
erosion, and perarticular sclerosis of the right sacroiliac
joint. The symphysis pubis joint was widened with mar-
ginal erosions and perarticular sclerosis. ACT-guided
biopsy sample was taken. Granulomatous lesions were
seen on histopathological examination and the patient
was treated with chemotherapy. She had no evidence
of other bony focal disease as noted by isotope scintig-
raphy. The second patient was a 37-year-old man who
presented with a 7-month history of suprapubic pain
with swelling, fever, weight loss, and difficulty in walk-
ing because of the pain. Axial CT with contrast showed
bilateral pubic bone erosions with subcutaneous
abscess formation. Close needle CT-guided drainage,
aspiration, and collection of biopsy sample were per-
formed on the abscess and the bone. Granuloma forma-
tions were seen on histopathological examination and
cultures were negative. The patient was given chemo-
therapy and responded well to treatment (Fig. 35.27).
Patients with tuberculosis of the symphysis pubis may
present with hypogastric mass at the onset of the dis-
ease. Manzaneque et al. (1992) reported a 79-year-old
woman with primary Sjogren's syndrome and immune
thrombocytopenia who was treated with systemic ste-
roid for 7 months. She represented with a 2 week history
ofhypogastric swelling. Computed tomography showed
cystic mass attached to the pelvic wall with destruction
of the right and left pubis. CT-guided aspiration of the
cyst yielded tubercle bacilli. In another report, Bal-
sarkar and Joshi (2001) described a 28-year-old man
who presented with a 6-week history of swelling in the
hypogastric area. Radiography showed bony erosion of
the symphysis pubis. Aspiration of the mass yielded the
bacilli. Chemotherapy alone for 14-18 months or with
drainage of abscess if present is sufficient treatment of
tuberculosis of the symphysis pubis.
34.3.3
Poncet's Disease
Named after Antonin Poncet, a Lyons surgeon born
in 1849, who described an inflammatory condition of
the joints or periarticular structures resulting from
the presence of a tuberculosis focus elsewhere in the
body and named it tuberculous rheumatism. The dis-
ease was described earlier by Charcot in 1864 and by
Lanereaux in 1871. It is a form of reactive arthritis,
which may be either polyarticular, oligoarticular, or
monoarticular in distribution. The disease is pre-
Extraspinal Musculoskeletal Tuberculosis 597

dominantly nonmigratory and morning stiffness is in 4, radius in 3, and ulna in 2 patients. The author
usually absent. The pathogenesis of this disease is stilldescribed four basic imaging patterns of long bone
not clearly understood, as localization of tuberculous lesions as follows: 26 cystic, 10 infiltrative, 8 focal
infection of these joints is not found. erosions, and 6 spina ventosa (which means tubular
Allergy or hypersensitivity to tubercle protein expansion of the shaft of short tubular bones of the
was the favorite hypothesis adopted by many authors hands and feet). Peritoneal reaction was seen in 4 and
(Wilkinson and Roy 1984). Chaudhuri et al. (1995) sequestrum formation in 4. Four lesions crossed the
suggested that Poncet's disease and erythema nodo- growth plate and one had a pathological fracture. Soft
sum could be a different expression of the immuno- tissue swelling was noted in all patients. All patients
pathogenic response to the bacilli. The disease may were cured of their active tuberculosis osteomyelitis.
be either primary, as the first manifestation of tuber- The author has also reported the sequelae of the tuber-
culosis, or secondary if it occurs during the course culous osteomyelitis after cure as follows: four patients
of active disease (Chaudhuri et al. 1995). Patients had short legs, one a short arm, and one a short thumb.
may present with arthralgia, and acute, subacute, A vascular necrosis occurred in two femoral heads and
or chronic arthritis. Constitutional symptoms, ery- one navicular bone. Coxa vera was noted in two hips.
thema nodosum, conjunctivitis, lymphadenopathy, In another large series of extraspinal tubercu-
and active pulmonary tuberculosis may be associ- lous osteomyelitis, Bahulkar and colleagues (2002)
ated findings (Chaudhuri et al. 1995). The disease reported 74 patients from all age groups that were
responds well to chemotherapy alone, with complete seen over a period of 25 years. The long tubular bones
remission with no residual joint deformities. were affected in 39 patients, short tubular bones in 24,
and flat bones in 11 patients. The three most common
sites reported in this series were; the humerus in 10,
34.3.4 tarsal bones in 10, and pelvic bones in 5.
Extraspinal Tuberculous Osteomyelitis These authors confirmed the diagnosis by histo-
pathology and/or microbiology of biopsy specimens.
Extraspinal tuberculous osteomyelitis is rare and All patients responded well to a full course of che-
comprises approximately of 2% to 3% of all cases of motherapy.
osteoarticular tuberculosis. Localization of tubercu-
lous infection in the bone without joint involvement
can occur but is rare, and it can affect any bone. Long 34.3.5
or short tubular bones, flat bones such as the clavicle Tuberculous Osteomyelitis of the Chest Wall
and pelvis, small round bones such as the patella,
talus, and navicular can be affected with tuberculosis. Tuberculous osteomyelitis of the ribs, sternum and
Tubular long bones such as the femur, tibia, humerus, lateral end of the clavicle are rarely reported. Tuber-
radius, and ulna can be affected in the metaphysis, culous osteomyelitis of the sternum may occur in the
diaphysis, or epiphysis. (Figs. 35.2, 3, 4,10,32) absence of an underlying pleural, lung, or mediasti-
The presenting clinical features include pain, swell- nal tuberculosis. (Fig. 34.4 and Fig. 36.7a-l).
ing of the bone with warmth and tenderness, swelling One of our patients, a 36-year-old woman, pre-
of the surrounding soft tissue with abscess and sinus sented with painless swelling in the anterior chest
formation. The initial diagnosis of such patients is wall over the sternum. Her symptoms had started
often difficult. It is commonly mistaken for all types 3 months before presentation. As the swelling gradu-
of bone pathology. Rasool (2001) reviewed 42 children ally increased in size, the patient noticed fever, chills,
with extraspinal tuberculous osteomyelitis seen over loss of appetite, weight loss, but she did not have any
a 16-year period and found 50 lesions. In five patients respiratory symptoms. The patient was under weight,
osteomyelitis was multifocal. The metaphysis of a long chronically unwell, with swelling over the manu-
bone was affected in 25 patients (60%), the diaphysis brium sterni. It was not acutely inflamed and not
in 3 (7%), the epiphysis in 4 (10%). Short tubular bones tender, but fixed with fluctuation. Chest radiography
of hands and feet were affected in 8 (19%). Flat bones was reported as normal. Right lateral radiography
were affected in 7 (17%) and small round bones in 3 of the manubrium showed soft tissue swelling and
(7%). (Khanna et al. 1980; Sinnott et al. 1990; Franco bone erosion (See Fig. 35.28a, b. The left lateral view
2001; Garcea et al. 1994). showed larger erosion with a pathological fracture.
Tubular long bones were affected in the following In another patient of ours, the sternum was affected
descending order: femur in 17, tibia in 6, humerus as part of a multifocal tuberculosis that involved the
598
Fig. 34.4. A sinus scar formation secondary to tuberculous
osteomyelitis of the ribs on the left chest wall in a 40-year-
old patient
chest wall, mediastinal lymph nodes, and vertebral
body erosion with paravertebral and spinal canal soft
tissue swelling. There was a small subcutaneous pre-
sternal cold abscess (Fig. 35.29a, b). In both patients
the diagnosis was confirmed by tissue biopsy and
abscess drainage showing granulomas and positive
culture of the bacilli.
Sarlak et al. (2001) described a 43-year-old Turk-
ish man with discharging sinuses over the sternum
for 18 months. The patient sought medical advice
during this period with no diagnosis but only surgical
debridements. Chest radiography and CT of the thorax
showed destruction of the anterior border of the ster-
num, while other thoracic organs were normal. The
diagnosis was eventually confirmed with tissue biop-
sies showing granulomatous lesions. Sternal tubercu-
losis was reported after coronary artery bypass graft
surgery in a 58-year-old man from Greece. Although
the pulmonary granulomatous nodule was excised
during the cardiac operation, the patient refused
further investigation of it. One year later, he presented
with weakness, fatigue, weight loss, and pus discharge
from the sternum which was sterile. The patient pre-
sented 10 months later with heart failure, and a CT of
the chest showed sternal osteomyelitis and right upper
lobe infiltrates. Sputum and pus cultures were both
positive for acid-fast bacilli.
Tuberculous osteomyelitis of the ribs, with dis-
charging sinuses and involvement of underlying
lungs and pleura is rare. We noted that the retro-
mammary region or areas close to it seems to be the
M. M. Madkour et al.
site commonly reported by other authors as well (see
Fig. 34.3). Huang et al. (2001) reported a 4-year-old
girl with fever, cough, and chills treated with antibiot-
ics as bronchopneumonia. Chest radiography and CT
showed a soft tissue lesion in the third rib, which was
not detected on physical examination.
Excisional biopsy of the lesion was diagnosed as
enchondroma and lymphadenitis, and tuberculosis
was not diagnosed. One year later, a painless nonery-
thematous mass near the right nipple area over the
previous third rib location was the presenting clini-
cal feature detected by the family. Cystic lesion was
found this time on physical examination. Incisional
biopsy showed granulomatous lesions and urine cul-
ture yielded acid-fast bacilli.
In New York, Adler and colleagues (l993)
reported tuberculous osteomyelitis of the chest wall
in four patients, of the rib in two, sternum in one,
and costal cartilage in one. Underlying pleural or
pulmonary parenchymal tuberculosis was noted
in two patients. Plain radiography looked normal
despite bone destruction and a diagnosis was only
made by CT scan. In a large series on eight women
with tuberculous osteomyelitis of the rib, Supe and
colleagues (2002) reported the role of CT scan in
the diagnosis. The age of these women ranged from
12 to 44 years, and they presented with a palpable
nontender mass in the posterior aspect of the breast
which had appeared 1 month to 2 years before
presentation. None of these women had constitu-
tional symptoms or had tuberculosis. The authors
reported that CT scan depicted an abscess in seven
patients, which was aspirated. The eighth patient
had an ulcer, which was biopsied. A well-margin-
ated abscess with rim enhancement was the CT scan
finding in all eight patients.
The lateral end of the clavicle is rarely reported as
a site for tuberculous osteomyelitis. Basanagoudar
and colleagues (2001) from India reported a 30-year-
old female with pain in the right shoulder after a
minor trauma. Shoulder radiography was misinter-
preted as fracture of the lateral end of the clavicle.
The same lesion was reported as a giant cell tumor
during a repeat of radiography. However, a discharg-
ing sinus developed later and the culture yielded
Staphylococcus aureus and the diagnosis was made as
chronic osteomyelitis with discharging sinus super-
imposing on preexisting giant cell tumor. Incisional
biopsy showed granulomatous lesions. The patient
responded to 15 months of chemotherapy. The sinus
healed and the full range of shoulder movement
was achieved 4 months after starting chemotherapy.
These authors reviewed the literature on tuberculous
Extraspinal Musculoskeletal Tuberculosis
osteomyelitis of the lateral end of the clavicle and
found 16 cases reported between 1932 and 1996.
The mandible is an extremely rare site of tuber-
culous osteomyelitis. Bhatt and Jayarkishnan (2001)
reported a 4-year-old girl who was referred to a dental
clinic with a I-month history of right cheek swell-
ing with submandibular lymph node enlargement
after dental extraction. Radiography revealed poorly
defined radiolucency at the optical region of the man-
dible. Biopsy showed granuloma formation. Chest
radiography did later revealed left lung tuberculosis
with mediastinal lymphadenopathy. The father of this
child had had spinal tuberculosis 1 year earlier.
34.3.6
Tuberculous Tenosynovitis, Bursitis, and Myositis
Tuberculous tenosynovitis and bursitis are rare con-
ditions and have a great propensity to mimic other
diseases. Because of its rarity, failure in considering
tuberculosis as a possible cause is common, leading
to misdiagnosis and delay in giving the appropriate
treatment. Clinicians have to have a high degree of
vigilance in order to make the diagnosis. Symposium
tendons, bursae, and muscle involvement secondary
to bone and joint tuberculosis are well recognized
but rarely occur without bone and joint involvement
(Hoffman et al. 1996; Asaka et al. 1996; Abdelwahab
and Kenan 2000; Esenyel et al. 2000; Toda et al. 1998;
Soler et al 2000; Hassan et al. 1993; Del Giglio 1997;
Bonomo et al. 1995).
34.3.7
Epidemiology
Tuberculous tenosynovitis and bursitis occur in
approximately 1% of all cases of osteoarticular
tuberculosis (Jaovisidha et al. 1996). Jaovisidha and
colleagues (l996) described the imaging findings of
21 patients with proven tuberculous tenosynovitis
and bursitis. Tenosynovitis was found in 12 patients,
and most commonly affected the tendon sheath of
the hands and wrists. The mean age was 46.3 years
with male predominance. Tuberculous bursitis was
found in 9 patients, most frequently around the hip,
particularly in the trochanteric bursa. The mean age
was 51.5 years with male predominance. Other authors
have reported rare cases of tuberculous bursitis (Ihara
et al. 1998).
In our own series, 80 patients had extraspinal
musculoskeletal tuberculosis; two patients had teno-
599
synovitis affecting the Achilles tendon and elbow
(see Figs.35.2a-d and 35.33). Tuberculous bursitis
of the greater trochanter was found in one patient
(see Figs. 35.5, 35.8, and 35.9). Discharging sinuses,
and cold abscesses in the gluteal region and groin
were seen in nine patients (see Figs. 35.6,35.17,35.19,
35.22,35.35, and 35.39).
34.3.8
Pathogenesis
Tuberculous tenosynovitis and bursitis are most
commonly caused by M. tuberculosis. Rarely, it
may be caused by nontuberculous mycobacterial
organisms. The mode of transmission of the bacilli
may be through a hematogenous route or by direct
spread from adjacent bone and joint tuberculosis
(Jaovisidha et al. 1996; Bocanegra 1994; Donovan
and Sosman 1940; Briede 1945; Chafetzet et al. 1982;
Hoffman et al. 1996). After the entry of bacilli, seri-
ous fluid exudate or thickened granulation tissue
proliferation with thickening of the sheath may take
place depending on the virulence of the organisms
and host resistance.
As the disease progresses, caseation and fibrosis
of the sheath may occur, and the tendon itself may
be affected leading to spontaneous rupture (Pimm
and Waugh 1957). Granulation tissue, caseation, and
abscess formation may spread along the tendon and
break through the skin, discharging a straw-colored
fluid. Secondary bacterial infection may superimpose
itself (Fig. 35.40). In the bursae, a similar inflammatory
process may take place leading to an increase in size
due to inflammatory fluid, with subsequent formation
of rice bodies and calcifications (Kim et al. 2002).
34.3.9
Clinical Features and Diagnosis
Tuberculous tenosynovitis has no distinct clinical
feature. Its onset is usually gradual and insidious
with slowly progressive swelling at the involved site.
The duration of symptoms before presentation may
be up to several years (Jaovisidha et al. 1996; Gold-
berg and Avidor 1985; Abdelwahab et al. 1993; Robins
1967). Symptoms of mild pain and slight limitation of
motion become apparent as the disease progresses.
Symptoms of carpal tunnel syndrome due to tuber-
culous tenosynovitis of the flexor tendons of the wrist
joints has been reported (Lee 1985; Suso et al. 1988;
AI-Qattan et al. 1994; Skoll and Hudson 1999). Clini-
600
cal presentation of de Quervain's-like disease due to
tuberculous tenosynovitis of the abductor pollicis
longus and extensor pollicis brevis and the first dorsal
retinacular compartment has been reported. Chen
and Eng (1994b) described a 35-year-old man with
pain, tenderness, and swelling over the radial styloid
for 2 months, for whom Finkelstein's test was positive.
He was diagnosed as having de Quervain's disease
and treated with intraretinacular injection of steroid.
Symptoms improved temporarily but the swelling
progressed. Tenosynovectomy was performed and
histology revealed granuloma while microbiological
examination was negative. Antituberculous treatment
was given for 12 months and the patient improved
with no recurrence at follow-up. Asaka and col-
leagues (1996) reported a 74-year-old woman who
presented with a mass lesion on the flexor side of the
right forearm near the elbow joint. She was other-
wise asymptomatic but had a strong family history of
tuberculosis and a positive tuberculin test. Her chest
CT showed right upper lobe infiltrate, and she was
started on antituberculous chemotherapy. The mass
did not change in size during therapy and increased
in size after discontinuation of treatment. Ultrasonog-
raphy, CT, and MRI of the mass showed encapsulated
cystic lesion. Aspiration biopsy was nonspecific and
the diagnosis was confirmed as tuberculous by doing
a PCR on the abscess material.
Tuberculous bursitis may present with persistent
and slowly progressive dull pain, commonly affecting
bursae that are subject to trauma. The most common
sites affected are bursae around the hip and subdel-
toid, but it may affect other bursae (Jaovisidha et al.
1996). Tuberculous bursitis over the greater trochan-
ter of the femur may present with insidious persistent
dull pain over the hip and a limp during walking. The
time interval between the onset of symptoms and
presentation may be months or years.
A soft tissue swelling with fluctuation may be
found. Constitutional symptoms may be lacking.
Imaging modalities may be helpful in determining
the extent of the lesion and are useful for guided tissue
biopsy purposes. In a report by Ihara and colleagues
(1998),it was found that for a 27-year-old woman with
tuberculous bursitis of the greater trochanter, imaging
modalities were helpful in revealing a multicystic soft
tissue mass. An open biopsy was performed to exclude
malignancy, and fluid and tissue biopsies confirmed
the diagnosis by positive culture of the bacilli.
Tuberculous myositis is extremely rare and its
diagnosis is confirmed by raised serum creatine
kinase (CK), EMG, and histopathological features
of tuberculous myositis (Derkash and Makley 1979;
M. M. Madkour et al.
Pouchot et al.1990; Abdelwahab and Kenan 2000).
Tuberculous myositis may be treated successfully
by chemotherapy alone, but may also be compli-
mented with surgical procedures such as surgical
decompression of carpal tunnel syndrome drainage,
tenosynovectomy, debridement or excision of bursae
(Cramer et al. 1991; AI-Qattan et al. 1994; Bonomo et
al. 1995; Del Giglio 1997; Hassan et al. 1993; Soler et
al. 2000; Toda et al.1998).
34.4
Multifocal Osteoarticular Tuberculosis
Multifocal or disseminated bone and joint tuberculosis
is a well-recognized pathological entity although rarely
reported. It is almost always misdiagnosed initially
even in the endemic areas, and the correct diagnosis
is often delayed (Aggarwal et aI. 2001; Morris et aI.
2002; Alexander and Mansuy 1950; Arslan 1999; Ip et
aI.1993; Lachenauer et aI.1991; Murray 1954; Ormerod
et aI. 1989; Tuli and Sinha 1969; Yip et aI. 1996; Eid et
aI. 1994; Wessels et aI. 1998). In the Indian population,
the incidence of multifocal osteoarticular involvement
is 7% to 10% of all cases of osteoarticular tuberculosis
(Kumar and Saxena 1988). In developed countries such
as those in North America, the incidence is not known
(Eid et aI. 1994; Tsui et aI. 1993; Aggarwal et aI. 2001;
Morris et aI. 2002).
A hematogenous mode of transmission of the
bacilli to the bone is thought to be the most common.
Tuberculous lymphadenitis with caseation and
abscess formation frequently found in patients with
multifocal osteoarticular disease favors the possibil-
ity of a lymph-born mode of transmission of the
bacilli (Morris et al. 2002). Multifocallocalization of
infection in the bone depends on host immunity. It
is suggested that the lesions occur at different times
and multifocallesions are seen at different stages of
development (Morris et al. 2002; Aggarwal et al. 2001;
Kumar and Saxena 1988). A primary source of infec-
tion is not always found in these patients. In their
series of 48 patients with multifocal osteoarticular
tuberculosis, Kumar and colleagues (1988) found that
plain chest radiography revealed a primary focus in
19% of cases, abdominal disease was present in 8%,
and for the remaining 73% no primary focus could
be found. It can affect all age groups particularly
children and young adults in endemic developing
countries and the elderly in developed countries
(Table 34.2). The disease affects both sexes equally
(Table 34.2).
Extraspinal Musculoskeletal Tuberculosis 601

Table 34.2. Multifocal musculoskeletal tuberculosis (analysis of 8 reports)

Author (year)

Country of study No. of Age in Sex Patient's country Sites of tuberculosis

cases (%) years of origin

Kumar and Saxena (1988) India 48 cases 4-42 Equal India

Kumar and Saxena (1988) India 13 (27%) numbers Multiple bone involvement
Kumar and Saxena (1988) India 14 (29%) Multiple joint involvement
Kumar and Saxena (1988) India 21 (44%) Multiple bone and joint involvement
Case Illustration 4 Female Long bones of upper and lower limbs, short
tubular bones of both hands
Case Illustration 31 Male Phalanx of (L) index finger, phalanx of (R) little
finger, (R) olecranon, (L) medial malleolus
Tsui et al. (1993) Hong Kong 13 Female China (L) scapula, (R) 3rd rib, (L) 5th rib, skull, (R)
middle ear, C7, TI, TI2, U vertebrae
Muradali et al. (1993) Canada 4 32 Female Philippines (R) rib, (L) 5th rib, sternum, (L) tibia, bilateral
sacroiliac joints, (L) sternoclavicular joint, T2,
T6, Til, TI2, Ll, L2 vertebral bodies, bilateral
pleural effusion
Muradali et al. (1993) Canada 38 Female Somalia (L) 5th rib, 3rd thoracic vertebra, bilateral medi-
astinallymphadenopathy
Muradali et al. (1993) Canada 16 Female Philippines Sternomanubrial joint, Ll vertebra
Muradali et al. (1993) Canada 26 Male Ghana (L) sacroiliac joint, (L) 1st rib, (L) 3rd rib, (L)
medial condyle, (L) kidney and ureter
Eid et al. (1994) USA 34 Male Yemen Lateral end of (L) clavicle, (R) iliac crest
Tsau et al. (1995) Taiwan 57 Male Taiwan Pubic bone, (L) sacroiliac joint, (L) medial con-
dyle, (L) kidney and ureter
Wessels et al. (1998) South Africa 6 Male S.Africa Skull, humeri, femora, tibiae, radii, ulnae, meta-
months tarsals, metacarpals, phalanges, liver, spleen lungs
Aggarwal et al. (2001) India 18 4-60 12 Males India 5 spine (3 patients)
Aggarwal et al. (2001) India 6 Females 2 long bones (2 patients)
Aggarwal et al. (2001) India 26 bones of hands and feet (15 patients)
Aggarwal et al. (2001) India 10 joints (9 patients)
Aggarwal et al. (2001) India 1 axial nonvertebral (1 patient)
Morris et al. (2002) India 9 Female India Skull, sternum, (R) first metatarsal, pulmonary

The metaphysis of long bones are often the ini-
tial site as an endarteritis particularly in children
(Rasool et al. 1994). The multiplicity of various sites
as reported by several authors is noted in Table 34.2.
The clinical and imaging manifestations are nonspe-
cific. Pain is the most frequent presenting feature and
may precede constitutional symptoms. A high level
of vigilance by the attending clinician is essential
in sorting out the long list of differential diagnosis.
Chemotherapy and surgical intervention can cure
this pathological entity of tuberculosis.
References
Abdelwahab IF, Kenan S (2000) Tuberculous abscess of the
brachialis and biceps brachii muscles without osseous
involvement. A case report. J Bone Joint Surg Am 82:446
Abdelwahab IF, Kenan S et al (1991) Atypical skeletal tubercu-
losis mimicking neoplasm. Br J RadioI64:551-555
Abdelwahab IF, Kenan S et al (1998) Tuberculous gluteal abscess
without bone involvement. Skeletal Radiol 27:36-39
Adler BD, Padley SPG et al (1993) Tuberculosis of the chest
wall: CT findings. JComput Assist Tomogr 17:271-273
Aggarwal AN, Dhammi IK et al (2001) Multifocal skeletal
tuberculosis. Trop Doctor 31:219-220
Al Matar MJ, Cabral DA et al (2001) Isolated tuberculous
monoarthritis mimicking oligoarticular juvenile rheuma-
toid arthritis. J RheumatoI28:204-206
Albers W, Siihler H et al (1984) Die Behandlung der Hand-
gelenkstuberkulose. Fortschr Med 102:947-951
Alexander GH, Mansuy MM (1950) Disseminated bone tuber-
culosis: So called multiple cystic tuberculosis. Radiology
55:839-843
Al-Qattan MM, Bowen V et al (1994) Tuberculosis of the hand.
J Hand Surg (Br) 19B:234-237
American Thoracic Society (1986) Treatment of tuberculosis
and tuberculosis infection in adults and children. Am Rev
Respir Dis 134:355-366
Arafiles RP (1981) A new technique of fusion for tuberculous
arthritis of the elbow. JBone Joint Surg (Am) 63:1396-1400
Arslan A, Ciftci E et al (1999) Multifocal bone tuberculosis
presenting as a breast mass: CT and MRI findings. Eur
RadioI9:1117-1119
602
Asaka T, Takizawa Yet al (1996) Tuberculous tenosynovitis in
the elbow joint. Intern Med 35:162-165
Babhulkar SS, Pande SK (2000) Unusual manifestations of
osteoarticular tuberculosis. Clin Orthop 398:114-120
Balsarkar DJ,Joshi MA (2001) Tuberculosis of pubic symphysis
presenting with hypogastric mass. J Postgrad Med 47:54
Basanagoudar PL, Gupta PN et al (2001) Tuberculosis of the
clavicle presenting as an expansile lytic lesion: a case
report. Acta Orthop Belg 67:505-509
Bass JB, Farer LS et al (1994) Treatment of tuberculosis and
tuberculosis infection in adults and children. American
Thoracic Society and The Centers for Disease Control and
Prevention. Am J Respir Crit Care Med 149:1359-1374
Benkedd~che Y, Gottesmann H (1982) Skeletal tuberculosis of
the hand and wrist: a study of 27 cases. J Hand Surg (Am)
7:593-600
Berney S, Goldstein M et al (1972) Clinical and diagnostic
features of tuberculous arthritis. Am J Med 53:36-42
Bhatt AP, Jayakrishnan A (2001) Tuberculous osteomyelitis of
the mandible: a case report. Int J Pediatr Dent 11:304-308
Bonomo RA, Graham R et al (1995) Tuberculous pyomyositis:
an unusual presentation of disseminated Mycobacterium
tuberculosis infection. Clin Infect Dis 20:1576-1577
Brashear HR, Winfield HG (1975) Tuberculosis of the wrist: a
report of ten cases. South Med J 1345-1349
Bush DC, Schneider LH (1984) Tuberculosis of the hand and
wrist. J Hand Surg (Am) 9:391-398
Cedidi C, Raff T et al (1998) The radial forearm fascial flap as
a salvage procedure for chronic tuberculosis of the hand
and wrist. J Hand Surg 23B:476-478
Centers for Disease Control (1997) Reported tuberculosis cases by
form of disease, US, 1997: reported tuberculosis in U.S., P 60
Chaudhuri MK, Singh Set al (1995) Poncet's disease: tubercu-
lous rheumatism. Indian J Pediatr 62:363-365
Chen WS, Eng HL (1994a) Posterior interosseous neuropathy
associated with tuberculous arthritis of the elbow joint:
report of two cases. J Hand Surg (Am) 19A:611-613
Chen WS, Eng HL (1994b) Tuberculous tenosynovitis of the wrist
mimicking de Quervain's disease. J RheumatoI21:763-765
Chen WS, Wang CJ et al (1997) Tuberculous arthritis of the
elbow.1nt Orthop 21:367-370
Chow SP, Yau A (1980) Tuberculosis of the knee: a long term
follow-up. Int Orthop 4:87-92
Chow SP, Stroebel AB et al (1983) Mycobacterium marinum
infection of the hand involving deep structures. J Hand
Surg 8:568-573
Chow SP, IP FK et al (1987) Mycobacterium marinum infections
of the hand and wrist. J Bone Joint Surg 69A:1161-1168
Chretien J (1990) Tuberculosis and HIV. The cursed duet.
Tuberc Lung Dis 65:25-32
Cohn DL, Catlin BJ et al (1990) A 62-dose, 6-month therapy
for pulmonary and extrapulmonary tuberculosis: a twice
weekly, directly observed, and cost-effective regimen. Ann
Intern Med 112:407-415
Combs DL, O'Brien RJ et al (1990) USPHS Tuberculosis Short-
Course Chemotherapy Trial 21: Effectiveness, toxicity, and
acceptability: The report of final results. Ann Intern Med
112:397-406
Cortez LM, Pankey GA (1973) Mycobacterium marinum infec-
tions of the hand. Report of three cases and review of the
literature. J Bone Joint Surg 55A:363-370
Cramer K, Seiler JG III et al (1991) Tuberculous tenosynovitis
of the wrist: two case reports. Clin Orthop 262:137-140
M. M. Madkour et al.
Cremin BJ (1971) Chronic bone infection in the young. J Bone
Joint Surg 53B:561
Dalldorf PG, Banas MP et al (1994) Tuberculosis of the foot: a
case report. Foot Ankle 15:157-161
Davidson PT,Horowitz I (1979) Skeletal tuberculosis: a review with
patient presentations and discussion. Am J Med 48:77-84
Del Giglio A, Pinczowiski H et al (1997) Tuberculous skeletal
muscle involvement in acute leukemia: report of two cases.
Tumori 83:618-620
Derkash RS, Makley JT (1979) Isolated tuberculosis of the
triceps muscle. J Bone Joint Surg (Am) 61:948
Dhillon MS, Nagi ON (2002) Tuberculosis of the foot and
ankle. Clinical Orthop 398:107-113
Dhillon MS, Rao SS et al (1998) Tuberculosis of the patella.
Skeletal Radiol 27: 40-42
Dhillon MS, Gupta R et al (2000) Bilateral sternoclavicular
joint tuberculosis. Arch Orthop Trauma Surg 120:363-365
Dhillon MS, Gupta RK et al (2001a) Tuberculosis of the sterno-
clavicular joints. Acta Orthop Scand 72:514-517
Dhillon MS, Tuli SM et al (2001b) Osteoarticular tuberculosis
of the foot and ankle. Foot Ankle 22:679-686
Dutt AK, Moers D et al (1986) Short-course chemotherapy for
extrapulmonary tuberculosis: nine years experience. Ann
Intern Med 104:7-12
Dutt AK, Stead WW (1989) Treatment of extrapulmonary
tuberculosis. Semin Respir Infect 4:225-231
Eid A, Chaudry N et al (1994) Multifocal musculoskeletal
cystic tuberculosis without systemic manifestations. Scand
J Infect Dis 26:761-764
Elliott AM, Luo N et al (1990) Impact of HIV on tuberculosis in
Zambia: A cross sectional study. Br Med J 301:412-415
Ellis ME, El-Ramahi KM et al (1993) Tuberculosis of periph-
eral joints: a dilemma in diagnosis. Tubercle Lung Dis 74:
399-404
Esenyel CZ, Bulbul Met al (2000) Isolated tuberculous tenosy-
novitis of the flexor tendon of the fourth of the hand. Case
report. Scand J Plast Reconstr Surg Hand Surg 34:283-285
Eskola A, Santavirta S, Konttinen YT et al (1988) Arthroplasty for
old tuberculosis of the knee. J Bone Joint Surg 70B:767-769
Farnell RD, Davies N et al (2001) Tuberculous compartment
syndrome of the forearm. Orthop 24:393-394
Feldmann JL, Menkes CJ, Weill B et al (1981) Multicentric
study of 214 cases. Rev Rhum 48:83-91
Franco M, Bendini JC et al (2001) Ischiatic tuberculous osteitis
and prolonged fever in a dialysis patient. Joint Bone Spine
68:178-182
Friedman B, Vishwa NK (1970) The management of tuberculo-
sis of the knee joint. Am Rev Respir Dis 101:265-273
Fukasawa H, Suzuki H et al (2001) Tuberculous arthritis mim-
icking neoplasm in a hemodialysis patient. Am J Med Sci
322:373-375
Gamboa F et al (1998) Comparative evaluation of initial and
new versions of the gen-probe amplified mycobacterium
tuberculosis direct test for direct detection of mycobac-
terium tuberculosis in respiratory and non-respiratory
specimens. J Clin Microbiol 36:684-689
Garcea S, Segur JM et al (1994) Tuberculosis of the ischium.
Acta Orthop Belg 60:238-239
Goldberg J, Kovarsky J (1983) Tuberculous sacroilitis. South
Med J 76:1175-1176
Gorbach S, Bartlett J et al (1992) Infectious diseases. Saunders,
Philadelphia, p 1243
Govinder S, Annamalai K et al (2000) Spinal tuberculosis
Extraspinal Musculoskeletal Tuberculosis
in HIV posItive and negative patients: Immunological
response and clinical outcome. Int Orthop 24:163-166
Gupta SK (1982) The treatment of synovial tuberculosis of
the knee by a method with unrestricted activities. Ind J
Orthop 16:14-18
Halsey JP, Reeback JS et al (1982) A decade of skeletal tuber-
culosis. Ann Rheum Dis 41:7-10
Hannanchi M, Martini M et al (1988) Comparison of 3 che-
motherapeutic regimens of short duration (6 months) in
osteoarticular tuberculosis: results after 5 years. Bull Int
Union Tuberc Lung Dis 57:46-47
Hasan N, Baithum S et al (1993) Tuberculosis of striated
muscle. Muscle Nerve 16:984-985
Hoffman KL, Bergman AG et al (1996) Tuberculous tenosyno-
vitis of the flexor tendons of the wrist: MR imaging with
pathologic correlation. Skeletal Radio125:186-188
Hoffmann EB,Allin J et al (2002) Tuberculosis of the knee. Clin
Orthop 398:100-106
Houston S, Fanning A (1994) Current and potential treatment
of tuberculosis. Drugs 48:689-708
Huang CY, Su WJ et al (2001) Childhood tuberculosis present-
ing as an anterior chest wall abscess. J Formos Med Assoc
100:829-831
Hunfeld KP, Rittmeister M et al (1998) Two cases of chronic
arthritis of the forearm due to Mycobacterium tuberculo-
sis. Eur J Clin Microbiol Infect Dis 17:344-348
Ihara K, Toyoda K et al (1998) Tuberculous bursitis of the
greater trochanter. J Orthop Sci 3:120-124
Indudhara R, Singh SK et al (1992) Tuberculous pyomyositis in
a renal transplant recipient. Tuberc Lung Dis 73:239-241
IP FK, Chow SP (1992) Mycobacterium fortuitum infections of
the hand. J Hand Surg 17B:675-677
Ip M, Tsui E et al (1993) Disseminated skeletal tuberculosis
with skull involvement. Tuberc lung Dis 74:211-214
Jacobs JC, Li SC et al (1994) Tuberculous arthritis in children.
Clin Paediatr 33:344-348
Jaovisidha S, Chen C et al (1996) Tuberculous tenosynovitis
and bursitis: imaging findings in 21 cases. Radiology 201:
507-513
Jellis JE (1996) Orthopaedic surgery and HIV disease in Africa.
Int Orthop 20:253-256
Jellis JE (2002) Human immunodeficiency virus and osteoar-
ticular tuberculosis. Clin Orthop 398:27-31
Katayama R, Itami Y et al (1962) Treatment of the hip and
knee-joint tuberculosis: an attempt to retain motion. J
Bone Joint Surg 44A:897-917
Kerri 0, Martini M (1985) Tuberculosis of the knee. Int Orthop
9:153-157
Khanna U, Chowdhary B, Roy D (1980) Tuberculous diaphysi-
tis of long bones. J Indian Med Assoc 75:71-72
Kim JY, Park YH et al (1999) MRI of tuberculous pyomyositis.
J Comput Assist Tomogr 23:454-457
Kim YH (1988) Total knee arthroplasty for tuberculosis arthri-
tis. J Bone Joint Surg 70A:1322-1330
Kim RS, Lee JY, Jung SR, Lee KY (2002). Tuberculous deltoid
bursitis with rice bodies. Yonsei Med J,43:539-542
Kozin SH, Bishop AT (1994) Atypical mycobacterium infection
of the upper extremity. J Hand Surg (Am) 19A:480-487
Kumar K, Saxena MBl (1988) Multifocal osteoarticular tuber-
culosis. Int Orthop 12:135-138
lachenauer CS, Cosentino S et al (1991) Multifocal skeletal
tuberculosis presenting as osteomyelitis of the jaw. Pediatr
Infect Dis J 10:940-944
603
Lee AS, Campbell JAB et al (1995) Tuberculosis of the knee in
children. J Bone Joint Surg 77B:313-3l8
Lee KE (1985) Tuberculosis presenting as carpal tunnel syn-
drome. J Hand Surg (Am) 10A:242-245
Leibert E, Schluger NW et al (1996) Spinal tuberculosis in patients
with human immunodeficiency virus infection: clinical pre-
sentation, therapy and outcome. Tuber lung Dis 77:329-334
Madkour MM (2001). Osteoarticular brucellosis. In: Madkour
MM (ed). Madkour's Brucellosis, 2nd edn. Springer, Berlin
Heidelberg New York
Malhan K, Kumar A et al (2001) Use of polymerase chain reac-
tion in diagnosis of occult tuberculosis of the fibula. Acta
Orthop Belg 67:510-512 .
Manzaneque l, Marin I et al (1992) Osteoarticular tubercu-
losis of the symphysis pubis presenting as a hypogastric
cystic mass in a woman with primary Sjogren's syndrome.
Br J Rheum 31:495-496
Martini M (1988) Tuberculosis of the bone and joints. Springer,
Berlin Heidelberg New York
Martini M, Adjrad A (1988) Tuberculosis of ankle and foot
joints. In: Martini M (ed) Tuberculosis of the bones and
joints. Springer, Berlin Heidelberg New York, pp 138-148
Martini M, Gottesman H et al (1980) Results of conservative
treatment of tuberculosis of the elbow. Int Orthop 4:83-86
Martini M, Adjrad A et al (1986a) Tuberculous osteomyelitis:
a review of 125 cases. Int Orthop 10:201-207
Martini M, Benkeddache Y et al (1986b) Tuberculosis of the
upper limb joints. Int Orthop 10:17-23
Martini M, Bondjeman A et al (1986c) Tuberculous osteomy-
elitis: a review of 125 cases. Int Orthop 10:202-208
Morris BS, Varma R et al (2002) Multifocal musculoskeletal
tuberculosis in children: appearances on computed tomog-
raphy. Skeletal Radio131:1-8
Muradali D, Gold WL et al (1993) Multifocal osteoarticular
tuberculosis: report of four cases and review of manage-
ment. Clin Infect Dis 17:204-209
Murray RO (1954) Observation on cystic tuberculosis of bone.
Proc R Soc Med 47:133
Negusse W (1993). Bone and joint tuberculosis in childhood in
a children's hospital, Addis Abeba. Ethip Med J, 31:51-61
Newton P, Sharp J et al (1982) Bone and joint tuberculosis in
greater Manchester 1969-79. Ann Rheum Dis 41:1-6
Nicklas BJ, latta DK et al (1996) Mycobacterial skeletal infec-
tion of the foot. JAPMA 86:280-287
Ong Y, Cheong PY et al (1998) Delayed diagnosis of tuber-
culosis presenting as small joint arthritis - a case report.
Singapore Med J 39: 177-179
Ormerod lP, Grundy M et al (1989) Multiple tuberculous bone
lesions simulating metastatic disease. Tubercle 70:305-307
Parkinson RW Hodgson SP et al (1990) Tuberculosis of the elbow:
a report of three cases. J Bone Joint Surg (Br) 72:523-524
Patel DA (2001) The supratrochlear lymph nodes: their diag-
nostic significance in a swollen elbow joint. Ann R Coll
Surg Engl 83:425-426
Pouchot J, Vinceneux P et al (1988) Tuberculosis of the sacro-
iliac joint: clinical features, outcome, and evaluation of closed
needle biopsy in 11 consecutive cases. Am J Med 84:622-628
Pouchot J, Vinceneux P et al (1990) Tuberculous polymyositis
in HIV infection. Am J Med 89:250-251
Ragni MV, Crosset lS et al (1995) Postoperative infection fol-
lowing orthopaedic surgery in human immunodeficiency
virus-infected hemophiliacs with CD4 counts < or = 2001
mm 3• J Arthroplasty 10:716-721
604
Rasool MN (2001) Osseous manifestations of tuberculosis in
children. J Pediatr Orthop 21:749-755
Rasool MN, Govender S et al (1994) Cystic tuberculosis of
bone in children. J Bone Joint Surg 76B:113-117
Resnick D (1988) Diagnosis of bone and joint disorders, vol 4,
2nd edn. Saunders, Philadelphia, pp 2661-2686
Rieder HL, Snider DE Jr, Cauthen GM (1990) Extrapulmonary
tuberculosis in the United States. Am Rev Respir Dis 141:
347-351
Rossmann MD, MacGregor RR (1995) Tuberculosis: clinical
management and new challenges. McGraw-Hill, New York,
pp 182-183
Ruggieri M, Pavone V et al (1997) Tuberculosis of the ankle
in childhood: clinical, roentgenographic and computed
tomography findings. Clin Paediatr 36:529-534
Sant M, Bajaj H (1991) Role of histopathology in the diagnosis
of tuberculous synovitis. J Indian Med Assoc 90:263-264
Sarlak AY, Giindes H et al (2001) Primary sternal tuberculosis:
a rare unhealed case treated by resection and local rota-
tional flap. Thorac Cardiovasc Surg 49:58-59
Shah AA, Ahmed S et al (1992) Bone tuberculosis in Abbot-
tabad. JPMA 42:180-181
Shanmugasundaram TK (1983) Tuberculosis of hip. In:
Shanmugasundaram TK (ed) Bone and joint tuberculosis.
Kothandaram, Madras, India, pp 59-72
Silva JF (1980) A review of patients with skeletal tuberculo-
sis treated at the University Hospital, Kuala Lumpur. Int
Orthop 4:79-81
Singh SB, Saraf SK et al (1992) Osteoarticular tuberculosis in
children. Indian Paediatr 29:1133-1137
Sinnott JT, Cancio MR et al (1990) Tuberculous osteomyelitis
masked by concomitant staphylococcal infection. Arch
Intern Med 150:1865-1868
Sipsas NV, Panayiotakopoulos GD et al (2001) Sternal tuber-
culosis after coronary artery bypass graft surgery. Scand J
Infect Dis 33:387-388
Skoll PJ, Hudson DA (1999) Tuberculosis of the upper extrem-
ity. Ann Plast Surg 43:374-378
Soholt ST (1951) Tuberculosis of the sacro-iliac joint. J Bone
Joint Surg (Am) 33:119-130
Soler R, Rodriguez E et al (2000) Magnetic resonance imaging
of pyomyositis in forty-three cases. Eur J RadioI35:59-64
Soler R,Rodriguez E et al (2001) MRI of musculoskeletal extra-
spinal tuberculosis. J Comput Assist Tomogr 25:177-183
Stecher DR, Gusis SE et al (1992) Tuberculous arthritis in the
course of connective tissue disease: report of 4 cases. J
RheumatoI19:1418-1420
Sueyoshi E, Uetani M et al (1996) Tuberculous tenosynovitis
of the wrist: MR findings in three patients. Skeletal Radiol
25:569-572
Sub JS, Lee JD et al (1996) MR imaging of tuberculous arthritis:
clinical and experimental studies. J Magn Res Imag 1:185-189
Supe AN, Prabhu RY et al (2002) Role of computed tomography
in the diagnosis of rib and lung involvement in tuberculous
retromammary abscesses. Skeletal RadioI31:96-98
Suso S, Peidro L et al (1988) Tuberculous synovitis with "rice
bodies" presenting as carpal tunnel syndrome. J Hand Surg
(Am) 13A:573-576
Tenth Report of the Medical Research Council Working Party
M. M. Madkour et al.
on Tuberculosis of the Spine (1986) A controlled trial of
six-month and nine-month regimens of chemotherapy in
patients undergoing radical surgery for tuberculosis of the
spine in Hong Kong. Tubercle 67:243-259
Toda K, Yasunaga Y et al (1998) MR image and CT scan of a
tuberculous abscess in the gluteus maximus muscle. Comput
Med Imag Graph 22:425-427
Tsay MH, Chen MC et al (1995) Atypical skeletal tuberculosis
mimicking tumor metastases: report of a case. J Formos Med
Assoc 94:428-431
Tuli SM (1997) Tuberculosis of the Skeletal System (Bones,
Joints, Spine and Bursal Sheaths), 2nd edn. Jaypee Brothers
Medical, New Delhi
Tuli SM, Sinha GP (1969) Skeletal tuberculosis: 'unusual lesion'.
Ind J Orthop 3:5-18
Valdazo JP, Ruiz FP et al (1990) Tuberculous arthritis. Report
of a case with multiple joint involvement and periarticular
tuberculous abscesses. J RheumatoI17:399-40 1
Veerken H, Bermejo A (1992) The impact of the HIV epidemic
on tuberculosis control programmes in developing coun-
tries. Trop Doc 22:64-67
Versefeld GA, Solomon A (1982) Diagnostic approach to tuber-
culosis of bones and joints. J Bone Joint Surg Br 64:446-449
Visuthikosol V, Kruavit A et al (1996) Tuberculous infection of
the hand and wrist. Ann Plast Surg 37:55-59
Vohra R, Kang HS (1995) Tuberculosis of the elbow: a report of
ten cases. Acta Orthop Scand 66:57-58
Wang CT, Sun JS et al (2000) Mycobacterial infection of the
upper extremities. J Formos Med Assoc 99:710-715
Watts HG, Lifeso RM (1996) Current concepts review: tuber-
culosis of bones and joints. J Bone Joint Surg (Am) 78:
288-298
Weir MR, Thorton GF (1985) Extrapulmonary tuberculosis:
experience of a community hospital and review of the lit-
erature. Am J Med 79:467-478
Wessels G,Hesseling PB et al (1998) Skeletal tuberculosis: dactyli-
tis and involvement of the skull. Pediatr RadioI28:234-236
Wilkinson AG, Roy S (1984) Two cases of Poncet's disease.
Tubercle 65:301-303
Wilkinson MC (1969) Tuberculosis of the hip and knee treated
by chemotherapy, synovectomy, and debridement: a follow-
up study. J Bone Joint Surg 51A:1343-1359 .
Williams CS, Riordan DC (1973) Mycobacterium marinum
(atypical acid-fast bacillus) infections of the hand: a report
of six cases. J Bone Joint Surg 55A:I042-1050
Wilson IN (1953) Tuberculosis of the elbow: a study of 31 cases.
J Bone Joint Surg (Br) 35:551-560
Wobeser WL et al (1996) Evaluation of Roche Amplicor PCR assay
for mycobacterium tuberculosis. J Clin MicrobioI34:134-139
Wolinsky E (1994) Conventional diagnostic methods for tuber-
culosis. Clin Infect Dis 19:396-401
Wright T, Sundaram M et al (1996) Tuberculous osteomyelitis
and arthritis. Orthopedics 19:700-702
Yao DC, Sartoris DJ (1995) Musculoskeletal tuberculosis. Radiol
Clin North Am 33:679-689
Yip KMH,Lin J et al (1996) Cystic tuberculosis of the bone mim-
icking osteogenic sarcoma. Tuberc Lung Dis 77:566-568
Yoon CT, Chung HW et al (2001) MR findings of tuberculous
dactylitis. Eur J RadioI39:163-167
3S Imaging of Musculoskeletal Tuberculosis
MOHAMMED ABD EL BAGI, MONA AL SHAHED, M. MONIR MADKOUR

CONTENTS 35.1
Introduction
35.1 Introduction 605
35.1.1 Overview 605
35.1.2 Pathology 606
35.1.1
35.1.3 Clinical Presentation 608 Overview
35.1.3.1 Laboratory Tests 609
35.1.4 Classification of Musculoskeletal Tuberculosis is one of the rare bone infections (RBI).
Tuberculosis 609 This term refers to conditions where only a small
35.1.4.1 Tuberculosis Arthritis 609
35.1.4.2 Tuberculosis Osteomyelitis 609
percentage of cases have bone manifestations or
35.1.4.3 Tuberculosis of Muscles 610 where the infective organism is altogether rare (Abd
35.1.4.4 Tuberculous Tenosynovitis 610 El Bagi et al. 1999). Incidence of RBI can increase if
35.1.4.5 Tuberculous Bursitis 610 predisposing factors prevail: for example, diabetes,
35.1.4.6 Cystic Tuberculosis of Bone 610 immunocompromise, drug abuse, steroids therapy,
35.2 Imaging 611
35.2.1 Plain Radiographs 611
immigration, undernourishment, and overcrowding.
35.2.2 Sinogram 612 Osteoarticular tuberculosis is rare, occurring in 1-3%
35.2.3 Isotope Scanning 614 of all tuberculosis patients and is present in 30% of all
35.2.4 Ultrasound 614 extrapulmonary tuberculosis cases (Engin et al. 2000).
35.2.5 Computed Tomography (CT) 614 Resurgence of tuberculosis is blamed on HIV infec-
35.2.6 Magnetic Resonance Imaging (MRI) 616
35.3 Radiologic Presentation 616
tion and the development of multidrug resistance.
35.3.1 Tuberculosis of the Knee 616 Following the decline of tuberculosis in the second
35.3.2 Tuberculosis of the Hip 617 half of the last century in response to new efficient
35.3.3 Tuberculosis of the Pelvis chemotherapeutic agents, osteoarticular tuberculosis
and Sacroiliac Joints 618 was often overlooked, leading to a delay in diagnosis
35.3.4 Chest Wall Tuberculosis 618
35.3.5 Tuberculosis of the Hands and Feet 620
sometimes of many years (Yao and Sartoris 1995).
35.3.6 Tuberculosis of the Elbow 621 Skeletal tuberculosis is almost invariably secondary.
35.3.7 Tuberculosis of the Shoulder 621 The original site is usually intrapulmonary or in the
35.3.8 Tuberculosis of the Ankle 622 mesenteric glands. Occasionally the bone infection is
35.4 Interventions 622 by contiguity from a nearby joint or infected soft tis-
35.4.1 Biopsies 622
35.4.2 Drainage 623
sues. In some instances there is no apparent primary
35.5 Summary 623 lesion. Skeletal involvement may follow the initial con-
References 624 stitutional symptoms by 1-2 years. However, we have
seen cases where the bone lesions preceded the chest
disease. In one of our patients, miliary tuberculosis
developed 6 months after the onset of knee infection.
M. ABD EL BAGI, MB BCh, DMRD, FSRRCSI Concomitant pulmonary tuberculosis was reported in
Senior Consultant, Department of Radiology, Riyadh Armed
Forces Hospital,
12-50% of cases (Hugosson et al. 1996). This is more
P.O. Box 7897, Riyadh 11159, Saudi Arabia common in children.
M. M. MADKOUR, MD, DM, FRCP Bone lesions are usually solitary because sensitiza-
Consultant, Department of Medicine, Riyadh Armed Forces tion of the patient to the tubercle bacilli occurs before
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia the onset of skeletal lesions (Kumar and Saxena 1988).
M. AI SHAHED, MBBS, FRCR
Senior Consultant Radiologist, Department of Radiology, Riyadh
Multiple lesions can occur in 1-15% of cases (Watts
Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi and Lifeso 1996). Any bone in the body can be affected.
Arabia The commonest bone involved is the spine in 50-70%

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
606 M. Abd EI Bagi et al.

of cases, particularly the thoracic vertebrae in 50%,
but lumbar vertebrae involvement is not uncommon
(Table 35.1). Actually the first lumbar vertebra was
considered to be the most common site (Resnick and
Niwayama 1988). Spinal tuberculosis is discussed in
detail elsewhere. The articulation of the lower limbs
is more commonly affected by tuberculosis than that
of the upper limps. The commonest peripheral site of
extraspinal tuberculosis is the knee (Abd EI Bagi et al.,
2002). Previously the hip was thought to be the most
commonest (Hugosson et al. 1996). These two large
joints of the hip and knee together represent 10% of all
cases of osteoarticular tuberculosis, while small joints
of the hand, ankle, and foot are infected in 2-4% of cases
(McGuiness 1999). Trauma has been associated with
tuberculosis in 30-50% of cases (Rasool et al.1994).
The pattern of osteoarticular tuberculosis has
changed over the years (Resnick and Niwayama 1988).
Initially the disease was usually encountered in chil-
dren and young adults. Currently patients of all ages
Fig.35.1. Multifocal TB arthritis. Frontal radiograph of the
pelvis on a 50-year-old female complaining of chronic pain
in low back and pelvic bones with progressive discomfort on
walking. There is widening, erosion, and periarticular sclero-
sis of the right sacroiliac joint (small arrows). Both sides-the
iliac and sacral-of the joint are involved. There is asymmetry
of the pubic symphysis associated with marginal erosions,
widening of the joint space, and periarticular reactive scle-
rosis (long arrow)
are affected if there are predisposing causes. The dis-
ease is rare in the first year oflife. All forms of tubercu-
10sis are more common in the young where the disease
is prevalent, and less common in the elderly where the
disease is rare. Osteoarticular tuberculosis affects
both sexes equally. Spinal disease is more common in
males. Monoarticular disease is the rule (Resnick and
Niwayama 1988), but we have seen cases with multiple
joint involvement (Fig. 35.1) (Muradali 1993).
In adults, tuberculosis usually originates in the
epiphyses and spreads to the neighboring joint.
Actual joint involvement is slower than in pyogenic
infection or brucellosis, due to the lack of proteolytic
enzymes (Fig. 35.2a-d) (Davidson and Horowitz
1970). Tuberculosis very rarely arises in the shaft of
long bones. In children it starts in the metaphyses due
to the increased vascularity in this region (Fig. 35.3)
(Edeikin et al.I963). It may however cross the physes
into the epiphyses (Fig. 35.4), which is an important
sign differentiating it from pyogenic infection.
Bone infection with acid-fast tuberculosis bacillus
causes a destructive, caseating necrosis leading to
cold abscess formation and sinus tracts. The radio-
logic signs of bone tuberculosis are variable. Tuber-
culosis can mimic pyogenic infections, traumatic
lesions, collagen, or degenerative disease (Goldblatt
and Cremin 1978).At times it could be confused with
tumors (Abdel Wahab et al. 1991). Although bone
tuberculosis may present in classical well-described
radiologic changes, a significant number may present
in a bizarre fashion taxing the diagnostic acumen of
the radiologist (Goldblatt and Cremin 1978).
35.1.2
Pathology
Various body organs respond to the tuberculosis bacil-
lus differently. The Mycobacterium usually reaches
the bone marrow via hematogenous seeding. Bone

Table 35.1. The distribution of musculoskeletal tuberculosis

Authors Davidson and Goldblatt and Watts and Hugosson

Horowitz (1970) Cremin (1978) Lifeso (1996) et al. (1996)

Geographic location USA Colored South Africans USA Middle East

Spine 50% 59.9% 50% 71.5%
Hip 15% 10.6% 5% 3.4%
Knee 15% 8.2% 5% 11.3%
Pelvis 4.2% 12% 4.5%
Femur and tibia 5% 10% 2.2%
Ankle and foot 2.6% 8% 4.5%
Others 20% 9.5% 8% 2.6%
Multiple sites 8.2%
 Imaging of Musculoskeletal Tuberculosis 607

a c

b ~----~"'d

Fig. 35.2. a TB osteomyelitis of the right tibia in an adult. Right ankle radiographs of a 63-year-old diabetic male complaining
of joint pain and swelling for 2 years. He received various medical treatments. There is a soft tissue swelling (long white arrows)
and periarticular osteoporosis. Note the large erosion above the medial malleolus (open arrowhead) and the smaller erosion in
the lateral malleolus and talus (small arrowheads). There is a faint periosteal reaction (small arrows). There is no sequestrum.
The joint space has been spared any damage. b Lateral radiograph. The soft tissue swelling is surrounding the tendo calcaneus
(long white arrows). Vascular calcification is present due to the diabetes. c Follow-up frontal radiograph after 6 months of treat-
ment. The cortex is consolidated (arrow) and the periostitis has disappeared. Yet, there is a residual multilobular cystic area in
the tibia and a smaller subarticular lesion in the talus. d Lateral radiograph showing resolution of the previous large soft tissue
swelling, which was surrounding the tendo calcaneus (a). Vascular calcification is more apparent

infection could be initiated by trauma. Either overt or
occult trauma can lead to local hemorrhage, vascular
stasis, or cellular effusion, creating a suitable medium
for bacterial growth. The initial process is an endar-
teritis. The tissue reaction to the tuberculosis bacillus
forms a small follicle, the "tubercle:' Centrally these
are multinucleated giant cells surrounded by clusters
of epithelial cells with elongated nuclei. Peripherally
there are abundant lymphocytes and mononuclear
cells whose presence is characteristic of tuberculosis.
The bacillus produces "tuberculin:' which incites
the process of caseating necrosis. This causes resorp-
tion of the trabecular lamellae. It should be noted
that epithelial granulomas or granulomatous bone
608
Fig. 35.3. TB osteomyelitis of the lower end of the right tibia
in a child. Frontal radiograph of a child complaining of pain
and swelling above the right ankle not responding to antibi-
otic treatment. There is a large metaphyseal lytic lesion with a
periosteal reaction, subperiosteal bone deposition, and a soft
swelling tissue. The periostitis is not florid and there are no
sizable sequestra
lesions are not pathognomonic for tuberculosis.
They represent a histologic nonspecific response
to antigenic stimuli and can be seen in brucellosis,
sarcoidosis, lymphoma, neoplasm, or autoimmune
disorders (Resnick and Niwayama 1988). Such gran-
ulomatous reactions are not always proven as tuber-
culosis. The bacilli are scarce at the infected sites.
Not uncommonly the diagnosis is made on clinical,
laboratory, and radiologic correlation. Yet, biopsy
and aspirations are mandatory. However, biopsies are
not always positive (Davidson and Horowitz 1970).
Some clinicians would repeat the biopsy to assess
the response of a doubtful granuloma to treatment.
Regression of a suspicious granuloma on treatment
is considered evidence of tuberculosis. In countries
where tuberculosis is highly prevalent and where
medical facilities are limited, findings suggestive of
the disease may not have a biopsy in order for the
physician to make the diagnosis and institute treat-
ment. There, biopsy is reserved for cases that do not
respond to treatment (Watts and Lifeso 1996).
M. Abd El Bagi et al.
Fig. 35.4. Metaphyseal TB focus eroding the physis into epiphy-
ses. Frontal radiograph of the knee of a child who presented
with knee pain and swelling. There is a small lytic lesion in
the lateral aspect of the lower femoral metaphyses surrounded
by minimal reactive sclerosis (black arrow heads). There is
another well-defined lytic lesion at the lateral margin of the
lower femoral epiphyses (long white arrow). The infection has
crossed the physeal plate, a feature which indicates TB rather
than pyogenic infection
Agranuloma takes a few days to develop and become
visible even to the naked eye as a white nodule in the
bone marrow (Dutchie and Nelson 1996). This may
slowly regress or heal by fibrosis and fatty replacement.
Progression leads to caseating necrosis, which may form
a cold abscess (Glassroth 1993). Periosteal hyperemia
occurs at a very early stage in the disease. This stage is
usually missed by the time imaging is performed. The
process ultimately leads to periosteal thickening and
bone deposition subperiosteally (Fig. 35.3).
35.1.3
Clinical Presentation
There is no typical clinical presentation for muscu-
loskeletal tuberculosis. The clinical signs and symp-
toms of osteoarticular tuberculosis could be:
1. Localized signs: pain, swelling, tenderness, limita-
tion of movements, joint stiffness, muscle wasting,
and sinus formation.
Imaging of Musculoskeletal Tuberculosis 609

2. Systemic symptoms of tuberculosis infection:
fever, chills, night sweats, weight loss, anorexia, or
malaise.
3. Signs due to concomitant disease: e.g., cough and
hemoptysis in pulmonary disease, or kyphosis
and paraplegia in spinal disease.
4. Atypical presentation: e.g., pyrexia of unknown
origin (PUQ).
protein (CRP) and polymerase chain reaction (PCR)
are more sophisticated tests and claim higher accu-
racy. Another sensitive new test is the enzyme-linked
immunosorbent assay (ELISA).
35.1.4
Classification of Musculoskeletal Tuberculosis

Pain and swelling are the most common present- 35.1.4.1

ing symptoms. Joint pains may be greater at night
when the protective muscle spasm is relaxed and
inflamed surfaces may rub over each other. Swell-
ings are usually cold in contrast to septic arthritis,
brucella, or pyogenic abscesses. Soft tissue swelling
could be due to a joint effusion, bursitis, inflamma-
tory mass, lymphadenopathy, or subcutaneous cold
abscess. Muscle weakness or wasting is noted in
chronic or neglected cases. Sinus formation is a sign
of skin disruption by expanding cold abscess. We had
cases where a sinus was the first presentation. Patho-
logic fractures are rare presentations. Joint stiffness,
limb shortening, and ankylosis are late sequelae.
Tuberculosis Arthritis
Tuberculosis arthritis is commonly monoarticular.
It usually presents as gradually worsening arthri-
tis (Fig. 35.5). Sometimes a soft tissue swelling is
noticed. In many cases the use of nonsteroidal
anti-inflammatory drugs offers some response and
a false sense of security (Vohra et al. 1997). The
reactive form of tuberculous arthritis, "Poncet's
rheumatism:' can mimic juvenile rheumatoid
arthritis (Wihlborg et al. 2001). This identification
was, however, previously questioned (Resnick and
Niwayama 1988).

35.1.3.1 35.1.4.2
Laboratory Tests Tuberculosis Osteomyelitis

These are discussed in detail elsewhere. The simplest Pain and swelling are the presenting features. Regional
are ESR and the Mantoux test. Special stain micros- lymphadenopathy can be seen. Peripheral osteoarticu-
copy and guinea pig cultures are routine. C-reactive lar tuberculosis predominantly involves the bones in

Fig. 35.5. a TB of the talus. Radiograph of a 29-year-old female who presented with
joint pain, limitation of movement, and fever. There was no obvious soft tissue swell-
ing. Frontal view showing a cystic erosion in the medial side of the talus (arrowhead)
associated with narrowing of the ankle joint. Note scalloping of the medial margin of
the distal end of the fibula (open arrows). There is no periosteal reaction. b Lateral
radiograph confirming presence of the cystic erosion in the talus (arrow). Patient
had previous temporary releif by steroid treatment

a b
610
combination with joints in 62-84% of cases, while
purely osseous lesions without joint involvement
occur in 16-38% (Hugosson et al. 1996). Tuberculous
osteomyelitis may not show radiographic changes in
the early stage. Radioisotope studies are more sensitive.
An eccentric osteolytic lesion with little or no reactive
bone is characteristic (Vohra et al.1997) (Fig. 35.5a, b).
Sclerosis and periostitis are less than those occurring
in pyogenic osteomyelitis. Presence of a sinus that does
not heal on antibiotic treatment should raise the pos-
sibility of tuberculous osteomyelitis. Biopsy is manda-
tory under all circumstances. Location of tuberculous
osteomyelitis could be epiphyseal, metaphyseal, or in
the diametaphysis.
35.1.4.3
Tuberculosis of Muscles
Tuberculosis of skeletal muscles is an unusual finding
even in patients with widespread disease (Derkash and
Makley 1979). The psoas muscle is the exception. A
cold abscess may form (Fig. 35.6). Primary hematog-
enous tuberculosis of the muscle is very rare. Pain is
not a feature. The disease is not usually associated with
miliary tuberculosis (Wilbur et al. 1995). Localized
swelling is the usual presentation. Slowly developing
cellulitis and muscle weakness are important. Tuber-
culous myositis should be considered in the expanding
spectrum ofHIV-infection manifestations in the pres-
ence of muscular symptoms (Pouchot et al.1990). The
differential diagnosis of tuberculous myositis includes
polymyositis, bacterial infections, lymphoma, and
Kaposi's sarcoma particularly in AIDS patients.
35.1.4.4
Tuberculous Tenosynovitis
Previously, invasive contrast studies like arthrogra-
phy, tenography, or bursography were the only ways
to examine tendons and bursae (Jaovisidha et al.1996).
Modern imaging enables explicit demonstration of
these soft tissue structures. Tuberculous tenosynovitis
has a gradual clinical onset. It presents as limitation
of movement and soft tissue swelling. The condition
can be confused with soft tissue tumors. High resolu-
tion ultrasound and MRI provide useful information.
Tendon sheath of the hands and wrist are the most
common site for tuberculous tenosynovitis.
Kanavel described the various stages of tubercu-
lous tenosynovitis (KanaveI1923). The earliest is the
hygromatous "watery phase" followed by the "case-
ous" serofibrinous phase. The fungoid "cheesy" form
is the third stage.
M. Abd El Bagi et al.
35.1.4.5
Tuberculous Bursitis
Primary tuberculous bursitis is an important clini-
cal entity. The most common site is at the greater
trochanter of the femur (Fig. 35.7). MRI is particu-
larly useful in demonstrating bursitis (Fig. 35.8). A
histologic diagnosis should be reached by aspiration
and biopsy to exclude nontuberculous infections
and collagen diseases (Fig. 35.9). Tuberculous bursi-
tis and tenosynovitis have seldom been described in
the literature Oaovisidha et al. 1996).
35.1.4.6
Cystic Tuberculosis of Bone
This type of osteomyelitis is reported more in chil-
dren than adults (Harisinghani et al. 2000). According
to Rasool the multicystic variety is more common
than the solitary cystic lesions (Rasool et al. 1994).
He described the confusion with Jungling's disease or
osteitis tuberculosa multiplex cystoids of sarcoidosis
which he considered to be a manifestation of tubercu-
losis. These lesions were renamed as osteitis multiplex
cystoids sarcoidosa by Komins in 1952 (Rasool et al.
1994). This multicystic type of presentation is rare
in our experience. We see more of the solitary cystic
from (Figs. 35.3, 35.5, 35.10). Cystic tuberculosis is in
the differential diagnosis of other benign cystic lesions,
eosinophilic granuloma, plasmacytoma, and tumors.
Tuberculous cysts are well defined and may not have
a sclerotic rim. They could be single or multiple and
may expand in size. In children these are common in
long bones while in adults they are commoner in flat
bones like the pelvis and the scapula. They tend to be
Fig. 35.6. Intramuscular cold abscess. A young adult male
who presented with pain and swelling in the left groin. Axial
postcontrast CT scan of the upper third of the thigh showed
multilocular cold abscess with enhancing rim and fluid con-
tent distending the adductor muscles
Imaging of Musculoskeletal Tuberculosis 611

Fig. 35.8. Subtle TB bursitis. T2-weighted MRI sequence of the

hips at the level of the greater trochanters. Well-defined hyper-
intense fluid collection in the left trochanteric bursa (black
Fig. 35.7. Subtle TB bursitis. A young adult male who presented arrow head). This MRI belongs to the same patient shown in
with chronic left hip pain. On the axial CT scan of the hips at Figs. 35.7 and 35.9
the level of the greater trochanters, there is slight distension,
thickening, and irregularity of the walls of the left greater tro-
chanter bursa (arrowhead). See Fig. 35.8 for MRI image and
Fig. 35.9 for diagnostic aspiration of the same patient

smaller and less expansile in adults. The bone expan-

sion is due to formation of granulation tissue. Cystic
tuberculosis does not necessarily mean that these are
fluid-filled lesions (Cremin and Jameism 1995). The
term "pseudocystic" tuberculosis was used previously
(Clinton and Young 1955).

35.2
Imaging

35.2.1
Plain Radiographs

Plain radiography remains the cornerstone for imag-
ing of tuberculosis infection of bones and joints
(Ridley et al. 1998). However, no specific radiographic
feature is pathognomonic for musculoskeletal tuber-
culosis. Therefore nonradiologic evidence is always
necessary to confirm diagnosis. Signs include soft
tissue swelling, osteoporosis, and osteosclerosis.
Appreciation of soft tissue swelling or displacement
of fat planes could be the earliest sign (Figs. 35.11-12).
Osteoporosis develops at an early stage due to hyper-
emia. Subtle narrowing of the joint space and faint ero-
sions are also early signs (Fig. 35.14). Osteosclerosis is
characteristic of the chronic form of the disease. The
joint space may be widened initially, but it is usually
normal in contrast to rheumatoid disease where joint
space narrowing is the rule. Erosions and cystic lesions
eventually develop (Fig. 35.15). Sequestra are usually
small when they occur and can be seen in oppos-
ing articular surfaces (so-called kissing sequestra,
Fig. 35.9. CT-guided aspiration biopsy for the left trochanteric
bursa for the patient shown in Figs. 35.7 and 35.8. The procedure
was performed under local anesthesia. Aspirate was positive for
TB culture. Note position of the needle (white arrows) within
the small infected trochanteric bursa (black arrowhead)
see Fig. 35.15). Periosteal reaction is usually limited
(Figs. 35.2, 35.3). Later in the disease the Phemister
triad becomes evident, namely osteoporosis, mar-
ginal erosions, and joint space narrowing (Figs. 35.15,
35.16) (Resnick and Niwayama 1988). Soft tissue cal-
cification can be seen on plain radiographs but CT is
more sensitive (see Fig. 35.22b). Bony ankylosis may
develop at a later stage. In rheumatoid disease, joint
space narrowing is early and predominant. Rheuma-
toid disease is usually symmetric and bilateral. Gout
does not provoke osteoporosis. Pyogenic infections are
more rapidly progressing than tuberculosis and lead
612
Fig. 35.10. Cystic TB osteomyelitis. Lateral radiograph of the
right knee of an adult male who presented with a discharging
sinus. There is a large well-defined lytic lesion eroding the
upper end of the tibia and sparing the epiphyses. Little reactive
sclerosis is present. No significant periosteal reaction, which
is commoner in tumors and pyogenic osteomyelitis. No large
sequestra. (see Fig. 35.18a-c)
Fig.35.11. Early radiographic signs of TB arthritis of the
left knee. Frontal radiograph of the knees of a child who
presented with limping, pain, and swelling. There was no his-
tory of trauma. Soft tissue swelling due to a joint effusion was
evident on the plain films (black arrow heads) before any bony
changes could be visible. This finding is nonspecific. Isotope
bone scan was suggestive of focal infection (see Fig. 35.12).
Diagnosis was confirmed by aspiration and culture
M. Abd El Bagi et al.
to early joint space narrowing and more periostitis.
In pigmented villonodular synovitis the joint space
is preserved and there is no osteoporosis. MRI shows
characteristic signs. Brucellosis may be coendemic
with tuberculosis but synovial infection is the earliest
and most significant manifestation (AI Shahed et al.
1994). In children, the late occurring synovial involve-
ment leads to chronic hyperemia with overgrowth of
the epiphyses and premature fusion. This finding may
simulate juvenile rheumatoid arthritis (Haygood and
Williamson 1994).
35.2.2
Sinogram
The skin overlying a superficial cold abscess becomes
progressively thinner and may ultimately yield, form-
ing a sinus (Fig.35.17). Injection of water-soluble
contrast media under fluoroscopy can demonstrate
the size, depth, and direction of the sinuses. These
may form fistulae if adjacent to a hollow viscus.
Fig. 35.12. Isotope bone scan for detecting early signs of
TB infection. A three-phase 99mTc MDP isotope scan for the
child shown in Fig. 35.11 was positive for infection. There is
increased activity in the vascular phase as well as in the blood
pool phase (black arrowhead)
Imaging of Musculoskeletal Tuberculosis 613

a"'__ b

Fig. 35.13. a Lateral radiograph of the left elbow for a 47-year-old male who presented with slowly developing joint swelling and limi-
tation of movement. There is a large swelling of soft tissue in the arm (arrows) associated with elevation of the anterior and posterior
fat pads (open arrowheads). The elevation of fat pads is commonly seen with post-traumatic joint effusion. Minimal haziness of the
olecranon tip was missed (open black arrow). b Same patient shown in Fig. 35.l3a, after 2 months. An erosion at the olecranon tip
has become apparent (arrow). The soft tissue swelling of the lower arm and the elevation of the fat pads have decreased

Fig. 35.15. Late findings on plain film of advanced TB arthritis.

Fig. 35.14. Early findings on plain film of TB arthritis. Frontal
radiograph of the left knee of an adult with chronic knee pain.
There was slight narrowing of the medial joint compartment.
Small marginal erosion is noted (arrow). Such findings can be
missed. A high level of clinical vigilance is necessary to initiate
relevant tests to exclude low-grade infection like TB
Frontal radiograph of the right knee of an adult who suffered
chronic knee pain. There is periarticular osteopenia,joint space
narrowing, and numerous marginal erosions. These together
constitute the Phemister triad. Small "kissing" sequestra are
noted at the lateral compartment of the knee joint originating
from erosion of the opposing articular surfaces
614
Fig.35.16. TB of the right hip. Hip radiograph of a 53-year-old
male who presented with right hip pain, limping, and night
sweats. The radiograph shows osteoporosis narrowing of the joint
space and numerous articular surface erosions (open arrows).
Note the large swelling of soft tissue (long arrows). These find-
ings represent the Phernister triad. Diagnosis was confirmed by
aspiration biopsy
35.2.3
Isotope Scanning
A three-phase technetium methylene diphosphonate
99m
( Tc MDP) scan is usually the first test performed
to diagnose osteoarticular infection whether due to
tuberculosis or otherwise (Figs. 35.12, 35.18). It is a
sensitive test but a negative scan cannot exclude pres-
ence of tuberculous bone infection. Indium-labeled
white blood cells in WBC scan, technetium hexa-
methylenepropyleneamine T2 oxime (99m Tc HMPAO),
and the nonspecific indium-labeled immunoglobulin
scan are also widely used to localize infections. When
bone infection is not the initial presentation in patients
known to have tuberculosis or where diagnosis of
tuberculosis is not established, e.g., cases of pyrexia
of unknown origin (POU), gallium citrate 67Ga scan is
the test of choice. The use of 67Ga in tuberculosis was
well described by Hamilton and Nabulsi (1999). It can
detect skeletal and all extraskeletal sites of tuberculosis
infection including lymph nodes and visceral lesions
but is nonspecific for infections and can cause these to
be confused with lymphoma and malignancies.
Isotope scanning is more sensitive than morphologic
imaging but findings are nonspecific. They should be
reviewed in conjunction with clinical signs and other
imaging findings. A specific advantage is the ability to
M. Abd El Bagi et al.
Fig. 35.17. Contrast sinogram. An adult male presented with
chronic knee pain and a discharging sinus. Lateral radiograph
of the knee after administration of water-soluble contrast
medium via a soft catheter advanced through the sinus open-
ing at the fossa poplitea. Note the large erosion in the upper
end of the tibia and multiple ill-defined erosions in the lower
end of the femur due to TB arthritis
demonstrate multiple sites of osteoarticular involve-
ment. In this respect, scintigraphy has totally replaced
radiographic skeletal surveys. Positive multiple foci
should be differentiated from metastatic disease.
35.2.4
Ultrasound
Ultrasound is a quick, simple, inexpensive, widely
available, non-ionizing imaging technique that can
easily be used in different clinical settings, including
the emergency department, radiology department,
and intensive care unit (Bureau et al. 1998). High
resolution probes are now available providing excel-
lent soft tissue details for detection of joint and peri-
articular pathology including joint effusions, bursitis,
cold abscesses, and soft tissue calcification (Fig. 35.19).
Ultrasound helps to exclude thrombophlebitis, which
can be confused with bone and joint infections.
35.2.5
Computed Tomography (CT)
Computed tomography (CT) is superior to MRI in
demonstrating details of bone destruction, detec-
 Imaging of Musculoskeletal Tuberculosis 615

a b

Fig.35.18a-c. Isotope bone scan. a Vascular phase of 99mTc

MDP isotope bone scan vascular phase showing increased
tracer accumulation at the TB infection site in the upper
end of the right tibia as early as the arterial phase first pass.
b Blood pool phase taken after 5 min of radiotracer administra-
tion showing a doughnut appearance of progressive increased
tracer accumulation. c Mediolateral view in the delayed static
phase showing persistent increased tracer accumulation at the
infected site in the upper end of the right tibia. This isotope
bone scan belongs to the patient shown in Fig. 35.10 c

Fig.35.19. Subcutaneous cold abscess. An adult male patient

presented with chronic swelling of the right buttock and
loss of weight. Ultrasound scan of the gluteal region show-
ing a well-defined cold abscess with thick fluid contents and
multiple wall calcifications. Note the acoustic shadow of the
calcified granuloma adherent to the posterior wall of the cold
abscess (similar to the effect of gall stones)
 616 M. Abd El Bagi et al.

tion of small sequestra, and minute soft tissue cal- 35.2.6

cification (see Fig. 35.22a, b below). It is particularly
important in patients where MRI is contraindicated.
CT has a broader spectrum than MRI by demonstrat-
ing subcutaneous, intraosseous, bowel involvement,
lymphadenopathy lesions, and lung disease in a
single extended exam.
Magnetic Resonance Imaging (MRI)
Magnetic resonance imaging (MRI) is superior in sen-
sitivity and specificity for infections when compared
to isotope scintigraphy due to the superb anatomical
detail and the high soft tissue contrast resolution.
Subtle changes of bone marrow signal or enhancement
can only be demonstrated by MRI. The use of contrast
enhancement helps to differentiate between noninfective
conditions ofinternal derangement, like osteochondritis
dissecans, and infective arthritis. Differentiation can be
difficult on plain films (Fig. 35.20) but easier on MRI
(Fig. 35.21a, b). Specific changes of arthropathies can be
shown by MRI, e.g., small signal void bodies or bloom-
ing effect in pigmented villonodular synovitis. Soft tissue
involvement is best demonstrated by MRI. Demonstra-
tion of periarticular disease by MRI is a useful asset. An
example is trochanteric bursitis (Figs. 35.8).

35.3
Radiologic Presentation

Fig. 35.20. TB mimics. Lateral radiograph of a knee for a young
adult patient who presented with painful knee. There are faint
subarticular erosions. These were initially diagnosed as osteo-
chondritis dissecans (see Figs. 35.21a, b)
35.3.1
Tuberculosis of the Knee
The knee is the most common site for tuberculous
and nontuberculous osteoarticular infections. It is

a b

Fig.35.21. a Noncontrast coronal Tl-weighted MRI demonstrating two very well defined subchondral cystic erosions with
little surrounding low signal change due to focal edema. This was still considered to be due to osteochondritis. b Postcontrast-
enhanced fat-saturated Tl-weighted image showing intense contrast enhancement all over the medial femoral condyle. This is
not a feature of osteochondritis. There is also a small joint effusion with significant synovial enhancement. Presence of infection
was suggested. This was proved to be due to TB after synovial biopsy. Note a small enhancing focus at the posterior margin of
the tibia under the tibial insertion of the posterior cruciate ligament. This MRI belongs to the patient shown in Fig. 35.20
 Imaging of Musculoskeletal Tuberculosis 617

also the most common site for malignant tumors, severe on the subchondral bone which erodes the car-
degenerative disease, and internal derangements. tilage from beneath (Ainslie and Bateman 1991). This
It is therefore important to diagnose the cause of in turn slowly leads to joint space narrowing. These
symptoms in the knee accurately. This is why MRI "arthritic" subchondral erosions (Fig.35.16, 35.23)
is commonly used. are different from the "osteomyelitic" cystic erosions,
Kerri and Martini classified the radiologic stages of which may not affect the joint or articular cartilage
knee tuberculosis (Kerri and Martini 1985). In stage (Fig. 35.10, 35.23) (Lee et al.1995).
1 there may be slight osteopenia with or without soft
tissue hypertrophy (Fig. 35.11). Stage 2 is described as
osteomyelitic stage with cystic erosions of the epiphy- 35.3.2
sis or metaphysis, but a normal joint space. In stage Tuberculosis of the Hip
3, arthritis ensues with joint space narrowing and
disorganization which may progress further to stage The hip is the most common extraspinal site of
4 (Fig. 35.15). We believe ankylosis should be consid- tuberculosis infection in children. It is also the most
ered stage 5. This classification is a good predictor of common site of septic arthritis in this age group.
the outcome of the disease (Lee et al. 1995). In stage 1 Scrutiny of plain films is essential to avoid missing hip
and 2, protective immunity causes synovial hypertro- infections (Fig. 35.24). Both conditions, namely tuber-
phy, synovitis, and granuloma formation. In stages 3 culosis and pyogenic hip infections, are debilitating if
and 4, there is tissue destroying hypersensitivity most not treated at an early stage (Figs. 35.25, 35.26).
Shanmugasundaram (1983) described the radio-
logic presentations of tuberculosis of the hips. In type
1, the hip looks normal. The disease is restricted to the
synovium. A traveling acetabulum is noted in type 2,
while a frankly dislocated hip was noted in type 3. In
type 4, a, Perthes' like configuration of the capital epiph-
yses with sclerosis and total head involvement is seen
but no associated metaphyseal changes like Perthes'
disease. Type 5 presents as protrusio acetabuli. Type 6
is an atrophic head with decreased volume, while type 7
shows mortar and pestle appearance (Fig. 35.25).

Fig. 35.22. a TB of the right sacroiliac joint. Bone setting of an

axial CT scan of the pelvis at the level of the sacroiliac joints.
There is widening and erosions of the articular surfaces with
formation of small "kissing sequestra" (long arrow) from either
of the articular surfaces. There is a large right gluteal abscess
(black arrowheads) in continuity with soft tissue swelling of
the right iliacus muscle (small arrow). b Soft tissue setting of
same patient showing sacroiliac joint erosions, iliacus muscle
swelling, right gluteal abscess, and soft tissue calcification
Fig. 35.23. TB of the right knee. Lateral radiograph of the
knee of a 37-year-old male who complained of knee pain and
swelling. There is extensive subarticular permeation of both
femoral condyles (arrows). Note patella erosions (arrowheads)
and the soft tissue swelling (open arrows)
618
Fig. 35.24. TB of the left hip, early changes. Frontal radiograph of
the hips of a teenager complaining of left hip pain and limping.
There is slight nonconcentric narrowing of the superior joint
space. This was associated with nonhomogeneous density of the
metaphysis in the left femoral neck. This stage of the TB infection
can be missed. Diagnosis was made after open biopsy
Fig.35.25. Advanced TB left hip (stage 7). Missed case of
severe TB arthritis of the left hip showing avascular necrosis
and fissuring of the left femoral head ending in the mortar
and pestle appearance. Peculiarly there was still good range
of hip joint movement
Campbell and Hoffman have reviewed, confirmed,
and slightly modified the abovementioned classifica-
tion (Campbell and Hoffman 1999). They considered
the irregular head of type 6 as type 1b and a destroyed
head of type 6 as type 2. Type 3 refers to fibrosis and
ankylosis while their type 4 refers to bone ankylosis.
We believe the Shanmugasundaram classification is
more elaborate and easier to apply. Despite the bizarre
appearance of the hips, both Shanmugasundram and
Campbell found good prognosis in the advanced cases
because of the good range of movement.
35.3.3
Tuberculosis of the Pelvis and Sacroiliac Joints
Primary tuberculous sacroiliitis is rare. Unlike anky-
losing spondylitis which is usually bilateral, tuber-
culosis generally involves only one sacroiliac joint
M. Abd El Bagi et al.
Fig. 35.26. Missed case of septic arthritis in the left hip.
Advanced articular surface erosions with avascular necro-
sis due to a missed diagnosis. Both septic arthritis and TB
arthritis have serious sequelae on the joints if untreated. Open
biopsy was negative for TB bacilli on staining and culture
(Chapman et al. 1979). It usually occurs due to spread
from the lower lumbar spine. The disease starts in the
synovial part of the joint at its lower third. The disease
may be missed on plain films due to bowel overlap. An
isotope scan is the examination of choice initially. CT
provides accurate delineation (Fig. 35.22). Differential
diagnosis includes pyogenic infection, collagen dis-
ease, and fungal infection. Aspiration biopsy is essen-
tial. The ilio-psoas muscle is usually involved and a
cold abscess is common. We encountered some cases
of multiple pelvic infection (Fig. 35.1). Involvement of
the pubic symphysis is rare. CT is a very useful tool for
examination of the pubic symphysis (Fig. 35.27).
35.3.4
Chest Wall Tuberculosis
Chest wall involvement usually results from hematog-
enous infection but more rarely from direct spread
of pleural or lung parenchymal disease. Tuberculosis
occasionally involves the sternoclavicular joint or a
rib, leading to bone destruction and abscess forma-
tion (Kim et al. 2001). This is seen in less than 2% of
musculoskeletal tuberculosis cases (Adler et al.I993).
Plain films are not always as successful in revealing
this condition, as can be seen in Fig. 35.28a, b. CT
is the examination of choice following a positive
nuclear scan (Fig. 35.29a, b). Findings include bone
 Imaging of Musculoskeletal Tuberculosis 619

Fig.35.27. TB pubic symphysis. An adult male patient pre-

sented with suprapubic pain and swelling. Axial postcontrast
CT scan of the pelvis at the level of the inferior margin of the
symphysis pubis. There are bilateral pubic bone erosions asso-
ciated with subcutaneous small cold abscess formation

a b

Fig. 35.28. a TB of the manubrium sterni. Right lateral view of the sternum of a 36-year-old female who presented with chest
wall swelling and fever. Chest X-ray was reported as normal. Lateral radiograph of the manubrium demonstrated a soft tissue
swelling (arrow) and a bone erosion (white arrowhead). Note presence of periostitis (long arrow). b Follow up left lateral view
of the sternum for the same patient in (a) showing the extent of soft tissue swelling (white arrow), large erosion (e) and a
pathologic fracture (open arrowhead)

Fig. 35.29. a Multifocal TB of the chest wall, mediastinal lymph nodes, and vertebral body. An adult male patient who presented with
fever. Anterior chest wall swelling and backache. Axial postcontrast CT of the thoracic inlet at the level of the sternoclavicular joints.
There is a small subcutaneous presternal abscess on the right side of the midline (black arrowhead). An enlarged left retroclavicular
lymph node is seen (long arrow). b Postcontrast CT scan of the upper chest at the level of the aortic arch in the same patient demonstrat-
ing D3 vertebral body erosion with a small paraspinal soft tissue component. There is epidural encroaclunent onto the spinal canal
620 M. Abd El Bagi et al.

or costal cartilage erosions, soft tissue enhancement,

lymphadenopathy, abscess formation, calcification,
and underlying lung diseases (Adler et al. 1993).
Good radiographs can depict the lesions (Fig. 35.28a,
b). In one series, tuberculosis of the ribs was as fre-
quent as 7% of all musculoskeletal tuberculosis cases
(Watts and Lifeso 1996). This high incidence was not
supported by other authors or by our own experience.
MRI is also sensitive for chest wall infections but may
be subject to motion artifacts (Fig. 35.30).

35.3.5
Tuberculosis of the Hands and Feet

Tuberculous mycobacterial infection of the short

tubular bones of the hand and foot, "dactylitis:' is
a disease of childhood (Chapman et al. 1979). It is
more common in the hands than the feet. The most
frequent locations are the proximal phalanx of the
index and middle fingers and the metacarpals of
the middle and ring fingers (Hardy and Hartman
1947). Painless swellings appear on the dorsum of Fig. 35.31. TB tarsal bones. Oblique radiograph of the right foot
the hands and feet. Soft tissue changes precede the of an adult female who presented with chronic foot pain. There
is multilocular expansile cystic lesion of the tarsal navicular
bone changes. Ballooning of the cortex, periostitis
and cortical destruction can lead to what is called
spina ventosa "wind filled sail sign" (Clark 1990). We
rarely see this appearance. Cystic lesions can be seen
(Figs. 35.5,35.31). An erosive dactylitis is now more
common (Fig. 35.32). (Abd El Bagi et al. 1999). Dif-
ferentiation from the hand-foot syndrome of sickle

Fig. 35.32. TB dactylitis. Frontal radiograph for the metatar-

Fig. 35.30. TB sternoclavicular joint. Postcontrast fat -sup-
pressed Tl-weighted MRI of the sternoclavicular joint. There
is erosion of the right clavicular head (arrowhead) with a sur-
rounding enhancing soft tissue swelling (long arrow). There
is also right paratracheal adenopathy (open arrow). This MRI
belongs to the same patient as that in Fig. 35.29a, b
sals and toes of an adult male who presented with pain and
swelling of the dorsum of the foot. Patient is not diabetic.
There is erosion of the base of the middle metatarsal (large
black arrowhead). Minimal periosteal reaction noted (small
black arrowheads). Diffuse soft tissue swelling is noted (open
white arrows)
Imaging of Musculoskeletal Tuberculosis 621

cell disease is important where both diseases are
endemic. Sickle cell dactylitis is characteristically
bilateral (Wessels et al. 1998). Other differential
diagnosis includes hemangioma and sarcoidosis.
Congenital syphilis is no longer prevalent.
becomes prominent as limitation of movement pro-
gresses. Tuberculous bursitis could be the presenting
feature. This can mimic a soft tissue mass (Figs. 35.13,
35.33, 35.34).

35.3.6
Tuberculosis of the Elbow
Tuberculous arthritis of the elbow was reported to
be more common than that of the wrist or shoul-
der (Resnick and Niwayama 1988). Muscle wasting
35.3.7
Tuberculosis of the Shoulder
A peculiar variety of "caries sicca" was described. It
presents with osteoporosis, muscle wasting, and pain.
Alternatively an erosive form develops (Fig. 35.35).
Bursitis could be the presenting feature.

Fig. 35.33. TB of the elbow. A 65-year-old male

presented with a slowly progressing tense mass
in the right cubital fossa. This was clinically sus-
pected to be a tumor. High resolution US demon-
strated that the mass is cystic and multilocular

a b

Fig. 35.34. a TB of the elbow. T2-weighted axial MRI of the right elbow confirmed the multilocular cystic nature of the cubital
fossa lesion as seen by ultrasound in Fig. 35.33. No solid component. Note the transcondylar hyperintense signal in the lower
humerus due to early bone involvement (open white arrows). b Sagittal postcontrast fat-saturated Tl-weighted image showed
enhancement of the walls and septations of the multilocular lesion. This is a sign of infection. US-guided aspiration for culture
was positive for TB
 622
Fig. 35.35. TB of the scapula. A 25-year-old male presented with
a chronic sinus at the scapular region. Axial CT scan demon-
strated a subglenoid bone erosion (black arrowhead) associated
with inflammatory changes and focal swelling of the subscapu-
laris muscle anterior to the neck of the scapula (long white
arrow). Diagnosis was confirmed by CT-guided biopsy
35.3.8
Tuberculosis of the Ankle
Being a weight-bearing joint, the ankle is subject to
repeat trauma. Tuberculosis infection starts in the
bones and later involves the joint. The talus, calca-
neum, or cuboid can be the source of infection. The
patient usually presents with soft tissue swelling.
a b
Fig. 35.36a, b. TB of the ankle. A 49-year-old male complained of chronic right ankle joint pain and swelling (open arrows) and
articular surface erosions (small arrowheads). There is diffuse decrease of bone density. No periosteal reaction
M. Abd El Bagi et al.
Cystic erosions are common (Figs. 35.2, 35.5), while
articular surface erosions are late manifestations
(Fig. 35.36a, b). Periostitis which is rare in tuber-
culosis can be provoked by weight-bearing stress
(Fig. 35.2a). Subluxation can occur (Fig. 35.37).
35.4
Interventions
35.4.1
Biopsies
Tissue diagnosis is not easily established in tuberculo-
sis, though it is imperative. The selection of biopsy sites
is important. Cold abscesses and necrotic tissue do not
usually yield bacilli. Stains from joint aspirates may also
be negative for bacilli. The presence of a high blood
cell count, low glucose, and pure mucin clot formation
is diagnostic of tuberculous joint aspirate. Guinea pig
cultures are more reliable than Ziehl-Neelsen stains
(Glassroth 1993). Either US or CT can be used for
biopsy or drainage procedures. Ultrasound does not
involve exposure to ionizing radiations. US-guided
procedures are quicker than CT-guided procedures
and allow the freedom of bedside use. US provides
real-time guidance. It is particularly suitable for
superficial lesions and synovial biopsies (Fig. 35.38).
CT is however superior for biopsies of intraosseous
and deeply seated lesions inaccessible for US guidance.
We have no experience with MRI-guided interventions,
Imaging of Musculoskeletal Tuberculosis
Fig. 35.37. Frontal radiograph of the right ankle of a 51-year-
old male who complained of joint pain and swelling. There is
a large soft tissue swelling (arrowheads). The lateral malleolus
is osteoporotic and nonhomogeneous. Note the slight medial
subluxation of the tibia over the talus
which require MRI-compatible nonferromagnetic
titanium needles and instruments. Recent reports of
large series indicate successful biopsy samples using
a closed configuration magnet (Salmonowitz 2001).
In general, unless a biopsy sample is accurately taken
from a granuloma, the margins of cystic lesions, or the
synovium, the histologic results are likely to be equivo-
cal. (Versefeld and Solomon 1982). Biopsies are very
important where multidrug-resistant organisms occur
(Watt and Lifeso 1996).
35.4.2
Drainage
Percutaneous drainage of tuberculous abscesses
was said to be not very common elsewhere (Dinc
et al. 1996). We frequently perform this procedure,
whenever a cold abscess is detected (Fig. 35.39). This
prevents sinus and fistula formation and relieves
local symptoms. A specimen is sent for microscopy
and culture. Common sites are psoas compartments,
the hips, and gluteal regions. The source of infec-
tion could be in the lumbar spine, pelvic bones, or
sacroiliac joints (Pouchot et al. 1988). We tend to use
the smallest size catheter that can possibly match
the fluid thickness, and we puncture the skin and
abscess at two different points. These precautions
help to avoid forming persistent sinus.
623
Fig.35.38. TB synovitis. Adult male presented with unex-
plained chronic knee pain and swelling. US-guided synovial
biopsy of the knee was performed. There was only a trace
of joint effusion (not shown here) but there was significant
synovial thickening (black arrowheads). Histopathology and
culture of tissue biopsies were positive for TE. This film dem-
onstrated the thickened hypoechoic synovium (black arrow-
heads). Note position of the biopsy needle (open white arrow)
tip at the thickest site of the synovium (black arrowheads)
Fig. 35.39. Percutaneous drainage of a cold abscess. Tense cold
abscess tracking from the left groin was distending the adduc-
tor muscles of the left thigh. CT scan of the upper third of the
thigh showing drainage catheter in situ
35.5
Summary
Musculoskeletal tuberculosis is an important cause of
morbidity in developing countries with resurgence in
industrial countries. Diagnosis of tuberculosis infec-
tion of bone joints and adjoining soft tissues is a real
challenge, which requires a high degree of clinical
vigilance. The combination of careful clinical history,
elaborate laboratory tests, and radiologic findings can
624
lead to an early diagnosis to prevent serious bone and
joint destruction. Diagnosis should be confirmed by
aspiration and biopsy samples for microscopy, culture,
and histopathology. These could be negative at times.
Alternatively, biopsy samples should be taken from the
primary site or other extraskeletallesions particularly
abdominal lymph nodes. We suggest an algorithm for
handling the various presentations of osteoarticular
tuberculosis using a structured diagnostic pathway.
References
Abd El Bagi M, Al Shahed MS, Sammak M (2002) In: Davies
AM, Cassar-Pullicino VN (ed) Imaging of the knee.
Springer, Berlin Heidelberg New York (Medical Radiology
series) pp 249-267
Abd El Bagi M, Sammak B, Al Shahed M, Yousef BA, Demuren
OA, AI Jared M, AI Thagafi M (1999) Rare bone infections
"excluding the spine:' Eur RadioI9:1078-1087
Abdel Wahab I, Kenan S, Hermann G, Lewis M, Klein M,
Rabinowitz J (1991) Atypical skeletal TB mimicking neo-
plasm. Br J RadioI64:551-555
Adler BD, Padley SPG, Muller NL (1993) Tuberculosis of
the chest wall CT findings. J Comput Assist Tomogr 17:
271-273
Ainslie GM, Bateman ED (1991) Immunological aspects of the
host response to Mycobacterium tuberculosis infection. In:
Cooradia HM, Benatar SR (eds.) A century of tuberculosis:
South African Perspectives. Cape Town, etc. Oxford Univer-
sity Press: 224-242
AI Shahed MS, Sharif HS, Haddad MC, Aabed MY, Sammak
BS, Mutairy M (1994) Imaging features of musculoskeletal
brucellosis. Radiographics 14:333-348
Bureau NJ, Ali SS, Chhem RK, Cardinal E (1998) Ultrasound of
musculoskeletal infection. Semin Musculoskeletal Radiol2:
299-306
Campbell JAB, Hoffman EB (1995) Tuberculosis of the hip in
children. J Bone Joint Surg Br 77:319-326
Chapman M, Murray R, Stocker D (1979) Tuberculosis of the
bones and joints. Semin RotentenoI14:266-282
Clarke JA (1990) Tuberculous dactylitis in childhood: the need
for continued vigilance. Clin RadioI42:287-288
Clinton Thomas CL, Young WB (1955) Multiple pseudocystic
tuberculosis of bone: report of a case. J Bone Joint Surg Br
37:624-631
Cremin BJ, Jamieson DH (1995) Imaging of skeletal tuberculo-
sis in childhood tuberculosis: modern imaging and clinical
concepts. Springer, Berlin Heidelberg New York
Davidson PT, Horowitz J (1970) Skeletal tuberculosis, a review
with patient presentation and discussion. Am J Med 48:
77-84
Derkash RS, Makley JT (1979) Isolated TB of the triceps
muscle. J Bone Joint Surg 61:948
Dine H, Sari A, Yu Lung C, Gumele H (1996) CT guided drain-
age of multilocular pelvic and tuberculous abscess. AJR
167:667-668
Duthie R, Nelson C (1996) Infections of the musculoskeletal
system. In: Duthie RB, Bentley G (eds) Mercer's orthopae-
dic surgery. Arnold Lonon, Sydney, Ankloud pp 596-621
M. Abd El Bagi et al.
Edeiken J, De Palma AF, Moskowitz H, Symthe V (1963) Cystic
tuberculosis of bone. Clin Orthop 28: 163-168
Engin G, Acunas B, Acunas G, Tunaci M (2000) Imaging of
extrapulmomary tuberculosis. Radiographics 20:471-488
Glassroth J (1993) Diagnosis of tuberculosis: a comprehen-
sive interventional approach. In: Reichman LB, Hershfield
ES (eds) Tuberculosis: a comprehensive international
approach, Dekker, New York, pp 149-165
Goldblatt M, Cremin B (1978) Osteoarticular tuberculosis, its
presentation in colored races. Clin Radiol 29:669-677
Hamilton D, Nabulsi J (1999) Radionuclides in the evaluation
of extrapulmonary tuberculosis. In: McGuiness F (ed)
Clinical imaging of non pulmonary tuberculosis. Springer,
Berlin Heidelberg New York, pp 176-187
Hardy JB, Hartman JR (1947) Tuberculous dactylitis in child-
hood: a prognosis. J Paediatr 30:146-156
Harisinghani MG, McLoud TC, Shepard JA, Ko JP, Shroff MM,
Mueller PR (2000) Tuberculosis from head to toe. Radio-
graphics 20:449-470
Haygood TM, Williamson SL (1994) Radiographic findings of
extremity tuberculosis in childhood: back to the future?
Radiographics 14:561-570
Hugosson C, Nyman RS, Brismar J, Larson SG, Luidahl S,
Lundstedt C (1996) Imaging of tuberculosis V: peripheral
osteoarticular and soft tissue tuberculosis. Acta Radiol37:
512-516
Jaovisidha S, Chen S, Ryu KN, Siriwongpairat P, Pekanan P,
Sartoris D (1996) Tuberculous tenosynovitis and bursitis
imaging findings in 21 cases. Radiology 201:507-513
Kanavel AB (1923) Tuberculous tenosynovitis of the hand. A
report of 14 cases of tuberculous tenosynovitis. Surg Gyn-
aecol Obstet 37:635-647
Kerri 0, Martini M (1985) Tuberculosis of the knee. Int Orthop
9:153-157
Kim HY, Song KS, Goo JM, Lee JS, Lee KS, Lim T-H (2001)
Thoracic sequelae and complications of TE. Radiograph-
ies 21:839-860
Komins C (1952) Multiple cystic tuberculosis: a review and
revised nomenclature. Br J RadioI25:1-8
Kumar K, Saxena MB (1988) Multifocal osteoarticular tuber-
culosis. Int Orthop 12:135-138
Lee As, Campbell JAB, Hoffman EB (1995) Tuberculosis of the
knee in children. J Bone Joint Surg 77B:313-318
McGuiness F (1999) Osteoarticular and soft tissue tubercu-
losis in clinical imaging of non-pulmonary tuberculosis.
Springer, Berlin Heidelberg New York, pp 176-187
Muradali D, Gold WL, Vellend H, Becker E (1993) Multifocal
osteoarticular tuberculosis: report of four cases and review
of management. Clin Infect Dis 17:204-209
Pouchot J, Vinceneux P, Barge J et al (1988) TB of the sacroiliac
joints clinical features: outcome and evaluation of needle
biopsy in 11 consecutive cases. Am J Med 84:622-628
Pouchot J, Vinceneux P, Barge J et al (1990) Tuberculous poly-
myositis in HIV infection. Am J Med 89:250-251
Rasool M, Governder S, Naidoo K (1994) Cystic tuberculosis of
bone in children. J Bone Joint Surg Br 76(B):113-117
Resnick D, Niwayama G (1988) Osteomyelitis, septic arthritis
and soft tissue infections: the organisms in diagnosis of
bone and joint disorders, vol 4, 2nd edn. Saunders, Phila-
delphia, pp 2661-2686
Ridley N, Shaikh MI, Remedios D, Mitchell R (1998) Radiology
of skeletal tuberculosis. Orthopaedics 21: 1213-1220
Salmonowitz E (2001) MR imaging guided biopsy and thera-
Imaging of Musculoskeletal Tuberculosis
peutic interventions in a closed configuration agent: single
centre series of 361 punctures. AJR 177:159-163
Shanmugasundaram TK (1983) Bone and joint tuberculosis.
Kothandaram and Co., Madras
Versfeld G, Solomon A (1982) A diagnostic approach to tubercu-
losis of bones and joints. J Bone Joint Surg Br 64(B):446-449
Vohra R, Kang H, Dogra S, Saggar R, Sharma R (1997)
Tuberculous osteomyelitis. J Bone Joint Surg Br 79(B):
562-566
Watts H, Lifeso R (1996) Tuberculosis of bones and joints. J
Bone Joint Surg 76(A):288-298
625
Wessels G, Hesseling P, Beyers N (1998) Skeletal tuberculosis
dactylitis and involvement of the skull. Paediatr Radiol 28:
234-236
Wihlborg C, Babyn P, Ranson M, Laxer R (2001) Radiologic
mimics of juvenile rheumatoid arthritis. Paediatr Radiol
31:315-326
Wilbur A, Gorodetsky A, Hibbeln J,Aizenstein R (1995) Tuber-
culous myositis: CT and sonographic findings in two cases
J Clin Ultrasound 23:495-499
Yao D, Sartoris D (1995) Musculoskeletal tuberculosis. RCNA
33:679-689
36 Tuberculosis of the Skin
AIDA J. AL KUDWAH

CONTENTS 36.1
Tuberculous and Nontuberculous Infections
36.1 Tuberculous and Nontuberculous Infections 627
36.1.1 History 627
36.1.2 Epidemiology 627
36.1.1
36.1.3 Classification of Mycobacteria 628 History
36.2 Tuberculosis of the Skin 628
36.2.1 Etiopathogenesis 628 Tuberculosis has been a recognized ailment of
36.3 Tuberculosis of the Skin Due mankind from ancient history. The most ancient
to M. tuberculosis / M. bovis 629
36.3.1 Clinical Variants 629
documentation is from the Pharaonic dynasties,
36.3.1.1 Primary Inoculation Tuberculosis 2000-3000 years B.C. It is one of the most important
(Tuberculous Chancre) 629 human infections because of its prevalence and its
36.3.1.2 Acute Miliary Tuberculosis of the Skin 631 being one which causes more suffering for humans
36.3.1.3 Lupus Vulgaris 632 than all bacterial genera combined (Tapeiner and
36.3.1.4 Tuberculosis Verrucosa Cutis 635
36.3.1.5 Scrofuloderma 636
Wolff 1999; Sehgal 1994). DNA of Mycobacterium
36.3.1.6 Tuberculosis Cutis Orificialis 637 tuberculosis has been detected in Peruvian mummies
36.3.1.7 Metastatic Tuberculous Abscess (Gumma) 639 (Gawkrodger 1998). Cutaneous tuberculosis was first
36.3.1.8 BCG Vaccination and Cutaneous Tuberculosis 641 documented by Laennec where he described his own
36.3.1.9 Congenital Tuberculosis 641 prosector's wart (Laennec 1962) in 1826.
36.4 The Tuberculids 642
36.4.1 Tuberculosis-specific Tuberculids 642
36.4.1.1 Papulonecrotic Tuberculids 642
36.4.1.2 Lichen Scrofulosorum 643 36.1.2
36.4.2 Tuberculosis Nonspecific Tuberculids 644 Epidemiology
36.4.2.1 Erythema Induratum 644
36.4.2.2 Erythema Nodosum 645
36.4.3 Nontuberculids 646
Tuberculosis of the skin has a worldwide distribution.
36.4.3.1 Lupus Miliaris Disseminatus Faciei It constitutes a small portion of all cases of extrapul-
(Acne Agminata) 647 monary tuberculosis which in itself is a small fraction
36.4.3.2 Rosacea-like Tuberculid 647 (1 0%) of all tuberculous infections. It has been more
36.4.3.3 Lichenoid Tuberculid 648 prevalent in regions with cold and humid climates
36.5 Cutaneous Tuberculosis
in the Immunocompromised 648 but is recognized in the tropics (Anonymous 1991;
36.6 Atypical Mycobacterial Cutaneous Infections 648 Marcoval et al. 1992). The improvement in socioeco-
36.6.1 M. marinum Infection 649 nomic standards and living conditions in European
36.6.2 Mycobacterium ulcerans Infections 649 and North American countries, immunization pro-
36.6.3 Mycobacterium kansasii Infection 650 grams, and the use of three or four drug regimens
36.6.4 Mycobacterium scrofulaceum Infection 651
36.6.5 Mycobacterium avium-intracellulare
of treatment have caused a steady decline in its inci-
Complex Infection 651 dence over the last decades; it parallels the decline in
36.6.6 Mycobacterium fortuitum Complex Infection 652 incidence of pulmonary tuberculosis. Resurgence and
36.6.7 Mycobacterium haemophilum Infection 653 increased incidence of cutaneous tuberculosis in the
36.7 Treatment of M. tuberculosis Skin Infections 654
United States is mainly due to the AIDS epidemic, and
36.7.1 Dermatologic Considerations 654
References 654
this increase could be explained either by endogenous
reactivation or increased susceptibility to reinfection
with M. tuberculosis or low virulence mycobacteria
A. J. AL KUDWAH, MD, DD, FRCP (Ed)
Consultant Dermatologist, Department of Dermatology,
(atypical) because of impaired cellular immunity. In
Riyadh Armed Forces Hospital, P.O. Box 7897, C-1l7, the United States, until the 1980s, skin tuberculosis
Riyadh 11159, Saudi Arabia was a rare disease; in Europe the incidence now is less

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
628
than 0.5%. In Germany only 300-400 new cases are
estimated to occur every year. In India the incidence
is 0.15% (Tapeiner and Wolff 1999). In our experience
over the last 10 years (1993-2002) a total of 23 cases
of cutaneous tuberculosis were seen among a total
of 20,000 new dermatologic cases attended to by our
hospital for the same period (0.12%).
36.1.3
Classification of Mycobacteria
Mycobacteria are slender, unencapsulated, non-
motile, nonsporulating, weakly positive acid-fast
anaerobic rods. The tubercle bacillus takes the form
of slender, curved, delicate, beaded or banded rods,
1-4 11m long and about 0.5 11m in diameter (Sehgal
and Wagh 1990a). They occur either singly, in pairs,
or in clumps. Their highly complex waxy coating
determines the mycobacterial physiologic properties
(e.g., resistance to degradation following phagocyto-
sis) and may influence the pattern of host response
(Citron and Girling 1987; Higuchi et al. 1981). They
are difficult to stain, also due to their cell wall high
lipid content, but once stained with basic dyes (e.g.,
carbolfuchsin), they retain the color and resist decol-
orization with acid and alcohol, thus the name "acid-
fast bacilli." Runyon (1965, 1974) classified the myco-
bacteria other than tuberculosis (MOTT) according
to their rate of growth and pigment-production
ability with or without light. Since then more stud-
ies have been made and the genus Mycobacterium
stands among the best classified bacterial genera.
Table 36.4 shows the classification of the atypical
mycobacteria (Runyon 1965,1974). Slow growers are
divided according to their pigment-forming proper-
ties in culture: group I photochromogens capable
of pigment production on light exposure, group II
scotochromogens which are able to produce pigment
without exposure to light, group III nonchromogens,
non-pigment producers, and group IV including all
rapid growers. A different classification could be used
for clinical purposes which divides mycobacteria into
obligate and facultative human pathogens and non-
pathogens (Tapeiner and Wolff 1999).
36.2
Tuberculosis of the Skin
Cutaneous tuberculosis has been described long
before Robert Koch identified M. tuberculosis (see
A. J. Al-Kudwah
Laennec 1962). It is caused by M. tuberculosis, M.
bovis, or under certain circumstances by the bacille
Calmette-Guerin (BCG). Various classifications of
the clinical variants exist but none of them has been
entirely satisfactory. These classifications were based
on the mode of the infection, immunologic state of
the host, and whether mycobacteria have been cul-
tured and identified from the skin lesions or not. This
resulted in a number of conditions with no cultur-
able connection to tuberculosis but some have shown
therapeutic response to the use of antituberculous
treatment, while others show spontaneous resolution,
response to other antibiotics, or to nonspecific mea-
sures such as bed rest. These are called tuberculids
(Table 36.2). In this chapter an attempt to take into
consideration variables that govern the outcome of
the infection as well as the route of infection will be
discussed, clinico-immunologic classification of the
various cutaneous manifestation of the tuberculous
process is seen in Table 36.1.
36.2.1
Etiopathogenesis
The clinical presentation is determined by the out-
come of the host immune responses, mycobacterial
virulence, and the route of entry of the bacilli. The
host has inherent or acquired ability to control infec-
tion. Susceptibility is decreased in populations where
there is contact with TB for many generations. Such
contact results in partially immunized individuals
through natural infections. Blacks as well as carriers
of HLA-BI5 major histocompatibility antigen are
more susceptible to the disease. The age, an intact
immune system, as well as socioeconomic and envi-
ronmental factors are also of importance (Tapeiner
and Wolff 1999).
Once the host is invaded by the mycobacterium,
the interaction between host immune responses and
mycobacterial factors such as its virulence and the size
of the inoculum will determine the outcome. Either
there is unchecked multiplication, with the disease
becoming apparent, or this process is arrested. Anti-
gen-presenting cells engulf the bacteria and through
the antigens displayed on their surface, they activate
the T lymphocytes which release cytokines. These in
turn activate macrophages that engulf the mycobac-
teria and kill them. Few organisms are killed in each
cycle, and in the very early stages resistance depends
on rapidly repeated cycles. Cytotoxic T cells lyse
infected phagocytic cells with the subsequent release
of surviving organisms which are phagocytosed again,
Tuberculosis of the Skin
Table 36.1. Clinico-irnmunologic classification of cutaneous tuberculosis
Clinical presentation Sensitization Route of infection
of the host
Exogenous Hematologic Lymphatic Contiguous
Primary tuberculous infection
Tuberculous chancre Nonsensitized +
Acute disseminated miliary TB Nonsensitized
Postprimary tuberculous infection
Reinfection
Lupus vulgaris Pre-sensitized +
Tuberculosis verrucosa cutis Pre-sensitized +
Reactivation
Scrofuloderma Pre-sensitized
Tuberculosis cutis orificialis Pre-sensitized
Reactive eruptions
"True" tuberculids Pre-sensitized
Table 36.2. Suggested classification of tuberculids
Tuberculosis-specific tuberculids (TSTs)
Conditions in which M. tuberculosislbovis plays a signifi-
cant role
Papulonecrotic tuberculid
Lichen scrofulosorum
Tuberculosis-nonspecific tuberculids (TnSTs)
Conditions in which M. tuberculosislbovis is one of sev-
eral other etiopathogenic factors
Erythema induratum
Erythema nodosum
Nontuberculids (conditions previously described as tubercu-
lids but for which recent evidence suggests no relationship
to tuberculosis)
Lupus miliaris disseminatus faciei
Rosacea-like tuberculids
Lichenoid tuberculids
and so on. A few organisms die also at the time of lysis
of infected macrophages (Tapeiner and Wolff 1999;
Gawkrodger 1998). The balance between lymphokines
that activate macrophages and lymphokines that are
toxic, necrogenic, or both will determine the final clin-
ical picture of the infection. Different clinical entities
have different bacterial numbers as seen histologically.
It is difficult to determine whether the small bacterial
number has caused, for example, lupus vulgaris or the
host immunity has checked or decreased the number
in that clinical entity. The route of infection and the
targeted tissue properties (e.g., structure and vascular
supply), determine, in addition to the hypersensitivity
status of the host, the clinical variant exhibited derma-
tologically. The skin involvement could be exogenous
or endogenous. Infection originating from an endoge-
nous source can reach its destination hematologically,
lymphatically, contiguously, or by autoinoculation.
Local injury may act as a localizing factor.
629
+
+ +
+
++
autoinoculation
+
36.3
Tuberculosis of the Skin Due
to M. tuberculosis / M. bovis
36.3.1
Clinical Variants
36.3.1.1
Primary Inoculation Tuberculosis
(Tuberculous Chancre)
Tuberculous chancre (TC) is the result of mycobacte-
rial infection of the skin at the site of inoculation in
an unsensitized host who lacks natural or acquired
immunity. The skin lesion and the affected regional
lymph nodes constitute the tuberculous primary
complex of the skin.
Incidence
TC is a rare form of skin tuberculosis nowadays, but
in 1930 it constituted 0.40% of all primary tubercu-
lous lesions. In Asia where socioeconomic and living
conditions are poor, the incidence remains high. Most
patients are children, but this condition may affect
adolescents and young adults, particularly those
working in the medical field. It can be an occupa-
tional hazard for those people.
Pathogenesis
M. tuberculosis gains access into the skin and rarely
the mucosa, through minor cuts, trauma, or pyo-
dermas - as it cannot penetrate intact skin. It has
been know to follow minor surgical procedures,
such as circumcision (Holt 1913) when performed
630 A. J. AI-Kudwah

Table 36.3. Etiologic factors in erythema nodosum

Infections Systemic Drugs Miscellaneous

diseases

Bacterial Benign Halides Pregnancy

Streptococcal infection Enteropathies Sulfonamides
M. tuberculosis Connective tissue Oral contraceptive
diseases pills
Yersinia enterocolitica Sarcoidosis
Viral and chlamydial Behyet's disease
Chlamydia psittaci Malignant
Epstein-Barr virus Myeloproliferative
disorders
Hepatitis B virus
Parapox virus
Mycoses
Superficial and deep mycoses

by a tuberculous rabbi, nose and ear piercing or tat- Table 36.4. Classification of atypical mycobacteria
tooing where unsterilized needles were used. It can A. Slow growers:
also be transmitted sexually through contact with Runyon group I - Photochromogens
individuals who have genitourinary tuberculosis. M. kansasii
Mouth-to-mouth resuscitation has been reported M. marinum
M. simiae
to cause primary mucosal inoculation tuberculosis,
Runyon group II - Scotochromogens
and unpasteurized milk infected with bovine myco- M. szulgai
bacteria has caused oral lesions following trauma or M. scrofulaceum
tooth extraction (Tapeiner and Wolff 1999; Sehgal M. xenopi
and Wagh 1990; Moschella 1985). Conjunctival and Runyon group III - Nonchromogens
M. avium-intracellulare complex
penile involvement has been reported. The skin
M. genavense
lesion develops 2-4 weeks after inoculation, and M. haemophilum
regional lymph nodes become infected 3-8 weeks M. malmoense
later. As the host's acquired immunity is built up, the M. ulcerans
process is localized to that area. B. Rapid growers:
Runyon group IV
M. fortuitum complex
Clinical Features

The initial lesion presents a small brownish papule
or nodule that rapidly enlarges and then breaks
down to form a sharply demarcated, ragged ulcer
with undermined reddish-blue edges, and a granu-
lar hemorrhagic base that is studded with miliary
abscesses or covered with a necrotic fragmented
membrane. As the lesion matures it becomes more
indurated with thick adherent crust. Spontaneous
but slow healing may take place. Apparent healing
may conceal active infection beneath, which will
eventually break through the surface giving granu-
lating ulcers or sinuses (Tapeiner and Wolff 1999).
Conjunctival lesions can either be shallow ulcers or
fungating edematous granulations clinically known
as phlyctenular conjunctivitis. Oral ulcers are char-
acteristically painless and affect teeth sockets or the
gum. Subcutaneous abscesses can result from the use
of unsterilized needles. Painless paronychia can be
caused by a primary tuberculous infection. Regional
lymph nodes enlarge slowly and occasionally may be
the only clinical symptom. They are firm and pain-
less, subside slowly, and often calcify. Cold abscesses
may occasionally develop after weeks or months, they
break through the skin forming discharging sinuses.
Rarely, this may progress to miliary tuberculosis with
fatal outcome. Constitutional symptoms may be pres-
ent. Occasionally the process assumes a more acute
course simulating a pyogenic infection (Moschella
1985).
Course
The chancre will heal slowly, taking up to 12 months,
with scarring. Rarely, lupus vulgaris or tuberculosis
verrucosa cutis develops at the site of a tuberculous
chancre. Scars mark the sites of previously discharg-
ing liquefied lymph nodes which are calcified in 50%
of cases. Calcification by itself is not a sign of cure.
Tuberculosis of the Skin 631

Hematogenous spread from breakdown of such foci
will cause infection of other organs (Wolff 1921;
Duken 1933), and depending on the size of the inocu-
lum, and the age and immune response of the host,
miliary tuberculosis may occur. Erythema nodosum
develops in 10% of patients (Tapeiner and Wolff 1999;
Gawkrodger 1998; Miller and Cashman 1955). The
primary tuberculous complex imparts satisfactory
immunity and high sensitivity in the host as shown
by a positive tuberculin skin test.
Histopathology
The early lesion of TC shows features of acute
inflammation with necrosis. Mycobacteria are easily
detected. The lesion, as well as the regional lymph
nodes, show granulomatous morphology with case-
ation. Bacilli disappear 3-6 weeks later.
Diagnosis and Differential Diagnosis
(intrauterine transmission) miliary tuberculosis has
been reported, but organisms have not been cultured
from lesions (McCray and Esterly 1981). Tuberculin
skin testing may be negative.
Clinical Features
The morphology of skin lesions in AMTS is often
deceptive. There are usually disseminated bluish red
to brown discrete pinhead papules, vesicles, pustules,
or hemorrhagic lesions in a severely ill patient. Pap-
ules are capped by tiny vesicles that rupture or dry up
forming a crust which upon removal shows a small
umbilication (Fig. 36.1 a, b). Vesicles as such may
become necrotic and form small ulcers (Sahn and Neff
1974). Erythematous subcutaneous nodules have been
reported (Kounis and Constantinidis 1979). A search
for other tuberculous infection should be made (Riet-
broek et al. 1991).

A high degree of suspicion is needed to avoid erro-

neous diagnosis. A persistent, painless, nonheal-
ing ulcer with unilateral lymph node involvement
should arouse suspicion of the tuberculous nature
of the lesion, especially in a child, medical, or para-
medical professional, particularly in endemic areas.
Earlier lesions will show acid-fast bacilli in histologic
specimens or smears. Diagnosis is confirmed by cul-
ture, though treatment should be initiated without
awaiting results. The chancre should be differentiated
from tularemia, primary complex of syphilis, sporo-
trichosis, cat scratch fever, and, when sinuses develop,
from actinomycosis and ulcerative lesions of other
mycobacterial infections (Pereira et a1. 1976).

36.3.1.2
Acute Miliary Tuberculosis of the Skin

Acute miliary tuberculosis of the skin (AMTS) is also

known as tuberculosis cutis miliaris disseminata. It
is a rare form of cutaneous tuberculosis subsequent
to an acute hematogenous dissemination of the
organism, occurring primarily in infants, children,
and immunocompromised patients (Tapeiner and
Wolff 1999).

Pathogenesis Fig.36.1. a Acute miliary tuberculosis of the skin. Multiple

The initial focus of infection is either pulmonary or
extrapulmonary. It may follow infection that reduced
immunologic defenses, like measles, other exanthems,
and AIDS (Tapeiner and Wolff 1999). Congenital
red to brown discrete scaly and crusted papules are seen on
the trunk. Central crusts can be seen in some of the lesions.
Courtesy of Dr. M. Enani, Department of Medicine, RKH.
b Postinfiammatory hyperpigmentation of the previously
involved skin of the same patient. Courtesy of Dr. M. Enani,
Department of Medicine, RKH
632
Course
AMTS is usually fatal if the diagnosis is missed or
delayed. Very young patients usually experience
a downhill course. Favorable outcome has been
reported and only happens if antituberculous ther-
apy is instituted immediately upon suspicion, with
supportive management. Cutaneous lesions heal
with white depressed scars that have a brownish halo
(Tapeiner and Wolff 1999; Schermer et al. 1969).
Histopathology
Acute phase histologic findings are nonspecific.
Necrosis, nonspecific inflammatory infiltrate, which
at times forms small abscesses, and occasionally signs
of vasculitis are seen. Mycobacteria may be present
in and around blood vessels. If the patient's immune
responses develop there will be lymphocytic cuffing
of blood vessels and even granuloma with tubercle
formation.
Diagnosis
An unusual exanthematous rash in a gravely ill
patient who has tuberculosis or contact with a patient
who has tuberculosis should arouse suspicion. These
patients are usually very ill, and the rash may go often
unnoticed (Tapeiner and Wolff 1999). Confirmation
of diagnosis may be achieved by histology and
microbiologic methods. PCR may be helpful in the
diagnosis when available.
36.3.1.3
Lupus Vulgaris
Lupus vulgaris (LV) is one of the most common
forms of cutaneous tuberculosis and the earliest rec-
ognized expression of the disease. It is a postprimary
tuberculous infection: in other words, it occurs in a
patient who has been sensitized to mycobacterium
through a previous infection or vaccination and has
developed high sensitivity but moderate immunity.
It is a chronic, progressive condition with variable
clinical expression and little tendency to spontane-
ous healing.
Incidence
Lupus vulgaris was reported early in the last century.
A predilection for the northern European countries
has been noted, as cool, humid climate favors its
development. Specialized hospitals were available
A. J. Al-Kudwah
solely for the treatment of patients with lupus vul-
garis because of the high prevalence of these cases in
the 1940s and 1950s. It is now uncommon in Europe,
infrequent in the East, and rarely reported in the
United States (Goh et al. 1974). Women appear to be
affected two to three times as often as men (Horwitz
1960; Kalkoff 1950). All age groups are uniformly
affected (Tapeiner and Wolff 1999; Gawkrodger
1998), but in developing countries the majority of
cases are in the 2nd and 3rd decade of life (Singh
1974; Sehgal et al. 1987; Pandhi et al. 1977).
Pathogenesis
Lupus vulgaris is a postprimary tuberculous infection
and thus the skin is affected through either direct inocu-
lation, hematogenous, lymphatic, or contiguous spread.
Hematogenous dissemination could be from a distant
tuberculous focus, this method could be implicated in
cases in which only old pulmonary primary complex is
found where disseminated mycobacteria are seeded in
the skin awaiting suppression of the immune system
of the host by various stimuli (Ustvedt and Ostensen
1951). Facial lesions, for example, might be explained
by lymphatic spread from the mucous membranes
of the nose and throat. Rarely, it may follow primary
inoculation tuberculosis or BCG vaccination. Contigu-
ous spread would explain lupus vulgaris appearing in
skin over tuberculous lymphadenitis or joint involve-
ment. It is preceded by scrofuloderma in 30% of cases
(Kalkoff 1950; Horwitz 1959). The organisms are usu-
ally difficult to culture and difficult to demonstrate in
histologic sections.
Clinical Features
Early lesion of LV may appear as is an asymptom-
atic solitary, tiny, reddish brown, soft, flat plaque.
The lesions gradually enlarge and become more
infiltrated with a darker brownish color (Fig. 36.2b).
Slow peripheral extension produces gyrate shapes.
The surface might remain smooth or become cov-
ered by fine scales. Apple-jelly nodules, a diagnostic
feature, can be demonstrated by diascopy and when
probed reveal a soft consistency. This feature might
not be appreciated in patients with dark complexion
(Banerjee 1956).The lesion runs an indolent course
of peripheral extension in some areas while involut-
ing with scarring in others. Fresh lesions may appear
in apparently healed, scarred areas (Horwitz and
Christensen 1960). The head and neck are affected
in 80-90% of patients (Horwitz 1959; Kanan and
Ryan 1976), particularly around the nose or on the
 Tuberculosis of the Skin
a b
Fig. 36.2. a Lupus vulgaris. A reddish brown infiltrated plaque
with superficial microabscesses and scales over the buttock of
a patient with multiple tuberculous foci. Atrophic depressed
scars of healed areas (arrows) are seen. Short arrow shows
tethered skin of healed fistula. b Close-up of the lesion. c
Lupus vulgaris. After treatment lesions appear less infiltrated
and areas of scarring can be seen. There is a rim of postinflam-
matory hypopigmentation surrounding the lesions
cheek. Ear lobules are frequently affected. The arms
and legs are affected to a lesser extent, while the trunk
is the least involved except in cases with disseminated
lupus vulgaris (Fig. 36.2a, c).
Although lesions are usually solitary, two or more
sites may be affected and at times satellite lesions
appear adjacent to the primary one. Five clinical pat-
terns can be observed in lupus vulgaris depending on
the local tissue response (Gawkrodger 1998):
1. Plane forms manifest as flat, smooth surfaced or
scaly plaques with polycyclic or serpiginous bor-
ders. Larger lesions may show irregular scarring
with islands of activity that might persist indefi-
nitely, and at times it might cover large areas of
the body. The edge often becomes thickened and
hyperkeratotic.
2. Hypertrophic tumor-like forms show deeply infil-
trating tumorous growths that are soft in consis-
tency or hyperkeratotic. Scaling and scarring are
absent. There is marked soft tissue destruction
with resultant edema, lymphatic stasis, vascular
dilatation, and recurrent erysipelas leading to
gross deformities (Tapeiner and Wolff 1999).
3. Ulcerative forms are mutilating because of exten-
sive tissue necrosis, scarring, and ulceration, and
if nasal or auricular cartilages are involved, exten-
sive destruction takes place with contractures and
deformities.
633
""_.. . .---c
4. Vegetating forms result from granulation tissue at
areas of extensive ulceration and necrosis. Infiltra-
tion is marked but scarring is minimal.
5. Papular and nodular forms occur in disseminate
lupus following a transient impairment of immu-
nity, e.g., after measles (lupus postexanthematus)
through hematogenous spread.
Mucosal Lupus Vulgaris
The buccal, pharyngeal, nasal, and conjunctival muco-
sae can be involved primarily through lymphatic
spread or by extension from neighboring skin lesions.
Small, soft, gray-pink papules, ulcers, or granulating
masses are often present, which bleed easily. Advanced
lesions may destroy the cartilage of the nasal or the
larynx, resulting in collapse of the nasal septum and
laryngeal stenosis, respectively. Cicatricial deformity
of the soft palate may also occur (Tapeiner and Wolff
1999; Moschella 1985; Ihin et al. 1983).
Course
Lupus vulgaris is a slowly progressive chronic infection
that might be neglected by the patient because ofthe ini-
tiallack ofsymptoms. Older patients may show extensive
lesions because they probably have had the disease for
longer periods or, if they were infected when older, the
634 A. J. Al-Kudwah

disease may have progressed more rapidly (Harrison
and Marks 1980). If untreated it progresses relentlessly,
leaving in its wake scarring, destruction, contractures,
and deformities. Joint immobility, cicatricial ectropion,
microstomia, nasal deformity, and laryngeal stenosis are
some examples of what happens when such cases are
neglected. The most serious complication in long-stand-
ing (l0-50 years) lupus vulgaris is the development of,
most frequently, squamous and, less commonly, basal
cell carcinomas. Sarcomas have also been described
and their incidence was estimated as approximately
10% (Hekel and Seyss 1951), and it is believed that they
have no relationship to previous X-ray therapy. An asso-
ciated pulmonary or extrapulmonary tuberculosis may
be present in up to 11 % in anyone series (Kanan and
Ryan 1975). Tuberculous lymphadenitis was reported in
up to 40% of patients with LV, and pulmonary, bone, and
joint tuberculosis was reported in 10-20% in another
series. LV may be regarded as a symptom of another site
of tuberculous disease running a serious course (Clasen
and Horwitz 1960).
Histopathology
LV lesions display typical tubercle formation with
sparse caseation that is normally present in the upper
dermis (Fig. 36.3a-c). The mycobacterium bacilli are
often hard to demonstrate in sections but occasionally
may be numerous. Secondary changes especially in
long-standing cases often supervene, such as epider-
mal thinning and atrophy, acanthosis, hyperkeratosis,
and occasionally pseudoepitheliomatous hyperplasia.
The dermis may show nonspecific inflammation dis-
guising the tuberculous nature of the lesion, especially
when necrosis and inflammation are present. Dormant
lesions are often indistinguishable from sarcoidal infil-
trates when epitheloid granulomas dominate in sec-
tions of tissue specimens (Tapeiner and Wolff 1999).
Fibrosis with small epitheloid foci, that eventually
disappear, may be seen in involuting lesions.
Diagnosis and Differential Diagnosis
Morphologically, the classic soft, brownish red lupus
vulgaris plaque with its apple-jelly nodules noted
on diascopy does not present a diagnostic problem.
Early lesions must be distinguished from discoid lupus
erythematosus, lymphocytoma, and Spitz nevus. In
older lesions sarcoidosis, tertiary syphilis, leprosy,
deep mycotic infections, lupoid leishmaniasis, and
chronic vegetating pyodermas should be considered.
Other differential diagnoses include psoriasis, Bowen's
disease, and Wegener's granulomatosis. Facial lesions
might resemble rosacea (Warin and Wilson-Jones
1977). Histologic examination and culture are man-
datory for confirming the diagnosis. Tuberculin test
is strongly positive except in early phases of postex-
anthematic lupus (Fig. 36.4) (Sundt 1925).

Fig. 36.3. a Skin biopsy exhibiting granulomatous inflammation involving the superficial and deep part of the dermis with
some accentuation around the adnexa. (H&E 4) Courtesy of Dr. A. Al-Ajlan, dermatopathologist, RKH. b High power shows a
granulomatous inflammation with pseudoepitheliomatous hyperplasia at the edge of the granuloma. (H&E 10). Courtesy of Dr.
A. AI-Ajlan, dermatopathologist, RKH. c Higher power shows a granulomatous inflammation with lymphocytic rimming. (H&E
40). Courtesy of Dr. A. AI-Ajlan, dermatopathologist, RKH
Tuberculosis of the Skin
Fig. 36.4. PPD skin test showing a strong reaction with 3 em of
induration. There is pus formation in the center
36.3.1.4
Tuberculosis VerrUCOSQ Cutis
Tuberculosis verrucosa cutis (TVC) is a verrucous skin
lesion that occurs as postprimary tuberculosis in the
previously sensitized host who has a moderate to high
degree of immunity, through an exogenous route. Rene
Laennec (1781-1826) described his "prosector's wart"
(Marmelzat 1962). The name "verruca necrogenica"was
coined by Samuel Wilson (1824-1911). Other synonyms
have been used, including anatomist tubercle, lupus ver-
rucosus, and butcher's wart (Sehgal and Wagh 1990).
Incidence
It is one of the common types of skin tuberculosis
encountered in the East, in Hong Kong it accounts
for 40% of the cases (Wong et al. 1968).
Pathogenesis
This infection can result from direct inoculation
of mycobacteria via minor skin trauma coming in
contact with soil that has been contaminated with
infected sputa (Mitchell 1954; Wong et al.I968).Acute
inoculation by the patient's own infected sputum is
another possibility. Accidental infection occurs in
certain professional groups: for example, physicians,
pathologists, medical students, and laboratory and
postmortem attendants. This constitutes an occu-
pational hazard in these professions. Butchers and
farmers can be similarly infected (Tapeiner and Wolff
1999; Gawkrodger 1998; Sehgal and Wagh 1990), but
in these cases, M. bovis is the most common cause.
Clinical Features
The lesions of TVC may occur on areas that are
subject to trauma, such as the hand, particularly
635
over the radial border of the back of the hand, and
fingers, as noted among European patients. The
ankles, knees, and buttocks are more likely to be
affected in children in Eastern endemic countries
(Tapeiner and Wolff 1999).The initial lesion may
appear as an asymptomatic indurated warty papule
or papulopustule with a purple inflammatory halo.
It gradually enlarges to form a hyperkeratotic warty
plaque that might assume an irregular shape because
of uneven spread with finger-like projections. The
verrucous hyperkeratotic surface might be traversed
by clefts and fissures. Though it is usually firm, areas
of softening, especially in the center, will discharge
pus and keratinous material through these fissures
that extend deep into the infiltrated brownish red
base (Hirsh and Johnson 1984; Platou and Lennox
1956; Tapeiner and Wolff 1992; Gawkrodger 1998).
Tuberculosis verrucosa cutis is usually a single
lesion but multiple lesions may occur. In contrast
to the cutaneous primary complex, lymph node
involvement is rare but may occur as the result of
secondary bacterial infection (Tapeiner and Wolff
1999).
A variation in the clinical expression of the disease
may be seen in a deeply destructive papillomatous and
sclerotic form, which may cause deformity of the limbs
(Tapeiner and Wolff 1999). An exuberant granuloma-
tous variant was described in patients from Eastern
countries (Mitchell 1954; Wong et al. 1968). Psoria-
siform and keloidal morphology has been observed
(Gawkrodger 1998).
Course
A slow and gradual resolution of the lesion may
take place and the course may extend over many
years if untreated. Spontaneous remission often
occurs with atrophic sunken scars (Tapeiner and
Wolff 1999; Gawkrodger 1998) and occasionally
shows central clearing with peripheral activity.
A secondary bacterial infection may give rise to
acute inflammatory changes. Active pulmonary,
osseous, or glandular disease may coexist in these
patients.
Histopathology
A striking pseudoepitheliomatous hyperplasia with
marked hyperkeratosis and dense inflammatory infil-
trate is common. Abscesses form in the upper dermis,
subepidermally or within the pseudoepitheliomatous
rete ridges. Typical tubercles are uncommon. Myco-
bacteria are found only occasionally.
636
Diagnosis and Differential Diagnosis
Early TVC lesions should be differentiated from warts
or seborrheic keratoses. Hyperkeratotic lupus vulgaris
elicits the apple-jelly nodules and a has predilection
to other sites. Hypertrophic lichen planus is usually
pruritic and other cutaneous or mucosal lesions might
be found. Localized lichen simplex chronicus is also
pruritic, and lesions mature as itching continues. Deep
mycotic infections such as blastomycosis, chromomy-
cosis, chromoblastomycosis, and actinomycosis, iodo-
derma, bromoderma, leishmaniasis, tertiary syphilis,
and pyodermas due to other organisms should be
considered and differentiating tests should be under-
taken. Nontuberculous mycobacteria as causative
agents (esp. M. marinum) can only be differentiated
by culture and PCR.
36.3.1.5
Scrofuloderma
Scrofuloderma occurs as a postprimary reactivation
tuberculosis. An underlying tuberculous lymph-
adenitis or osteomyelitis may lead to cold abscess
formation with affection of overlying skin by con-
tiguous spread leading to its breakdown and sinus
formation. It was initially described by French writ-
ers who termed it scrofulous gumma (Grange 1988;
Segal and Wagh 1990). The ancient Egyptian doctors
described scrofuloderma about 5000 years ago (see
chapter "Tuberculosis in Ancient Egypt")
Incidence
In patients with tuberculous lymphadenitis, scrofulo-
derma remains relatively common in endemic coun-
tries. It is the most common form of tuberculosis
of the skin in Mexico (Amezquila 1963) and other
tropical countries, and in the immigrants from these
places. All age groups are affected but the prevalence
is higher among children, adolescents, and the elderly
(Tapeiner and Wolff 1999).
Pathogenesis
As the skin over a tuberculous focus get involved,
it breaks down discharging seropurulent discharge.
The most common focus is a lymph node, but tuber-
culosis of bones and joints, tuberculous epididymitis,
and lacrimal gland or duct tuberculosis have been
implicated. Cervical tuberculous lymphadenopathy
is the most common site, but other groups of lymph
nodes can be the source of such focus, like paraster-
A. J. Al-Kudwah
nal, axillary, inguinal, epitrochlear, and retroauricu-
lar. It can result also from accidental or intentional
introduction of bacilli in the subcutaneous tissue by
trauma or injections in patients with previous latent
or manifest infection (Chien and Wiggins 1954).
Clinical Features
The initial cutaneous lesion of scrofuloderma is an
asymptomatic bluish red, firm subcutaneous nodule or
deep subcutaneous infiltrate that is freely mobile in the
earlier stages of the disease. As the infiltrate enlarges,
the tissue becomes matted and doughy. After a period
of time liquefaction occurs, and the lesion becomes
fluctuant. The skin breaks down to form linear or
serpiginous ulcers with granulating, uneven floors
and undermined, inverted (rolled in) bluish edges
(Fig. 36.5). The discharge could be watery and purulent
or caseous. Many intercommunicating fistulae lead to
the formation of subcutaneous soft pockets filled with
discharge. These pockets alternate with slightly firmer
gummatous nodules. Scarring develops with dense
fibrous bands that bridge ulcerative areas or occasion-
ally normal skin, giving on examination a plague with
knobby fibrosis at some areas and discharging sinuses
or fluctuant masses in others. The face and neck are the
most common sites affected, in particular the parotid,
submandibular, and supraclavicular regions and the
lateral aspects of the neck corresponding to the groups
of cervical lymph nodes. Other groups oflymph nodes
could be affected (Fig. 36.6a, b). Lesions on the extremi-
ties or on the trunk accompany tuberculous disease of
pharyngeal bones and joints, the sternum, the ribs, and
the spine (Fig. 36.7a-i).
Course
Spontaneous resolution does occur but the process
is protracted (indolent). Inflammatory and ulcerative
Fig.36.5. Scrofuloderma. A broken down skin with bluish
rolled-in margins and granulating uneven floor
 Tuberculosis of the Skin
a
b 1939). Men are more frequently affected, particularly
Fig. 36.6. a Scrofuloderma. Bilateral inguinal tuberculous
lymphadenopathy which shows on the right side an enlarged
lymph node and above it a scar of a lymph node biopsy while
on the left there is a puckered scar of a previously draining
liquefied lymph node. On the left upper thigh there is an area
of scrofuloderma with dense fibrosis. b The left inguinal area
of same patient as (a) after antituberculous treatment
lesions take years before complete healing with scar-
ring. The thick fibrotic scars are so characteristic that
diagnosis could be made even after healing. Lupus
vulgaris may complicate preexisting scrofuloderma
(Tapeiner and Wolff 1999).
Histopathology
The typical caseating granuloma could be seen at
the periphery of an abscess or the margin of a sinus.
The central changes are nonspecific showing massive
necrosis and abscess formation (Fig. 36.8a, b). Tuber-
culosis bacilli are usually seen and can be cultured.
Diagnosis and Differential Diagnosis
Scrofuloderma has to be differentiated from deep
fungal infections, e.g., sporotrichosis and blastomy-
cosis (which yield typical fungal elements on culture);
lymphogranuloma venereum (in the presence of nega-
tive LVG complement fixation test); from actinomyco-
637
sis and tularemia by culture; syphilitic gumma which
form deep craters when ulcerated; and severe forms
of acne conglobata and hydradenitis suppurativa.
Infection by M. scrofulaceum must be excluded as
it is poorly responsive to antituberculous treatment
and affected lymph nodes have to be surgically excised.
Diagnosis is confirmed by bacterial culture.
36.3.7.6
Tuberculosis Cutis Oriticialis
Tuberculosis cutis orificialis (TCO) is a rare postpri-
mary form of cutaneous tuberculosis affecting the
mucous membranes and adjacent skin due to auto-
inoculation of mycobacteria from an advanced tuber-
culous infection of internal organs in a pre-sensitized
individual who has impaired cellular immunity.
Incidence
It is a rare form of cutaneous tuberculosis affecting
only 0.2% of patients with internal disease (Bryant
among middle-aged or elderly patients.
Pathogenesis
The underlying tuberculous disease is usually an
advanced pulmonary, intestinal, or rarely a genito-
urinary infection. Large numbers of bacilli are shed
from the foci, and trauma determines sites of inocu-
lation. Histopathologic examinations at autopsy of
patients with pulmonary tuberculosis revealed a
higher incidence of oral involvement than expected,
indicating that many mucosal lesions are undetected
clinically (Katz 1941).
Clinical Features
The oral cavity is commonly affected in patients with
pulmonary tuberculosis, the tongue, particularly the
tip and lateral margins, being the most frequently
involved. The soft and hard palate, the lips, teeth
extraction sockets and the nose are other commonly
affected sites. In patients with intestinal tuberculosis
the area at and around the anus is involved while in
genitourinary tuberculosis the vulva, glans, penis, and
urinary meatus might be affected to varying degrees
(Fisher 1977; Banerjee 1956; Nepomuceno 1971). The
initial lesion is a small yellowish or reddish edematous
nodule that rapidly breaks down to form an exquisitely
painful shallow ulcer with a typical "punched out"
look, undermined bluish edges and soft consistency.
e f

Fig.36.7. a Scrofuloderma. An indurated area with ulcers is seen

over the tuberculous sternum and medial end of the left clavicle.
b Abscess formation before breakdown. c The infiltrated plaque of
scrofuloderma. d The necrotic purulent base of the ulcer is seen with
granulation tissue evident in one. e Postcontrast-enhanced axial CT
image of sternoclavicular region demonstrates destruction of the
medial end of the left clavicle and associated abscess formation
(arrow). fAxial CT image distal to (e) demonstrates destruction of
the manubrium sterni and the left sternoclavicular joint (arrow).
g Axial postgadoliniurn Tl-weighted MRI image demonstrates
enhancement of the irregular wall of the tuberculous cavity (arrow).
h Sagittal Tl-weighted MR image of the same patient demonstrates
low signal intensity tuberculous abscess collection in the region of
the left sternoclavicular joint (arrows). i Sagittal TZ-weighted MR
image of the same level demonstrating high signal intensity collec-
tion (arrow) caused by tuberculous abscess
 Tuberculosis of the Skin
Fig. 36.8. a Histologic section of a scrofulodermatous tuberculous skin lesion exhibiting suppurative process with scattered multi-
nucleated giant cells and aborted attempt at granuloma formation. b Close-up of the section showing epithelioid and multinucle-
ated giant cells
The floor is covered by a pseudomembrane and often
exhibits multiple yellowish tubercles and broken ves-
sels. The surrounding mucosa is swollen, edematous,
and inflamed. It seldom exceeds 2 em, may be single
or multiple and show little tendency toward healing.
These lesions are painful and may result in dysphagia.
They signify poor prognosis.
Course
patients with orificial tuberculosis usually run a down-
hill course. The enlarging lesions reflect progression of
internal disease and herald a most unfavorable prog-
nosis. Anergy develops with loss of tuberculin skin
sensitivity in terminal stages.
Histopathology
There is severe nonspecific inflammatory infiltrate
and necrosis. Granulomatous tubercle formation is
not seen except may be in the deep dermis. Myco-
bacteria are easily demonstrated.
Diagnosis and Differential Diagnosis
Aphthous ulcers, syphilitic lesions, and carcinoma
should be excluded. There is usually abundant evi-
dence of the disease elsewhere and the development
of painful ulcers at mucocutaneous junctions will help
in making the diagnosis (Tapeiner and Wolff 1999).
Smears and histologic specimens show large numbers
of bacilli. Bacterial culture confirms the diagnosis.
36.3.1.7
Metastatic Tuberculous Abscess (Gumma)
Tuberculous gumma (TG) is a postprimary tubercu-
lous process that results from hematogenous spread
639
of mycobacteria from a distant focus during periods
of lowered or altered immunity, leading to the devel-
opment of a single or multiple cutaneous-subcutane-
ous abscesses.
It occurs mainly in malnourished children and
adults who are immunocompromised.
Pathogenesis
The bacilli lodge in subcutaneous tissue following
periods of silent bacteremia and altered immunity
particularly at sites of previous trauma resulting
in the formation of one or more subcutaneous
cold abscesses. Progressive organ tuberculosis and
miliary tuberculosis usually precede abscess devel-
opment, but they can develop without an obvious
tuberculous focus.
Clinical Features
Early in the course of the development of the tuber-
culous gumma, a firm subcutaneous nodule forms
that slowly softens becoming a fluctuant nontender
subcutaneous abscess. The lesion may be single or
multiple, appearing on the trunk, extremities, or head
(Fig. 36.9a, b) (Tapeiner and Wolff 1999). Abscesses
break down through the overlying skin to form ulcers
with undermined edges and fistulas (Fig. 36.lOa, b).
The discharge could be either purulent or serosan-
guinous (Fig. 36.11). The surrounding skin is bluish
and tethered to the underlying mass resembling
scrofuloderma. Rarely, secondary lesions develop
along the draining lymphatics (Gawkrodger 1998).
Tuberculous Abscesses of the Breast
Mammary involvement is rare in either primary
or postprimary cases of tuberculosis. The largest
 640 A. J. Al-Kudwah

a .....- - - b

Fig. 36.9. a Metastatic tuberculous abscess. An abscess formation with induration of the affected tissue and pusy discharge from
the broken down areas. Courtesy of Dr. A. Al-Talaq. b A close-up view of the abscess. Courtesy of Dr. A. Al-Talaq

Fig. 36.10. a A healed tuberculous fistula showing tethering of

the skin onto the underlying previously affected rib. b Close-
a up view of the area

numbers of these cases came from India and Turkey

(Kakkar et al. 2000; Khanna et al. 2002; Goksoy et
al. 1995). The clinical features include a breast lump,
breast lump with sinus formation, or persistent
abscess of the breast with or without associated
lymph adenopathy. An underlying tuberculous pro-
cess in the chest wall, pleura, or adjacent lymph nodes
is often present.

Course

Depending on the general condition of the patient,

Fig. 36.1 I. A tuberculous abscess. A serosanguinous discharge
is seen
the integrity of the cellular immunity, and the
presence or absence of an advanced pulmonary or
extrapulmonary tuberculous process, the patient
may either be gravely ill or fairly asymptomatic apart
from lesion-related complaints. Tuberculin skin test
Tuberculosis of the Skin
positivity is often lower than with other forms of
postprimary tuberculosis but becomes absent in the
very ill.
Histopathology
There is abscess formation with necrosis in the center.
Tubercles with granuloma formation are found at the
periphery of the lesion and bacilli are usually seen by
acid-fast staining.
Diagnosis and Differential Diagnosis
The diagnosis of TG depends on microbiologic and
histopathologic findings of specimens obtained
from affected tissues. The differential diagnosis
includes all forms of panniculitis, deep fungal infec-
tions including sporotrichosis, syphilitic gumma,
and hydradenitis suppurativa. Tuberculous mastitis
should enter in the differential diagnosis of any
breast lump especially in areas where tuberculosis
remains highly prevalent.
36.3.7.8
BCG Vaccination and Cutaneous Tuberculosis
BCG vaccination has substantially reduced the inci-
dence of childhood tuberculosis (Curtis et al. 1984).
Two weeks after the intracutaneous introduction
of bacille Calmette-Guerin, an infiltrated papule
develops. It gradually enlarges to 1 cm in size at
6-12 weeks, it ulcerates and slowly heals leaving a
scar. Regional lymph nodes may enlarge but usu-
ally do not break down. Positive skin testing usually
appears 5 to 6 weeks later.
Complications related to BCG vaccination may
occur and can be divided into nonspecific and spe-
cific (Dostrovsky and Sagher 1963; Jorgensen and
Horwitz 1956). Nonspecific complications include
keloid formation, epithelial cysts, granuloma,
eczema, generalized hemorrhagic rashes, and
erythema nodosum. The specific complication is
the development of a localized or systemic tuber-
culous process caused by the attenuated bovine
bacillus. The majority of cases are lymphoglan-
dular and/or skin reactions that usually simulate
"naturally" acquired mycobacterial infections.
Horwitz and Meyer (1957) reported the develop-
ment of nonfatal general complications in 1 or
2 persons/million, perforating regionallymphad-
enitis in 2% of vaccinated children in Denmark
while postvaccinal lupus vulgaris developed in
5-10 persons/million (Horwitz and Meyer 1957;
641
Waaler and Rouillon 1974). BCG reactions are usu-
ally of milder nature than "spontaneous" tubercu-
losis of the skin and more often occur after revac-
cination (Jorgensen and Horwitz 1956; Izumi and
Matsunaga 1982).
Among these specific reactions are:
1. Lupus vulgaris:
This form doesn't differ from the usual lupus
vulgaris morphologically and follows the same
course and responds to the same treatment. It
develops near the vaccination site after a latent
period of several months to years. Organisms
can be recovered from the lesion in a quarter of
patients.
2. Koch's phenomenon with necrosis and ulceration
may develop in previously sensitized individuals
and takes on a shorter time course. Regional lymph-
adenitis is common and constitutional symptoms
may be present (Horwitz and Meyer 1957).
3. Local subcutaneous abscesses:
This is a complication of faulty vaccination tech-
nique, where the bacillus is injected too deeply into
the skin. Severe ulceration may ensue (Dostrovsky
and Sagher 1963).
4. Severe regional lymphadenitis is definitely the
most commonly occurring complication in the
younger age group.
5. Scrofuloderma may develop as a complication
of regional lymphadenitis, and suppuration may
persist for 6-12 months.
6. Generalized tuberculid-like eruption has rarely
been reported (Dostrovsky and Sagher 1963;
Jorgensen and Horwitz 1956).
7. A tuberculous infection in a distant organ (lymph
nodes, bone, and joints) has occurred occasionally
(Rouillon and Waaler 1976).
Other rare but serious reactions are anaphylactic
reaction and hepatic dysfunction (Aungst et al.1975).
Fatal disease secondary to generalized BCG tubercu-
losis is very rare (l per million vaccinations) among
immunocompromised patients.
36.3.7.9
Congenital Tuberculosis
Congenital tuberculosis is still being reported in
endemic sub-Saharan African countries (see chapter
on childhood tuberculosis). It should be considered
in an infant with febrile systemic illness who exhibits
discrete, erythematous lesions with central necrotic
dell. The diagnosis can be confirmed by histologic
and microbiologic means.
642
36.4
The Tuberculids
These are generalized eruptions that develop in
response to an internal focus of tuberculosis giving rise
to lesions that are usually transient, symmetric, and dis-
seminated. Darier (1896) designated the name initially
to a dissimilar group of eruptions that were thought
to be tuberculous in origin (Darier 1896). Many of the
previously called "tuberculids" do not fulfill the criteria
for "true tuberculids" and they are not related to tuber-
culosis and represent small indolent granulomas.
The following criteria have to be fulfilled for a
condition to be a true tuberculid:
1. The lesion should show a tuberculoid histology.
2. Mycobacteria should not be demonstrated in the
lesion.
3. Tuberculin skin testing is highly positive.
4. Lesions resolution following treatment with anti-
tuberculous drugs.
5. Presence of a concomitant focus of a tuberculous
infection elsewhere in the body.
Pathogenesis
Tuberculids are probably due to hematogenous spread
of tubercle bacilli in patients with a moderate to high
degree of immunity. The underlying tuberculous
process might not be evident and the general health
is usually excellent. Changes in the immune status of
the patient, stasis, skin temperature, and blood supply
affect the development of the eruptions. Bacilli are not
seen in the lesions because of their small number or
due to their rapid destruction (Miescher 1955). Myco-
bacterial DNA has been demonstrated in papulone-
crotic tuberculids and erythema induratum of Bazin
using polymerase chain reaction (PCR) (Victor et al.
1992; Degitz et al. 1993).
36.4.1
Tuberculosis-specific Tuberculids
36.4.1.1
Papulonecrotic Tuberculids
Papulonecrotic tuberculids (PnT) are symmetric
necrotizing papules appearing in crops, affecting
the extremities, and healing with scars.
Incidence
PnT have become rare in recent literature. It is not so
uncommon in areas with high prevalence of tubercu-
A. J. AI-Kudwah
losis. Morrison and Fourie (1974) reported 91 cases
seen in South Africa over a 17-year period (Degitz
et al. 1993). Children and young adults are usually
affected.
Pathogenesis
The bacilli enter the blood circulation periodically
where they are opsonized and settle preferentially
in skin capillaries. Immune complexes form in an
Arthus-like reaction to be followed by a delayed
hypersensitivity response to the mycobacteria.
Clinical Features
Papulonecrotic tuberculid involves recurrent, sym-
metric, firm, dusky red, pinhead- to pea-sized
papules that appear in crops with predilection to
extensor surfaces of the extremities, in particular the
knees, tips of elbows, buttocks, and lower trunk, and
the dorsal surfaces of hands and feet. The face, ears,
and penis may be involved sometimes alone (Kumar
and Sharma 1987; Nishigori et al. 1986).
Papules show a central depression which is followed
by crust formation that upon removal leaves a crater-
like ulcer. It probably signifies localized necrosis. In
some cases the papules heal rapidly (within weeks)
while in others they form ulcers that heal after many
months. Healing ends with the development of white
punched-out scars that might show marginal hyper-
pigmentation. Fresh outbreaks follow healing/healed
lesions and the disease process may last for years.
Spontaneous resolution may occur. It affects young
adults predominantly but has been seen in infants
and young children. Often there are no symptoms
related to other organ involvement but fever, malaise,
and other constitutional symptoms might precede the
eruption. In some cases it has followed BCG vaccina-
tion (de Bruyne et al. 1953) and in others transition
into lupus vulgaris has occurred and has an associa-
tion with erythema induratum (Morrison and Fourie
1974; Jordaan et al. 1994). Tuberculous arteritis with
subsequent gangrene has been reported.
Course
The clinical picture is governed by the individual
immune responsiveness and the number and viru-
lence of the mycobacteria. The underlying tuberculous
focus may not be apparent and extracutaneous tuber-
culosis has been found in only 30-40% of cases (Simon
1959). Prompt response to antituberculous therapy has
been described (Morrison and Fourie 1974).
 Tuberculosis of the Skin
Histopathology
In a well-developed PnT lesion there is a wedge-
shaped necrosis in the upper dermis extending to and
involving the epidermis, the dermis, and sometimes
the subcutaneous tissue. The inflammatory infiltrate
surrounding the area is usually tuberculoid, but may
be nonspecific. Obliterative and sometimes granu-
lomatous vasculitis is a cardinal feature leading to
thrombosis and complete occlusion of the vascular
channel. Recanalization has been observed (Tapeiner
and Wolff 1999).
Diagnosis and Differential Diagnosis
The clinical features of PnT are classical, but because
of the rarity of the condition today and the frequent
absence of an obvious focus of tuberculosis, it may
deceive the unwary (Tapeiner and Wolff 1999).
Diagnosis should be confirmed histologically
or by using PCR when available. A therapeutic test
is usually decisive. Differential diagnoses include
insect bites that are usually itchy and less sym-
metric. Prurigo nodules are also itchy and differ
in distribution. Papular secondary syphilis is more
widely distributed with palm and sole involvement.
Serologic testing for syphilis is essential. Leuko-
cytoclastic vasculitis rash is less indurated, more
red in color, might show dark central vesiculation
depending on severity, but histopathologic findings
may help in reaching a diagnosis despite some simi-
larities. Pityriasis lichenoides et varioliformis acuta
(PLEVA) may present a diagnostic challenge but is
more wide spread, involving the trunk, palms, and
soles.
Fig. 36.12. a Lichen scrofulosorum grouped follicular and perifollicular scaly papules.
a b Close-up view of the grouped papules
643
36.4.1.2
Lichen Scrofulosorum
This is a lichenoid eruption of 2-S-mm erythema-
tous, grouped, scaly often perifollicular papules in a
patient with high sensitivity to M. tuberculosis.
Incidence
Lichen scrofulosorum (LSc) is a rare form of tuber-
culosis. It was first recognized by Hebra (Rauschkolb
1934). Children are usually affected as well as adoles-
cents and adults.
Pathogenesis
Lichen scrofulosorum is usuallyassociated with chronic
tuberculous infection of lymph nodes, bones, or both.
Rarely, it is associated with pulmonary tuberculosis. It
results from hematogenous spread of mycobacteria. It
has been observed following BCG vaccination (Curtis
et al. 1984).
Clinical Features
Lesions of lichen scrofulosorum consist of small,
firm, yellow to reddish brown follicular or perifollicu-
1ar papules. They may be flat topped or scaly, or bear a
minute horny spine. Rarely, they may be surmounted
by a tiny pustule. Grouping of individual lesions
results in the formation of discoid plaques usually
on the trunk where they may persist unchanged for
long periods, involuting slowly over months without
scarring (Fig. 36.12a, b) (Tapeiner and Wolff 1999;
Goldschmidt and Grekin 1990).
644
Course
Patients are usually well, due to the lack of consti-
tutional symptoms. Chronic plaques may persist
unchanged for months. Antituberculous treatment
results in complete resolution, though slower and
less dramatically than with other tuberculids.
Histopathology
Superficial dermal tuberculoid granulomas usually
develop around hair follicles and sweat ducts, but may
also develop independent of the adnexa (Montgomery
1937). It often abuts on and erodes the epidermis. There
is no caseation usually, and bacilli are not seen.
Diagnosis and Differential Diagnosis
Biopsy will show the tuberculoid pathology. Lichen
nitidus lesions are smaller, shinier, and more periph-
eral. Keratosis pilaris and lichen spinulosa papules
are usually skin colored and truly follicular. Lichen
planus, secondary syphilis, and sarcoidosis should
also be excluded.
36.4.2
Tuberculosis Nonspecific Tuberculids
36.4.2.7
Erythema Induratum
Erythema induratum (EI) is a chronic recurring
nodular and ulcerative disorder that affects the calves
of the legs, occurring secondary to tuberculosis else-
where in the body. Other stimuli such as ischemia or
infection can provoke this reactive pattern. These are
better called nodular vasculitis (Montgomery et al.
1945; Eberhartinger 1963).
Incidence
Cases of nodular vasculitis in general comprise 0.1-
0.2% of dermatologic patients seen in a university
hospital in Europe (Montgomery et al. 1945; Eber-
hartinger 1963). It affects predominantly women with
peaks in adolescence and at menopause. There is also
a seasonal prevalence in winter and early spring.
Pathogenesis
As most patients with this condition have eryth-
rocyanotic lower extremities and thick, firm, but
A. J. AI-Kudwah
not edematous, skin, it has long been accepted that
transient mycobacteremia produces a reaction that
is both induced by and superimposed upon the basic
circulatory disorder. The pathogenic significance of
the vasculature is clearly demonstrated by the vascu-
litic histologic pattern. Vessels of these patients react
abnormally to changes in ambient temperature, thus
cutis marmorata is common and the association of
the rash with cold exposure. Mycobacterial DNA
using PCR has been demonstrated in 20-50% of the
cases. The peripheral T lymphocytes of patients with
EI show an exaggerated response to PPD skin testing.
T cells predominate in biopsy specimens suggesting
aT-cell mediated response to a tuberculous antigen
(Kuramato et al. 1990; Oller et al. 1993).
Clinical Features
In the initial stages, there are slightly tender ery-
thematous indurations appearing on the back of
the lower legs of young to middle-aged women on a
background of erythrocyanotic circulation. Affected
legs tend to be plump and follicular perniosis may
be present. Regression of these lesions occurs during
the warmer months at first, but as the condition pro-
gresses the indurations give rise to indolent, mildly
tender, ill-defined, dull red, deep nodules. These
nodules often persist for months and may either dry
up and heal or soften to form deep persistent, ragged
ulcers with bluish margins (Goldschmidt and Grekin
1990). Resolution is usually slow because of the cir-
culatory dysfunction. It may take weeks and ends in
scarring (Forstrom and Hannuksela 1970).
Course
At any given time, all stages of activity of current
lesions as well as the depressed white scars of previ-
ous ulcers can be seen in anyone case. The disease
may persist for years. A search should be made for
active tuberculosis elsewhere, e.g., at nasopharyn-
geal, renal, or endometrial sites.
Diagnosis and Differential Diagnosis
The constellation of indurated ulceration of the skin
over the calves on a background of erythrocyanotic
circulation is characteristic. Occasionally the circu-
latory disturbance is not obvious. A high index of
suspicion is required to make an early diagnosis, and
confirmation can be achieved by histology and PCR.
Differential diagnoses include nodular vasculitis of
nontuberculous origin which lacks response to anti-
Tuberculosis of the Skin
tuberculous treatment. Erythema nodosum affects
primarily the shins of the legs, and is of short dura-
tion and rapid resolution. Syphilitic gumma is usually
unilateral, single or forming a small group of lesions;
serologic tests and histology are decisive. Nodose
lesions due to iodide and bromide resemble erythema
nodosum or cause ulcers with vegetating bases.
Histopathology
The histologic features of EI are those of nodular vascu-
litis with areas of tuberculoid granulomas, fat necrosis,
and foreign body giant cell reaction. Changes due to
perniosis can usually be seen (Wall and Smith 1981).
36.4.2.2
Erythema Nodosum
Erythema nodosum (EN) is a reaction pattern
manifesting as acute erythematous tender nodules
affecting extension aspects of legs especially over the
shins and less commonly the thighs and forearms.
Tuberculosis is only one of many infections that can
lead to its development and infections are among
many other etiologic factors.
Incidence
The average hospital incidence of EN (regardless
of it etiology) in the UK is about 0.5% of new skin
cases seen. The prevalence of certain infections will
be reflected in the incidence of this reactive pattern
according to geographic differences and seasonal
variations (Ellis et al. 1982; Riska and Selnoos 1964;
Ross et al. 1971). The age of onset is between 20 and
45 years with a peak from 20 to 30. There is a female
preponderance with a female to male ratio of 3-6: 1
(depending on geographic location and differences in
etiologic factors) that is less dramatic in children.
Pathogenesis
A delayed hypersensitivity reaction results in the
formation of immune complexes (IC). IC are depos-
ited in and around venules of the deep dermis. The
slow-flowing, rich but highly permeable capillary
bed of the adipose tissue favors their deposition.
This sequence of events results in the development
of erythema nodosum lesions at sites of predilection.
This "reaction" pattern is elicited by many triggering
factors. There is a predilection for the shins because
of the combination of (1) relatively sparse arterial
supply, (2) sluggish flow in the venous system because
645
of gravitational effect, cooling, and the lack of mus-
cular pump, (3) a lymphatic system that is unable to
meet any increase in fluid load and has no mechani-
cal advantage. Immune complexes are deposited in
the blood vessel wall causing physical damage mani-
festing itself in the characteristic "bruising" seen in
many EN lesions. Granulomas are formed around
vascular channels. The damage to the blood vessel
wall is not complete, so that a degree of blood flow
is maintained, thus avoiding the necrosis, ulceration,
and scarring that is seen in erythema induratum.
Once immune complexes are cleared, healing takes
place with initial postinflammatory hyperpigmenta-
tion that also eventually disappears.
Etiology
Antigenic stimuli for the development of erythema
nodosum can be categorized into infections, drugs,
benign and malignant systemic disease, and a few
miscellaneous disorders. See Table 36.3.
Clinical Features
The skin lesions appear as erythematous, raised, hot,
tender nodules on the anterior aspect of the lower
legs (shins) which are fairly classical (Fig. 36.13). The
elevation at the center is more pronounced than that
at the periphery. Their color changes from red to
dusky red with a purplish tinge, during the 2nd week
of the eruption, to yellowish green like a bruise,
most marked during the 3rd week, before subsiding
with postinflammatory hyperpigmentation without
ulceration or scarring. The inflammation subsides
between the 4th and 6th week. The lateral aspects of
the legs, thighs, and the extensor aspects of the arms,
face, and any area of the body with subcutaneous fat
can be involved but less frequently. The eruption is
Fig. 36.13. Erythema nodosum. Classic lesions are seen over
the shins
646 A. J. Al-Kudwah

often preceded by fever of 38-39°C, chills, malaise, Histopathology

and arthralgia. Lesions continue to erupt for about
10 days. Bed rest is important to minimize the aching The pathology is that of septal panniculitis. There
of the legs and the ankle swelling. Recovery may take is usually a septal lymphohistiocytic infiltrate with
several weeks. Episcleral lesions (phlyctenular con- few neutrophils, but rarely do neutrophils predomi-
junctivitis), hilar lymphadenopathy and arthropathy nate. There is often small as well as medium-size
(in 70% of patients) may develop during the course of vessel invasion with inflammatory infiltrate with or
the eruption, which may persist for a long time after without extravasation of erythrocytes. Involvement
resolution of the skin lesions. Familial cases occur in of fat lobules is limited to their periphery, so fatty
tuberculous families and streptococcal cases. Upper tissue necrosis does not occur. At a later stage, giant
respiratory tract infection may precede the eruption cells are present either alone or in association with
by 7-14 days and medications given for it might be a few histiocytes but true granulomas are very rare.
erroneously blamed for the eruption. Attention to In the chronic migratory type, granulomas are more
specific points in the history might lead to identifi- pronounced and vascular changes not as much, but
cation of the possible etiologic factor. there is no caseation.
Erythema nodosum migrans is a variant of the
condition that is also known as subacute nodular Diagnosis and Differential Diagnosis
migratory panniculitis (Bafverstedt 1954; Vilanova
and Pinol-Aguade 1959; Rostas 1980). The eruption is Post-traumatic phlebitis and cellulitis might
often unilateral, and lesions are fewer in number, less resemble erythema nodosum but they are usually
tender, and persist longer than in classic erythema unilateral. Nodular vasculitis elicits smaller, tender,
nodosum, affecting the lateral aspect of the leg rather and chronic lesions that resolve slowly and lead to
than the anterior shin. The lesion progresses and depressed scars on the calves usually on the back and
extends laterally over a period of weeks or months sides of the legs. The very early infiltrated plaques of
changing its configuration. Its outline becomes pyoderma gangrenosum may be difficult to differen-
irregular, forming crescents or arciform patterns tiate without the presence of an associated disease.
with bright erythematous borders and brownish Insect bites with abnormal reaction are irregular in
violaceous centers (Rostas 1980). Bruising does not distribution. Granulomas secondary to Trichophyton
occur and trauma might determine the initial site infection of the feet are situated lower over the legs.
(Gawkrodger 1998). In familial Mediterranean fever a erythema nodo-
sum-like lesion may occur.
Course
Treatment
Crops of new lesions can occur for up to 6 weeks.
Mild cases resolve in 3 weeks and severe ones take Spontaneous resolution of EN usually occurs. Bed
longer. Aches and pain in the legs, and ankle edema rest is important and supportive bandaging should
especially towards the evening, may occur especially be used when the patient is ambulatory. If eruption
if the patient remains ambulant. is severe, anti-inflammatory drugs may be of use.
The use of systemic steroids is potentially danger-
Laboratory Investigations ous because of a possible underlying infection, and
may only be necessary if the eruption is extensive and
Laboratory abnormality in EN may be lacking but an severe. Oral potassium iodide has been used as well
elevated erythrocyte sedimentation rate is common. as hydroxychloroquine (Horio et al.1983; Schulz and
Depending on the clinical history, physical findings, Whiting 1976; Jarrett and Goodfield 1996).
age of the patient, and the geographic location, a
decision is made on what further tests to do. A full
blood count, blood chemistry as well as a chest x- 36.4.3
ray are needed in all erythema nodosum patients. A Nontuberculids
skin biopsy is mandatory. The standard punch biopsy
is inadequate as frequently the subcutaneous fat is In these conditions the tuberculous etiology has not
not included. An excisional biopsy deep enough to been proven and no mycobacteria have been recov-
include subcutaneous fat is needed to confirm the ered from the lesions although they exhibit tuber-
diagnosis. culoid features on histology. Tuberculin skin testing
Tuberculosis of the Skin
sensitivity is low and in the absence of an underly-
ing tuberculous infection they are not an id reaction.
The incidence of past tuberculous infections among
patients with these conditions does not exceed that
of the general population. Three clinical entities are
recognized and discussed below.
36.4.3.7
Lupus Miliar;s D;ssem;nafus Fac;e; (Acne Agm;nafa)
This is a chronic popular eruption of the face with
spontaneous involution after a variable period of
12-24 months.
Pathogenesis
It appears that this is either due to a reaction to
a factor yet unidentified or due to many etiologic
triggers giving rise to a single morphologic entity.
Implicated factors include zirconium, sebum, follicu-
lar contents, and even Demodex folliculorum (Ueki
and Masuda 1979; Pinkus and Mehregan 1981). Most
cases are currently believed to represent a sarcoidal
form of rosacea.
Clinical Features
Lupus miliaris disseminatus faciei (LMF) occurs pre-
dominantly in young adults and adolescents of both
sexes, particularly of Japanese origin. It consists of
multiple discrete indolent, dull brown follicular or
nonfollicular dome-shaped papules with yellowish
center. Their size ranges between 1 and 3 mm in
diameter and they are symmetrically distributed
preferentially in the central "muzzle" area of the face:
the lower portion of the forehead, temples, periorbital
areas, bridge of the nose, cheeks, nasolabial folds,
perioral areas, and chin. Occasionally dissemination
occurs to involve arms, legs, trunk, and axillae. The
papules are initially soft and translucent, become
firm, may crust or pustulate before resolving with
scar formation. Apple-jelly nodules similar to those
of lupus vulgaris are demonstrated by diascopy.
Course
The natural history is that of spontaneous resolution
with scarring.
Histology
Tuberculoid follicles with central caseation are seen
in the mid-dermis in relation to the pilosebaceous
647
unit with a striking similarity to tuberculosis.
Damage to the vessel wall with extravasation of red
blood cells and leukocytoclasia (nuclear dust) are
seen in the early lesions.
Diagnosis and Differential Diagnosis
LMF needs to be differentiated from classic rosacea
which shows background erythema and telangiecta-
sia with polymorphic lesions, and from acne vulgaris
with its comedonal hallmark and pleomorphism.
Lewandowsky's rosacea-like eruption has more
peripheral distribution and smaller papules; perioral
dermatitis does not usually involve periorbital areas;
sarcoidosis; lupus vulgaris is less symmetric and less
florid; and papular granuloma annulare involves a
more generalized papular eruption. The combination
of the brownish red papules and the pitted atrophic
scars in the center of the face is characteristic. Histol-
ogy will help to establish the diagnosis.
Treatment
There is no single effective treatment. Tetracyclines
may be beneficial in some patients. Dapsone has been
used successfully in others (Kumanu et al. 1983) and
appears to shorten the expected natural duration and
prevent the eruption of new crops of lesions, which
has also been achieved with the use of oral steroids
(O'Driscoll and Morgan 1974).
36.4.3.2
Rosacea-like Tuberculid
This is also known as Lewandowsky's rosacea, and
it is largely agreed today that this is a micropapular
form of rosacea with histologic tuberculoid features
(Lewandowsky 1917).
Clinical Features
Women aged 20-50 are affected with these numerous
tiny, slightly indurated, flattened papules on a back-
ground of persistent erythema. The lesions are more
peripheral than classic rosacea affecting the outer
aspect of the cheeks and forehead. Occasionally the
rash involves the whole face.
Treatment
Gradual and slow spontaneous resolution is common.
Tetracyclines may be of help but the response is slug-
gish compared with rosacea.
648
36.4.3.3
Lichenoid Tuberculid
Lichenoid tuberculid (LT) was originally described
by Ockuly and Montgomery (1950), and it is likely
that the majority of their cases were sarcoidal reac-
tions. Only a few additional cases have been described
(Ockuly and Montgomery 1950; Kowalenko 1963;
Schuhmachers 1967).
Clinical Features
The skin lesions of LT present as symmetric, small,
reddish purple flat-topped papules that suddenly
appear, affecting the extremities rather than the
trunk, with a tendency to dissemination in middle-
aged patients. The papules are 3-5 mm in diameter
and might be surmounted by a fine scale. Grouping of
lesions has been noted and when coalesced, annular
configuration ensues. It resolves spontaneously leav-
ing brownish macules with no scarring.
Diagnosis and Differential Diagnosis
The clinical picture can only be differentiated from
other similar eruptions histologically. It should be
distinguished from lichen planus, pityriasis lichen-
oides chronica, and sarcoidosis.
36.5
Cutaneous Tuberculosis
in the Immunocompromised
Patients who have impaired immune responses due
to various causes, as well as patients with HIV infec-
tions, are at the risk ofexhibiting unusual, atypical, and
severe forms of cutaneous tuberculosis which do not
fit in the aforementioned categories. It is imperative
that any cutaneous lesions developing in immunosup-
pressed patients should be biopsied for histologic, bac-
terial, and fungal studies. Histologically well-formed
granulomas are uncommon, but AFBs are seen in
abundance. The skin biopsy culture may yield either
M. tuberculosis, M. bovis, or atypical mycobacteria.
Patients usually respond to antituberculous drugs
but some die despite appropriate treatment. BCG vac-
cination of the patients has resulted in disseminated
disease, and targeted benefits as well as definite risks
have to be carefully weighed in immunocompro-
mised patients (Watson and Gill 1990; Tarantola and
Mann 1987; Lane and Fauci 1985).
A. J. Al-Kudwah
36.6
Atypical Mycobacterial Cutaneous
Infections
Atypical mycobacteria are widely distributed organ-
isms and are present in nature as environmental
saprophytes found in animal and human feces,
soil, vegetation, and water in lakes, rivers, swim-
ming pools, and aquariums. They are usually non-
pathogenic to man, eliciting an immune response
without overt infection, as they are of low virulence.
A degree of impaired host immunity is needed for
the development of the disease. The outcome of the
infection is determined by the organisms' virulence,
the inoculum load, and the host immunity. The most
common target organs are the lungs, bones, joints,
and less frequently the skin, where trauma is essen-
tial for the introduction of the infective organism.
Infection with the atypical mycobacteria runs a
more benign and limited course than that caused
by M. tuberculosis.
Transmission from person to person is exceed-
ingly rare and as the reservoir of the organism is
large, control methods known to be effective in
tuberculosis are ineffective in the control of atypi-
cal mycobacterial infections. With the decreased
incidence of mycobacterial infection of the skin
especially in developed countries there seems to be a
relative increase in infections caused by other myco-
bacteria (Palenque 2000).
History
Only recently did the pathogenicity of slow-
growers other than M. leprae and M. tuberculosis
become recognized: Infections clinically mimick-
ing those caused by M. tuberculosis and the strict
and often unusual culture requirements are in
part responsible for this delay in recognition. In
1948, M. ulcerans was identified as the cause of an
ulcerating skin condition in Australia. Ten years
later it was recognized to be the etiologic agent of
the "Buruli ulcer" in the Buruli district of Uganda.
M. marinum was isolated from patients with swim-
ming pool granulomas in 1954, although had been
known since 1926 (Clancey et al. 1961; MacCullum
et al. 1948; Linell and Norden 1954). The identi-
fication of rapid-growers atypical mycobacteria
as human pathogens was reported earlier, when
DaCosta isolated an organism from a postinjec-
tion abscess in 1938 and named it M. fortuitum
(DaCosta Cruz 1938).
Tuberculosis of the Skin
Pathogenesis
An alteration in the host immune status, or trauma or
damage to a particular organ including the skin, is a
prerequisite for the development of infection in most
cases. Pulmonary infections are frequent in patients
with preexisting lung disease and are pathologi-
cally indistinguishable from M. tuberculosis-caused
infections. Skin and lymph node infections can be
granulomatous, suppurative, or mixed, mimicking
tuberculosis, sporotrichosis, leishmaniasis, or other
diseases (Tapeiner and Wolff 1999).
The most commonly encountered atypical myco-
bacterial skin infections are discussed below.
36.6.1
M. marinum Infection
Also known as "swimming pool granuloma" or "fish
tank granuloma;' this is a granulomatous eruption
caused by M. marinum which is found in fresh water
and salt water (Findlay 1980). It may be also found in
heated water in temperate climates, in the sea or natural
ponds in warmer regions, in diseased fish, mud, and
even cWorinated water where chlorination appears suf-
ficient (Beurey et al.1981). It can occur sporadically or in
small community outbreaks depending on the source.
Clinical Features
M. marinum is only pathogenic to abraded skin, and
areas of predilection are the back of the hands and on
the fingers of fish fanciers, and on the feet, elbows, and
knees of swimmers (Hautmann and Lotti 1994). The
incubation period is usually 2-3 weeks but ranges from
1 week to 2 months (Huminer et al. 1986). The usual
initial presentation is that of a solitary, tender, bluish
red nodule or pustule that may ulcerate and break
down to form a crusted ulcer, or may suppurate into
an abscess or remain warty. Lesions are usually solitary
but occasionally there are multiple lesions along the
draining lymphatics giving rise to a sporotrichoid pat-
tern. Dissemination may occur in immunosuppressed
patients (Gombert et al. 1981). Regional lymph nodes
may enlarge slightly but usually do not break down.
Spontaneous healing usually takes place within a few
months to 2 years with residual scarring.
Histopathology
The histologic features vary from nonspecific acute
or chronic inflammation to a more specific tuber-
649
culoid inflammatory infiltrate. Fibrinoid rather than
caseation necrosis is seen and Langerhans giant cells
are not always present. Intracellular acid-fast bacilli
may be seen.
Diagnosis and Differential Diagnosis
A high index of suspicion is needed. The clinical
history and histopathology are suggestive of the
diagnosis. Confirmation is obtained by demonstrat-
ing the organism in culture from biopsy specimen.
PPD testing with M. marinum antigen is of little diag-
nostic value because of its antigenic cross-reactivity
(Tapeiner and Wolff 1999). Other granulomatous
infections should be considered in the differentials:
for example, other mycobacterial infections, deep
fungal infections, tertiary syphilis and yaws, and
leishmaniasis. Verruca vulgaris and cutaneous neo-
plasia are other possibilities.
Treatment
M. marinum is poorly susceptible to antituberculous
drugs. Three drug regimens have proved successful in
the treatment: (I) tetracyclines 500 mg 3-4 times daily,
(2) rifampicin 600 mg and ethambutol 1,000 mg daily,
or (3) sulfamethoxazole and trimethoprim 960 mg
twice daily. Minocycline (or doxycycline) 200 mg
daily is considered the drug of choice by many clini-
cians (Brown and Sanders 1987). Treatment is recom-
mended for many months. Surgical excision when
feasible is effective (Wolinsky et al. 1972).
36.6.2
Mycobacterium ulcerans Infections
This is a cutaneous ulcerative process caused by M.
ulcerans that is also known as "Buruli ulcer." The dis-
ease is most common in wet, moldy, or swampy areas
of the subequatorial regions with cases clustered in
the Nile riverbed areas of Uganda. Person-to-person
transmission does not seem to occur (Yeager 1985).
Pathogenesis
The organism is introduced into the skin by micro-
abrasions, pricks, or cuts from certain plants (Feldman
1974). The infection is confined to the skin as the
organism is unable to survive the warmer core body
temperature (like M. marinum). A toxin produced by
M. ulcerans mayor may not be responsible for the
ulceration and necrosis typical of this infection. Heal-
650 A. J. Al-Kudwah

ing is always preceded by the reversal of the patient's dermal necrosis. There is no typical caseation or
anergic state when tested by M. ulcerans antigens. tubercle formation in these lesions. Acid-fast bacilli
are commonly noted in tissue biopsies.
Clinical Features
Diagnosis and Differential Diagnosis
M. ulcerans affects mostly African children and
young adults, with a female predominance. A pain- The diagnosis in endemic areas is not difficult
less subcutaneous swelling develops after a nearly because of a high index of suspicion. The histologic
3-month incubation period. The swelling becomes features and culture of the organism from tissue
larger and well defined, forming a nodule that is firm specimens confirm the diagnosis. The differential
and mobile. It eventually ulcerates forming a shallow diagnosis is made according to the presenting clinical
necrotic ulcer with deeply undermined, sometimes picture. The subcutaneous nodule has to be differen-
hyperpigmented, edges. There is rapid and irregular tiated from lipomas, subcutaneous cysts, panniculi-
extension with the ulcer reaching a diameter of sev- tis, nodular vasculitis, foreign body granuloma, and
eral centimeters over a course of a few weeks. The appendageal tumors. Ulcers have to be differentiated
floor of the ulcer is formed by necrotic fat and there from necrotizing cellulitis, pyoderma gangrenosum,
may be mucoid clear discharge (Gawkrodger 1998). deep fungal infections, suppurative panniculitis, or
The ulceration may extend into deeper structures, cutaneous neoplasia.
involving the muscle and bone. It is of interest to note
that despite the large size of the ulcer, the patient con- Treatment
tinues to feel well with no constitutional symptoms
or lymphadenopathy (Grange 1980). In the early stages, simple excision is the treatment
Itching is occasionally a feature, thus the local of choice. Larger ulcers require a wider excision with
name "mputa matadi" (the itching stone) in Ghana. skin grafting. Hyperbaric oxygen (Krieg et al. 1975)
The lesions affect the head, neck, and trunk in chil- and local heat therapy (Meyers et al. 1974) in order
dren, and extremities in adults. Van der Werf (1990) to increase affected limb temperature to levels above
made the observation that affection of the left leg in that needed for survival of the viable organism have
particular in adult Ghanaian farmers has to do with been tried. Therapeutic attempts with rifampicin
the common farming stance with the left leg forward. and trimethoprim-sulfamethoxazole can be of value
The introduction of the infection takes place near the separately or in combination (Standford and Philipps
ground where feet are vulnerable to pricks, thorns, 1972). Clofazimine has been shown to be ineffective
and grasses, while the right arm is significantly more (Revill et al. 1973).
affected in children because the use of the right hand
is a cultural imperative (van der Werf and van der
Graaf 1990). Ulcers are usually single, and satellite 36.6.3
lesions may develop. The course is prolonged with Mycobacterium kansasii Infection
ulceration persisting for months or years. Eventual
healing is with scarring Contractures and severe M. kansasii is the atypical mycobacterium that is
deformity may result from fibrosis, dystrophic calci- closely related to M. tuberculosis. It is frequently
fication, as well as lymphedema (Gawkrodger 1998; associated with chronic pulmonary disease in adult
Tapeiner and Wolff 1999). white men, but any age, sex, or race can be affected
(Lichtenstein et al. 1965). Its prevalence is higher
Histopathology in temperate areas. Cutaneous involvement usually
occurs after minor trauma, in immunocompetent
The histologic features are those of necrosis of the as well as immunocompromised hosts in whom an
dermis and the subcutaneous tissue with a charac- extracutaneous focus is more likely to be also present
teristic extensive involvement of the fat as in septal (Hirsch and Saffold 1976).
panniculitis. The subcutaneous necrotic fat becomes
surrounded by granulation tissue and giant cells. Clinical Features
Muscles are usually spared probably because of
their higher temperature. The deep dermis usually Skin lesions due to M. kansasii infection are clinically
shows leukocytoclastic vasculitis affecting small and variable and most frequently present as verrucous
medium-size vessels which probably account for the nodules with sporotrichoid spread. Papulopustules,
Tuberculosis of the Skin
abscesses, crusted ulcers or cellulitis-like lesions have
been reported (Hirsch and Saffold 1976; Dore et al.
1979; Owens and McBride 1969; Banker 1974). Dis-
seminated disease may occur in immunosuppressed
patients (Beyl et al. 1980; Owen and Toone 1970),
and in these patients periorificial skin infection may
develop as the organism may be present in nasopha-
ryngeal secretions (Ahn et al. 1976). These lesions
might progress slowly, be chronic and persistent,
or spontaneously regress (Tapeiner and Wolff 1999).
Histopathology
The histologic appearance is indistinguishable from
that of M. tuberculosis.
Differential and Differential Diagnosis
Confirmation of the diagnosis is by identification of
the organism by bacterial culture. PPD skin testing
is not helpful in the diagnosis. The differential diag-
noses include tuberculosis, sporotrichosis, and other
atypical mycobacterial infections.
Treatment
M. Kansasii is more susceptible to antituberculous
drugs than other atypical mycobacteria but not
as susceptible as M. tuberculosis. Care should be
taken to avoid the emergence of resistant strains,
and rifampicin has to be included in any regimen
used with at least the standard dosage and duration
recommended. Persistent isolates are susceptible
to slightly higher concentrations (Ahn et al. 1981;
Hobby et al. 1967; Pezzia et al. 1981). The currently
recommended regimen for pulmonary and extrapul-
monary M. kansasii infection is isoniazid (300 mg),
ethambutol (15 mg/kg), and rifampicin (600 mg)
daily for 18 months (Hautmann and Lotti 1994). In
disseminated cases, in addition to the three-drug reg-
imen, streptomycin 1 gm twice weekly can be added
for the first 3 months (Snider et al. 1987). Minocycline
200 mg daily has been used successfully in one case
(Dore et al. 1979).
36.6.4
Mycobacterium scrofulaceum Infection
Mycobacterium scrofulaceum infection commonly
affects the cervical lymph nodes in children and
resembles scrofuloderma. The organism is widely
distributed in the environment and has been isolated
651
from tap water and soil. Children probably acciden-
tally inhale or ingest the organism while playing.
Clinical Features
Children between the ages 1 and 3 years are primar-
ily affected. The submandibular and submaxillary
lymph nodes are targeted rather than the tonsillar
and anterior cervical lymph nodes typical of M.
tuberculosis infection. Constitutional symptoms are
usually lacking and mild neck pain is usually the
most common presenting symptom. The involve-
ment is often unilateral where the glands gradually
enlarge over many weeks then soften and ulcerate
draining their contents. Occasionally there are sporo-
trichoid lesions due to lymphatic spread and multiple
skin abscesses following hematogenous seeding of
the organism (Murray-Leisure et al. 1897).
Histopathology
The histologic features are similar to lymphadenitis
caused by M. tuberculosis with tubercle granuloma
formation.
Diagnosis and Differential Diagnosis
The clinical picture of unilateral lymphadenopathy
with a normal chest roentgenogram is suggestive of
the diagnosis. Confirmation may be achieved by bacte-
rial culture from tissue biopsy. Differential diagnosis
includes all infective and noninfective causes of cervi-
cal lymphadenopathy. Appropriate hematologic, sero-
logic, and histopathologic tests should be employed.
Treatment
Surgical excision is the treatment of choice of affected
lymph nodes as M. scrofulaceum is not very sensitive
to antituberculous-drug treatment. Multiple antitu-
berculous drugs should be given to patients with
Widespread disease until the sensitivity tests results
are available.
36.6.5
Mycobacterium avium-intracellulare Complex
Infection
These organisms may cause lung involvement in
patients with preexisting pulmonary disease, but
osteomyelitis and cervical lymphadenitis may also
occur. Disseminated infections are rare but their
652
incidence is rising sharply as this infection is one of
the most common causes of opportunistic infections
in AIDS patients (Collins 1989; Lerner and Tapper
1984). However, primary cutaneous infection is rare.
It is ubiquitous in the natural world, and more than
20 subtypes have been identified with immunologic
techniques. It has been isolated from normal human
feces in as many as 30% of tested cases.
Pathogenesis
M. avium-intracellulare complex is of low virulence
for the immune-competent host but causes dissemi-
nating disease in the immunocompromised patients.
It is transmitted through air into the lungs, or through
water and food into the gut. Trauma is needed for the
rare cases of primary cutaneous infection to occur.
Clinical Features
Post-traumatic inoculation of the skin is rare and
may cause purulent leg ulcers, folliculitis, ulcerated
nodules, panniculitis, abscesses, or even scaly non-
descript plaques (Cox and Stransbough 1981). More
commonly, the cutaneous involvement is secondary
to disseminated disease with clinical presentation
similar to the primary infection but more generalized
and severe. The survival time of AIDS patients who
are infected with M. avium-intracellulare complex is
about 4 months (Horsbugh et al. 1991).
Histopathology
Histologic features are those of noncaseating granu-
lomas. Acid-fast bacilli may be found in the specimen
and inside giant cells.
Diagnosis and Differential Diagnosis
Culture of the organism from various body fluids and
tissue specimens such as urine, sputum, skin, lymph
nodes,bone marrow, and liver is mandatory for the diag-
nosis. The differential diagnoses include other atypical
mycobacterial infections, deep fungal infections, and
other chronic granulomatous skin conditions.
Treatment
Purely cutaneous-limited lesions are best treated with
surgical resection whenever feasible because of poor
susceptibility to chemotherapeutic agents. Response
to treatment with tetracycline or minocycline has
been reported (Karinuma and Suzuki 1994; Noel et
A. J. AI-Kudwah
al. 1988). Antituberculous drugs are used in dissemi-
nated cases but are less effective because of 10-100
times less activity against M. avium-intracellulare
complex (compared with M. tuberculosis), larger
dosage and more drugs are used concomitantly but
with the risk of higher toxicity (Hautmann and Lotti
1994). Ciprofloxacin, clarithromycin, and azithromy-
cin have been reported to be effective (Mandell and
Sande 1990; Frank and Cabie 1995). Acquired drug
resistance develops 2-7 months after commencing
treatment, so a combination with other antituber-
culous drugs is recommended. Treatment is for life.
Prophylaxis in HIV-positive patients with rifampicin
300 mg daily is recommended (Gawkrodger 1998).
36.6.6
Mycobacterium fortuitum Complex Infection
This is the only rapid grower which has been found
to cause human disease in the form of subcutaneous
abscesses following trauma or surgery. M. fortuitum,
M. chelonae, and M. abscessus (which make up the
complex) are commonly found in water, milk, soil,
dust, marine life, fish, frogs, cows, and the saliva of
healthy humans.
Pathogenesis
Infection may occur following trauma or surgi-
cal intervention. Infection can be disseminated in
immunocompromised patients. The organism has
the ability to infect a wide variety of human tissue
(Hautmann and Lotti 1994). All subtypes share the
affinity for the skin and soft tissues, in addition M.
abscessus has an affinity for the lungs.
Clinical Feature
In the primary cutaneous infection, following trauma
by 1-2 months, a painful erythematous nodule forms,
which matures into an abscess that may rupture drain-
ing a clear fluid. The process may remain localized,
or in cases of compromised immunity, disseminated
cutaneous disease may result in either multiple recur-
ring abscesses on the extremities or generalized macu-
lopapular eruption, a sporotrichoid pattern of spread
has been noted. Constitutional symptoms are either
mild or absent. Postinjection abscesses are probably the
most common skin lesions caused either by M. chelonae
or M. fortuitum. It has followed injections of vitamins,
insulin, steroids, and inoculations (DaCosta Cruz 1938;
Inman et al. 1969; Lau 1986; Borghans and Stanford
Tuberculosis of the Skin
1973). The source of infection seems to be contaminated
injection solutions (Tapeiner and Wolff 1999). Surgical
procedures (e.g., mammoplasty, ophthalmic operations,
median sternotomy, or percutaneous catheterization),
or other procedures (e.g., hemodialysis and endoscopic
examinations) can be followed by the introduction of
the organism into the tissues and subsequent develop-
ment of infection. Awide range of clinical presentations
has been reported, including cellulitis, subcutaneous
abscesses, osteomyelitis, lymphadenitis, synovitis, men-
ingitis, mastoiditis, hepatitis, keratitis, prosthetic valve
endocarditis, and bacteremia, thus eliciting the affinity
of this organism to many body tissues. It is possible that
trauma and surgical procedures may produce localized
disease in the immune-competent patients. Impairment
of immunity due to diabetes, chronic renal failure, or
corticosteroid and other immunosuppressive agents
will promote dissemination either as a disseminated
cutaneous infection or as a systemic disease (Baack
and Brown 1991; Wallace et al. 1992). HIV infection
does not seem to impart an increased susceptibility to
this infection (Palenque 2000).
Histopathology
The histologic features are those of the simultaneous
occurrence of polymorphonuclear microabscesses
and granuloma formation with foreign-body-type
giant cells and are considered characteristic for this
infection. This is known as «dimorphic inflamma-
tory response" (Tapeiner and Wolff 1999). Caseation
is usually absent, but necrosis occasionally may be
present. Acid-fast bacilli can be seen in approximately
one third of the patients.
Diagnosis
An unusual cold abscess that is resistant to conven-
tional treatment should arouse suspicion of atypical
mycobacterial infection especially in the tropics.
These abscesses should be differentiated from for-
eign-body reactions and deep mycosis. Culture of
the organism from biopsy material serves a dual
purpose: it confirms the diagnosis and helps the
clinician make an informed decision on the choice
of which antimicrobials to use as different subspecies
have markedly different susceptibilities.
Treatment
Surgical debridement or excision of a localized disease
may be sufficient, but is probably better combined with
appropriate antibiotics (e.g., amikacin or doxycycline).
653
The M. fortuitum complex organisms are resistant to
most standard antituberculous drugs (Wallace 1989).
M. fortuitum is more susceptible to amikacin, doxycy-
cline, sulfonamides, cefoxitin, imipenem, or ciproflox-
acin. M. chelonae is usually sensitive to erythromycin,
cefoxitin, and tobramycin, while M. abscessus is usually
sensitive to amikacin, cefoxitin, and clarithromycin. In
systemic disease an initial empirical treatment with
the combination of intravenous amikacin and cefoxi-
tin is recommended until the result of an in vitro sen-
sitivities test is available. Oral treatment can then be
commenced (Palenque 2000). Triple therapy is usually
used with immunosuppressed patients.
36.6.7
Mycobacterium haemophilum Infection
The organism is a rare cause of disease with the
skin being most commonly involved. It requires
culture media supplemented with hemin or ferric
ammonium citrate, hence the name. Its prevalence is
increasing, and it has been identified as the causative
organism of subcutaneous granulomatous absc .sses
or ulcers in immunocompromised hosts.
Clinical Features
M. haemophilum infection is manifested by multiple,
tender, often violaceous cutaneous nodules situated
commonly over joints of the extremities. Facial and
chest lesions have been described, as well as joint
effusions, tenosynovitis, and weight loss.
Histopathology
Tissue specimens exhibit dimorphic inflammatory
response similar to M. fortuitum complex with no
caseation necrosis.
Diagnosis and Differential Diagnosis
Culture of the organism from tissue specimens with
particular attention to the temperature and media
requirement will confirm the diagnosis. The differential
diagnosis should include other atypical mycobacterial
infections, panniculitis, and foreign-body granulomas.
Treatment
Treatment is often disappointing as lesions may per-
sist despite therapy or relapse when medications are
stopped. The severity of immunosuppression often
654
influences the outcome (Kristjansson et al. 1991).
Sensitivity tests suggest that the organism may be
sensitive to rifampicin, ciprofloxacin, clarithromycin,
and p-aminosalicyclic acid (Tapeiner and Wolff 1999;
Darling et al. 1994).
A number of other mycobacteria have been
implicated in a diversity of cutaneous/subcutaneous
lesions such as nodules, abscesses, ulcers, panniculi-
tis, and draining sinuses usually in the immunocom-
promised host. They are primarily opportunistic pul-
monary pathogens with the skin secondarily infected
through hematogenous spread. These include M.
simiae, M. malmoense, M. smegmatis, M. gordonae,
M. thermoresistible, M. flavescens, and M. szulgai.
Most are resistant to antituberculous drugs, and cul-
tures from various specimens should aid in using the
appropriate chemotherapeutic agents.
36.7
Treatment of M. tuberculosis
Skin Infections
The treatment of cutaneous tuberculosis is not dif-
ferent from that of tuberculosis elsewhere, although
additional measures might need to be taken to opti-
mize patient care. Patients have to be fully evaluated
in search of an underlying tuberculous process.
Attention has to be given to the patient's general con-
dition as well as to any coexisting illnesses that might
adversely affect the ongoing tuberculous process: for
example, diabetes, immunosuppressive medications,
and superadded infections.
36.7.1
Dermatologic Considerations
Whenever an extracutaneous infection coexists with
a cutaneous infection, a full antituberculous regimen
must be followed. This will help avoid emergence of
multidrug-resistant organisms. Because of that danger,
the older, single-drug therapy with isoniazid, of lupus
vulgaris and tuberculosis verrucosa cutis is better
avoided, despite high cure rates. Surgical excision of
isolated small lesions of lupus vulgaris or tuberculosis
verrucosa cutis may be effective. Plastic surgery is
needed to correct disfigurement left by healed exten-
sive lupus vulgaris. Surgery in scrofuloderma is needed
to debride dead and infected tissue so as to reduce
morbidity and shorten the duration of chemotherapy
(Gawkrodger 1998; Tapeiner and Wolff 1999).
A. J. Al-Kudwah
Acknowledgements. I would like to thank my son M.
Monir Madkour for his diligent help with MedLine
searches and computer skills. Also I would like to
thank Amr, Rasha, and Reem Madkour for their help
and support, each one in his or her own way. Above
all, I would like to thank my husband for giving me
this opportunity.
References
Ahn CH et al (1976) Ventilatory defects in atypical mycobac-
teriosis. Am Rev Respir Dis 113:273
Ahn CH,Loweli JR,Ahn SA et al (1981) Chemotherapy for pul-
monary disease due to mycobacterium kansasii efficacies
of some individual drugs. Rev Infect Dis 3:1028-1034
Amezquila R (1963) Tuberculosis cutanea: aspectos clinicos y
epidemiologicos en Mexico. Acta Lepro116:1
Anonymous (1991) Epidermiologic notes and reports: Myco-
bacterium haemophilum infection, New York City Metro-
politan Area, 1990-1991. Arch Dermato1127:1628
Aungst CW et al (1975) Complications of BCG vaccinations in
neoplastic disease. Ann Intern Med 82:666
Baack BR, Brown RE (1991) Atypical mycobacterium soft-
tissue infection on the dorsal radial wrist: a possible com-
plication of steroid injection for de Quervain's disease. Ann
Plast Surg 27:73-76
Bafverstedt B (1954) Erythema nodosum migrans. Acta Derm
Venereol (Stockh) 34:181
Banerjee BN (1956) Common clinical manifestations of tuber-
culosis of the skin in the tropics. Ind J DermatoI2:7-13
Banker DJP (1974) Mycobacterial skin ulcers. Br J Dermatol
91:473-474
Beurey J, Weber M, Vignaud JM et al (1981) Mycobacterioses
cutanees: enquete epidemiologique. Ann Dermatol Vene-
reol 108:439-442
Beyl BE et al (1980) Cutaneous mycobacteriosis: analysis of
34 cases with a new classification of the disease. Medicine
(Baltimore) 60:95
Borghans JGA, Stanford JL (1973) M. Chelonei in abscesses
after injection of diphtheria-pertussis tetanus-polio vac-
cine. Am Rev Respir Dis 107:1-8
Brown JW, Sanders CV (1987) Mycobacterium marinum infec-
tions: a problem of recognition not therapy (editorial)?
Arch Intern Med 147:817-818
Bryant JC (1939) Oral tuberculosis. Am Rev Tubercle 39:738
Chien JTT, Wiggins ML (1954) Self-inoculation with M. tuber-
culosis and P. aeruginosa by a diabetic woman. Am Rev
Tuberc 69:818
Citron KM, Girling DJ (1987) Tuberculosis. In: Weatherhall
DJ, Ledinghan JGG, Warrell DA (eds) Oxford textbook of
medicine, vol I, 2nd edn. ELBS/Oxford University Press,
Oxford, 5.278-5.299
Clancey JK et al (1961) Mycobacterial skin ulcers in Uganda.
Lancet 2:951
Clasen K, Horwitz 0 (1960) The morbididty for pUlmonary
tuberculosis in patients suffering from extrapulmonary
tuberculosis disease, lupus vulgaris cutis. Adv Tuberc Res
10:237
Collins FM (1989) Mycobacterial disease, immunosuppression
Tuberculosis of the Skin
and acquired immunodeficiency syndrome. Clin Microbiol
Rev 2:360-377
Cox SK, Stransbough LJ (1981) Chronic cutaneous infection
caused by mycobacterium intercellulare. Arch Dermatol
117:794
Curtis HM et al (1984) Incidence of childhood tuberculosis
after neonatal BCG vaccination. Lancet 1:145
DaCosta Cruz J (1938) "Mycobcterium fortuitum" un novo
bacilo acido resistente patogenico par 0 homem. Acta Med
Rio de Janeiro 1:297-301
Darier MJ (1896) Les "tuberculids" cutanees. Ann Der
Syphiligr (Paris) 7:1431-1436
Darling TN, Sidhu-Malik N, Corey GR et al (1994) Treatment
of Mycobacterium haemophilum infections with an anti-
biotic regimen including clarithromycin. Br J Dermatol
131 :379-390
De Bruyne JI, van Creveld S, Prakken JR (1953) Papular
tuberculids after BCG vaccination. Acta Derma Venereol
(Stockh) 33:385-390
Degitz K, Messer G et al (1993) Successful treatment of ery-
thema induratum of Bazin following rapid detection of
mycobacterial DNA by polymerase chain reaction. Arch
DermatoI129:1619-1620
Dore N, Collins JP, Mankiewicz E (1979) A sporotrichoid-like
M. kansasii infection of the skin treated with minocycline
hydrochloride. Br J Dermatoll0l:75-79
Dostrovsky A, Sagher F (1963) Dermatological complications
of BCG vaccination. Br J Dermatol 75:181
Duken J (1933) Ober Verlaufsarten der extrapulmonalen Pri-
martuberkulose. Z Kinderheilkd 55:687
Eberhartinger C (1963) Das Problem des Erythema induratum
Bazin. Ein Beitrag zur Kenntnis der rezidivierenden suba-
kuten nodosen Gefassprozesse am Unterschenkel. Arch
Klin Exp Dermato1217:196
Ellis ME, Pope J, Mokashi A et al (1982) Compylobacter colitis
associated with erythema nodosum. Br Med J 285:937-940
Feldman RA (1974) Primary mycibacterial skin infection: a
summary. Int J Dermatol 13:353
Findlay GH (1980) Mycobacterial infections other than leprosy.
In: Rook A, Savin JA (eds) Recent advances in dermatology,
5th edn. Churchill Linvingstone, Edinburgh, pp 59-81
Fisher JR (1977) Miliary tuberculosis with unusual cutaneous
manifestations. JAMA 238:241-242
Forstrom L, Hannuksela M (1970) Anti-tuberculous treatment
of erythema induratum of Bazin. Acta Derm Venereol
(Stockh) 50:143-147
Franck N, Cabie A (1995) The reemergence of tuberculosis
and atypical abscesses due to mycobacterial diseases. Curr
Opin Dermatol 249-253
Gawkrodger DJ (1998) Mycobacterial infections. In: Rook,
Wilkinson, Ebling (eds) Textbook of Dermatology,vol 28,
6th edn. Blackwell Science, Oxford, pp 1181-1214
Goh YS, Ong BH, Rajan VS (1974) Tuberculosis cutis in Singa-
pore: a two year experience. Sing Med J 15:223-226
Goksoy E, Duren M, Durgun V, Uygun N (1995) Tuberculosis
of the breast. Eur J Surg 161:471-473
Goldschmidt H, Grekin RC (1990) Mycobacterial diseases. In:
Andrew's diseases of the skin. W. B. Saunders Company,
Philadelphia, London, Toronto, Montreal, Sydney, Tokyo.
Clinical dermatology, chap 16, 8th edn. pp 375-388
Gombert MD et al (1981) Disseminated mycobacterium mari-
num infection after renal transplantation. Ann Intern Med
94:435-486
655
Grange JM (1980) Mycobacterial disease. Arnold, London, pp
80-82
Grange JM (1988) Lupus and lepros. Int J Dermatol 59:
111-112
Harrison PV, Marks JM (1980) Lupus vulgaris and cutaneous
lymphoma. Clin Exp DermatoI5:73-77
Hautmann G, Lotti T (1994) Atypical mycobacterial infections
of the skin. Dermatol Clin 657-668
Hekel K, Seyss R (1951) Die malignen tumoren in lupo vul-
garis. Hautarzt 2:349
Higuchi M, Suga M, Dannenberg AM et al (1981) Persistence
of protein, carbohydrate and wax components of tubercle
bacilli in dermal BCG lesions. Am Rev Respir Dis 123:
397-401
Hirsch FS, Saffold OE (1976) M Kansassii infection with der-
matologic manifestations. Arch Dermatol 112:706-708
Hirsh BC, Johnson WC (1984) Pathology of granulomatous
diseases. Int J DermatoI23:237-246
Hobby GL, Redmond WB, Runyon EH et al (1967) A study of
pulmonary disease associated with mycobacteria other
than M. tuberculosis; identification and characterization
of the Mycobacteria. Am Rev Respir Dis 95:954-971
Holt LE (1913) Tuberculosis acquired through ritual circumci-
sion. JAMA 61:99
Horio T, Danno K, Okamoto H et al (1983) Pottassium iodide
in erythema nodosum and other erythematous dermato-
ses. J Am Acad DermatoI9:77-81
Horsburgh C,Harlick J, Ellis E (1991) Survival of patients with
acquired immune deficiency syndrome and disseminated
mycobacterium avium complex infection with and without
anti-mycobacterial chemotherapy. Am Rev Respir Dis 144:
557-559
Horwitz 0 (1959) The localization of lupus vulgaris of the
skin. Acta Tuberc Scand 47:175
Horwitz 0 (1960) Lupus vulgaris cutis in Denmark 1895-1954:
its relation to epidemiology of other forms of tuberculosis.
Acta Tuberc Scand 49: 1
Horwitz 0, Christensen S (1960) Numerical estimates of the
extent of the lesion of lupus vulgaris cutis and their sig-
nificance for epidemiologic and clinical research. Am Rev
Respir Dis 82:862-872
Horwitz 0, Meyer J (1957) The safety record of BCG vaccina-
tion and untoward reactions observed after vaccination.
Adv Tuberc Res 8:245
Huminer D, Pitlik SD, Block C et al (1986) Aquarium borne M.
marinum skin infection. Arch DermatoI122:698-703
Ihin CW, Choi CK, Suh JI (1983) LUpus tumidus involving
facial skin, nasal cavity, throat and eye. Dermatologica
166:38-39
Inman PM, Beck A, Brown AE et al (1969) Outbreak of infec-
tion abscesses due to M. abscessus. Arch Dermatol 100:
141-147
Izumi AK, Matsunaga J (1982) BCG vaccine-induced lupus
vulgaris. Arch Dermatol 118: 171
Jarrett P, Goodfield MJD (1996) Hydroxychloroquine and
chronic erythema nodosum (letter). Br J Dermatol134:373
Jordaan HF, van Niekerk DJ et al (1994) Papulonecrotic tuber-
culid. A clinical histopathological and immunohistochemi-
cal study of 15 patients. Am J DermatopathoI16:474-485
Jorgensen BB, Horwitz 0 (1956) Dermatological complications
of BCG vaccination. Acta Tuberc Scan 32:179
Kakkar S, Kapila K, Singh MK, Verma K (2000) Tuberculosis of
the breast: a cytomorphologic study. Acta eytol 44:292-296
656
Kalkoff KW (1950) Die Tuberculose der Haut. Thieme, Stutt-
gart
Kanan MW, Ryan TJ (1975) Endonasallocalization of blood
borne viable and non-viable particulate matter. Br J Der-
matol 92:475-478
Kanan MW, Ryan TJ (1976) The localization of granulomatous
diseases and vasculitis in the nasal mucosa. In: Ryan TJ (ed)
Microvascular injury. Saunders, London, pp 195-220
Karinuma H, Suzuki H (1994) Mycobacterium avium complex
infection limited to the skin in a patient with systemic
lupus erythematosus. Br J Dermatol130:785-790
Katz HL (1941) Tuberculosis of the tongue. Q Bull Seaview
Hosp 6:239
Khanna R, Prasanna GV, Gupta P, Kumar M, Khanna S, Khanna
AK (2002) Mammary tuberculosis: report on 52 cases. Post-
grad Med J 78:422-424
Kounis NG, Constantinidis K (1979) Unusual tuberculous skin
manifestations. Practitioner 222:390-393
Kowalenko W (1963) Ober ein lichenoides Exanthem unklarer
Pathogenese (zur Kenntnis des "Lichenoid Tuberculid" von
Montgomery). Dermatol Wochenschr 147:13
Krieg RE et al (1975) The treatment of mycobacterium
ulcerans infection by hyperbaric oxygenation. Aviat Space
Environ Med 46:1241
Kristjansson M, Bieluch VM, Byeff PD (1991) Mycobacterium
haemophilum infection in immunocompromised patients
case report and review of the literature. Rev Infect Dis 13:
906-910
Kumanu K, Tani M, Murata Y (1983) Dapsove in the treatment
of miliary lupus of the face. Br J Dermatoll09:57-62
Kumar B, Sharma VK (1987) Papulonecrotic tuberculides on
glans penis. Dermatologica 174:151-152
Kuramato Y, Aiba S, Tagami H (1990) Erythema induratum
of Bazin as a type of tuberculid. J Am Acad Dermatol 22:
612-616
Laennec RTH (1962) Traite de L'auscultation mediate et des
maladies des peumons et du coevr, vol 1. Asselin and Cie,
Paris, 1826, pp 649-650. Quoted in Marmelzat WL. Laennec
and the «prosector's wart». Arch DermatoI86:122-124
Lane HC, Fauci AS (1985) Immunologic aspects of acquired
immuno-deficiency syndrome. Adv Host Def Mech 5:
131-148
Lau JH (1986) Hand infection with M. chelonei. Br Med J 292:
444-445
Lerner CW, Tapper MC (1984) Opportunistic infection com-
plicating acquired immunodeficiency syndrome. Medicine
63:155-164
Lewandowsky F (1917) Ober rosacea-ahnliche Tuberkulide
des Gesichtes. Korresp Bl Schweiz Arz 47:1280
Lichtenstein MR, Takimura Y, Thompson JR (1965) Photochro-
mogenic mycobacterial pulmonary infection in a group of
hospitalized patients in Chicago. Am Rev Respir Dis 91:
592-595
Linell F, Norden A (1954) Mycobacterium balnei: a new acid-
fast bacillus occurring in swimming pools and capable of
producing skin lesions in humans. Acta Tuberc Scand 33:
1 Pereira CA, Webber B, Orson JM (1976) Primary tuberculous
MacCullum P et al (1948): New mycobacterial infection in
man: clinical aspects. J Pathol Bacteriol 60:93
Mandell GL, Sande MA (1990) Antimicrobial agents. Drugs
used in the chemotherapy of tuberculosis and leprosy. In:
Goodman, Gilman et al (eds) The pharmacological basis of
therapeutics, 8th edn. Pergamon, New York, p 1146
A. J. Al-Kudwah
Marcoval J, Servitje 0, Moreno A et al (1992) Lupus vulgaris:
clinical, histopathologic and bacteriologic study of 10
cases. J Am Acad Dermatol 26:404-407
Marmelzat WL (1962) Laennec and the prosector's wart. Arch
DermatoI86:74-76
McCray MK, Esterly NB (1981) Cutaneous eruptions in con-
genital tuberculosis. Arch Dermatol 117:460
Meyers WM et al (1974) Heat treatment of mycobacterium
ulceran infections without surgical excision. Am J Trop
Med Hyg 23:924
Miescher G (1955) Uber katamnestische Untersuchungen bei
Fallen mit Tuberkulid. Dermatologica 110:23
Miller FJW, Cashman JM (1955) The natural history of periph-
eral tuberculosis lymphadenitis associated with a visible
primary focus. Lancet 1:1286
Mitchell PG (1954) Tuberculosis verrucosa cutis among Chi-
nese in Hong Kong. Br J Dermatol 66:444-448
Montgomery H (1937) Histopathology of various types of
cutaneous tuberculosis. Arch Dermatol Syphilol 35:698
Montgomery H et al (1945) Nodular vascular diseases of the
legs. JAMA 128:335
Morrison JGL, Fourie ED (1974) The papulonecrotic tubercu-
lid from Arthus reaction to lupus vulgaris. Br J Dermatol
91:263-270
Moschella SL (1985) Mycobacterial infections. In: Moschella
SL, Hurley HJ (eds) Dermatology, 2nd edn. Saunders,
London, pp 921-946
Murray-Leisure KA, Egan N et al (1897) Skin lesion caused
by mycobacterium scrofulaceum. Arch Dermatol 123:
369-370
Nepomuceno DR (1971) Tuberculosis of the anal canal: report
of a case. Dis Colon Rectum 14:313-316
Nishigori C, Taniguchi S, Hayakawa M (1986) Penis tuber-
culosis papulo-necrotic tuberculides on the glans penis.
Dermatologica 17:93-97
Noel SB, Ray MC, Greer DL (1988) Cutaneous infection with
M. avium intracellulare scrofulaceum intermediate: a new
pathogenic entity. J Am Acad DermatoI19:492-495
O'Driscoll T, Morgan G (1974) Acne agminata of the eye lid.
Proc Roy Soc Med 67:869
Ockuly OE, Montgomery H (1950) Lichenoid tuberculid: a
clinical and histopathologic study. J Invest Dermatol 14:
415
Oller MW, Thomas P, Korting HC et al (1993) Erythema indu-
ratum of Bazin: evidence of T-lymphocyte hyper-respon-
siveness to purified protein derivative of tuberculin. Arch
DermatoI127:496-473
Owen DS, Toone E (1970) Soft tissue infection by group I
atypical mycobacteria. South Med J 63: 116
Owens DW, McBride ME (1969) Sporotrichoid cutaneous
infection with M. kansasii. Arch Dermatoll00:54-58
Palenque E (2000) Skin disease and nontuberculous atypical
mycobacteria. Int J Dermatol 39:659-666
Pandhi RK, Bedi TR, Kanwar AJ et al (1977) Cutaneous tuber-
culosis: a clinical and investigative study. Ind Dermatol 22:
99-107
complex of the skin. JAMA 235:942
Pezzia W, Raleigh JW, Bailey MC et al (1981) Treatment of
pulmonary disease due to mycobacterium kansasii: recent
experience with rifampin. Rev Infect Dis 3:1035-1039
Pinkus H, Mehregan AH (1981) Guide to dermato histopathol-
ogy, 3rd edn. Appleton-Century-Croftis, New York, p 231
Tuberculosis of the Skin
Platou RV, Lennox RH (1956) Tuberculous cutaneous com-
plexes in children. Am Rev Tuberc 74:160-172
Rauschkolb JW (1934) Tuberculosis of the skin. Arch Dermatol
Syphilol 29:398
Revill WD et al (1973) A controlled trial of the treatment of
mycobacterium ulceras infection with clofaxzimine. Lancet
2:873
Rietbroek RC, Dahlmans RP, Smedts F et al (1991) Tuberculo-
sis cutis miliaris disseminata as a manifestation of miliary
tuberculosis: literature review and report of a case of recur-
rent skin lesions. Rev Infect Dis 13:265-269
Riska N, Selnoos 0 (1964) Clinically diagnosed sarcoidosis in
Finland in 1960. Acta Tuberc Scand 44:267-275
Ross PW, Christy MK, Knox JDE (1971) Sore throat in children:
its causation and incidence. Br Med J 624-626
Rostas A (1980) Erythema nodosum migraus in a young
woman. Arch Dermatol116:325
Rouillon A, Waaler H (1976) BCG vaccination and epidemio-
logical situation. A decision making approach to the use of
BCG. Adv Tuberc Res 19:65
Runyon EH (1965) Pathogenic mycobacteria. Adv Tuberc Res
14:235
Runyon EH (1974) Ten mycobacterial pathogens. Tubercle 55:
235-248
Sahn SA, Neff TA (1974) Miliary tuberculosis. Am J Med 56:
459-505
Schermer DR, Simpson CG, Haserick JR et al (1969) Tuber-
culosis cutis miliaris acute generalisata. Arch Dermatol
99:64-69
Schuhmachers R (1967) 2 Falle eines lichenoiden Tuberculids
(Lichen scrophulosorum). Hautarzt 18:81
Schulz EJ, Whiting DA (1976) Treatment of erythema nodo-
sum and nodular vasculitis with potassium iodide. Br J
DermatoI94:75-78
Sehgal VN (1994) Cutaneous tuberculosis. Contemp Trop
DermatoI12:645-668
Sehgal VN, Wagh SA (1990a) Cutaneous tuberculosis. Current
concepts. Int J DermatoI29:237-251
Sehgal VN, Wagh SA (1990b) The history of cutaneous tuber-
culosis. Int J Dermatol 29:666-668
Sehgal VN, Srivastava G, Khurana VK et al (1987) An appraisal
of epidemiologic, clinical bacteriologic, histopathologic
and immunologic parameters in cutanneous tuberculosis.
Int J DermatoI26:521-526
Simon VN (1959) Die Rolle des peripheren Lymphknoten in
der Pathogenese der Tuberkulide. Hautarzt 10:303-309
Singh G (1974) Lupus vulgaris in India. Ind J Dermatol Vene-
reoI40:257-260
Snider DE, Hopewell PC, Millis Jet al (1987) Mycobacteriosis
and the acquired immunodeficiency syndrome . Am Rev
Respir Dis 136:492-496
Standford JL, Philipps I (1972) Rifampicin in experimental
657
mycobacterium ulcerous infection. J Med Microbiol 5:
39
Sundt H (1925) A case of lupus disseminatus (post-exanthe-
matic miliary tuberculosis cutis). Br J Dermatol37:316
Tapeiner G, Wolff K (1999) Tuberculosis and other myco-
bacterial infections. Fitzpatrick's Dermatol Gen Med 191:
2370-2394
Tarantola D, Mann JM (1987) Acquired immunodeficiency
syndrome (AIDS) and expanded programmes on immu-
nization. Special programme on AIDS. WHO, Geneva
Ueki H, Masuda T (1979) Lupus miliaris disseminatus faciei.
Hautarzt 30:353-355
Ustvedt HJ, Ostensen IW (1951) The relation between tubercu-
losis of the skin and primary-infection. Tubercle 32:36
Van der Werf TS, van der Graaf WTA (1990) Buruli ulcer in
West Africa. Lancet 336:1440
Victor T, Jandann HF, VanNiekerk DJT et al (1992) Papulo-
nerotic tuberculid: identification of mycobacterial tuber-
culosis DNA by polymerase chain reaction. Am J Derma-
topathoI14:491-492
Vilanova X, Pinol-Aguade J (1959) Subacute nodular migra-
tory panniculitis. Br J Dermatol 71:45
Waaler H, Rouillon A (1974) BCG vaccination policies accord-
ing to the epidermiological situation. Bull Int Union Tuberc
29:166
Wall LM, Smith NP (1981) Perniosis: a histopathological
review. Clin Exp DermatoI6:263-271
Wallace RJ (1989) The clinical presentation, diagnosis and
therapy of cutaneous and pulmonary infections due to
the rapidly growing mycobacteria, M. fortuitum and M.
chelonai. Clin Chest Med 10:419-429
Wallace RJ Jr, Brown BA, Onyi GO (1992) Skin, soft tissue and
bone infections due to mycobacterium chelonei: impor-
tance of prior corticosteroid therapy, frequency of dis-
seminated infections and resistance to oral antimicrobials
other than clarithromycin. J Infect Dis 166:405-412
Warin AP, Wilson-Jones E (1977) Cutaneous tuberculosis of
the nose with unusual clinical and histological features
leading to delay in the diagnosis. Clin Exp Dermatol 2:
235-242
Watson JM, Gill ON (1990) HIV infection and tuberculosis. Br
Med J 300:63-65
Wolff E (1921) Ober Zirkumzisionsuberkulose. Berlin Klin
Wochenschr 58:1531
Wolinsky E et al (1972) Sporotrichoid M marinum infection
treated with rifampin-ethambutol. Am Rev Respir Dis 105:
964
Wong Ko, Lee KP, Chin SF (1968) Tuberculosis of the skin in
Hong Kong. Br J Dermtol 80:424-429
Yeager H Jr (1985) Other mycobacterium species. In: Mendell
GL et al (eds) Principles and practice of infectious diseases,
chap 207, 2nd edn. Wiley, New York
37 Abdominal Tuberculosis
MOHAMMAD SULTAN KHUROO and NAIRA SULTAN KHUROO

CONTENTS small intestinal obstruction and stricture. However,

by the middle of the century all forms of tuberculosis
37.1 Epidemiology 659
had declined dramatically. This decline was caused by
37.2 Tuberculosis of Small Bowel and Colon 661
37.3 Tuberculous Peritonitis 666
a number of factors, which included an increased stan-
37.4 Tuberculosis of Mesenteric Lymph Nodes 670 dard of living, pasteurization of milk, control of bovine
37.5 Tuberculosis of Solid Abdominal Organs 671 tuberculosis. and introduction of antituberculous treat-
37.6 Tuberculosis of Other Gastrointestinal Sites 672 ment (O'Reilly and Daborn 1995). In fact, frequency of
37.7 Treatment 674 abdominal tuberculosis in the United States in 1960s and
References 675
1970s dropped to such low levels that the disease was
classified as a "rare" or Third World disease. However,
since 1985 the number of reported cases of abdominal
Tuberculosis remains one of the major health tuberculosis has dramatically increased. This was due
problems in the world. WHO estimates that each to two reasons: (1) an increased incidence of all cases
year 8 million new cases of tuberculosis occur and of tuberculosis (Brudney and Dobkin 1991; Cantwell
approximately 3 million people die from the disease et al. 1994) and (2) an increased proportion of extra-
(WHO 1996). Tuberculosis is a disease of develop- pulmonary disease, especially abdominal tuberculosis
ing countries; however, its incidence is increasing in (Farer et al.1979; Alvarez and McCabe 1984). From 1980
developed countries as well, mainly in the immigrant onward, reported cases of tuberculosis in the United
population and in patients with AIDS (McKenna et States increased. The majority of these cases were in
al. 1995; Barnes et al. 1991). Hispanics, blacks, prisoners, immigrants, refugees, and
nursing home patients (McKenna et al. 1995; Cantwell
et al. 1994; Nardell et al. 1986; Raviglione and O'Brien
2001; Bradney and Dobkin 1991). Multidrug-resistant
37.1 tuberculosis in AIDS patients contributed significantly
Epidemiology to this increase in the occurrence of the disease (Edlin
et al. 1994; Bloch et al. 1994; Gordin et al. 1996; Frieder et
Abdominal tuberculosis is still prevalent in developing al.1993; Selwyn et al.1989; CDC 1990, CDC 1991; Small
countries (Tandon and Prakash 1972; Bhansali 1977; et al.1993; Anand 1956). The impact of the disease was
Kapoor 1998). There is confusion on the exact inci- seen particularly in urban areas. In 1979, there were
dence of abdominal tuberculosis in such countries due 1,530 new cases of tuberculosis in New York City, and
to problems of actual reporting, difficulty in diagnosis, by 1991 the city had 3,673 new cases of tuberculosis, a
and inability to separate tuberculosis from Crohn's yearly increase that is three times the national average.
disease, which can closely resemble it in its clinical The number of cases continued to increase and peaked
manifestations. Abdominal tuberculosis was common in 1992. As a result of aggressive health care control
in the United States early in the 20th century (Horvath policies, the number of cases has shown a gradual
and Whelan 1998). It was the cause of most cases of downward trend.
Another reason for high occurrence of abdominal
M. S. KHURoo, MD, DM, FRCP (Edin), MACP tuberculosis was high proportion of extrapulmonary
Professor, Consultant Gastroenterologist, Department of Med- disease (Farer et al. 1979; Alvarez and McCabe 1984).
icine - MBC 46, King Faisal Specialist Hospital and Research
In 1960s only 8% of patients with tuberculosis had
Centre, P.O. Box 3354, Riyadh 11211, Saudi Arabia
N. S. KHuRoo, MBBS extrapulmonary manifestations. By 1986, extrapul-
Department of Radiology, King Faisal Specialist Hospital and monary disease constituted 25% of all cases of tuber-
Research Centre, P.O. Box 3354, Riyadh 11211, Saudi Arabia culosis. The lung is the commonest site (over 85%) of

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
660
involvement in immunocompetent persons, while the
disease predominantly affects extrapulmonary sites
(over 50%) in patients with AIDS. As the urban epi-
demic of tuberculosis in the United States occurred
in AIDS patients, abdominal tuberculosis revealed a
significant resurgence.
Mycobacterium tuberculosis is the pathogen for
most cases of abdominal tuberculosis. Mycobacte-
rium bovis, an organism transmitted by unpasteur-
ized diary products, is the cause of a small percent-
age of cases in developing countries (Raviglione and
O'Brien 2001; Marshall 1993).
The route of infection occurs by one of the follow-
ing mechanisms (Kapoor 1998; Horvath and Whelan
1998; Marshall 1993):
1. Swallowing of infected sputum. This occurs
in patients with sputum-positive pulmonary
disease and those with laryngeal involvement.
This was the most important route of infection
before the era of effective treatment. Autopsies in
patients with pulmonary tuberculosis in the pre-
treatment era demonstrated intestinal disease in
55% to 99%. The frequency of intestinal disease
was related to the severity of pulmonary involve-
ment: 1% of patients with minimal pulmonary
tuberculosis, 4.5% with moderately advanced
disease, and 25% with far-advanced disease. In
modern series, this mode of infection is less
important and chest radiograph is completely
normal in the majority of patients with intestinal
tuberculosis.
2. Hematogenous spread from active pulmonary,
miliary tuberculosis or silent bacteremia during
the primary phase of tuberculosis. Most cases
of abdominal tuberculosis occur as a result of
reactivation of a latent focus in the small bowel
or peritoneum. This focus is established perhaps
previously, because of hematogenous spread from
a primary focus in the lungs that subsequently
healed completely (as they usually do without
leaving any radiologic evidence of a lung lesion).
Less commonly, hematogenous spread can occur
from active pulmonary focus tuberculosis. Hepa-
tosplenic tuberculosis almost always follows mili-
ary seeding and is a manifestation of dissemina-
tion throughout the body.
3. Ingestion of contaminated milk or milk products.
This mode of infection had been a common cause
of spread of bovine tuberculosis in the past. How-
ever, pasteurizing and/or boiling milk has con-
trolled this mode of transmission (O'Reilly and
Daborn 1995). At present, Mycobacterium bovis is
involved in a small percentage of intestinal disease
M. S. Khuroo and N. S. Khuroo
in developing countries, and this form of disease
is rare in the West (Anand 1956).
4. Contiguous spread from adjacent organs. Occa-
sional cases of abdominal tuberculosis are related
to contiguous spread from tuberculous lesions of
adjacent organs. Peritoneal spread can occur from
lesions in the fallopian tubes and intestines. Recent
data showed that this is an infrequent mechanism
in most patients with abdominal tuberculosis.
More often, lymph node lesions spread the infec-
tion to the bowel wall or pancreas.
5. Tuberculosis in patients with AIDS. Tuberculosis
occurs with increased frequency in AIDS patients
as the CD4 count drops below 400 cells per III
(Jones et aI.1993).An autopsy study in West Africa
found that 50% of adults dying of AIDS had active
tuberculosis and in 85% of them the liver was
involved. In fact tuberculosis is the most common
specific hepatic HIV-associated lesion in such
patients (Lucas 1994). The pathology of tubercu-
losis varies with the immune status of the patient
(Bhargava et al. 1984; Edwards and Kirkpatrick
1986). In patients with intact immune systems,
granulomas with Langhans giant cells and case-
ation or non-giant cell epitheloid granulomas are
usually seen. In patients with extreme immune
deficiency as commonly seen in terminal AIDS
patients, the histologic pattern is that of nonre-
active tuberculosis. Foci of granular necrosis are
surrounded by degenerate swollen macrophages,
and a large number of acid-fast bacilli are seen.
An analysis using restriction-fragment-Iength
polymorphisms to study the mode of infection of
tuberculosis in patients with AIDS has shown that
the disease is readily spread from index patients
and progresses rapidly to active disease. There
was no evidence that disease occurs from reacti-
vation of a latent focus (Daley et al. 1992; Small et
al. 1994).
6. Liver disease and tuberculous peritonitis. Patients
with cirrhosis of the liver with ascites have a higher
chance (around 10%) of concomitant tuberculous
infection (Aguado et aI.1990). In the United States,
half of the patients with tuberculous peritonitis
have underlying alcoholic cirrhosis as a cause of
ascites formation (WHO 1990; Raviglione and
O'Brien 2001; Lucas 1994). The mechanism of
this infection in patients with liver disease is not
known. It may be due to reactivation of a latent
tuberculous focus in the peritoneum facilitated by
lower immunity and coexistent ascites.
7. Tuberculous peritonitis in patients undergoing
long-term or continuous ambulatory peritoneal
Abdominal Tuberculosis
dialysis (CAPD). Tuberculous peritonitis has been subject. Abdominal tuberculosis may affect the gas-
reported as a complication of CAPD (Holley and
Piraino 1990; Cheng et al.1989; Lui et al.1996; Lam
et al. 2000). Talwani and Horvath (2000) reviewed
the English-language literature and found 51
reported cases of CAPD-associated tuberculous
peritonitis and added a 52nd case from their
own experience (Lui et al. 1996). Defects in local
immunity unique to CAPD may predispose to
active tuberculosis in such patients. Removal of
the CAPD catheter is not considered necessary for
cure of the infection.
Abdominal tuberculosis can occur at any age and
is equally prevalent in males and females (Bhansali
1977; Kapoor 1998; Horvath and Whelan 1998). The
majority of patients have symptoms present for
1 month to 1 year; however, around 20% of patients
have symptoms for 1 month or less at the time of
presentation. Low-grade fever, night sweats, anorexia,
weight loss, general lassitude, and weakness occur in
around two thirds of patients. Symptoms of disease
at other sites occur in patients with active disease in
extraabdominal organs. This is of particular signifi-
cance in patients with active pulmonary tuberculosis
or disseminated tuberculosis. Laboratory results
reveal mild normocytic or microcytic anemia and
normal white blood cell count (Pouchot et al. 1997).
PPD is positive in most of the patients; however, it
may be negative in immunosuppressed and malnour-
ished patients (Bass et al. 1985; Huebner et al. 1993;
American Thoracic Society 1981; Markowitz et al.
1993). Chest X-rays show active disease in about one
fifth of patients.
The evolution ofthe disease in a patient with abdom-
inal tuberculosis depends upon route of infection, site
of involvement, and underlying immune status of the
gastrointe tinal
II
35%
.•j ....."...
u·'"
lIroccal
7%
Lymph node
Fig.37.1. Sites of organ involvement in abdominal tuber-
culosis. The data are based on 596 patients with abdominal
tuberculosis
661
trointestinal tract, peritoneum, lymph nodes, liver and
spleen and pancreas singly or in combination.Abdom-
inal tuberculosis in immunosuppressed patients poses
special problems as disease has distinct bacteriologic
and clinical characteristics. Gastrointestinal tubercu-
losis affects, in order of frequency, the ileocecal region,
jejunum/ileum, colon, anorectum, stomach, appendix,
duodenum, and esophagus (Marshall 1993). Reported
sites of involvement of abdominal tuberculosis are
shown in Fig. 37.1.
37.2
Tuberculosis of Small Bowel and Colon
Pathogenesis (randon and Prakash 1972; Anand
1956). After the tubercle bacilli enter the gastroin-
testinal tract, they traverse the mucosa to lodge in
the submucosa. There, the presence of the bacilli
induces inflammatory changes, including serosal
and subserosal edema, cellular infiltrate, and lym-
phatic hyperplasia. Eventually, the appearance of
granulomata causes small papillary mucosal eleva-
tions. Lymphangitis, endarteritis, and fibrosis ensue,
which lead to mucosal ulceration, caseating necro-
sis, and narrowing of the intestinal lumen. Mucosal
ulceration may occur as a result of endarteritis of
submucosal vessels. Infection can spread to mesen-
teric lymph nodes.
As mentioned earlier, the most common site of
involvement is the ileocecal region. The affinity of
the bacilli for this site may be due to its relative stasis
and abundant lymphoid tissue. The macroscopic
appearances of intestinal lesions can follow one of
the below-mentioned patterns. Such lesions are usu-
ally segmental, and multiple sites of involvement are
common. Rarely, diffuse colonic involvement may
% simulate ulcerative colitis and Crohn's disease. Other
100
n characteristics include increased mesenteric fat and
7S mesenteric lymphadenopathy, which can cause trac-
tion diverticula with narrowing, local fixation, and
so sinus tract development.
a. An ulcerative lesion is characterized by multiple
2~
superficial ulcers. Ulcers are circumferential and
usually surrounded by inflamed mucosa. This is
the most common lesion, occurring in around
60% of such patients and is associated with a
virulent clinical course.
b. A hypertrophic lesion is characterized by scar-
ring, fibrosis, and pseudotumor formation. This
is seen in around 10% of such patients.
662 M. S. Khuroo and N. S. Khuroo

c. An ulcerohypertrophic lesion is characterized by
an inflammatory mass with thickened and ulcer-
ated mucosa. The lesion is most commonly seen in
the ileocecal region. It causes cone-shaped defor-
mity of the cecum, shortening of the ascending
colon, and thickening of the ileocecal valve, where
a wide gape is created. Overall this lesion is seen
in 30% of such patients.
d. Fibrous stricture occurs in some patients as a
result of healed ulceration causing luminal nar-
rowing and gut obstruction. In some cases this
occurs after effective antituberculous therapy.
Luminal narrowing may also occur due to extrain-
testinallymph node involvement without intrinsic
intestinal lesions.
Clinical Manifestations. Abdominal symptoms depend
upon the site of involvement of disease, pattern of
pathologic changes, and underlying immunologic
status of the host (Bhansali 1977; Kapoor 1998;
Horvath and Whelan 1998; Marshall 1993). Involve-
ment of small bowel and colon leads to single or
multiple strictures through a number of underlying
pathogenic mechanisms (see as above). Abdominal
pain in such patients is characteristically described
as a "ball of wind" moving around the umbilicus. It is
associated with abdominal distension, inability to pass
wind, and borborygmi. Following an episode of pain,
diarrhea usually ensues. Steatorrhea and significant
weight loss can occur due to bacterial overgrowth.
Right lower quadrant mass and pain can occur in
patients with hypertrophic ileocecal tuberculosis.
Rarely, diffuse colonic disease can simulate symptoms
of ulcerative colitis. Perforation and fistulae occur in
a small percentage of patients. Massive bleeding from
the lesion in the gut has been reported. Clinical exami-
nation reveals distended bowel loop and exaggerated
bowel sounds. Plain X-ray of the abdomen reveals
distended bowel loops with multiple fluid levels.
Diagnosis. Intestinal tuberculosis can be difficult
to diagnose. The reasons for this include absence
of a particular pattern of symptoms and signs. In
fact, symptoms of the disease may be vague and
signs nonspecific. Thus, a high degree of suspicion
is needed. Even with adequate imaging, endoscopic
examination and bacteriologic tools, diagnosis can
correctly be made in only around 50% of patients
with intestinal tuberculosis. The dominant reason for
this is the inaccessibility of common sites of disease
segments of the bowel, namely the ileum and ileoce-
cal region. Moreover, the hallmark of tuberculous
pathology, namely caseating granulomas, may be
absent in the bowel wall and present in the drain-
ing lymph nodes (Tandon and Prakash 1972; Anand
1956). Laparotomy and resection of the involved
segments with culture and animal inoculation of
the organisms have been performed to make a diag-
nosis with precision in endemic areas. Therapeutic
trial with antituberculous drugs is commonly used
in developing countries to make a diagnosis.
A number of clinical conditions closely simulate
intestinal tuberculosis. These include Crohn's disease,
amebiasis, carcinoma colon, Yersinia enterocolitis,
gastrointestinal histoplasmosis, and periappendiceal
abscess. A number of features may help to differentiate
intestinal tuberculosis from Crohn's disease and Yer-
sinia infection. These have been detailed in Table 37.1.
The diagnostic algorithm to be followed for intes-
tinal tuberculosis may vary with the exact site of dis-
ease involvement (Fig. 37.2). X-ray chest, PPD skin
test, and flat abdominal films are usually used for the

Table 37.1. Differentiating features of abdominal tuberculosis from Crohn's disease and Yersinia
infection
Feature Tuberculosis Crohn's disease Yersinia enterocolitica

Clinical course Prolonged Long intermittent Several weeks

Stool culture Negative Negative Positive
Serology for Yersinia Negative Negative Positive
PPD Positive Negative Negative
X-ray of chest Positive Negative Negative
Ileal disease Short Long Short
Ulcers Circumferential Linear Normal endoscopy
Fistulae Unusual Common Nil
Granulomas Large, many, caseating Small, few, Intramural, multiple, large,
noncaseating with satellite Abscess
Anal lesions Rare Frequent Nil
Strictures Usually <3 cm Long Localized
Nodal involvement Often, independent Only with trans- In children with ileitis
of mural disease mural disease
Abdominal Tuberculosis 663

I Suspect abdominal tuberculosis

X-ray chest

PPO

Luminal disease I Lymphadenopathy I

'colonOSCO~y, enter ileum, biopsy
!
'Barium enema •Ascitic fluid analysis •Fine needle aspiration
'Barium follow through 'Laparoscopy/ biopsy biopsy
'Enteroclysis

Fig. 37.2. An algorithm which is useful

Culture & Direct Amplification Tests to investigate most cases of abdominal
tuberculosis

initial investigations. Active pulmonary disease may

help; however, it is seen in only a minority of patients.
PPD skin test is positive in the majority of patients
with abdominal tuberculosis, but is of limited value
because it does not differentiate between active dis-
ease and previous exposure or vaccination. Further-
more, the PPD skin test may be negative in older or
immunosuppressed patients (Lui et al. 1996; Lam et
al. 2000). Careful examination of the flat abdominal
films may give important clues to the nature and
site of underlying pathology. Calcification of lymph
nodes is of the speckled type, and rarely calcification
of peritoneum may coexist. Episodes of abdominal
pain are usually associated with dilated bowel loops
with air fluid levels proximal to the site of stricture.
Abdominal imaging by ultrasound, computed
tomography, or magnetic resonance imaging is useful Fig. 37.3. Tuberculosis of colon. A 28-year-old male presenting
with fever, abdominal pain, and loose motions of 18 months
to define the bowel wall, abdominal lymph nodes, duration. He had a strong history of contact with an open case
and changes in peritoneum, mesentery, and omen- of pulmonary tuberculosis. General physical examination was
tum. CT, with its ability to provide a comprehensive unremarkable. Abdominal examination revealed fullness and
overview of abdominal structures, is the imaging vague tender mass in the right iliac fossa. ESR was 10 mm and
modality of choice for such evaluations (Suri et al. PPD skin test was 7 mm. Plain X-rays of chest and abdomen
were normal. Computed tomography of abdomen with oral
1999; Balthazar et al. 1990; Ha et al. 1999). The most
and IV contrast shows thickened cecum (short arrow), a mass
common CT findings are mural thickening affecting below the cecum (long arrow) and 2 lymph nodes of 1 em
the ileocecal region, either limited to the terminal diameter each (arrowhead). A barium enema (Fig. 37.4a),colo-
ileum or cecum or, more commonly, simultaneously noscopy (Fig. 37.5c), colonic biopsies from the lesion in the
involving both regions (Fig. 37.3,38.10,38.12). This cecum (Fig. 37.6b) were performed. Colonic biopsy samples
grew Mycobacterium tuberculosis on culture. He made rapid
mural thickening is usually concentric, but is occa-
clinical improvement with antituberculous treatment
sionally eccentric, and it predominantly affects the
medial wall. In some patients, low-density areas, most
likely to represent necrosis, may be noted within the may be seen elsewhere in the small bowel, usually
thickened wall. Ileocecal involvement is usually asso- affecting the ileal loops. These segments may also
ciated with enlarged hypodense nodes in the adjacent show luminal narrowing with or without proximal
mesentery. Skip areas of concentric mural thickening dilatation. The presence of such lesions in combina-
 664
tion with ileocecal involvement should strongly sug-
gest the diagnosis of tuberculosis.
Barium enema and small bowel follow-through
may show mucosal ulceration, strictures, deformed
cone-shaped and retracted cecum, incompetent ileo-
cecal valve, a wide gap between a thickened ileocecal
valve and a narrowed ileum (Fleischner's sign), and a
fibrotic terminal ileum that empties into a rigid con-
tracted cecum (Stierlin's sign) (Fig. 37.4, 38.9) (Suri
et al.1999; Balthazar et al. 1990; Ha et al.1999). Small
bowel enema (enteroclysis) has a special advantage
in defining the site and number of small bowel stric-
tures (Fig. 38.5, 38.6, 38.7)
Colonoscopy has been used in patients with
colonic and ileocecal tuberculosis (Singh et al. 1998;
Bhargava et al. 1992; Shah et al. 1992; Misra et al.
1999; Kalvaria et al. 1988). It has the advantage that
targeted biopsies from endoscopic abnormalities
can be taken for histology, culture, and molecular
techniques (Kochhar et al. 1991; Pulimood et al.
1999; Jost et al. 1995; Anand et al. 1994; Kashima et
al. 1995; Pfyffer et al. 1996; Yajko et al. 1995; Tevere et
al. 1996; Rich et al. 1996; Simon et al. 1993; Schluger
et al. 1994; Bradley et al. 1996; Wobeser et al. 1996;
Carpentier et al. 1995; Vlaspolder et al. 1995; Shah et
al.1998). Colonoscopic examination in 50 patients of
colonic tuberculosis revealed ileocecal disease in 16,
ileocecal and contiguous ascending colon disease in
14, segmental colonic disease in 13, ileocecal disease
and nonconfluent involvement of another part of the
a b
Fig. 37.4a, b. Tuberculosis of colon. a Barium enema shows filling defect (arrowhead) and cone shaped deformity (long arrow) of
the cecum. AppendiX is normally filled. b Barium enema shows lack of distensibility and nodular defects of the hepatic flexure,
transverse and splenic flexure (arrowheads)
M. S. Khuroo and N. S. Khuroo
colon in 5, and pancolitis in two patients (Singh et al.
1996). The colonoscopic appearances include muco-
sal nodules and ulcers, stricture with nodules and
ulcerations, and mucosal nodules with or without
pseudopolypoid folds (Fig.37.5). Nodules vary in
size from 2 to 6 mm and have a pink surface. These
are scattered and at places densely packed. Friabil-
ity of mucosa over nodules is unremarkable. Ulcers
may be from a few millimeters to 2 em long and are
superficial with sharply defined irregular margins.
Ulcers are covered with slough, which is difficult to
wash away. The surrounding mucosa is nodular and
hyperemic and blends imperceptibly with normal
mucosa. When the ileocecal valve is involved, it is
edematous, deformed, patulous, and easily admits
the endoscope into the diseased terminal ileum. With
diffuse colonic involvement, mucosa from rectum to
cecum is hyperemic and friable and shows areas of
circumferential ulcerations of different sizes along
the entire length of the colon. Biopsy samples should
be taken from ulcer edge, ulcer base, nodules, and
from adjacent normal mucosa.
Endoscopic mucosal biopsies from the colon and
terminal ileum may show a mixture of pathologic
changes and include (l) characteristic and diagnostic
caseating granulomas in about 25% of patients, (2)
noncaseating granulomas in about 35%, (3) ulceration
with nonspecific granulation tissue and infiltration
with polymorphs forming microabscesses in around
60%, (4) variable mucosal reparative changes in around
 Abdominal Tuberculosis 665

Fig. 37.5a, b. Tuberculosis of colon. Colonoscopic views in

three patients. a A transversely placed ulcer in the descend-
ing colon (arrows). b An infiltrative and nodular lesion in the
transverse colon (arrows). c Hypertrophic nodular mass in the
b cecum (arrows)

20%. Characteristic granulomas show caseous necrosis
in the center, are often large, with marked variations
in size, and usually tend to be confluent (Fig. 37.6).
The granulomas seem to be enlarged by expansion of
individual granulomas or by confluence of numerous
satellite granulomas. This is in sharp contrast to sarcoid
granulomas seen in Crohn's disease which are small in
size, closely adjacent but discrete, and do not become
confluent (Tandon and Prakash 1972).
Endoscopic mucosal biopsy rarely shows M.
tuberculosis organisms on smear, and routine culture
yields a growth of bacilli in only 6% to 40% of speci-

a b

Fig. 37.6a, b. Tuberculosis of colon. a Histologic examination of colonic biopsy revealed dense lymphoplasmocytic infiltrate
in lamina propria and a well-formed granuloma (arrows) consisting of epitheloid histiocytes and multinucleated giant cells.
b Histologic examination of colonic biopsy showing moderate lymphoplasmocytic infiltrate in the lamina with infiltration and
destruction of crypts (arrow). No granulomas were seen
666
mens (Singh et al. 1996; Bhargava et al. 1992; Shah et
al.1992; Misra et al.1999; Kalvaria et al.1988). Recent
technologic developments have introduced a number
of improvements in the ability of clinical laboratories
to cultivate and identify Mycobacterium tuberculosis
complex more quickly and precisely than previously.
These developments include more rapid detection of
growth Oost et al. 1995) and tests to identify RNA
or DNA of M. tuberculosis complex directly in clini-
cal samples (Anand et al. 1994; Kashima et al. 1995;
Pfyffer et al. 1996; Yajko et al. 1995; Tevere et al. 1996;
Rich et al.1996; Simon et al.1993; Schluger et al.1994;
Bradley et al. 1996; Wobeser et al. 1996; Carpentier
et al. 1995; Vlaspolder et al. 1995; Shah et al. 1998).
Exploitation of such tools for intestinal tuberculosis
will make the diagnosis easier and more frequent.
37.3
Tuberculous Peritonitis
Pathogenesis (Marshall 1993; Singh et al. 1969). Peri-
toneal seeding by tubercle bacilli causes granulomas,
which appear as multiple, whitish miliary nodules
«5 mm) scattered over the visceral and parietal perito-
neum. In addition, the peritoneal lining along with the
omentum and mesentery is thickened and adhesions
develop with abdominal organs. A majority (>95%)
of patients develop exudative free or loculated ascites;
however, a small group of patients may have a more
advanced dry fibroadhesive (plastic) or purulent form
of disease. Plastic peritonitis causes adhesions and mat-
ting of bowel loops, mass formation due to matting
of bowel loops, adenopathy, mesenteric and omental
thickening (omental cake). Purulent peritonitis is usu-
ally secondary to tuberculous salpingitis and causes
abscess formation due breakdown of caseous lesions in
lymph nodes, mesentery, or omentum. These abscesses
are present within matted bowel loops and thickened
omentum and mesentery. Fistulae, both cutaneous and
enteric, are common when such abscesses rupture
either through the skin or into the bowel.
Clinical Manifestations (Marshall 1993; Singh et al.
1969; Manohar et al. 1990). Tuberculous peritonitis
in its ascitic form presents insidiously with progres-
sive abdominal distension. Diffuse abdominal pain
(65%), fever (71%), and weight loss (38%) are seen in
a variable percentage of patients. Clinical examination
reveals shifting dullness, abdominal tenderness, and
transverse solid epigastric intra-abdominal mass. The
last is caused by rolled-up, thickened omentum infil-
M. S. Khuroo and N. S. Khuroo
trated with tubercles. The encysted form of the disease
produces a localized cystic mass usually in the central
or lower abdomen, resembling a mesenteric cyst in
children and ovarian cyst in females. Plastic peritonitis
produces matted small bowel loops with thickening of,
and adhesions with omentum and mesentery. Patients
often present with recurrent attacks of subacute intes-
tinal obstruction. Acute intestinal obstruction may
sometime supervene. Dilated bowel loops produce
bacterial overgrowth and cause steatorrhea and wast-
ing. Abdominal examination reveals single or mul-
tiple bowel masses which are resonant to percussion
(thickened and matted bowel loops). Solid mass may be
caused by thickened mesentery. Patients with purulent
peritonitis are very sick, wasted, and in moribund clini-
cal status. Abdomen examination reveals tenderness,
guarding, multiple bowel masses, and usually a fecal
fistula commonly near the umbilicus.
Patients with tuberculous peritonitis with cirrho-
sis of the liver present with similar clinical features
to those without liver disease. However, patients with
liver disease are younger (42±8 years vs 54±15 years,
p<O.Ol) and have a higher maximum-recorded tem-
perature (l02±107 vs 100.5±1.3,p<0.0l). In addition,
clinical examination reveals hepatomegaly (48%)
and splenomegaly (20%) due to underlying liver
disease and portal hypertension (Aguado et al. 1990;
Shakil et al. 1996).
Tuberculous peritonitis in patients with long-term
or continuous ambulatory peritoneal dialysis present
with fever, abdominal pain, and cloudy dialysate. Peri-
toneal fluid has predominance of polymorphonuclear
cells as against lymphocytic predominant cells in
tuberculous peritonitis associated with other condi-
tions. Diagnosis is made at culture of the fluid, which
grows tubercle bacilli in two thirds of such patients
(Holley and Piraino 1990; Cheng et al. 1989; Lui et al.
1996; Lam et al. 2000; Talwani and Horvath 2000).
Diagnosis. Diagnosis of tuberculous peritonitis is
mainly focused on the differential diagnosis of ascites
and a well-established algorithm has been developed
in clinical practice to do so (Runyon et al. 1992). The
index of suspicion of tuberculous peritonitis should
be high in following circumstances:
a. Residence in developing countries or immigration
to a Western country from a developing country
b. Recent exposure to open tuberculosis
c. Underlying cirrhosis
d. Patients on long-term or continuous ambulatory
peritoneal dialysis
e. Immunosuppressed patients, especially AIDS, and
patients with liver or renal transplants
 Abdominal Tuberculosis 667

The value of a chest X-ray, PPD, and flat abdominal
films has been discussed {Fig. 37.7). Abdominal imag-
ing, especially CT scan, is useful for an initial investi-
gation to give a comprehensive view of the abdominal
organs. Ascitic fluid analysis gives an important lead
to the possibility of infectious etiology. Laparoscopy
and peritoneal biopsy is the investigation of choice to
confirm the diagnosis of tuberculosis.
CT findings include changes in the peritoneal
lining and cavity, mesentery, and omentum (Suri et al.
1999). Peritoneal lining shows smooth uniform thick-
ening. Nodular implants with irregular thickening of
the peritoneum are unusual and more often sug-
gest peritoneal carcinomatosis {Fig. 37.8, 38.l9a,b).
Peritoneal fluid may be free or loculated and shows
high-density signals (25-45 HU), possibly explained
by high protein and cellular contents of the fluid.
However, tuberculous ascites may also be near water
density, perhaps reflecting an earlier transudative
stage of immune reaction. Mesenteric infiltration can
range from mild involvement in the form of linear
soft tissue strands, thickened and crowded vascular
bundles, a "satellite" appearance, and subtle increase
in mesenteric fat density, to more extensive involve-
ment resulting in diffuse infiltration with soft tissue
density masses involving the leaves of mesentery
surrounding the adjacent bowel loops. Omentum
infiltration may cause thickening, smudged appear-
ance, or omental "cake" formation. Retroperitoneal
and mesenteric nodes may be enlarged and caseate to
form large mesenteric abscesses (Fig. 37.9).
Ascitic fluid may be collected from either flank or
centrally below the umbilicus with a blind peritoneal
needle puncture and aspiration (Runyon 1986). In
patients with minimal fluid collection or those with
thick abdominal wall due to obesity, ultrasound-
guided fluid collection may be done {Goldberg et al.
1970). Ascitic fluid examination should include gross
inspection, biochemical tests, cytology, and smear
and culture for tuberculosis. Fluid for a cell count
should be sent to the laboratory in an anticoagulant
tube (i.e., containing heparin/ethylenediaminetetra-
acetic acid) to prevent clotting (Hoefs 1990). Before
the 1980s, the ascitic fluid total protein concentration

b
Fig. 37.7a, b. Tuberculous peritonitis. A 50-year-old woman presenting with low-grade fever, weight loss, diffuse abdominal
pain, and abdominal distension of 6 months duration. Clinical examination revealed abdominal tenderness and free fluid in the
peritoneum. ESR was 60 mm and PPD skin test was 25 mm. X-ray chest revealed right apical infiltration and scarring. Ascitic
fluid analysis revealed low-gradient lymphocytic exudate. a Laparoscopic examination revealed adhesions, peritoneal exudates,
and multiple small (3 to 5 mm), whitish, elevated lesions on the visceral and parietal peritoneal surface. In this photograph
multiple such lesions are shown on the liver surface. The results of a peritoneal biopsy from this patient are shown in Fig 37.10
b Plain X-ray of abdomen showing plaque-like calcification in the right and left upper quadrant (peritoneum - thick arrow),
nodular calcification in the abdomen (lymph nodes - arrowheads), and incidental atherosclerotic linear calcification along the
aortic wall (arrow).
 .•
668 M. S. Khuroo and N. S. Khuroo

.,'
• 4f.
.
.'.. ";~'
. ,.'. '~.":
'

til"
.
"I" ,

:." :".; ",',~.',~c.'.

~~',••11:'· .~,
't •

.'
~I'I-'~
• : ' tJA

G ".
' , .-.' i .'
. ':: ' #
. ".:
., I . .. ~,.,
~
. .
'~"l"'/
,'ll

.
__ ",,- ,.'

'I
,'=
. '

-:..~. I .~, f~ .
~! I' ...

a b

Fig. 37.8a, b. Tuberculous peritonitis. Computed tomography (CT) scans of two patients with documented tuberculous peritoni-
tis. a Contrast-enhanced CT of abdomen shows thickening and fat infiltration of the omentum (omental plaque--arrowheads).
b Contrast-enhanced CT of abdomen showing irregularity and fatty infiltration of the mesentery (arrowheads)

Fig. 37.9a-c. Tuberculous retroperitoneal mass. A 30-year-old woman presented with fever and night sweats of 3 months dura-
tion. Clinical examination was unremarkable. Plain X-ray of chest was normal. PPD skin test was 15 mm. Contrast-enhanced
spiral CT of abdomen (a and b) showed 9x5 cm mass (long arrows) behind the stomach and displacing it anteriorly. a Arterial
phase revealed celiac artery (thin arrow) within the mass, which was not involved by the lesion. b Late venous phase revealed
minimal contrast uptake by the lesion (arrows). c Histologic examination of the resected specimen revealed extensive necrotiz-
ing granulomatous lesion (arrow) with giant cells (arrowhead)
Abdominal Tuberculosis
was used to classify ascites into exudate (>25 gIl) and
transudate «25 gil). This classification does catego-
rize ascites into various etiologic groups with a high
degree of precision. Attempts at using combinations
of lactic dehydrogenase (LDH) and serum-ascitic
fluid ratio of LDH and protein have not been shown
to improve accuracy of classifying ascitic fluid into
exudate and transudate. The serum-ascites albumin
gradient (SAAG) has proven in multiple studies to
categorize ascites better than total protein concen-
tration and better than other parameters (Runyon
et al. 1992; Mauer and Manzione 1988). An SAAG of
11 gIl or more is classified as high gradient ascites,
and an SAAG of less than 11 gIl is classified as low
gradient ascites. If the SAAG is high, the patient has
portal hypertension as the cause of ascites, with 97%
accuracy. If the SAAG is low, portal hypertension
can be excluded as the cause of portal hypertension,
with 97% accuracy. The accuracy of this test is not
influenced by ascitic fluid infection, diuretic therapy,
therapeutic paracentesis, albumin infusion, and eti-
ology of liver disease.
Tuberculous ascitic fluid is usually opalescent due
to high protein content and cell count (Marshall 1993;
Holley and Piraino 1990; Singh et al.1969; Manohar et
al.1990; Runyon et al.1992). However fluids with a cell
count ofless than I,OOO/cm 3 (1.0 x109/1) may be almost
clear. Fluids with counts of over 50,OOO/cm 3 (50 x 109Il)
look purulent. Fluid may be sanguineous (RBC> 10,0001
mm 3) or frankly hemorrhagic (RBC>20,OOO/mm 3).
Bloody ascitic fluid due to underlying disease should
be differentiated from traumatic tap. The latter is only
streaked with blood and frequently clots. In contrast,
nontraumatic blood-tinged ascitic fluid is homoge-
neous and does not clot. Tuberculous ascitic fluid is
uniformly of low gradient variety with a SAAG of less
than 11 gil and has a high cell count (150 to 4,OOO/mm 3 )
with lymphocytic predominance. Tuberculous fluid in
patients with chronic peritoneal dialysis is typically
neutrocytic rather than lymphocytic.
Ascitic fluid adenosine deaminase is an enzyme
involved in the catabolism of purine bases (conver-
sion of adenosine to inosine) (Marinez-Vazques et al.
1986; Pettersson et al. 1984; Hillebrand et al. 1996).
Levels of ascitic adenosine deaminase are increased
in tuberculous peritonitis as a result of stimulation of
T lymphocytes in response to cell-mediated immu-
nity to mycobacterial antigens. A number of studies
have shown that at a cut-off of >33 VII, the sensitivity
and specificity in tuberculous ascites are about 100%
and 95%, respectively. A cut-off of >50 VII may even
be preferable because sensitivity remains excellent
and false positives are almost eliminated. Ascitic
669
adenosine deaminase has been proposed as a useful
test in detecting peritoneal tuberculosis. However, in
the Vnited States, where more than half of patients
with tuberculous peritonitis have underlying cirrho-
sis, ascitic adenosine deaminase has been found to be
too insensitive to be helpful.
Examination of an acid-fast, stained smear of
ascites will identify the organism in less than 3% of
cases. The chances of culturing M. tuberculosis from
the ascitic fluid are less than 20%. Culturing of ascitic
fluid concentrated by centrifugation may increase
the yield of culture. Culture reports are available by
4 to 8 weeks, which limits their diagnostic usefulness
(Runyon et al. 1992).
Laparoscopy with directed biopsies is an excellent
study for diagnosis of tuberculous peritonitis and
should be done in all patients with low-gradient asci-
tes, lymphocytic ascites, and in those with high risk
or high index of suspicion of tuberculosis (Singh et
al.1969; Bhargara et al.1992; Geake et al.I98l).
The laparoscopic appearances include:
a. Thickened peritoneum with loss of usual shiny
luster, miliary yellowish tubercles of uniform size
(about 4-5 mm) diffusely distributed over pari-
etal peritoneum and loops of the bowel, multiple
adhesions between organs and peritoneum: this
pattern is seen in around 66% of patients.
b. Thickened parietal peritoneum with loss of luster,
multiple adhesions between liver, peritoneum, and
loops of the bowel: this pattern is seen in around
21 %of patients.
c. Fibroadhesive pattern with marked thickening of
parietal peritoneum, peritoneum may show yel-
lowish nodules and cheesy material, thick adhe-
sions may fix the viscera to anterior abdominal
wall, sometimes it may not be possible to enter
the peritoneal cavity: this pattern is seen in 13%
of patients.
Laparoscopic biopsies from abnormal-appearing
lesions detect caseating granulomas in 85-90% of
patients (Fig. 37.10). Mycobacterium can be cultured
in 40% of patients. Laparoscopy in patients with
peritoneal tuberculosis appears to be relatively safe;
complications occurred in around 3% of patients,
including bowel perforation, intraperitoneal bleed-
ing, and subcutaneous hematoma.
Blind peritoneal biopsies in patients with free asci-
tes can be performed to obtain tissue for histology and
culture. The procedure is reasonably safe and incidence
of complications is low. The yield for positive diagnosis
is lower than with laparoscopic-targeted biopsies. It
is recommended in centers where laparoscopy is not
670
Fig. 37.10. Tuberculous peritonitis. Peritoneal biopsy showing
extensive necrosis (thick arrow) with giant cells (long arrow)
available or in patients who refuse for laparoscopy.
Minilaparotomy with peritoneal biopsies is recom-
mended for patients with extensive peritoneal adhe-
sions and for those patients for whom laparoscopy is
nondiagnostic (Levine 1968; Shukla et al.1982).
37.4
Tuberculosis of Mesenteric Lymph Nodes
Pathogenesis (Tandon and Prakash 1972; Marshall
1993). Tuberculosis of the mesenteric lymph nodes
is considerably less common than intestinal or peri-
toneal involvement. Tubercle bacilli reach the nodes
by way of Peyer's patches. Single or multiple lymph
nodes along the mesentery may be involved. Lymph
nodes are rounded or oval and readily mat together
and calcify. Breakdown may occur giving rise to
tuberculous pus in the mesentery. The bowel loop
may become adherent, or disease may spread to the
bowel wall or pancreas.
Clinical Manifestations (Mann et al. 1997). Tuber-
culosis of the mesenteric lymph nodes may present
with systemic symptoms of the disease only (see
above) without abdominal complaints. Sometimes
nonspecific abdominal pain may accompany these
symptoms. Abdomen lymph nodes may be palpable
on deep palpation as firm, discrete, tender, bean-like
masses most frequently to the right of and near the
umbilicus. Sometimes reactivation of infection in
the lymph nodes may cause pain and tenderness in
the right iliac fossa resembling subacute appendici-
tis. Subacute intestinal obstruction may result from
adhesions with a bowel loop or stricture as a result of
M. S. Khuroo and N. S. Khuroo
involvement of the bowel wall by the caseous lymph
node. A tuberculous mesenteric abscess gives rise to
a palpable cystic mass. Enlarged lymph nodes in the
ileocecal lymph nodes give rise to a palpable mass in
the right iliac fossa.
Diagnosis (Suri et al. 1999; Batra et al. 2000). Abdomi-
nal imaging (ultrasound and CT scan) is the inves-
tigation of choice for patients with tuberculosis of
mesenteric lymph nodes (Fig. 37.11, 38.24a-c). In
fact, imaging is often the tool which first points to
this possibility when patients with vague abdominal
symptoms are being investigated. Differential diagno-
sis includes other causes of lymph node enlargement
including lymphoma, metastases, Whipple's disease,
etc. Mesenteric and peripancreatic lymph nodes are
commonly affected sites, reflecting the lymphatic
drainage of commonly affected sites in the small
bowel and liver. Isolated tuberculous retroperitoneal
lymphadenopathy is uncommon and most patients
also have affected lymph nodes at other sites. CT scan
appearances of enlarged lymph nodes are as follows:
a. Enlarged nodes with hypodense centers and
peripheral hyperdense rims. This is the commonest
appearance and occurs in around 70% of patients.
b. Conglomerate mixed density nodal masses, most
likely representing multiple confluent nodes due
to perinodal spread of inflammation.
c. Enlarged nodes of homogeneous density, most
often associated with low density nodes at other
sites.
Fig.37.11. Tuberculous lymphadenopathy. Ultrasound abdo-
men in A 30-year-old woman with abdominal pain and pal-
pable abdominal masses showing multiple pre-aortic lymph
nodes. Histology of the resected lymph nodes is shown in
Fig. 37.12
Abdominal Tuberculosis 671

d. Increased number (>3 in one CT section) of

normal sized or mildly enlarged mesenteric nodes
of homogeneous density, usually located along the
mesenteric vessels or adjacent to the bowel loops.
e. Nodal calcification with characteristic distribu-
tion and appearance.

Whenever tuberculous lymphadenopathy is sus-

pected, diagnosis is confirmed by ultrasound or CT-
guided fine-needle aspiration biopsy and examining
the aspirated material and/or core biopsy sample for
histology, smear, and culture (Fig. 37.12). A biopsy
can also be performed with a laparoscopy or mini-
laparotomy if radiologically guided biopsies cannot Fig.37.12. Tuberculous lymphadenopathy. Histologic exami-
be done for reasons of access through a bowel loop. nation of the resected lymph node showing caseating granulo-
matous inflammation (thick arrows) with multiple giant cells
(thin arrows) and surrounding lymphocytic infiltration

37.S 4. Nodular disease. In this entity, single or multiple

Tuberculosis of Solid Abdominal Organs focal masses develop in the liver and are seen as
low-density nonenhancing lesions with or without
Hepatobiliary Tuberculosis. Hepatobiliary tuberculo- peripheral rim enhancement. Such appearances
sis is seen in number of situations: need to be differentiated from lymphoma, fungal
1. Incidental. Liver involvement is seen at autopsy in infection, and metastasis. Diagnosis is confirmed
25 to 50% of patients dying from active pulmo- by image-guided fine-needle aspiration biopsy of
nary tuberculosis. An autopsy study from West the lesion (Herman et a!. 1995; Achem et a!. 1992;
Africa found that 50% adults dying of AIDS have Buxi et a!. 1992).
active tuberculosis and in 85% of them the liver 5. Tuberculous liver abscess. Tuberculous abscesses
is involved (Lucas 1994). in the liver are extremely rare. The clinical pic-
2. Miliary form. This occurs due to hematogenous ture and imaging resemble those of a pyogenic
spread of the tubercle bacilli. The liver is involved or amebic liver abscess. Culture of the aspirated
by multiple granulomas. Miliary tuberculosis material confirms the diagnosis by growth of
presents with fever, night sweats, anorexia, weak- tubercle bacilli (Rahmatulla et al. 2001).
ness, and weight loss. Physical signs include 6. Tubular disease. These patients present with obstruc-
hepatomegaly, splenomegaly, and lymphadenopa- tive jaundice due to involvement of the bile ducts.
thy. Elevation of serum alkaline phosphatase and Bile ducts may be involved by an enlarged tubercu-
other abnormal values in liver function tests are lous lymph node compressing the bile duct or diffuse
detected in patients with severe hepatic involve- involvement of the intrahepatic ducts by tubercle
ment. Liver biopsy revealed granulomas in a high bacilli. ERCP reveals multiple intrahepatic biliary
percentage of patients and usually gives a clue to strictures, areas of dilatation, beading and ectasia,
diagnosis (Raviglione and O'Brien 2001). resembling sclerosing cholangitis or cholangiocarci-
3. Granulomatous hepatitis. Patients present with noma. Biliary stricture may occur at hilar region or
unexplained fever, jaundice, hepatomegaly, ele- distal common bile duct with dilatation of the intra-
vated alkaline phosphatase, and abnormality of hepatic ducts (Alvarez 1998; Hickey et al. 1999).
other liver function tests. Imaging of the liver may
be normal or reveal nonspecific abnormalities. Splenic Tuberculosis. Tuberculosis of the spleen pres-
Laparoscopy is useful and shows cheesy, white, ents with fever, night sweats, asthenia, and loss of
irregular nodules on the liver surface. Biopsy from weight. The spleen is enlarged and is clinically pal-
such lesions reveals caseating granulomas. Granu- pable. Portal hypertension may ensue. Tuberculous
lomatous hepatitis with multiple granulomas in the splenic abscess is rare and presents as splenomegaly
liver may also be seen following vaccination with with a heterogeneous mass on ultrasound or CT.
bacille Calmette-Guerin, especially in persons with Splenic puncture will yield cold abscess and culture
impaired immune response (Campos et al. 1996). grows the organisms (Mann et al. 1997).
 672
Pancreatic Tuberculosis. Pancreatic tuberculosis
presents with a wide spectrum of symptoms such
as abdominal pain, weight loss, fever, and obstruc-
tive jaundice. Abdominal ultrasound or CT detects
pancreatic masses closely mimicking pancreatic car-
cinoma (Fig. 37.13). Diagnosis is only revealed with
a fine-needle aspiration biopsy and culture of the
aspirated material(Harland and Varkey 1992).
37.6
Tuberculosis of Other Gastrointestinal Sites
Esophageal, gastric, duodenal, and isolated appen-
dicular tuberculosis are rare. Anal tuberculosis is
rare in the West; however, it comprises 16% of the
cases of fistulae-in-ano in developing countries
(Marshall 1993).
c without displacement and involvement
M. S. Khuroo and N. S. Khuroo
Esophageal tuberculosis usually results from exten-
sion of the disease from mediastinal lymph nodes or
spread from a pulmonary focus. Rarely, disease may
occur without a primary contiguous focus of disease.
Esophagus reveals ulceration, nodularity, strictures,
sinus-track formation, and fistulae with trachea or bron-
chus. Esophageal tuberculosis presents with dysphagia,
odynophagia, choking, and aspiration due to tracheo-
esophageal or bronchoesophageal fistula and upper
gastrointestinal bleeding. Bleeding from esophageal
infiltration and ulceration is usually of no major conse-
quence; however, massive bleed from aortoesophageal
fistula complicating tuberculosis has been reported.
Chest X-ray and CT scan of the chest are helpful in
identifying active pulmonary lesion and mediastinal
masses. Barium-swallow findings include ulcerations,
stricture, pseudotumor masses, fistulae, sinuses, and
traction diverticula. Upper gastrointestinal endoscopy
with biopsy is the diagnostic procedure of choice and
Fig. 37.13a-c. Pancreatic tuberculosis. Contrast-enhanced
spiral computed tomography of abdomen (a and b) shows a
hypodense mass (long arrow) in the head of pancreas. Arterial
phase (a) revealed displacement of the hepatic artery (arrow-
head), and portal venous phase (b) revealed displacement of
the portal vein (arrowheads). c Spiral computed tomography
of the abdomen in another patient revealed a well-defined
hypodense mass (long arrow) in the pancreas. Hepatic artery
(arrowhead) and portal vein (small arrow) are well defined
Abdominal Tuberculosis 673

Fig. 37.14. Esophageal tuberculosis. Endoscopic view of Fig.37.15. Gastric tuberculosis. Endoscopic view of stomach
esophagus shows ulceration, narrowing, and whitish exudates showing ulceration (arrowhead), nodularity (thick arrow), and
and pseudomembrane (arrow) bridging mucosal lesions (long arrows)

usually reveals the etiologic nature of the esophageal causes duodenal stricture and presents as abdominal
disease (Fig. 37.14). In patients with mediastinallyrnph- pain, vomiting, and gastric stasis. Barium meal exami-
adenopathy, endoscopic ultrasound has made a major nation (Fig. 37.16, 38.1, 38.2, 38.3) and upper gastroin-
advance in identifying the nodes, and biopsies can be testinal endoscopy define the type of involvement of
taken from these lesions for histology and culture under the duodenum. Biopsies are usually not helpful and
endoscopic ultrasound guidance (Tassios et al. 1995; show nonspecific changes. Diagnosis is confirmed at
Eng and Sabanathan 1991; Sutton 1990). histologic examination of resected diseased segment
Gastric tuberculosis usually occurs in the absence (Marshall 1993).
of pulmonary disease. This has been attributed to the
presence of acid and paucity of lymphoid tissue in
the stomach. Disease causes ulceration, nodularity of
the mucosa, a tumor-like mass, and extensive submu-
cosal infiltration and fibrosis causing linitis plastica.
The antrum is the most common site of involvement.
Gastric tuberculosis presents with nonspecific symp-
toms including abdominal pain, nausea, vomiting,
and gastrointestinal bleeding. Such symptoms are
usually confused with peptic ulcer or gastric neo-
plasm. Fever, night sweats, and weight loss may point
to the possibility of tuberculosis. Barium meal and
upper gastrointestinal endoscopy reveal ulceration,
gastric outlet obstruction, nodular masses, and rigid
nondistended stomach suggestive of linitis plastica
(Fig. 37.15). Diagnosis is usually confirmed at histo-
logic examination of the resected stomach (Lin et al.
1999; Rathnaraj et al. 1997; Quantrill et al. 1996; Goh
et al. 1994; Raskin 1976).
Duodenal tuberculosis commonly occurs second-
ary to extraintestinal lymph node involvement and
Fig. 37.16. Duodenal tuberculosis. Barium-meal examination
causes segmental narrowing of the lumen. However of stomach and duodenum showing duodenal loop irregular-
ulceration, nodularity, and masses may develop due ity and narrowing (arrow) and an ulcer in the duodenal bulb
to mucosal disease. Duodenal tuberculosis usually (arrowhead)
674 M. S. Khuroo and N. S. Khuroo

Appendix involvement is common in patients

with ileocecal tuberculosis. Isolated appendicular
tuberculosis causes subacute inflammation. Isolated
appendicular tuberculosis presents as nonspecific
abdominal pain and right iliac fossa tenderness and
simulates other clinical conditions involving organs
in this region (AI-Hilaly et al.1990).
Anal tuberculosis causes ulcers, fissures, fistu- .=.:;====::::;......;.._...;;...;;". a
lae, abscesses, and warty or hypertrophic growths.
Active pulmonary tuberculosis is found in around
15% of such patients. Anal tuberculosis presents
with a variety of appearances namely ulcers, fis-
sures, fistulae, abscesses, and wart or hypertrophic
growths (Fig.37.17, 38.16, 38.17). Crohn's disease
causes similar appearances and in view of its higher
occurrence in the West is the usual clinical diagnosis
(Candela et al. 1999; Chung et al. 1997; Harland and
Varkey 1992).

37.7
Treatment
Standard Drug Regimen. Patients with abdominal
tuberculosis should receive a standard antitubercu-
lous drug regimen (Raviglione and O'Brien 2001).
Therapy is highly successful and cure occurs in
around 90% patients with tuberculosis of ileum
and colon and over 95% patients with tuberculous
peritonitis and tuberculosis of lymph nodes. Careful
consideration needs to be given to selection of initial
drugs for therapy, compliance issues, modification of
drug regimen based upon drug susceptibility test-
ing, and monitoring of the therapy and drug toxicity.
These do not differ in any way from those of treating
pulmonary tuberculosis. Short-course regimens are
divided into an initial or bactericidal phase and a
continuation or sterilizing phase. The initial phase
consists of 2-month therapy with isoniazid, rifampi-
cin, and pyrazinamide, followed by a 4-month ther-
apy with isoniazid and rifampicin. If drug resistance
is suspected on epidemiologic or other grounds,
ethambutol (or streptomycin) should be included
for the initial2-month therapy or until the drug sus-
ceptibility tests become available. Treatment may be
given daily throughout the course of therapy, or three
times weekly throughout the therapy, or daily for the
first 2 months followed by twice weekly for the next
4 months. Direct observed therapy (DOT) requires an
intermittent dosage schedule and increases compli-
ance. Provision of drugs in combined formulations
is useful, as patients have to swallow only one tablet
Fig. 37.17a, b. Anorectal tuberculosis. a Resected specimen of
the rectum showing mucosal hypertrophy, nodularity (arrow-
heads), and strictures (arrows). b Histologic examination of
the lymph node from the specimen revealed extensive caseat-
ing granuloma (arrow) with giant cells (arrowhead)
daily. It eradicates drug formulation errors, increases
compliance and reduces chances of drug resistance.
Pyridoxine should be added to the regimen given to
persons at high risk of vitamin deficiency, namely
alcoholics, malnourished persons, pregnant and
lactating mothers, and patients with such conditions
as chronic renal failure, diabetes, and HIV infection,
because such patients are prone to neuropathy. A
significant proportion of patients with abdominal
tuberculosis are given therapy without positive
cultures for tubercle bacilli. These patients need
treatment with a standard regimen. Monitoring of
response to treatment in abdominal tuberculosis is
usually assessed by clinical parameters.
Course of Action When the Diagnosis Is Unclear. The
diagnosis of tuberculosis may remain unclear despite
the diagnostic efforts described above (Marshall
1993). This can occur in as many as 50% patients
in developing countries, where diagnostic tools are
not freely available and culture techniques have not
been refined. Either of two possible actions can be
taken in such cases: namely, therapeutic antituber-
culous trial or diagnostic exploratory laparotomy.
In patients with clinical disease highly suggestive of
Abdominal Tuberculosis
tuberculosis, a history of exposure to tuberculosis,
a strong positive PPD, evidence of tuberculosis on
chest X-ray, and residence or origin from develop-
ing countries, therapeutic antituberculous therapy is
feasible. Rapid clinical response to medical therapy is
seen. If the patient fails to respond within 2 weeks,
a laparotomy is indicated. However many clinicians
suggest prompt diagnostic laparotomy in absence of
definite nonoperative diagnosis, since diseases like
Crohn's disease, lymphoma, and malignancy can
mimic tuberculosis in every possible way. Moreover
the criteria for response are based on clinical criteria
and fraught with errors.
Indications for Surgery. Surgery is indicated in a select
group of patients with complications of abdominal
tuberculosis (Bhansali 1977; Kapoor 1998; Marshall
1993; Shah et al. 1992). The most common indication
for surgery is multiple and/or long strictures which
are unlikely to respond to medical therapy. Rarely,
strictures may become critical during antitubercu-
lous therapy. The surgical resection of strictures
should be conservative. Strictureplasty is the stan-
dard treatment for single and especially multiple ileal
strictures. An alternative is to do an endoscopic bal-
loon dilatation of colonic or accessible terminal ileal
strictures (Bhasin et al. 1998). Resection or bypass
should be avoided as these may precipitate short
bowel or blind loop syndrome. For hypertrophic
ileocecal tuberculosis needing surgery, right hemi-
colectomy has been the treatment of choice in the
past. Other reasons for surgical intervention include
free perforation, confined perforation with abscess
and fistula, massive bleeding and intra-abdominal
cold abscesses.
Therapy in Patients with AIDS and Other Immu-
nosuppressed Patients. Patients with immunosup-
pressed conditions including AIDS respond well
to the standard 6-month regimen (Raviglione and
O'Brien 2001; Small et al. 1991). Treatment may be
prolonged to 9 months if the response is graded
as suboptimal. Rifampicin shortens the half life of
HIV protease inhibitors, indavir or nelfinavir, and
therefore is contraindicated when antiviral therapy
is being given concomitantly. In such situations,
rifabutin (150 mg/day or 300 mg twice weekly) is
given instead of rifampicin.
Therapy in Patients with Liver Disease. Patients
with liver disease pose a special problem because
of the hepatotoxicity of isoniazid, rifampicin, and
pyrazinamide (Raviglione and O'Brien 2001). These
675
drugs should be avoided and if absolutely essen-
tial should be used in reduced dosage under close
supervision. Patients with severe hepatic dysfunction
may be treated with ethambutol and streptomycin
and if required with reduced doses of isoniazid and
rifampicin under close monitoring for drug toxicity.
Pyrazinamide in patients with severe hepatic disease
is contraindicated.
Adjunctive Glucocorticoid Therapy. Some clinicians
administer glucocorticoids for 2 to 3 months along
with antituberculous treatment on the assumption
that this therapy will decrease fibrosis during heal-
ing. This may in turn reduce stricture formation in
intestinal tuberculosis and adhesions in peritoneal
tuberculosis. However, this has not been substanti-
ated by any well-conducted therapeutic trial. In fact,
large series of patients with intestinal tuberculosis
have been treated without glucocorticoids and no
strictures showed up in the follow-up. In other series,
patients treated with or without adjunctive glucocor-
ticoids did equally well (Dooley et al.).
References
Achem SR et al (1992) Pseudotumoral hepatic tuberculosis.
Atypical presentation and comprehensive review of the
literature. J Clin Gastroenterol 14:72-77
Aguado JM et al (1990) Tuberculosis peritonitis: a study com-
paring cirrhotic and non-cirrhotic patients. J Clin Gastro-
enterol 12:with perforation. Br J Clin Prac 44:632-633
Alvarez S, McCabe WR (1984) Extrapulmonary tuberculosis
revisited: a review of experience at Boston City and other
hospitals. Medicine 63:25
Alvarez SZ (1998) Hepatobiliary tuberculosis. J Gastroenterol
Hepatol 13:833-839
American Thoracic Society/American Lung Association (1981)
The tuberculin skin test 1981. Am Rev Respir Dis 124:1
Anand BS et al (1994) Diagnosis of intestinal tuberculosis by
polymerase chain reaction on endoscopic biopsy speci-
mens. Am J Gastroenterol 89:2248
Anand SS (1956) Hypertrophic ileocecal tuberculosis in India
with record of 50 hemicolectomies. Ann R Coll Surg Engl
19:205-222
Balthazar EJ, Gordon R, Hulnick D (1990) Ileocecal tubercu-
losis: CT and radiologic evaluation. AJR Am J Roentgenol
154:499
Barnes PF et al (1991) Tuberculosis in patients with human
immunodeficiency virus infection. N Engl J Med 324:1644
Bass JB, Sanders RV, Kirkpatrick MB (1985) Choosing an
appropriate cutting point for conversion in annual tuber-
culin skin testing. Am Rev Respir Dis 132:379
Batra A et al (2000) Sonographic appearances in abdominal
tuberculosis. J Clin Ultrasound 28:233-245
Bhansali SK (1977) Abdominal tuberculosis: experience with
300 cases. Am J Gastroenterol 67:324
676
Bhargara DK et al (1992) Peritoneal tuberculosis: laparoscopic
palleons and its diagnostic accuracy. Am J Gastroenterol
87:109-112
Bhargava DK et al (1984) Cell mediated immune response in
intestinal tuberculosis. Ind J Med Res 80:264-269
Bhargava DK et al (1992) Endoscopic diagnosis of segmental
colonic tuberculosis. Gastrointest Endosc 38:571
Bhasin DK et al (1998) Endoscopic balloon dilation of ileal
stricture due to tuberculosis. Endoscopy 30:S44
Bloch AB et al (1994) Nation-wide survey of drug-resistant
tuberculosis in the United States. JAMA 271:665
Bradney K, Dobkin S (1991) Resurgent tuberculosis in New
York City: human immunodeficiency virus, homelessness
and the decline of tuberculosis control programs. Am Rev
Respir Dis 144:745
Bradley S, Reed S, Catanzaro A (1996) Clinical efficacy of the
amplified mycobacterium tuberculosis direct test for the
diagnosis of pulmonary tuberculosis. Am J Respir Crit
Care Med 153:1606
Brudney K, Dobkin J (1991) Resurgent tuberculosis in New
York City. Am Rev Respir Dis 144:745
Buxi TB et al (1992) CT apperances in macronodular hepa-
tosplenic tuberuclosis: a review with five additional new
cases. Comput Med Imag Graph 16:381-387
Campos JM et al (1996) Disseminated Bacillus Calmette-
Guerin infection in HIV-infected children: case report and
review. Pediat AIDS HIV Infect 7:429-432
Candela F et al (1999) Perianal disease of tuberculous origin:
report of a case and review of the literature. Dis Colon
Rectum 42:110-112
Cantwell MF et al (1994) Epidemiology of tuberculosis in the
United States. JAMA 272:535
Carpentier E et al (1995) Diagnosis of tuberculosis by Ampli-
cor Mycobacterium tuberculosis test: a multicenter study.
J Clin Microbio133:3106
Carr WJ, Kyle RA, Bowie EW (1964) Hematologic changes in
tuberculosis. Am J Med Sci 248:709
CDC (1990) Outbreak of multidrug-resistant tuberculosis-
Texas, California, and Pennsylvania. MMWR Morb Mortal
Wkly Rep 39:369
CDC (1991) Transmission of multidrug-resistant tuberculosis
from an HIV-positive client in a residential substance-
abuse treatment facility-Michigan. MMWR Morb Mortal
Wkly Rep 40:129
Centers for Disease Control and Prevention (1998) Pre-
vention and treatment of tuberculosis among patients
infected with human immunodeficiency virus: Principles
of therapy and revised recommendations. MMWR Morb
Mortal Wkly Rep 47:1
Cheng IK, Chan PC, Chan MK (1989) Tuberculous peritonitis
complicating long-term peritoneal dialysis: report of 5
cases and review of the literature. Am J Nephrol9:155
Chung CC et al (1997) Anal and perianal tuberculosis: a report
of three cases in 10 years. J R Coli Surg Edinb 42:189-190
Daley CL et al (1992) An outbreak of tuberculosis with accel-
erated progression among persons infected with human
immunodeficiency virus: an analysis using restriction-
fragment-Iength-polymorphisms. N Engl J Med 326:231
Dooley DP, Carpenter JL, Rademachar S Adjunctive corticoste-
roid therapy for tuberculosis: a critical reappraisal of the
literature. Clin Infect Dis 25:872-887
Edlin BR et al (1994) An outbreak of multidrug-resistant
tuberculosis among hospitalized patients with the acquired
M. S. Khuroo and N. S. Khuroo
immunodeficiency syndrome. N Engl J Med 330:1710
Edwards D, Kirkpatrick CH (1986) The immunology of Myco-
bacterial diseases. Am Rev Respir Dis 134:1062
Eng J, Sabanathan S (1991) Tuberculosis of the esophagus. Dig
Dis Sci 36:536-540
Farer LS, Lowell LM, Meador MP (1979) Extrapulmonary
tuberculosis in the United States. Am J Epidemiol 109:205
Frieder IR et al (1993) The emergence of drug-resistant tuber-
culosis in New York City. N Engl J Med 328:521
Geake TMS et al (1981) Peritoneoscopy in the diagnosis of
tuberculous peritonitis. Gastrointest Endosc 27:66-68
Goh SH et al (1994) Gastric ulceration with acute bleeding
from tuberculosis of the stomach-a case report. Ann Acad
Med 23:903-906
Goldberg BB, Goodman GA, Clearfield HR (1970) Evaluation
of ascites by ultrasound. Radiology 96:15
Gordin FM et al (1996) The impact of human immunodefi-
ciency virus infection on drug-resistant tuberculosis. Am J
Respir Crit Care Med 154:1478
Ha HK et al (1999) Intestinal tuberculosis with abdominal
complications: radiologic and pathologic features. Abdom
Imaging 24:32
Harland RW, Varkey B (1992) Anal tuberculosis: report of
two cases and literature review. Am J Gastroenterol 87:
1488-1491
Herman P et al (1995) Nodular form oflocal hepatic tubercu-
losis: case report. J Trop Med Hyg 98:141-142
Hickey N et al (1999) Acute hepatobiliary tuberculosis: a
report of two cases and a review of the literature. Eur
Radiol 9:886-889
Hillebrand DJ et al (1996) Ascitic fluid adenosine deaminase
insensitivity in detecting tuberculous peritonitis. Hepatol-
ogy 24:1408
Hoefs JC (1990) Diagnostic paracentesis: A potent clinical tool.
Gastroenterology 98:230
Holley HL, Piraino BM (1990) Complications of peritoneal
dialysis: Diagnosis and management. Semin Dial 3:245
Horvath KD, Whelan RL (1998) Intestinal tuberculosis: return
of an old disease. Am J Gastroenterol 93:692
Huebner RE, Schein MF, Bass JB (1993) The tuberculin skin
test. Clin Infect Dis 17:968
Jones B et al (1993) Relationship of the manifestations of
tuberculosis to CD4 cell counts in patients with human
immunodeficiency virus infection. Am Rev Respir Dis
148:1292
Jost KC et al (1995) Identification of Mycobacterium tubercu-
losis and M. avium complex directly from smear-positive
sputum specimens and BACTEC 12B cultures by high-per-
formance liquid chromatography with fluorescence detec-
tion and computer-driven pattern recognition models. J
Clin Microbiol 33:1270
Kalvaria I, Kottler RE, Marks IN (1988) The role of colonos-
copy in the diagnosis of tuberculosis. J Clin Gastroenterol
10:516
Kapoor VK (1998) Abdominal tuberculosis. Postgrad Med J
74:459
Kashima K et al (1995) Detection of anti-cord factor antibod-
ies in intestinal tuberculosis for its differential diagnosis
from Crohn's disease and ulcerative colitis. Dig Dis Sci 40:
2630
Kim KM et al (1998) Intestinal tuberculosis: clinicopathologic
analysis and diagnosis by endoscopic biopsy. Am J Gastro-
enterol 93:606
Abdominal Tuberculosis
Kochhar R et al (1991) Colonoscopic fine needle aspiration
cytology in the diagnosis of ileocecal tuberculosis. Am J
Gastroenterol 86: 102
Lam MF, Tang SC, Lai KN (2000) Tuberculous peritonitis in
patients on continuous ambulatory peritoneal dialysis
(editorial). Int J Artif Org 23:154
Levine A (1968) Needle biopsy diagnosis of tuberculous peri-
tonitis. Am Rev Respir Dis 97:889-894
Lin OS et al (1999) Isolated gastric tuberculosis of the cardia.
J Gastroenterol Hepatol 14:258-261
Lucas SB (1994) Other viral and infectious diseases and
HIV-related liver disease. In: MacSween RNM, Anthony
PP, Scheur PJ, Burt AD, Portmenn B (eds) Pathology of
the liver, 3rd edn. Churchill Livingstone, Edinburgh, pp
269-316
Lui SL et al (1996) Optimal treatment and long-term outcome
of tuberculous peritonitis complicating continuous ambu-
latory peritoneal dialysis. Am J Kidney Dis 28:747
Mann CV, Russell RCG, Wilhams NS (1997) Barley and Love's
Short practice of surgery, 22nd edn. Chapman and Hall, pp
708,728,772-773,777,799
Manohar A, Simjee AA, Pettengill KE (1990) Symptoms and
investigative findings in 145 patients with tuberculous
peritonitis diagnosed by peritoneoscopy and biopsy over
a five year period. Gut 31:1130
Marinez-Vazquez JM et al (1986) Adenosine deaminase activity
in the diagnosis of tuberculous peritonitis. Gut 27:1049
Markowitz N et al (1993) Tuberculin and anergy testing in
HIV-seropositive and HIV-seronegative persons. Ann
Intern Med 119:185
Marshall JB (1993) Tuberculosis of the gastrointestinal tract
and peritoneum. Am J Gastroenterol 88:989
Mauer K, Manzione NC (1988) Usefulness of the serum-ascites
albumin gradient in separating transudative from exuda-
tive ascites: another look. Dig Dis Sci 33:1208
McKenna MT, McCray E, Onorato I (1995) The epidemiology
of tuberculosis among foreign-born persons in the United
States, 1986 to 1993. N Engl J Med 332:1071
Misra SP et al (1999) Colonic tuberculosis: clinical features,
endoscopic appearance and management. J Gastroenterol
Hepatol 14:723
Nardell E et al (1986) Exogenous reinfection with tuberculosis
in a shelter for the homeless. N Engl J Med 315:1570
O'Reilly L, Daborn C (1995) The epidemiology of Mycobacte-
rium bovis infections in animals and man: a review. Tuberc
Lung Dis 76 [Suppll]:1
Pettersson T, Ojala K, Weber TH (1984) Adenosine deaminase in
the diagnosis of pleural effusions. Acta Med Scand 215:299
Pfyffer GE et al (1996) Diagnostic performance of amplified
Mycobacterium tuberculosis direct test with cerebrospinal
fluid, other non-respiratory, and respiratory specimens. J
Clin Microbiol 34:834
Pouchot J et al (1997) Reliability of tuberculin skin test mea-
surement. Ann Intern Med 126:210
Pulimood AB et al (1999) Endoscopic mucosal biopsies are
useful in distinguishing granulomatous colitis due to
Crohn's disease from tuberculosis. Gut 45:537
Quantrill SJ, Archer GJ, Hale RJ (1996) Gastric tuberculosis
presenting with massive hematemesis in association with
acute myeloid leukemia. Am J GastroenteroI91:1259-1260
Rahmatulla RH et al (2001) Tuberculous liver abscess: a case
report and review of literature. Eur J Gastroenterol Hepatol
13:437-440
677
Raskin MM (1976) Some specific radiological findings and
consideration oflinitis plastica of the gastrointestinal tract.
CRC Crit Rev Clin Radiol Nucl Med 8:87-106
Rathnaraj S, Singh SK, Verghese M (1997) Gastric tuberculosis
presenting with hematemesis. Indian J Gastroenterol 16:
110-111
Raviglione MC, O'Brien RJ (2001) Tuberculosis. In: Braunwald
E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL
(eds) Harrison's Principles of internal medicine, 15th edn.
McGraw-Hill, New York, pp 1024-1035
Rezeiq MA et al (1998) Pancreatic tuberculosis mimicking
pancreatic carcinoma: four case reports and review of the
literature. Dig Dis Sci 43:329-331
Rich JA et al (1996) Polymerase chain reaction detection of
Mycobacterium tuberculosis in formalin-fixed tissue. Am
J Respir Crit Care Med 153:1419
Runyon BA (1986) Paracentesis of ascitic fluid: a safe proce-
dure. Arch Intern Med 146:2259
Runyon BA et al (1992) The serum-ascites albumin gradient is
superior to the exudate-transudate concept in the differen-
tial diagnosis of ascites. Ann Intern Med 117:215
Runyon BA Ascites and spontaneous bacterial peritonitis. In:
Feldman M, Sleisenger MH, Scharschrnidt BF (eds) Gastro-
intestinal and liver disease; pathophysiology, diagnosis and
management, 6th edn. Saunders, Philadelphia, pp 1310-1333
Schluger NW et al (1994) Clinical utility of the polymerase
chain reaction in the diagnosis of infections due to Myco-
bacterium tuberculosis. Chest 105:1116
Selwyn PA et al (1989) A prospective study of the risk of
tuberculosis among intravenous drug users with human
immunodeficiency virus infection. N Engl J Med 320:545
Shah S et al (1992) Colonoscopic study of 50 patients with
colonic tuberculosis. Gut 33:347
Shah S et al (1998) Rapid diagnosis of tuberculosis in vari-
ous biopsy and body fluid specimens by the AMPLICOR
Mycobacterium tuberculosis polymerase chain reaction
test. Chest 113:1190
Shakil AO et al (1996) Diagnostic features of tuberculous peri-
tonitis and presence of chronic liver disease: a case control
study. Am J Med 100: 179
Shukla AS et al (1982) Peritoneal biopsy for diagnosis of
abdominal tuberculosis. Postgrad Med J 58:226-228
Simon VA, Hale Y, Taylor A (1993) Identification of Mycobac-
terium tuberculosis and related organisms by HPLC with
fluorescence detection. LCGC-Magazine of Separation
Science 11:144
Singh MM, Bhargava AN, Jain KP (1969) Tuberculous peri-
tonitis: evaluation of pathogenic mechanisms, diagnostic
procedures and therapeutic measures. N Engl J Med 20:
1091-1094
Singh V et al (1996) Clinicocolonoscopic profile of colonic
tuberculosis. Am J Gastroenterol 91:565
Small PM et al (1991) Treatment of tuberculosis in patients
with advanced human immunodeficiency virus infection.
N Engl J Med 324:289
Small PM et al (1993) Exogenous reinfection with multidrug-
resistant mycobacterium tuberculosis in patients with
advanced HIV infection. N Engl J Med 328:1137
Small PM et al (1994) The epidemiology of tuberculosis in San
Francisco: a population-based study using conventional
and molecular methods. N Engl J Med 330:1703
Suri S, Gupta S, Suri, R (1999) Computed tomography in
abdominal tuberculosis. Br J Radiol 72:92-98
678
Sutton D (1990) A textbook of radiology and imaging. 6th edn.
Churchill Livingstone, Edinburgh, p 808
Talwani R, Horvath JA (2000) Tuberculous peritonitis in
patients undergoing continuous ambulatory peritoneal
dialysis and review. Clin Infect Dis 31:70-75
Tandon HD, Prakash A (1972) Pathology of intestinal tuberculo-
sis and its distinction from crohn's disease. Gut 13:260-269
Tassios P et al (1995) Tuberculous esophagitis: report of a case
and review of modern approaches to diagnosis and treat-
ment. Hepatogastroenterology 42: 185-188
Tevere VJ et al (1996) Detection of Mycobacterium tuberculo-
sis by PCR amplification with pan-Mycobacterium prim-
ers and hybridization to an M tuberculosis-specific probe.
J Clin Microbiol34:918
M. S. Khuroo and N. S. Khuroo
Vlaspolder F, Singer P, Roggeveen C (1995) Diagnostic value
of an amplification method (Gen-Probe) compared with
that of culture for the diagnosis of tuberculosis. J Clin
Microbiol 33:2699
Wobeser WL et al (1996) Evaluation of roche amplicor PCR
assay for Mycobacterium tuberculosis. J Clin Microbiol
34:134
World Health Organization (1996) Groups at risk: WHO report
on the tuberculosis epidemic. World Health Organization,
Geneva
Yajko DM et al (1995) Quantitative culture of Mycobacterium
tuberculosis from clinical sputum specimens and dilution
endpoint of its detection by the AMPLICOR PCR assay. J
Clin Microbiol 33:1944
38 Imaging of Gastrointestinal Tuberculosis
MONA S. AL SHAHED and MOHAMMED ABD EL BAGI

CONTENTS
38.1
38.1 Introduction 679 Introduction
38.1.1 Scope 679
38.1.2 Overview 680 38.1.1
38.1.3 Pathogenesis 680 Scope
38.1.4 Determinants of Imaging Strategies 680
38.1.4.1 Status of Clinical Doubts or Certainty 680
38.1.4.2 Population Status 680 This chapter addresses the imaging of abdominal
38.1.4.3 Organ-specific Localization 681 tuberculosis, which specifically encompass the gas-
38.1.5 Clinical Presentations 682 trointestinal tract, the peritoneum and its reflections,
38.2 Imaging Techniques 682 the lymphatic system, and the solid visceral organs
38.2.1 Plain Radiographs 682
38.2.2 Barium Studies 682
(Akhan and Pringot 2002). Imaging of the genito-
38.2.3 Ultrasound 682 urinary system organs is discussed elsewhere. The
38.2.4 Computed Tomography 682 retroperitoneum is unavoidable in this review, but no
38.2.5 Magnetic Resonance Imaging 682 dedicated description is provided.
38.2.6 Radioisotope Scans 683
38.3 Radiologic Manifestations
of Gastrointestinal TB 683
38.3.1 Alimentary Tract Tuberculosis 683 38.1.2
38.3.1.1 TB of Esophagus 683 Overview
38.3.1.2 Tuberculosis of the Stomach 683
38.3.1.3 TB of the Duodenum 683 Abdominal tuberculosis (TB) remains common in
38.3.1.4 TB of the Jejunum and Ileum 684
38.3.1.5 Ileocecal TB 684
developing countries (MacGregor 1993; Aston 1990;
38.3.1.6 TB of the Appendix 687 Rohweddler 1989; Suri et aI.1999), and there has been
38.3.1.7 Tuberculous Colitis 687 a recent resurgence in developed counties (Marshall
38.3.1.8 Anal TB 688 1993). TB is probably as old as humanity itself (Mac-
38.3.2 Peritoneal TB 688 Gregor 1993). Abdominal TB has long been known in
38.3.3 TB Lymphadenitis 690
38.3.4 TB of the Parenchymal Organs 691
history. Correct and early diagnosis of abdominal TB is
38.3.4.1 TB of the Liver and Spleen 691 crucial because untreated disease carries a 50% mortal-
38.3.4.2 Pancreatic TB 693 ity rate. Unfortunately, sometimes the disease may not
38.4 Imaging of Complications be diagnosed before the patient's death (Lingenfelser
of Gastrointestinal TB 694 et al. 1993; Bankier et al. 1995; Hulnick et al. 1985).
38.5 Mimics and Differential Diagnosis 694
38.6 Summary 696
However, once diagnosed, it is a curable disease with
38.7 Algorithms 697 a favorable prognosis (Bernhard et al. 2000). The
References 697 problem of multidrug-resistant strains is discussed
elsewhere. In spite of the advances in imaging modal-
ities, the diagnosis of abdominal TB remains a real
challenge even to the experienced physician because
many patients present with vague, nonspecific signs or
M. AL SHAHED, MBBS, FRCR symptoms (Sheikh et al. 1995; Haddad et al. 1987).
Senior Consultant Radiologist, Department of Radiology, Riyadh Some unfounded concepts have led physicians
Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi
to overlook the diagnosis of abdominal TB. Such
Arabia
M. ABD EL BAGI, MB BCh, DMRD, FSRRCSI concepts include the assumption that abdominal TB
Senior Consultant, Department of Radiology, Riyadh Armed is rare and a disease of the poor or that it is always
Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi Arabia associated with pulmonary TB (AI Quorain et al.

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
680
1993). Clinical tests are often inconclusive or mis-
leading (Bankier et al. 1995). Intestinal TB is diag-
nosed when histologic tests reveal caseating granu-
lomas or acid-fast bacilli. However, the sensitivity for
revealing these granulomas with caseating necrosis
or acid-fast bacilli is low at approximately 32% and
50%, respectively, by histologic tests (Kim et al.1998).
This augments the role of imaging in diagnosis. Even
laparoscopy may be resorted to for diagnosis, which
renders the clinical examination an invasive expe-
rience for the patient (Lam et al. 1999). Growth of
tuberculosis in cultures of tissue or ascites fluid takes
a long time. Clinicians may wish to make an earlier
decision depending on clinicoradiologic correlation.
A clinical diagnosis of intestinal TB is sometimes
made by therapeutic trial of antituberculous treat-
ment, especially in endemic areas (Park et al. 2000).
This does not exclude the contribution of imaging
because the response to the therapeutic test must be
monitored by clinical and radiologic follow-up.
According to Manahour, abdominal TB develops
in 2% to 4% of patients with pulmonary tuberculosis
(Manahour et al.1990).At our own institution, Karawi
and colleagues reported an incidence of 130 cases of
alimentary tract tuberculosis (16%) out of 820 patients
with TB seen over a period of 8 years (AI Karawi et al.
1995). Incidence of gastrointestinal TB is higher (50%)
in patients with AIDS (Chaisson and Sul Kin 1989).
Abdominal TB is the second most common extra-
thoracic site of involvement in the body (AI Karawi et
al. 1995). One of the major historical and diagnostic
problems with abdominal TB is that it is a great mimic
Oadvar et al. 1997; Panton et al. 1985). TB is a differ-
ential diagnosis in a variety of abdominal conditions.
These will be highlighted below.
Due to awareness of the resurgence of the "new
tuberculous disease" (Marshall 1993; Snider and
Roper 1992) the characteristics of abdominal TB have
been well described in different patient populations
(Kim et al. 1998; Marshall 1993; Gilinsky et al. 1983).
38.1.3
Pathogenesis
The pathologic mechanism and route of infection
determine the manifestations of gastrointestinal TB.
Swallowing infected sputum or contagious unpasteur-
ized milk can manifest as ulcerative disease particularly
in the terminal ileum and colon due to the abundance of
lymphatic tissue (Marshall 1993). The stomach esoph-
agus and duodenum are least involved due to rapid
transit and the low pH. There is an apparent affinity of
M. S. Al Shahed and M. Abd EI Bagi
tubercle bacilli to lymphoid tissue, and in most cases CT
scans are positive for lymphadenopathy. Hematogenous
spread can reach any part of the gastrointestinal tract.
The last method of local spread is by direct invasion
of nearby organs either microscopically or by gross
fistulae and sinuses. The gross pathology conforms
easily with radiologic findings in the various types and
presentations of TB including (1) ulcerative disease, (2)
hypertrophic, (3) ulcerohypertrophic (Paustian and
Bockus 1959), and (4) stricture (Bhansali 1977).
38.1.4
Determinants of Imaging Strategies
38.7.4.7
Status of Clinical Doubts or Certainty
The index of clinical suspicion dictates the pathway, the
end point, and the choice of radiologic examinations
(Fig. 38.1). When gastrointestinal TB is suspected, the
whole spectrum of radiographic procedures are acti-
vated in a sequential manner starting with plain films,
ultrasound (US), and computed tomography (CT). This
may end up with image-guided biopsies.When gastroin-
testinal TB is the most probable diagnosis, which entails
the presence of specific clinical signs (e.g., ascites, mass,
or bowel obstruction), a choice of an imaging technique
has to be made accordingly. For lymphadenopathy you
may proceed directly to CT, whereas if there is ascites,
you may start with US. For suspected alimentary tube
lesions, barium studies are necessary although CT may
suffice. When the diagnosis is confirmed by laboratory
tests, or endoscopic or percutaneous biopsies, the aim of
imaging is to assess the extent of disease, which entails
the use of CT. When patients are put on a therapeutic
trial of antituberculous drugs, a baseline investiga-
tion is documented. Treatment response is judged by
a comparative follow-up examination using the same
technique that was used in the baseline study, in cor-
relation with clinical progress.
38.7.4.2
Population Status
The type of community setting dictates the level and
index of clinical suspicion.
The availability of imaging tools in the community
is a determinant of the diagnostic test choice and
preference (Fig.38.2). We do not have experience
with TB in developed countries, nor do we have much
experience with gastrointestinal TB in immunocom-
promised patients. We actually see more fungal infec-
Imaging of Gastrointestinal Tuberculosis 681

Fig.38.1. Gastroduodenal tuberculosis. Adult male patient
presented with gastric outlet obstruction. Barium meal show-
ing widening of duodenal loop, narrowing of the descending
duodenum and thickening of the mucosal fold (arrow). There
is partial obstruction at the pylorobulbar region
Fig.38.2. Post-contrast-enhanced CT of the same patient at
the level of the stomach antrum, demonstrating transmural
thickening at the stomach outlet "hypertrophic nodular form"
(arrow). Note striation of the peritoneal fat (open arrow)

tions in patients who receive intensive therapy with

immunosuppressants. In our clinical settings, patients
who have pulmonary TB are not examined for gastro-
intestinal tuberculosis unless there are grave or overt
abdominal symptoms of a surgical nature. Most of the
abdominal complaints in these patients are usually
attributed to drug intolerance. Patients who are diag-
nosed to have gastrointestinal TB by endoscopic biop-
sies are usually scheduled for CT to assess the extent
of extraluminal disease and to exclude complications.
The last group of patients are those investigated for
obscure symptoms, for example, pyrexia of unknown
origin or suspicion of a malignant disease (Lundstedt
et al. 1996). Because TB is a great mimicker (Jadvar et
al. 1997), the diagnosis might even come as a surprise
Fig.38.3. CT at the level of the third part of the duodenum
from the biopsy results. Ga67 scanning plays a useful showing diffuse wall thickening and mucosal edema secondary
role in the screening of such patients supplemented by to direct infiltration (arrow). Note enlarged lymph node (curved
CT for the suspected site of tracer accumulation. arrow), between the aorta (a) and inferior vena cava (v)

38.1.4.3 ever, confusion with Crohn's disease or malignancy

Organ-specific Localization
As far as initial imaging is concerned, an anatomi-
cal localization of a single or multiple lesions is the
starting point in building up evidence for diagnosis
of TB (Fig. 38.3).
Barium studies are important in patients with
gastrointestinal TB (Palmer et al. 1985). Most proxi-
mal and most distal "hollow viscus" parts of the
alimentary tube are usually amenable to endoscopic
examination by gastroscopy or colonoscopy. How-
is most likely to happen (Marshall 1993). This is also
true with barium studies or a combination of the
two (Kim et al. 1998). If a chest X-ray or CT shows a
concurrently active TB, then the diagnosis is settled
radiographically (Marshall 1993). Parenchymal organ
or abdominal wall TB are amenable to ultrasound
examination. This can be supplemented by CT.
Mesenteric TB imaging depends on whether we
are dealing with a wet type, for which the ease and
high yield of US is helpful, or the dry type, which
is usually easier to assess by CT. For the mixed or
682
fibrotic type, a combination of the two techniques
is complementary. Lymphadenopathy is more accu-
rately assessed by contrast CT because overlapping
bowel gas hinders a US examination.
38.1.5
Clinical Presentations
Abdominal TB may present in many ways, ranging from
nonspecific abdominal symptoms to acute surgical
emergencies or nonspecific clinical signs (Fig. 38.4).
Patients presenting with any of the spectrum of
abdominal symptoms may end up going through a
whole series of tests before a diagnosis is reached.
Subacute intestinal obstructions are more common
than acute obstruction, which gives time for proper
contrast studies by barium or CT. Abdominal masses
are first examined by US to see if solid or cystic, but
CT remains the mainstay for characterization and
depiction of the extent of the disease.
38.2
Imaging Techniques
38.2.1
Plain Radiographs
Plain abdominal radiographs are usually considered
unhelpful because they are less sensitive compared
with other imaging modalities (Denton and Hossain
Fig. 38.4. TB jejunum. CT at the level of the superior mesen-
teric artery origin. Note the diffuse wall thickening of jejunal
loops with marked mucosal edema (arrows)
M. S. Al Shahed and M. Abd £1 Bagi
1993). Ascites may not be large enough to be detect-
able by radiographs. Calcified granulomas in solid
visceral organs or lymph nodes are fairly specific but
not pathognomonic. Where basic radiographic set-
ups are barely available, as in the peripheral clinics
of deprived communities, and TB is prevalent, plain
radiographs can be the first line of investigation.
38.2.2
Barium Studies
These are useful in patients with gastrointestinal TB
(Palmer et al. 1985). Fluoroscopic barium studies
are available in most developing countries. Specific
radiologic signs in barium studies have been well
described (Kolawole and Lewis 1975; Werbelloff et
al. 1973). These are, however, not diagnostic on their
own. A disadvantage of barium studies is the inability
to assess extraluminal disease (Hulnick et al. 1985).
Filling defects, masses, ulcerations, strictures, and
fistulous tracts can be well outlined by barium meal,
swallow, follow-through, or enema. At times, small
bowel enemas may be resorted to.
38.2.3
Ultrasound
As a simple and quick examination, ultrasound is
usually recommended as the first examination in
abdominal TB, particularly if fluid or a mass is sus-
pected (Gompels and Darlington 1978). Presence of
ascites enhances the ultrasound examination. Ultra-
sound is the initial screening test for solid organ
involvement.
38.2.4
Computed Tomography
CT is the most sensitive and specific tool for detecting
a wide spectrum of lesions of abdominal TB (Denton
and Hussain 1993). CT is essential in many situations
where endoscopic studies and or barium examina-
tions are nonspecific (Makanjoula 1998).
38.2.5
Magnetic Resonance Imaging
The use of MRI in diagnosis of gastrointestinal TB is
limited. It is particularly useful in solid organ disease
Imaging of Gastrointestinal Tuberculosis 683

(Mercusot et al. 1995) and when multiple organs are presence of a preexisting gastric mucosal ulceration
involved especially the combination of abdominal or tumor paves the way for TB infection (Denton
disease with spinal lesions (See Fig. 38.31b). and Hossain 1993; Al Hadeedi et al. 1990). Radio-
logic features of gastric TB are nonspecific (Akhan
and Pringot 2002). Ulcerations are the most common
38.2.6 findings. Gastric TB may present with signs of gas-
Radioisotope Scans tric outlet obstruction due to inflammatory changes,
fibrosis, or enlarged lymph nodes (Thoeni and
Gallium 67 citrate is considered the best radiophar- Margulis 1979). The hypertrophic form can mimic
maceutical for imaging of mycobacterium infections malignancy. Gastric TB should be differentiated
(McAfee 1996; Yang et al.1992). However, the specific- from gastric carcinoma, non-Hodgkin's lymphoma,
ity of gallium scintigraphy is low, although it is sensi- leiomyoma, Kaposi's sarcoma, gastritis, syphilis, or
tive for occult TB and should be followed by CT or sarcoidosis (Denton and Hossain 1993; Jadvar et al.
US. Gallium 67 citrate scanning is useful in obscure 1997). Exclusion of Crohn's disease from ulcerative
presentations like pyrexia of unknown origin (PUO), TB needs scrutiny and clinical correlation, with
unexplained weight loss, and bone disease. preference for TB where the disease is endemic (AI
Karawi et al. 1995). Rarely, gastric TB can present as
linitis plastica, which can lead to confusion with scir-
rhous carcinoma, sarcoidosis, syphilis, lymphoma,
38.3 radiation injury, or corrosive ingestion (Jadvar et
Radiologic Manifestations al.1997).
of Gastrointestinal 18 Sinus formation and fistula are rare in gastric TB
(Thoeni and Margulis 1979; Denton and Hossain
38.3.1 1993). The chest radiograph can be normal in over
Alimentary Tract Tuberculosis 50% of cases. Gastric TB is rare (about 2%) in patients
with pulmonary disease (Tromba et al. 1991). This is
38.3.7.7 due to the relatively low pH of gastric juice and the
TB of Esophagus rapid gastric emptying. Barium studies can demon-
strate ulcers, masses, or outlet obstruction (Fig. 38.1).
Primary esophageal TB is rare even in endemic US or CT can demonstrate other visceral lesions,
areas (Marshall 1993). However, it has recently been lymphadenopathy, or ascites. Although the diagno-
reported in AIDS patients (de Silva et al. 1990). sis can be confirmed by endoscopic biopsy, gastric
Esophageal TB is usually secondary to compres- TB is often not suspected until the time of surgery
sion or fistulation by mediastinal lymph nodes (Mc (Tromba et al.1991). Antropyloric disease can extend
Namara et al. 1987). Such patients complain of pain- into the duodenum and vice versa (Fig. 38.2).
ful dysphagia (Akhan and Pringot 2002).
Barium swallow can readily demonstrate esopha- 38.3.7.3
geal narrowing, ulceration, or fistulae. CT can depict TB of the Duodenum
the extraluminallesion and the extent of the disease
in the mediastinum and lungs (Willifort et al. 1983; Duodenal TB is rare, occurring in about 1-3% of
Dantew et al. 1987). The reasons for the rarity of patients suffering gastrointestinal TB (Lundstedt et
esophageal TB have previously been described, al.1996). Barium studies may demonstrate duodenal
including the lack of stasis due to the rapid transit of narrowing, dilatation, or distortion (Fig. 38.1). Nar-
food and the protective salivary coat over the strati- rowing is most common at the third or fourth part of
fied epithelium (McGuinness 2000). the duodenum (Akhan and Pringot 2002). The pylo-
robulbar region can be deformed in continuation of
38.3.7.2 the duodenal disease (Fig. 38.2). Thickening of the
Tuberculosis of the Stomach overlying mesenteric root can be demonstrated by
US or CT (Denton and Hossain 1993; Hulnick et al.
Gastric TB is also rare (Tromba et al.1991), though a 1985). Duodenal obstruction is a likely presentation
little more common than that of esophagus (Marshall (Bhansali 1977; Gupta et al. 1988).
1993). The antropyloric region and distal body are Duodenocolic fistula can occur but duodenoaor-
the sites usually involved (Leder and Low 1995). The tic fistula is exceedingly rare (Edie and Pollack 1968).
684
Mesenteric adenopathy at the root of the mesentery
can cause compression or invasion of the third and
fourth parts of the duodenum. CT is an excellent
examination for these lymph nodes (Fig. 38.3).
38.3.1.4
T8 of the Jejunum and Ileum
The jejunum and ileum are the second most common
GIT site after ileocecal TB (Marshall 1993). It is usu-
ally associated with peritonitis (Leder and Low
1995). Isolated jejunal stricture can occur. Diffuse
jejunal wall thickening is readily demonstrated by
CT (Fig. 38.4). Barium studies can demonstrate sites
of strictures, separation of bowel loops, bowel wall
thickening, or fistulae (Fig. 38.5).
Ulcerations can sometimes be detected in barium
studies (Pringot et al. 1984). Skip lesions with
intervening normal or dilated bowel segments are
common. Internal fistulae (Fig. 38.5) are a cause of
malabsorption (Bhansali 1977).
38.3.1.5
Ileocecal T8
Ileocecal involvement is the most common presenta-
tion of gastrointestinal TB occurring in as much as
80-90% of cases (Jadvar et al.1997; Makanjuola 1998;
Fig. 38.5. Multiple skip lesions of tuberculous enteritis: Barium
follow-through demonstrating diffuse jejunal wall thickening
causing separation of bowel loops (small arrows). Note areas
of circumferential luminal narrowing (curved arrow). Also
note an unusual barium location delineating a fistula tract
(long arrow). The terminal ileum is incompetent, fixed, and
irregular (arrowheads). The cecum shows multiple polypoid
nodules due to small granulomas (medium size arrow)
M. S. Al Shahed and M. Abd EI Bagi
Balthazar et al. 1990) (Fig. 38.6). This is attributed to
the abundance of lymphatic tissue, as well as the rela-
tive stasis (Leder and Low 1995). Early signs of the
disease include mucosal edema, spasm, and hyper-
peristalsis (Park et al. 2000). This stage can be missed,
especially if there is no associated ascites or obvious
lymphadenopathy. We have previously reported this
stage of "dry type" ileocecal TB, which was missed
in barium studies and even CT (Abd El Bagi and Al
Karawi 1997) (Fig. 38.7). Ultrasound is a useful test
because it allows simultaneous clinical examination
by the radiologist for localization of any tender spot,
which gives a clue to the site to be examined critically
by ultrasound (Fig. 38.8).
Ileocecal TB has characteristic signs, which can be
demonstrated by barium studies, US, or CT (Jadvar
1997; Balthazar et al. 1990; Hulnick 1985). The ter-
minal ileum can be thickened (Fig. 38.8) or patulous
(as shown in Fig. 38.7) (Jadvar 1996). At a later stage
the terminal ileum becomes irregular and immobile.
Deep ulcerations can be demonstrated, which could
cause confusion with Crohn's disease. However, TB
ulcers tend to be larger than those of Crohn's dis-
ease. The confusion can be exacerbated when there
is a fistula or skip lesions (Figs. 38.5, 38.6). Eventually
the terminal ileum becomes narrowed with thick
walls, "Fleischner's sign". The cecum becomes coned
and contracted on barium examinations, "Stierlin's
Fig. 38.6. TB ileum, cecum, terminal ileum, and appendix. This
image demonstrates fistulous communication with an abscess
cavity in the right iliac fossa as a result of perforated appen-
dicular abscess (small arrows). Note tethering of bowel loops
(open arrows), nodularity of the cecum (medium-size arrows)
and the fistula tract (long arrow)
Imaging of Gastrointestinal Tuberculosis
Fig. 38.7. Ileocecal TB missed on barium follow-through. Note
the cut-off of the lower pole of the cecum (arrow), which was
considered due to fecal residue. The terminal ileum was con-
sidered normal. A patent terminal ileum does not exclude
presence of TB
Fig. 38.8. Transabdominal US of the right iliac fossa showing
edematous waIls of a patent terminal ileum (arrowheads) and
a contracted lower pole of the cecum (arrow). No ascites was
present: "dry type TB"
sign" (Leder and Low 1995) (Fig. 38.9). US can also
show the cecal wall thickening (Lee et al. 1993) see
(Fig. 38.11). Cecal wall thickening can be dismissed
on CT as incomplete filling of the lower pole due to
colonic residue (Fig. 38.10). The presence of associ-
ated signs like lymphadenopathy should raise the
alarm (Fig. 38.10).
685
Fig. 38.9. Ileocecal tuberculosis: barium enema showing ulcer-
ation and narrowing of the terminal ileum with marked thick-
ening of the ileocecal valve (Fleischner's sign) (arrowheads).
The cecum is shrunken, stenosed, and classically conical in
shape (Stierlin's sign) (open arrow). Note submucosal polyp-
oid pattern (long arrows) and deep ulcers (small arrows)
Fig. 38.10. CT of the cecum showing incomplete filling of the
cecum and waIl thickening (large arrow). This was initially
considered normal. Note presence of mesenteric lymph node
enlargement (small arrow)
Contraction of the cecum may be complete, mim-
icking a lower pole mass (Fig. 38.12). The lower pole of
the cecum may be amputated by filling defect or mass
(Fig. 38.7). Marked cecal involvement is said to be more
common in TB than in Crohn's disease (AI Karawi
1995). A combination of Crohn's disease and TB in the
same patient has been reported (AI Karawi et al.1991).
 686 M. S. Al Shahed and M. Abd El Bagi

Fig. 38.12. CT of cecal TB presenting as a small mass. Con-

tracted cecum (c) mimicking a mass. Note enhancing mesoco-
lon (small arrows). Presence of a patent terminal ileum (open
arrow) does not exclude presence of TB
Fig. 38.11. US for TB of the cecum. Contracted thick coned
cecum

Endoscopy and endoscopic biopsy are popular
methods for diagnosis of TB, but double-contrast
barium enema can provide detailed information
(Ferentzi et al. 1988; Han et al. 1996). Radiologic
features of TB can be similar to Crohn's disease, lym-
phoma, amebiasis, carcinoma, and even sarcoidosis
(Thoeni and Margulis 1979). Crohn's disease tends to
produce a more uniform pattern than the irregular
changes of TB (Fig. 38.13). These can affect a longer
segment. Engorgement of the mesenteric vessels can
give the peculiar "comb sign" (Fig. 38.14). On CT,
mural stratification with symmetric, concentric wall
thickening (Fig.38.14) is characteristic of Crohn's
disease (Meyers and McGuire 1995; Makanjuola
1998). Amebiasis does not usually involve the small
bowel (Akhan and Pringot 2002). Cecal carcinoma
is always limited by the ileocecal valve (Reeder.and
Palmer 1994). The ambiguous situation of coexist-
ing adenocarcinoma and TB of the cecum has been
reported (Isaacs and Zissis 1997).

a b

Fig. 38.13. a Barium follow-through showing extensive uniform changes of Crohn's disease involving a very long segment of the
ileum (arrowheads). b Magnified view of the terminal ileum. Note cobblestone edematous mucosa (large arrowheads) and the
rose thorn ulcerations (small arrowheads). Compare the depth of the ulcers with the deeper ulcers in Fig. 38.9
 Imaging of Gastrointestinal Tuberculosis 687

a b

Fig. 38.14. a CT of the terminal ileum showing extensive long segment involvement with Crohn's disease. Note the prominent
mesenteric vasculature producing the "comb sign" (arrowheads). b Spiral CT of the terminal ileum showing characteristics of
Crohn's disease. Note the transmural layering of "stratification" sign (open arrow). Note also the enlarged, beaded appearance
of the vasa recta (small arrows)

38.3.1.6
TB of the Appendix

The appendix was reported to be involved in 5%

patients in one series (Al Karawi et al. 1995). A
lower incidence of 0.11% has been reported for
nonendemic countries (Jaffe 1951). Involvement of
the appendix is usually due to direct extension from
the more common ileocecal tuberculosis (Singh et
al. 1987). Tuberculosis limited to the appendix is rare
and has no characteristic features. It can be mistaken
for appendicitis or pelvic inflammatory disease, as it
can perforate spreading the infection into the peri-
toneum (Reeder and Palmer 1994). An abscess may
form (Fig. 38.6) with recurrent pain of mild or severe
degree (Marshall 1993).

38.3.1.7
Tuberculous Colitis

Infection of the colon without involvement of the
small bowel is reported in 9% of gastrointestinal TB
cases (Thoeni and Margulis 1979). Colonic tuberculo-
sis can take one of several forms including segmental
ulcers, inflammatory strictures, hypertrophic lesions
resembling polyps and masses (Chawla et al. 1971;
Stock and Li 1964; Peh 1988).
The early features of tuberculous colitis include
spasm, hypersecretion, increased motility, lymphoid
hyperplasia, thickened folds, and shallow ulcers (Park
et al. 2000). Small aphthous ulcers are considered spe-
cific of Crohn's disease but has also been described in
TB (Shah et al. 1992). Most of the radiologic signs of
colonic TB were described in single-contrast studies.
These include Stierlin's and Fleischner's signs. Lately,
findings with double-contrast enemas (Fig. 38.15)
Fig.38.15. Double-contrast barium enema of a 45-year-old
male adult patient presenting with fever, malaise, night sweats,
and discharging sinus in the anterior abdominal wall. Barium
enema demonstrates a track of barium originating from
descending colon (large arrow). Note focal area of symmetric
annular stenosis, napkin-ring appearance, in the descending
colon above the sinus tract (small arrows)
have been reported (Nakano et al. 1992). Advanced
features of tuberculous colitis in double-contrast
studies include transverse ulcers, nodularity, polyps,
and narrowing. Colonic lymphoid tissue follicles are
oriented transversely, which is why the ulcers fre-
quently adopt this orientation (Carrera et al. 1976).
Rose thorn ulcer can occur and may not be visible on
colonoscopy (Park et al. 2000). Symmetric narrowing
of the colon gives the "napkin ring" sign (Leder and
688 M. S. Al Shahed and M. Abd EI Bagi

Low 1995) (Fig. 38.15). TB strictures show smooth

transition to the nonaffected area with preservation of
the haustral pattern (Park et al. 2000). Sometimes TB
can simulate colonic tumors (Carrera et al. 1976). The
changes of tuberculous colitis can be seen with other
causes of inflammatory bowel disease. More extensive
involvement can also be demonstrated by CT including
edema, stricture formation, and other associated con-
ditions like ascites (Fig. 38.16). Ha and colleagues used
a combination of CT signs to differentiate TB peritoni-
tis from peritoneal cardiomatosis "omental cakes" (Ha
et al. 1996). Presence of little intervening thick ascites
fluid can give the appearance of "sliced bread" in US.
The walls of pocketed ascitic fluid or the lymph nodes
may calcify (Fig. 38.20). For everyday practice, there is Fig.38.16. TB of the rectosigmoid region. CT of the pelvis
a considerable overlap between the different types of demonstrating involvement of the rectosigmoid. Note sig-
nificant narrowing (curved arrow) and edema of the wall
peritoneal involvement with TB (Suri et al.1999). Com-
(arrowheads). Also note creeping peritoneal fat to the area of
monlya wide spectrum of multiple organ involvement infected bowel (small arrow). This also seen in Crohn's disease.
is encountered in each case (Fig. 38.6). Ascites is present (a)

38.3.1.8
Ana/TB

TB of the anal canal is rare, representing less than

4% of abdominal TB (AI Karawi et al. 1995). How-
ever, higher incidence has been reported from India
(Shukla et al. 1988). The disease may present with
ulcers, fissures, abscess, hypertrophic lesions, or fistula
(Marshall 1993).Anal TB can be confused with Crohn's
disease due to the common belief that the latter is more
common at this specific site. Biopsy and laboratory
tests should be obtained, because access is easy.

38.3.2
Peritoneal T8

Fig. 38.17. Wet TB peritonitis (pocketed). CT scan of the mid-

Involvement of the peritoneum is a very frequent find-
ing in abdominal TB and was reported in 30% of the AI
Karawi series and 57% of the Gilinsky series (AI Karawi
et al. 1995; Gilinsky et al. 1983). However, isolated peri-
tonitis is a rare manifestation of TB, occurring in less
than 4% of patients (Leder and Low 1995).
There are three well-recognized types of perito-
nitis TB (Hanson and Hunter 1985): (1) a wet ascitic
type (Epstein and Mann 1982), (2) a dry plastic type
(Abd EI Bagi and Al Karawi 1997), and (3) a fibrotic
fixed type (Suri et al. 1999).
The wet type is the most common type seen in
as much as 90% of cases, with excess high attenu-
ation ascites fluid, which can be free or pocketed
(Fig. 38.17). CT measurement of ascitic fluid density
is not always reliable, however (Bankier et al. 1995).
abdomen showing large, high density ascites localized anteri-
orly, displacing bowel loops posteriorly (a). Aspirate revealed
acid-fast bacilli, and tuberculous peritonitis was diagnosed
Sonographically echogenic debris is seen as fine
strands or particulate matter within the fluid (Lee et
al. 1991). The dry or plastic type is seen in 10% of
cases (Fig. 38.7). It can be missed due to lack of excess
peritoneal fluid, which when present, enhances
both the CT and US examinations (Fig. 38.18). The
fibrotic fixed type is characterized by hypervascu-
lar peritoneum, matting of the loops, and omental
masses referred to as omental cakes (Lee et al. 1991)
(Fig. 38.18). The walls of an encysted ascitic fluid
pocket may calcify (Fig. 38.20).
 Imaging of Gastrointestinal Tuberculosis 689

Fig.38.18. US of the ileocecal region showing dry-type TB

with focal thickening of the medial wall of the cecum (C) and
a trace of fluid (F), behind the terminal ileum (TE)

Fig. 38.19a, b. TB enteritis with lymphadenopathy and omental

cakes. a CT of lower abdomen demonstrating enlarged lymph
nodes (arrow) and thickened wall bowel loops (large arrow-
heads). Note the thickened peritoneum with mottled low-
density masses and nodular soft tissue thickening along with
hypervascularity of the peritoneal surface "pseudomyxoma
peritoni appearance" (small arrowheads). b CT at more caudal
levels demonstrates the pseudomyxoma peritoni appearance
of the involved peritoneum (arrowheads), bowel wall thicken-
ing (arrows), and loculated ascites (a) b

a b

Fig. 38.20a, b. Tuberculous peritonitis (wet type). a Plain abdominal X-ray demonstrates generalized ground-glass density
throughout the abdomen indicating presence of ascites. Note curvilinear calcification (arrowheads) in the upper abdominal
region. b CT scan of the same patient showing lobulated fluid collection surrounded by calcified wall, initially mistaken for
hydatid disease (arrowheads)
690 M. S. Al Shahed and M. Abd EI Bagi

38.3.3
TB Lymphadenitis

Lymph node involvement in gastrointestinal and other

abdominal infections with TB is a common finding. It
has been reported in 33% of cases in one series (Lund-
stedt et al. 1996). With the frequent use of CT, lymph-
adenopathy was found to be the most common finding
in abdominal TB, occurring in two thirds (Suri et al.
1999) of patients. An even higher incidence of 93%
was reported (Demirkazik et al.1996). The mesenteric
and peripancreatic group oflymph nodes are the most
often involved (Suri et al. 1999) (Fig. 38.3).
Various patterns of lymph node involvement have a
been observed in abdominal TB (Pombo et al. 1992).
The most common pattern, in 40-70% of cases, is a
hypodense center of lymph nodes with a peripheral
enhancing rim on CT scanning (Leder and Low 1995)
(Figs. 38.21, 38.30). These can be confluent due to
perinodal inflammation or remain discrete.
The second type is relatively homogenous with a
low density on nonenhanced CT and shows relatively
homogenous enhancing lymph nodes after intrave-
nous-contrast administration. These are usually small,
multiple, and rather discrete (Fig. 38.22). This pattern
is probably a different stage of pathology representing
noncaseating granulomas (Suri et al. 1999). Retro-
peritoneal lymph node involvement was previously b
considered rare in TB. But it can occur as a part of dis-
seminated disease (Fig. 38.21,38.23,38.24). Sometimes Fig. 38.22a, b. Tuberculous adenopathy. a CT of the abdomen
demonstrates multiple small,discrete lymph nodes with homog-
a third pattern of lymph node involvement is seen. enous density (arrows). b Postcontrast CT of abdomen showing
The lymph nodes, may show a heterogeneous central multiple small, discrete homogenously enhancing modes in a
enhancement pattern due to the presence of central, known TB patient after 4 months of treatment (arrow)

Fig. 38.21. Tuberculous adenopathy: CT scan of a patient demon-
strating the characteristic appearance of tuberculous adenopa-
thy manifesting as low attenuation center of caseous necrosis,
surrounded by an enhancing rim reflecting peripheral highly
vascular inflammatory reaction (n)
Fig.38.23. Post-enhanced CT of the abdomen demonstrat-
ing multiple enlarged lymph nodes with nonhomogeneous
ring enhancement centrally, due to presence of necrotic foci
(arrows)
 Imaging of Gastrointestinal Tuberculosis
a
b lobe (arrow)
scattered, necrotic tissue (Fig. 38.23). CT is believed
to be more accurate than US for assessment of lymph
node enlargement in TB (Denton and Hossain 1993).
However, US is usually a first line of investigation
for abdominal conditions. US can detect discrete or
confluent lymph node enlargement (Fig. 38.25). The
application of color Doppler US can demonstrate the
increased perinodal vascularity due to inflammatory
periadenitis (Fig. 38.26), as well as the hyperemia of
the omental layers. This phenomenon is not pathogno-
monic for TB and can occur in other benign or malig-
nant conditions, including metastases, lymphoma,
Whipple's disease, and Crohn's disease (Hulnick et
al. 1985; Bankier et al. 1995). Although abdominal
lymphadenopathy can reach a large size (Fig. 38.26,
38.27), obstruction due to extrinsic compression of
the gastrointestinal tract is unlikely. Actually, intesti-
nal obstruction in TB is commonly due to peritoneal
adhesions (AI Karawi et al. 1995). Different stages of
lymph node involvement can be demonstrated in the
same patient (Fig. 38.28).
691
c
Fig.38.24a-c. Aggressive presentation of abdominal tuber-
culosis. a Post-enhanced CT of the abdomen demonstrates
massive para-aortic adenopathy (n) surrounding the superior
mesenteric artery (s). The inferior vena cava (v) is displaced
by enlarged retrocaval lymph nodes. Note presence of ascites
(small arrows). Also note the enhanced vascularity of the peri-
toneum (p). This appearance is consistent with fibrotic fixed
type of peritonitis. b CT scan of the chest of the same patient
demonstrating enlarged posterior and anterior mediastinal
lymph nodes (arrows). c Chest X-ray of the same patient (b)
showing fibronodular tuberculous lesion in the right upper
38.3.4
TB of the Parenchymal Organs
38.3.4.1
T8 of the Liver and Spleen
Hepatobiliary TB has been reported in 20% of
abdominal TB patients (Lundstedt et al. 1996). As
much as 80-100% of patients having miliary TB
were proven to have hepatosplenic involvement
at autopsy (Thoeni and Margulis 1979). Infection
can reach the liver either by hematogenous spread
of disseminated TB or locally via the portal vein
from gastrointestinal lesions. The disease com-
monly presents as hepatosplenomegaly (Hulnick
et al. 1985; Ramaiya and Walter 1993). There are
three types ofhepatosplenic TB. The miliary spread
is the most common presentation. The liver and
spleen may become extensively infiltrated by small
lesions in the order of 0.5-2 cm. The smaller size
micronodular pattern may be below the resolution
 692 M. S. Al Shahed and M. Abd EI Bagi

a b

Fig. 38.25a, b. Tuberculous adenopathy. a Ultrasound examination of the abdomen showed lobulated hypoechoic homogenous
mass in the right side of the lower abdomen, caused by enlarged matted lymph nodes (between cursers). b CT scan of the same
patient demonstrates characteristic appearance of peripheral ring enhancement of the enlarged lymph nodes, which are totally
matted together (arrowheads)

Fig. 38.26a, b. TB abdominal lymphadenopathy in a 50-year-old male who presented with diarrhea, weight loss, and malaise.
On clinical examination he had a soft "doughy" abdomen and multiple masses were felt. a US examination of the abdomen
showed multiple hypoechoic enlarged lymph nodes with necrotic center and thick wall, matted together in the para-aortic
region (arrows). b On Doppler US these lymph nodes were avascular. Enhanced vascularity of the blood vessels surrounding
the lymph nodes and the adjacent peritoneal/omental vessels was demonstrated (arrows)

of a US scanner and present as a diffusely bright
liver or spleen (Andrew et al. 1982; Jadvar et al.
1997). On CT, a honeycomb pattern (Fig. 38.29)
of uncountable low attenuation lesions with mini-
mal enhancement can be seen (Choi et al. 1989).
A larger macronodular pattern is less commonly
encountered, while medium-size multiple lesions
are easier to discern on US or CT (Fig. 38.30). These
lesions are hypoechoic on US and hypodense on
CT. Lately, increased echogenicity or higher attenu-
ation may be noted, and calcification may be the
end result (Denton and Hussain 1993; Choi et al.
1989). The diffusely nodular TB of the liver and
spleen should be differentiated from metastatic
disease, lymphoma, and other infections particu-
larly fungal infections.
The second form of hepatosplenic TB is the
solitary lesion, which can be a granuloma or a frank
abscess formation (as shown in Fig. 38.31a) (Wilde
and Kueh 1991; Choi et al. 1989). The appearance
of the solitary lesion may show variation of density
according to progression or regression of the dis-
Imaging of Gastrointestinal Tuberculosis
Fig. 38.27. Postenhanced CT examination of the abdomen
showed massive adenopathy with characteristic appearance
of low attenuation center and peripheral rim enhancement
(n). Note small bowel wall thickening (arrowheads), ascites
(a), and thickening of the peritoneum (small arrows)
Fig. 38.29. Hepatosplenic tuberculosis, miliary type: Post-
enhanced CT of the abdomen demonstrating multiple tiny foci
oflow attenuation, widely scattered throughout the spleen and
the liver (arrowheads)
ease. On MRI imaging a solitary lesion in the liver
or spleen demonstrates a hypointense signal on TI-
weighted images and an isointense or hyperintense
nodule with a less intense rim on T2-weighted images
(Mercusot et al. 1995). After IV-contrast adminis-
tration, rim enhancement may be demonstrated
(Fig.38.31b). Hepatosplenic TB can be a part of
disseminated disease involving the spine, retroperi-
toneal structures like the adrenal glands or the psoas
muscle (Fig. 38.31,38.32), or the gastrointestinal tract
(Fig. 38.27).
693
Fig. 38.28. CT shows difference patterns of abdominal tuber-
culous lesions including lymphadenopathy (arrows). Note the
thickened infiltrated peritoneum (large arrowheads), ascites
(a), and the dilated bowel loops with edematous mucosa
(small arrowheads). (Fibrotic-fixed type of peritonitis)
Fig. 38.30. Hepatosplenic tuberculosis, macronodular type:
post-enhanced CT of the abdomen demonstrates splenomegaly
with multiple low attenuation lesions of variable size through-
out its parenchyma (arrowheads). Note associated adenopathy
with necrotic center (arrow)
38.3.4.2
Pancreatic TB
Tuberculosis of the pancreas is rarely reported
(Hulnick et al. 1985; Desai et al. I991). This could be
a part of miliary tuberculosis or a site of reactivated
disease (Desai et al. 2000). The pancreas may be
inseparable from surrounding infected retroperito-
neal lymph nodes (Lundstedt et al. 1996). However,
focal pancreatic lesions in TB can be readily dem-
onstrated by US or CT (Fig. 38.33). Confusion with
 694 M. S. Al Shahed and M. Abd EI Bagi

Fig.38.31a-c. Multisystemic tuberculosis. Male patient pre-

sented with weight loss, back pain, fever, and night sweats.
a US of the spleen showed a well-defined solitary hypoechoic
mass consistent with tuberculous granuloma (arrow) (macro
nodule presentation). b MRI of the spleen: Tl-weighted image
after contrast enhancement demonstrates low signal intensity
mass with rim enhancement (arrow). The examination also
demonstrates tuberculous spondylitis and paraspinal abscess
(large arrowheads). There is extension of the abscess into the
vertebral canal (small arrowheads). c MRI of dorsolumbar
spine of the same patient demonstrates tuberculous spondy-
litis, paraspinal abscess (large arrows) and the extension of the
abscess into the vertebral canal (arrowheads) c

pancreatic carcinoma can occur (Fig. 38.34). There
is a tendency to involvement of the pancreatic head
but the tail can, rarely, be involved (Lundstedt et al.
1996). This could be due to the abundance oflymph
nodes in the region of the pancreatic head. In cases of
pancreatic abscesses, clinical correlation and aspira-
tion are necessary to differentiate between a pyogenic
and tuberculous abscess.
tulous tract (Fig. 38.36). Intestinal complications of
abdominal TB have been well described (Ha et al.
1999). Intestinal obstruction is considered the most
common complication (Makanjuola et al. 1998).
38.5
Mimics and Differential Diagnosis

Despite pattern recognition and imaging charac-

38.4
Imaging of Complications
of Gastrointestinal T8
Cold abscess formation is a common complication of
gastrointestinal TB. The most serious complication
of gastrointestinal TB is fistula formation. These can
be demonstrated by a simple barium follow-through
examination (Fig. 38.5, 38.6, 38.35), but a fistulogram
can delineate the exact length and course of the fis-
teristics, TB is mimicked by and is a mimicker of
many other conditions (Yassawy et al. 1987; Jadvar et
al. 1997). In particular, differentiation from Crohn's
disease is a problem. There are no radiologic or colo-
noscopic features of Crohn's disease that may not be
mimicked by colonic TB (Lingenfelser et al. 1993).
Invasive procedures using imaging-guided aspiration
biopsies are necessary. However, the frequency of pos-
itive acid-fast smears is low (0-45%) (Leder and Low
1995). Measurement of ascitic fluid adenosine deam-
 Imaging of Gastrointestinal Tuberculosis 695

a b

c d

Fig.38.32a-d. CT manifestation of abdominal tuberculosis with multisystemic involvement. a Post-enhanced CT of the upper
abdomen demonstrating ill-defined macronodular hypodense lesions in the liver displacing blood vessels (arrowheads). Note
paraspinal abscess (small arrowheads). Also note tuberculous involvement of the right adrenal gland, which appeared enlarged with
area of caseation (arrow). b CT scan at lower level demonstrates multiple hypodense lesions in the spleen (arrows). Note the left
adrenal mass due to tuberculous involvement (arrowhead) appearing as a solid tumor with no evidence of necrosis. c CT scan at
more caudal level demonstrates cold abscess in the left psoas muscle (arrowheads). d CT of the same patient at more caudal level
demonstrating paraspinal abscess (arrows). Also note thickened bowel wall (small arrowheads) and large necrotic lymph node

Fig. 38.34. CT scan of the same patient demonstrating a homo-

Fig. 38.33. Tuberculosis of the pancreas. US examination dem- geneous mass of low attenuation in the region of the head of
onstrates a well-defined hypoechoic mass in the region of the pancreas (arrows), mistaken for carcinoma. Tuberculosis was
head of pancreas (arrow) diagnosed after surgical intervention
696
Fig. 38.35. Extensive tuberculous peritonitis and enteritis:
small bowel follow-through demonstrates widely separated
bowel loops due to peritoneal lymph adenopathy and due
to direct infiltration to the bowel loops. Opacification of the
rectum (large arrow), has occurred before barium reached the
left side of the colon indicating fistulous communication. The
cecum is conical in shape due to stenosis (open arrow). Note
multiple small bowel loops demonstrating wall edema and
narrowing (small arrows)
ine activity sensitivity above 95% has a specificity of
96-98% (Voigt et al.I989). Adenosine deamine activ-
ity is a quick test, which together with histologic tests
and serum PCR level determination have improved
the diagnosis of abdominal TB (Marshall 1993). Lapa-
roscopy is more invasive than imaging-guided aspira-
tion biopsy but has a high yield of 72-79% (Menzies
et al. 1985). Ha and colleagues described CT criteria
for differentiation between tuberculous peritonitis
and peritoneal carcinomatosis, including mesenteric
modulation, and enhancement of the infiltrated
omentum, omental line and splenic involvement. The
overall sensitivity of CT for prediction of TB perito-
nitis was 69% and for peritoneal carcinomatosis was
91 % (Ha et al.1996). Other conditions like lymphoma
and peritoneal mesothelioma should be excluded
(Akhan and Pringot 2002). Castleman disease can
cause abdominal lymphadenopathy but tends to
cause dense homogenously enhanced large lymph
nodes (Ferreiros et al. 1989). Whipple's disease tends
to cause an exceedingly low attenuation of lymph
nodes due to excessive fat deposition. Calcification
is more commonly seen in old, healed lymph node
infections than in the acute disease. It is important
M. S. Al Shahed and M. Abd El Bagi
Fig.38.36. Fistulogram using a soft catheter through a sinus
tract in the left iliac fossa demonstrates the tract (arrowheads),
which opens into the descending colon (arrow)
to note that a multitude of intra-abdominal lesions is
common in gastrointestinal TB, which helps to clarify
the nature of the disease.
38.6
Summary
Gastrointestinal TB has a wide spectrum of clinical
and radiologic manifestations. There are no pathog-
nomonic radiologic signs. Gastrointestinal TB can
mimic many other diseases. A high index of clini-
cal suspicion is necessary. The radiologic approach
should be systematic and structured in an objective
and practical manner. We propose criteria and path-
way algorithms to determine the imaging strate-
gies, which should be adopted in various clinical
situations and according to the type of patient. This
entails assessing the clinical presentation (Algorithm
1), the level of clinical suspicion ofTB (Algorithm 2),
the population setting (Algorithm 3), and the organ
expected to be involved (Algorithm 4). Accordingly,
the diagnostic pathway and the end point for all
investigations should be decided. Modern serologic
tests are sensitive and provide speedy assistance to a
diagnosis. Image-guided microbiologic or histologic
proof is essential for a definitive diagnosis.
Imaging of Gastrointestinal Tuberculosis
38.7
Algorithms
38.7.1
Algorithm 1. Clinical Presentation of Abdominal
18
38.7.2
Algorithm 2. Degree of Clinical Suspicion of
Abdominal 18
38.7.3
Algorithm 3. Population Setting
38.7.4
Algorithm 4. Organ Involved with 18
References
Abd El Bagi ME, AI Karawi MA (1997) Ultrasound diagnosis
of "Dry Type" ileocecal tuberculosis: a case report. Hepa-
togastroenterology 44:1033-1036
Akhan 0, Pringot J (2002) Imaging of abdominal tuberculosis.
Eur RadioI12:312-323
Al Hadeedi S, Walia HS,AI Sayer HM (1990) Abdominal tuber-
culosis. Can J Surg 33:233-237
AI Karawi MA, Shariz S, Mohamed AE et al (1991) Crohn's
disease associated with Mycobacterium infection. J Indian
Med Assoc 23:87-90
Al Karawi MA, Mohamed AE, Yassawy MI et al (1995) Protean
manifestations of gastrointestinal tuberculosis, report on
130 patients. J Clin Gastroenterol 20:225-232
Al Quorain A, Satti MB, Al Freihi H et al (1993) Abdominal
tuberculosis is Saudi Arabia: a clinicopathological study of
65 cases. Am J GastroenteroI88:75-79
Andrew WK, Thomas RG, Gollach BL (1982) Miliary tuber-
culosis of the liver: another cause of the "bright liver" on
ultrasound examination. South Afr Med J 62:808
Aston NO (1990) Abdominal tuberculosis. Br J Clin Pract 44:
58-61,63
Balthazar EJ, Gordon R, Hulnick D (1990) Ileocaecal tubercu-
losis CT and radiologic evaluation. AJR 154:499-503
Bankier AA, Fleischmann D, Weismayr Met al (1995) Update:
abdominal tuberculosis: unusual findings on CT. Clin
RadioI50:223-228
Bhansali SK (1977) Abdominal tuberculosis. Am J Gastroen-
terol 67:324-337
Bernhard JS, Bhatia G, Knauer CM (2000) Gastrointestinal
tuberculosis: an 18 patients experience and review. J Clin
GastroenteroI4:397-402
Carrera GF, Young S, Lewiki AM (1976) Intestinal tuberculosis.
Gastrointestinal Radiol 1:147
Chaisson RE, Sul Kin G (1989) Tuberculosis and human immu-
nodeficiency virus infection. J Infect Dis 59:96-100
Chawla S, Mukerjee P, Bery K (1971) Segmental tuberculosis of
the colon ca.: a report of ten cases. Clin RadioI22:104-109
697
Choi BI,Im JG,Han Mc et al (1989) Hepatosplenic tuberculosis
with hypersplenism: CT evaluation. Gastrointest Radiol14:
265-267
Dantew B, Frengley D, Wolinsky E et al (1987) Oesophageal
tuberculosis mimicking gastrointestinal malignancy. Rev
Infect Dis 9:140-146
Denton T, Hossain J (1993) A radiological study of abdominal
tuberculosis in a Saudi population with special reference
to ultrasound and computed tomography. Clin Radiol 47:
490-414
Dermirkazik FB,Akhan 0, Ozmen MN et al (1996) US and CT
findings in the diagnosis of tuberculous peritonitis. Acta
RadioI37:517-520
Desai DC, Swaroop VS, Mohandas KM et al (1991) Tuberculo-
sis of the pancreas: report of three cases. Am J Gastroen-
teroI86:761-763
Desai SR, Bhanthunmavin K, Hollands M (2000) Primary
pancreatic tuberculosis: presentation and diagnosis. Aust
NZ Surg 70:141-143
De Silva R, Stoopack PM, Roufman JP (1990) Oesophageal fis-
tulas associated with mycobacterium infection in patients
at high risk for AIDS. Radiology 175:449-443
Edie DGA, Pollack DS (1968) A complicated aorto-duodenal
fistula. Br J Surg 53:314-317
Epstein BM, Mann JH (1982) CT of abdominal tuberculosis.
AJR Am J Roentgenol139:861-866
Ferreiros J, Gomez-Leon N, Mata MI et al (1989) Computed
tomography in abdominal Castleman's disease. J Comput
Assist Tomogr 13:433-436
Ferentzi CV, Sieck JO, Ali MA (1988) Colonoscopic diagnosis
and medical treatment of ten patients with colonic tuber-
culosis. Endoscopy 20:62-65
Gilinsky NH, Marks IW, Kottler RE et al (1983) Abdominal
tuberculosis: a 10 year review S. Afr Med J 64:849-857
Gompels BM, Darlington LG (1978) Ultrasonic diagnosis of
tuberculous peritonitis. BJR 51:1018-1019
Gupta SK, Jaain AK, Agrawal AK et al (1988) Duodenal tuber-
culosis Clin Radiol 39: 159-161
Ha HK, Jung JI, Lee MS et al (1996) CT differentiation of
tuberculous peritonitis and peritoneal carcinomatosis.
AJR 167:743-748
Ha HK, Ko GY, Ya ES et al (1999) Intestinal tuberculosis with
abdominal complications: radiologic and pathologic fea-
tures. Abdom Imaging 24:32-38
Haddad FS, GhossainA, Saway E et al (1987) Abdominal tuber-
culosis. Dis Colon Rectum 30:724-735
Han JK, Kim SH, Choi KY et al (1996) Tuberculous colitis:
findings at double contrast barium enema examinations.
Dis Colon Rectum 39:1204-1209
Hanson RD, Hunter TB (1985) Tuberculous peritonitis CT
appearance. AJR 144:931-932
Hulnick DH, Megibow AJ, Naidich DP et al (1985) Abdominal
tuberculosis: CT evaluation. Radiology 157:199-204
Isaacs P, Zissis M (1997) Colonic tuberculosis and adeno-
carcinoma: an unusual presentation. Eur J Gastroenterol
HepatoI9:913-915
Jafee FA (1951) Tuberculous appendicitis: a clinicopathologic
study of 17 cases. Am Rev Tuberc 64:182-191
Jadvar H, Mindelzun R, Eric W et al (1997) Still the great mim-
icker: abdominal tuberculosis. AJR 168:1455-1460
Kim KM, Lee A, Choi KY et al (1998) Intestinal tuberculosis
clinicopathologic analysis and diagnosis by endoscopic
biopsy. Am J Gastroenterol 93:606-609
698
Kolawole TM, Lewis EA (1975) A radiologic survey of tubercu-
losis of the abdomen (gastrointestinal tract). Am J Roent-
genol Rad Ther Nucl Med 123:348-358
Lam KN, Rajasoorya C, Mak PK et al (1999) Diagnosis of
tuberculous peritonitis. Singapore Med J 9:601-604
Leder RA, Low VHS (1995) Tuberculosis of the abdomen.
Radiol Clin North Am 33:691-705
Lee DH, Lim JH, Ko YT et al (1991) sonographic findings in
tuberculosis peritonitis of wet ascitis type. Clin Radiol 44:
306-310
Lee DH, Ko YT, Yoon Y et al (1993) Sonographic findings in
intestinal tuberculosis. J Ultrasound Med 12:537
Lingenfelser T, Zak J, Marks IN et al (1993) Abdominal tuber-
culosis: still a potentially lethal disease. Am J Gastroenterol
88:744-750
Lundstedt C, Nyman R, Brismar J et al (1996) Imaging of
tuberculosis II. Abdominal manifestations in 112 patients.
Acta Radiol 37:489-495
MacGregor (1993) Tuberculosis from history to current man-
agement. Semin Roentgenol XXVIII(2):101-108
Makanjuola D (1998) Is it Crohn's disease or intestinal tuber-
culosis? CT analysis. Eur J RadioI28:55-61
Makanjuola D, AI Oraini I, Al Rashid R et al (1998) Radiologi-
cal evaluation of complications of intestinal tuberculosis.
Eur J RadioI26:261-268
Manahour A, Simjee AE, Haffejee AA et al (1990) Symptoms
and investigative finding in 145 patients with tuberculous
peritonitis diagnosed by peritoneoscopy and biopsy over a
five year period. Gut 31:1130-1132
Marshall JB (1993) Tuberculosis of the gastrointestinal tract
and periotoneum. Am J Gastroenterol 7:989-999
McAfee JG (1996) The year book of nuclear medicine. Editorial
comment. Mosby, London, PP 94-95
McGuiness F (2000) Tuberculousis of the gastrointestinal tract
and peritoneum: clinical imaging of non-pulmonary tuber-
culosis. Springer, Berlin Heidelberg New York, pp 105-137
McNamara M.,Williams CE, Brown TS et al (1987) Tuberculo-
sis affecting the oesphagus. Clin RadioI8:419-422
Menzies R, Fitzgerald JM, Mulpeter K (1985) Laparoscopic
diagnosis in ascites in Lesotho. Br Med J 291:473-475
Mercusot B, Arrive L, Rottenberg et al (1995) Imaging of
hepatic tuberculoma. J Radiol 76:277-279
Meyers MA, Mc Guire PV (1995) Spiral CT demonstration of
hypervascularity in Crohn's disease: vascular jejunization
in Crohn's disease or the comb sign. Abdo Imaging 20:
327-332
Nakano H, Jaramillo E, Watanabe M (1992) Intestinal tubercu-
losis: findings on double contrast barium enema. Gastroin-
test Radioll08-114
Palmer KR, Patil DH, Basran GS et al (1985) Abdominal
tuberculosis in urban Britain: a common disease. Gut 2:
1296-1305
Panton ON, Sharp R, English RA et al (1985) The great mimic
still at large. Dis Colon Rectum 28:446-4450
Park SJ, Han JK, Kim JS et al (2000) Tuberculous colitis: radio-
logic colonoscopic correlation. AJR 175:121-128
M. S. Al Shahed and M. Abd EI Bagi
Paustian FF, Bockus HL. (1959) So-called primary ulcerohy-
pertrophic ileocaecal tuberculosis. Am J Med 27:509-518
Peh WCG (1988) Filiform polyposis in tuberculosis of the
colon. Clin RadioI39:534-536
Pombo F, Rodriguez E, Mato J et al (1992) Patterns of enhance-
ment of tuberculous lymph nodes demonstrated by com-
puted tomography. Clin RadioI46:13-17 .
Pringot J, Concette L,Ponette E et al (1984) Non stenotic ulcers
of the small bowel. Radiographics 4:357-375
Ramaiya LI, Walter DF (1993) Sonographic features of tuber-
culous peritonitis. Abdom Imaging 18:23
Reeder M, Palmer P (1994) Miscellaneous disorder of the
gastrointestinal tract. In: Freeny Rohweddler JJ (1989)
Abdominal tuberculosis: a disease poised for reappear-
ance. NY State J Med 89:252-254
Shah S, Thomas V, Mathan M et al (1992) Colonoscopic study
of 50 patients with colonic tuberculosis. Gut 33:347-351
Sheikh M,Abu-Zeidan F,AI Hilaly M et al (1995a) Abdominal
tuberculosis: comparison of sonography and computed
tomography. J Ultrasound 23:413-417
Sheikh MY, Yosuf AH, Khalid TR et al (1995b) Co-existent
tuberculosis and carcinoma of the colon. J Pakistan Med
Assoc 45:305-306
Shukla HS, Gupta SC, Singh G et al (1988) Tubercular fistula
in anal. Br J Surg 75:38-39
Singh MK, Arunabh S, Kapoor VK (1987) Tuberculosis of the
appendix: a report of 17 cases and a suggested aetiopatho-
logical classification. Postgrad Med J 63:855-857
Snider DE, Roper WL (1992) The new tuberculosis. N Engl J
Med 326:703-705
Stock F, Li FWP (1964) Granulomas of the large bowel simulat-
ing malignant disease. Br J Surg 51:898-901
Suri R, Gupta S et al (1999) Computed tomography in abdomi-
nal tuberculosis. BJR 72:92-98
Thoeni RF, Margulis AR (1979) Gastrointestinal tuberculosis.
Semin RoentgenoI14:283-284
Tromba JL, Inglese R, Rieders B et al (1991) Primary gastric
tuberculosis presenting as pyloric outlet obstruction. Am J
Gastroenterol 86: 1820-1822
Voigt MD, Kalvaria I, Trey C et al (1989) Diagnostic value of
ascites adenosine deaminase in tuberculous peritonitis.
Lancet 1:751-754
Werbeloff L, Novis BH, Bank S et al (1973) The radiology of
tuberculosis of the gastrointestinal tract. BJR 46:329-336
Wilde CC, Kueh YK (1991) Case report: tuberculous hepatic
and splenic abscess. Clin RadioI43:215-216
Willifort ME, Thompson WM, Hamilton JD et al (1983)
Oesophageal tuberculosis: findings on barium swallow and
computed tomography. Gastroentrol RadioI8:119-112
Yang S-O, Lee YI, Chung DH et al (1992) Detection of extrapul-
monary tuberculosis with gallium 67 scan and computed
tomography. J Nucl Med 33:2118-2123
Yassawy MI, Al Karawi MA, Mohamed AE (1987) Alimentary
tract tuberculosis: a continuing challenge to gastroenter-
ologists. Report of 55 cases. J Gastroenterol Hepatol 2:
137-147
39 Genitourinary Tuberculosis
M. MONIR MADKOUR

CONTENT 39.2.6 Tissue Biopsy and Fine-Needle Aspiration 716

39.2.7 Obstructive Azoospermia and Tuberculosis 717
39.1 Urinary Tract Tuberculosis 699 39.3 Female Genital Tuberculosis 717
39.1.1 Introduction 699 39.3.1 Epidemiology of Female Genital Tuberculosis 717
39.1.2 Epidemiology 700 39.3.2 Pathogenesis of Female Genital Tuberculosis 718
39.1.3 Pathogenesis and Pathology 39.3.3 Clinical Features of Female Genital
of Renal Tuberculosis 701 Tuberculosis 719
39.1.4 Other Modes of Transmission 702 39.3.3.1 Presenting Symptoms 719
39.1.4.1 Renal Allografts 702 39.3.3.2 Physical Signs 719
39.1.4.2 Sexual Transmission 702 39.3.3.3 Tuberculosis of the Fallopian Tube and Ovary
39.1.4.3 Intrauterine Transmission 702 (Features and Investigations) 719
39.1.4.4 Direct Spread from Abdominal Tuberculosis 702 39.3.3.4 Tuberculosis of the Uterus
39.1.5 General Clinical Features (Features and Investigations) 720
of Urinary Tuberculosis 702 39.3.3.5 Synechia Uteri: Asherman's Syndrome
39.1.5.1 Symptoms 702 (Features and Investigations) 721
39.1.5.2 Signs of Urinary Tuberculosis 703 39.3.3.6 Tuberculosis of the Uterine Cervix
39.1.5.3 Other Clinical Features 703 (Features and Investigations) 721
39.1.6 Diagnosis of Urinary Tuberculosis 703 39.3.3.7 Tuberculosis of the Vagina
39.1.7 Imaging Modalities in Urinary Tuberculosis 703 (Features and Investigations) 722
39.1.7.1 Plain Radiography 703 39.3.3.8 Tuberculosis of the Vulva
39.1.7.2 Intravenous Excretory Urography 704 (Features and Investigations) 722
39.1.7.3 Ultrasound 704 39.3.3.9 Investigation of Female Genital Tuberculosis 722
39.1.7.4 Computed Tomography (CT) Scan 705 39.3.4 Microbiology and Histopathology 723
39.1.7.5 Magnetic Resonance Imaging (MRI) 707 39.3.5 Laparoscopy 723
39.1.7.6 Renal Scintigraphy 707 39.3.6 Treatment of FGTB 723
39.1.7.7 Cystoscopy 707 39.4 Tuberculosis in Patients with Chronic Renal Failure,
39.1.8 Polymerase Chain Reaction (PCR) 707 on Hemodialysis, or with Renal Transplant 723
39.1.8.1 A Large Series of One Thousand Patients 39.4.1 Epidemiology 724
with Suspected GUTB 707 39.4.2 Clinical Features of Tuberculosis
39.1.9 Treatment of Renal Tuberculosis 709 in these Groups 724
39.1.10 Surgery of Urinary Tuberculosis 709 39.4.3 Investigations 724
39.2 Male Genital Tuberculosis 710 39.4.4 Treatment of Genitourinary Tuberculosis 725
39.2.1 Epidemiology 710 39.4.5 Dose Modification 725
39.2.2 Pathogenesis 711 References 725
39.2.3 Clinical Features 712
39.2.3.1 Clinical Features of Tuberculous
Epididymo-orchitis 712
39.2.3.2 Clinical Features of Tuberculous Prostatitis
and Seminal Vesiculitis 712
39.2.3.3 Clinical Features of Penile Tuberculosis 713
39.2.4 Diagnosis of Male Genital Tuberculosis 713 39.1
39.2.5 Imaging Features 713 Urinary Tract Tuberculosis
39.2.5.1 Ultrasound and Color Doppler 713
39.2.5.2 CT and MRI 716 39.1.1
Introduction

Genitourinary tuberculosis (GUTB) is a common site

M. M. MADKouR, MD, DM, FRCP of extrapulmonary tuberculosis (EPTB). It accounted
Consultant, Department of Medicine, Riyadh Armed Forces for 10% to 33% of extrapulmonary TB cases in various
Hospital, P.O. Box 7897, C-1l9, Riyadh 11159, Saudi Arabia series (Garcia-Rodriguez et aI. 1994; Chattopadhyay et

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
700
al.1997; Alvarez and McCabe 1984; Weir and Thornton
1985). Clinical features of GUTB are often vague and
insidious, and clinicians have to have a high index of
suspicion to avoid the disastrous effect if this pathology
is ignored. In developing countries, GUTB is a particu-
lar diagnostic problem and in developed countries it
is expected to be on the rise with the resurgence of
tuberculosis due to the spread of HIV (Nzerue et al.
2000). GUTB will eventually occur in 4% to 20% of all
cases of pulmonary tuberculosis (Gow 1992). GUTB
can be treated successfully with a short course of
antituberculous drugs for 4-6 months. Such success
is due to the relatively smaller number of organisms
in renal tissue compared with those present in pul-
monary disease. The kidney has a very good blood
supply with high concentrations of the drugs in the
urine, and there is good penetration of these medica-
tions to closed cavities (Gow 1979a,b).Although GUTB
may affect one or more sites of the genitourinary tract,
renal, male genital, and female genital disease will be
discussed separately.
39.1.2
Epidemiology
The epidemiology of urinary tract tuberculosis is dif-
ficult to determine because of the sparsity of symp-
toms related to the kidney, which is the most common
site of involvement. Bacilliuria is also intermittent in
nature therefore the prevalence of urinary tuberculo-
sis is grossly underestimated. It has been estimated by
many authors that among all patients with pulmonary
tuberculosis, the genitourinary tract will be affected in
4% to 20% (Gow 1992, Lenk and Schroeder 2001).
Mycobacterium tuberculosis is the most common
cause of urinary tuberculosis. Occasionally Myco-
bacterium bovis may be the causative organism par-
ticularly in areas where raw milk is still consumed
(Stoller 1985; Yates et al. 1993; Dankner et al. 1993;
Cosivi et al. 1998). Other nontuberculous mycobacte-
ria may also cause urinary tract infection particularly
in HIV-infected patients. Urinary tract tuberculosis
is more common among high-risk groups including
Australian Aborigines, Native American Indians,
immigrant health care workers, alcoholics, drug
addicts, those living in shelters for the homeless, and
those on immunosuppressive agents or with AIDS
(CDC 1986,1985,1987,1995; Beck-Sague et al. 1992;
Cantwell et al.1994; McKenna et al.1995; Nardell et al.
1986; Pearson et al. 1992; Selwyn et al. 1989,1992).
Urinary tuberculosis occurs in all age groups but is
most commonly found in the third and fifth decades
M. M. Madkour
of life at the time of its presentation. It is relatively
uncommon among children and young adults (Ferrie
and Rundle 1983; Chang et al. 1998).
In India, Chattopadhyay and colleagues (1997)
reported nine children (five boys and four girls) with
genitourinary tuberculosis. Association with, or past
history of, tuberculosis and was noted in three of these
children (33%). Ferrie and Rundle (1983) reviewed
230 patients with genitourinary tuberculosis seen at
the Royal Infirmary in Glasgow between 1970 and
1979. The authors reported 29 patients under the age
of 25 years (10.8%). Past history of tuberculosis was
found in 6.9% and positive family history in 27.6%.
Both sexes were affected, with male predominance in
most series at a ratio of2:1. The incidence of genitouri-
nary involvement among all patients with tuberculosis
may range between 2% and 5%, but the incidence is
higher among patients with active pulmonary tuber-
culosis, ranging between 5% and 10% or higher.
In Spain, genitourinary tuberculosis was noted
to involve the most common site of extrapulmonary
disease. Garcia-Rodriguez et al. (1994) reported 198
cases of extrapulmonary tuberculosis and found 81
cases to have genitourinary tuberculosis (40.9%), with
male predominance (63%). Genitourinary tubercu-
losis accounted for 13.4% of all cases of tuberculosis
and history of tuberculosis was present in 27.2%. The
age distribution was more homogenous among differ-
ent age groups in this series and was not restricted to
older age groups. The authors noted that despite the
decline in the prevalence of pulmonary tuberculosis
in Spain, extrapulmonary localizations have increased
from 18% in the late 1970s to 24% in 1986. In Saudi
Arabia, the incidence of genitourinary tuberculosis
has increased recently from 5% to 16% among those
with extrapulmonary disease (see Figs. 3,4, and 5 in
the chapter «Epidemiology of TB in Saudi Arabia").
Ethnic and racial differences in the incidence of
genitourinary tuberculosis have been noted in dif-
ferent parts of the world. The Tuberculosis and Chest
Diseases Unit of the British Medical Research Council
conducted a survey in 1983 and reported in 1987 on
tuberculosis based on notifications and bacteriology
(National Survey - BMRC 1987). The survey indicated
that genitourinary localization accounted for 27%
of nonrespiratory tuberculosis among the European
(white) patients and only accounted for 5% among
those born in the Indian subcontinent (ISC): India,
Pakistan, and Bangladesh. Ormerod (1993) reported
a similar lower incidence of genitourinary tubercu-
losis among the ISC population living in the United
Kingdom than among the European (white) popula-
tion. He reviewed 1,163 notifications of tuberculosis
Genitourinary Tuberculosis 701

in Blackburn, Hyndburn, and Ribble Valley betweenmonly located in the lung. M. bovis and nontubercu-
1978 and 1990 inclusive. Extrapulmonary tuberculosis
lous mycobacteria can also affect the genitourinary
was more common in the ISC ethnic group than in the
system (Stoller 1985; Dahl et al. 1996). During pri-
white ethnic group (74% versus 24%, respectively). The
mary pulmonary tuberculosis development, many
ethnic white group accounted for 82% of genitourinary
authors believe that bacillemia occurs and metastatic
tuberculosis while patients in the ISC group accounted
seeding via hematogenous spread may occur to other
only for 15%. In another survey from Southeast Eng-
sites (Balasubramanian et aI.1994). Even after healing
land, Grange and colleagues (l995) conducted a survey
of the primary complex in the lung, the bacilli may
to determine the ethnic differences in the incidence of
remain dormant in the new seeded areas in the lung
genitourinary tuberculosis. These authors based their
and other organs. The exact mechanisms of protec-
survey on positive bacteriologic culture rather than on
tive immunity response against the development of
the total notifications. A total of 7,536 new tuberculosis
the disease in humans have not been totally clari-
patient records of ISC and European ethnic origins
fied (Ellner 1997). Reactivation of dormant granu-
were reviewed. The authors found the incidence oflomatous foci in the kidney or reinfection of both
genitourinary tuberculosis in the ISC group was only
kidneys from postprimary pulmonary tuberculosis
8% while in the European white ethnic group it was
via hematogenous spread is the most common route
25% with highly significant differences. However, they
of infection. The renal cortex of both kidneys at the
concluded that with elimination of the confoundedperiglomerular regions is the most common site of
factor of age, the occurrence of genitourinary tubercu-
initial localization of microscopic miliary granulo-
losis was similar in both ethnic groups. mas because of the high oxygen tension and rich
In the United States, ethnic and racial differences
blood supply. When the immune system is altered
in the incidence of genitourinary tuberculosis were
because of HIV, old age, diabetes mellitus, chronic
also reported. Reider and colleagues (l990) reported
renal failure, or the use of immunosuppressive agents,
the epidemiology of extrapulmonary tuberculosis these dormant foci of granulomata containing viable
according to age, sex, racial, and ethnic origin on
organisms proliferate (Medlar 1926; Medlar et al.
data obtained from the Centers for Disease Control1949; Berman et al. 1960; Christensen 1974; Psihramis
(CDC), Atlanta. The authors noted that genitourinary
and Donahoe 1986; Lazarus and Peraino 1984) and
tuberculosis was less common in blacks than in non-
granulomas may enlarge in size.
Hispanic whites, and more common among those Enlarging granulomas may coalesce and form a
born outside the United States. In Australia, there
large nodule that may in turn rupture and spread
was definite evidence of ethnic and racial differences
deeper into the renal tubules and produce medul-
in the incidence of genitourinary tuberculosis. Dwyer
lary granulomas. Caseation and necrosis of these
and his colleagues (l987) reviewed 51 patients with
nodules may cause sloughing of the papillae with
extrapulmonary tuberculosis seen between 1980 and
formation of a cavity or cavities. Debris passed from
1985. Ten patients had genitourinary tuberculosis (six
this process may obstruct a calyx, renal pelvis, pel-
were European immigrants and four were Austra- viureteric junction, or the ureter. Such cavitations
may be augmented by ureteric obstruction to form
lians) and none from Southeast Asia or among other
immigrants to Australia. The authors also noted that
a cystic mass lesion. Tubercle bacilli are shed from
in 1984, the incidence of extrapulmonary tuberculo-
these cavities and infect surrounding renal paren-
sis among all patients with tuberculosis in Australia
chyma and the collecting system. The ureter, uri-
was 24.3%. Recently, patients with AIDS were foundnary bladder, and urethra are infected through this
to have a higher incidence of genitourinary tubercu-
antegrade method of spread, and bacilli may appear
losis than occurred in non-HIV-infected patients.in the urine. Renal parenchymal healing with fibro-
Shafer and colleagues (l99l) from New York reported
sis, scarring, and calcification usually occurs in
genitourinary tuberculosis in 37% of patients with
part of the kidney or all of it. Calyceal renal pelvis,
AIDS and in 25% among non-HIV-infected patients. ureteric, and bladder ulceration may occur and usu-
ally associated with fibrosis, scarring, calcifications,
reduction in capacity, deformities, or obstruction.
39.1.3 As the disease advances, the renal cortex may be
Pathogenesis and Pathology of Renal Tuberculosis totally lost particularly with ureteric obstruction
and it becomes hydronephrotic nonfunctional
Mycobacterium tuberculosis spread to the kidney (autonephrectomy) (Psihramis and Donahoe 1986;
via hematogenous route from a primary focus com- Lazarus and Peraino 1984).
702
39.1.4
Other Modes of Transmission
Although hematogenous spread of infection to the
urinary tract is the most common route, there is
documented evidence that other modes of transmis-
sion can occur.
39.1.4.1
Renal Allografts
Transmission of Mycobacterium tuberculosis bacilli
to recipients by renal allograft obtained from cadav-
eric or living donors has been reported. Wajeh and
his colleagues (1990) from Saudi Arabia reported one
patient who developed tuberculosis after receiving a
kidney donated by his living brother. The brother was
found later to have active pulmonary tuberculosis.
These authors also reported their experience of 403
renal transplant recipients at the King Faisal Special-
ist Hospital and Research Center in Riyadh over a
9-year period. They found 14 patients (3.5%) who
developed tuberculosis after renal transplant. Tuber-
culosis was disseminated in nine (64.3%), pulmonary
in four (28.6%), and genitourinary in one patient
(7.1 %). A similar report from France indicated the
development of disseminated tuberculosis in two
allograft recipients (Mourad et al. 1985). The donor
was a 46-year-old French man who was hospitalized
for head injury and for whom a history of tuberculo-
sis could not be obtained. The two recipients were on
maintenance hemodialysis and did not have tuber-
culosis before the transplant. Both recipients devel-
oped tuberculosis 2 and 5 months after transplant,
respectively. Mycobacterium tuberculosis grew from
the urine of both patients in the absence of clinical
evidence of genitourinary tuberculosis.
39.1.4.2
Sexual Transmission
Transmission of tuberculosis via sexual intercourse
from a husband with renal tuberculosis to his wife
has been reported (Richards and Angus 1998).
These authors reported a 49-year-old Caucasian
man with urinary bladder tuberculosis who was
diagnosed retrospectively. His wife was being in-
vestigated for secondary infertility and during
a laparoscopy, left tuberculous pyosalpinx was
found. Subsequently, the husband was investigated
and found to have urinary bladder tuberculosis. The
authors found no other way for the transmission of
tuberculosis except through coitus.
M. M. Madkour
39.1.4.3
Intrauterine Transmission
Congenital renal tuberculosis in infants that has
been transmitted via the intrauterine route has been
reported. <;askurlu and colleagues (1998) from Turkey
reported a male infant with persistent pyuria from the
age of 1 year. Investigations showed a nonfunctioning
right kidney as well as hydronephrosis in the left
kidney. Tuberculosis was found in the resected right
kidney. Family screening for tuberculosis showed no
evidence of active tuberculosis. Because of the short
duration for such severe destructive damage to occur
in kidneys, right ureter, and urinary bladder with vesi-
coureteral reflux at this age, and in the absence of other
sources, the authors considered intrauterine transmis-
sion was the presumed mode of transmission.
Congenital or prenatally acquired tuberculosis
may occur through hematogenous route with spread
of the bacilli to the placenta leading to placental
tuberculosis and then tuberculosis in the fetus. Hema-
togenous spread of tuberculosis from the infected
placenta to the fetus via the umbilical cord vein to
the liver may occur. Rupture of placental tuberculous
caseation and necrosis into the amniotic fluid with
subsequent in uterofetal aspiration is another mode
of transmission of congenital tuberculosis (Myers et
al.1981; Cantwell et al.1994). (See also the chapter on
pregnancy and tuberculosis.)
39.1.4.4
Direct Spread from Abdominal Tuberculosis
Spread of infection to the genitourinary tract may
occur via direct contact with abdominal organs and
tissue tuberculosis including adhesions due to tuber-
culous peritonitis or intestinal tuberculosis. It may
also spread via intestinal or appendicular fistulas to
the genitourinary tract (Silva et al. 1988; Tripathy and
Tripathy 1990).
39.1.5
General Clinical Features of Urinary Tuberculosis
39.1.5.1
Symptoms
Renal tuberculosis is asymptomatic in up to 80% of
patients (Simon et al. 1977) in comparison with the
remaining 20% of cases of genital tuberculosis that
are more symptomatic, with a better chance of being
diagnosed earlier. Therefore, the treating clinician in
Genitourinary Tuberculosis
endemic areas should have a high index of suspicion
to diagnose it at its early stage. History of recurrent
urinary tract infections may be the only presenting
feature even when the causative pyogenic organism,
such as E. coli, is identified or sterile pyuria is pres-
ent. This should alert the clinician to the possibility
of urinary tract tuberculosis as a cause.
Constitutional symptoms, such as the fever, sweat-
ing, loss of appetite, and weight loss described in
14-60% of patients, nut may be absent in patients
with urinary tract tuberculosis (Garcia-Rodriguez
et al. 1994). Symptoms of recurrent attacks of pain-
less hematuria, or flank or abdominal pain may
occur in isolation or in association with mild con-
stitutional symptoms. However, the most common
urinary symptoms at the time of presentation are
related to urinary bladder involvement. Recurrent
or persistent dysuria, frequency, urgency, nocturia,
or suprapubic pain may be present. The incidence
of these urinary symptoms with bladder irrita-
tion varies from one series to another with aver-
age occurrence in 20% to 95% of patients with
urinary tuberculosis (Garcia-Rodriguez et al. 1994;
Psihramis and Donahoe 1986). Rarely, acute urinary
retention, weak urinary stream, periurethral abscess,
sinus, or fistulas may be the presenting symptoms
when the urethra is involved (Raghavaiah 1979;
Indudhara et al. 1992). Symptoms related to pulmo-
nary tuberculosis, or that in other systems or organs,
may be also present. Past history or family history of
tuberculosis should be obtained.
39.1.5.2
Signs of Urinary Tuberculosis
Physical signs in patients with urinary tract tubercu-
losis are often lacking. Occasionally flank tenderness
on percussion, enlarged palpable renal mass, cold
abscess with discharging nephrocutaneous fistula
may be found. Suprapubic tenderness, perineal and
urethral abscesses or fistulas, or mass at urethral
meatus, may be seen.
39.1.5.3
Other Clinical Features
Clinical features of tuberculosis affecting various
organs of the urinary and male and female genital
tract will be discussed separately. Although these
organs are often affected in more than one site of
the genitourinary tract, it is my intention to discuss
each organ separately in this chapter for the purpose
of clarity.
703
39.1.6
Diagnosis of Urinary Tuberculosis
The diagnosis of early urinary tract tuberculosis is
difficult as it has no distinctive clinical, hematologi-
cal (anemia) or biochemical features (raised urea in
1.2% to 2.6%) that could raise the suspicion of the
treating clinician to the possibility of tuberculosis as a
cause. At the time of diagnosis, most patients will have
advanced disease. Unsuspected accidental discovery of
urinary tuberculosis, while investigating other illnesses
or during surgery has been noted in 7.5% to 56%
of patients (Garcia-Rodriguez et al. 1994). The most
important initial test that may lead to requesting further
urinary tract investigations is urine analysis, done as a
routine test. It is abnormal in 59.3% to 84% (pyuria and
or hematuria) of patients with urinary tract tuberculo-
sis (Webester and Wright 1985, Gow 1992).
The second most important investigation is intra-
venous urography OVU), which may have features that
raise the suspicion of tuberculosis as a cause. Urine
smear and culture for acid-fast bacilli is the mainstay
for diagnosing urinary tract tuberculosis. Ziehl-Neelsen
staining of urine deposits is positive in 3-5 morning
samples on separate days in about 0.2% to 50% of cases
(Webster and Wright 1985; Gow 1992).
Smear staining of urine sediments requires a mini-
mum of 10,000 organisms/rnl to be able to identify the
bacilli on microscopic examination. However, urine
deposit staining may give misleading results due to con-
tamination by saprophytic mycobacteria particularly M.
smegmatis. Urine culture yield has been reported with
a wide range of positivity from 25% to 100% (Garcia-
Rodriguez et al. 1994; Moussa et al. 2000; Ferrie and
Rundle 1983). These variations may be related to the
intermittent bacilliuria. Bucholz, Salahuddin and Haque
(2000) from Pakistan reported 55 patients with genito-
urinary tuberculosis. Positive urine culture was found in
57% and urinary bladder biopsy changes in 54%.
Positive culture can also be obtained from draining
perineal and urethral sinus discharge or abscesses.
Concomitant presence of active pulmonary tubercu-
losis may provide another source for detecting bacilli
in the sputum.
39.1.7
Imaging Modalities in Urinary Tuberculosis
39.1.7.1
Plain Radiography
In the early stage of the disease process, plain radiog-
raphy appears normal. In advanced renal tuberculosis,
 704 M. M. Madkour

calcifications may be seen as stippling, curvilinear,
lobar in pattern or may affect the entire kidney (Leder
and Low 1995; Cohen 1986) (Fig. 39.1a, b). Calcifica-
tions may be so extensive that they affect the kidney,
ureter, and urinary bladder (Figs. 39.la, b, 39.3a, b).
39.1.7.2
Intravenous Excretory Urography
Early in the disease, IVU may look normal. Imaging
abnormalities were depicted in IVU in 65% to 95% of
patients in different reported series (Garcia-Rodriguez
et al. 1994).
Although bilateral renal tuberculosis pathology
is the rule (Figs. 39.1-39.3), unilateral abnormalities
in IVU are not rare (Fig. 39.5). With early calyceal
involvement, the outline of the calyces may look
hazy, but later, ulceration and fibrosis deformities
may appear more obvious. A cavity or multiple cavi-
ties may be seen when necrosis and sloughing of the
papillae occurs. Parenchymal fibrosis and scarring
may be depicted by IVU as well as by ultrasound.
Obstruction of the pelvicalyceal and pelviureteric
junctions leading to hydronephrosis can be easily
discerned. In the advanced stages, autonephrectomy
of the kidney may not be visualized (nonfunction-
ing) while the opposite side will be clearly seen by
IVU and may also show evidence of involvement
(Figs.39.1, 39.2, 39.4, 39.5). The ureter ulceration
and fibrosis may be seen as straightened ureter, and
hydroureter may be seen if ureterovesical reflux or
obstruction at the lower end is present. Reduced
bladder capacity and deformity can be noted by IVU
(Fig. 39.2c) (McGuinness 2000).
39.1.7.3
Ultrasound
Ultrasonography may not be helpful in the early stage
of renal tuberculosis. As the disease progresses, the
size of the renal pelvis may be reduced by fibrosis or
hydrocalyx, and strictures at various sites may cause
hydronephrosis. Mass lesions and even pelviureteric
obstructions can be demonstrated by ultrasound.
The thickness of the urinary bladder wall, its capacity,
and the caliber of the lower part of the ureter can be

a -~~--b

Fig. 39.la, b. A 53-year-old male presented with 2 years of dysuria and micturition. He was told in the past that he had renal
stones but these were not causing problems for him. At presentation he denied fever, chest symptoms, or contact with tubercu-
losis patients. His blood urea and creatinine were raised. a Abdominal radiograph shows lobar type of parenchymal calcification
of the left kidney (putty kidney), amorphous calcification at the lower pole of the right kidney (white arrowhead). Calcifica-
tion of the left ureter and the bladder (white arrows). The last two findings are infrequent in tuberculous involvement of these
structures. Note calcified pelvic lymph node due to abdominal tuberculosis (curved arrow). b Intravenous pyelogram shows
nonfunctioning left kidney and hydronephrotic right kidney. Note marked dilatation of the pelvis with irregular round filling
defects due to granuloma (arrow). Granulomas causing filling defects also seen in the dilated lower pole calyx (arrowhead).
Three morning urine cultures grew M. tuberculosis
 Genitourinary Tuberculosis 705

Fig. 39.2a-c. A 49-year-old man presented with recurrent urinary tract infec-
tion for 6 months before presentation. He had had many antibiotics. He was
referred to our hospital. He had fever, dysuria, and micturition. He had a past
history of pulmonary tuberculosis 10 years earlier and was treated. a Chest
radiography shows multiple calcified right paratracheal and right hilar lymph
nodes (arrowheads). Parenchymal fibrotic bands are also seen radiating from
the hilar regions (arrows). b Abdominal X-ray shows multiple irregular calci-
fications in a small right kidney (black arrowheads). Small calcific density is
also seen in the left kidney (white arrowhead in b and c). c Intravenous pyelo-
gram demonstrates nonfunctioning of the right kidney. Left hydronephrosis
and hydroureter. Shrunken small-capacity urinary bladder with marked wall
thickening (white arrow)

a __ ~ -",".,.J

assessed with ultrasound. However, ultrasound does
not assess renal function or calcifications. These are
better depicted by plain radiography IVU or CT scan-
ning (Premkumar et al. 1987; Cremin 1987).
39.1.7.4
Computed Tomography (CT) Scan
Axial CT with contrast can provide similar findings
to those of IVU, in depicting features of renal tuber-
culosis particularly the pelvicalyceal ulcerations and
wall thickening, focal caliectasis, and deformities
due to fibrosis and strictures. It can provide more
detailed morphology and some assessment of the
renal function through the degree of cortical per-
fusion and enhancement (Premkumar et al. 1987;
Birnbaum et al. 1990). Renal calcification patterns
are better depicted by axial CT than by the use of
IVU or ultrasound. An enhancement of an inflamma-
tory granuloma seen on axial CT with contrast will
differentiate it from caseating and necrotic material
that does not enhance (Goldman et al. 1985; Feeney
et al. 1994). Perinephric abscesses are much better
depicted with greater detail by axial CT than by the
use of IVU (Fig. 39.5) or ultrasound. Ureteric fibro-
sis, dilatation, wall thickening, and multiple strictures
are better seen on CT than by IVU. A small, shrunken,
deformed urinary bladder with a thick, irregular wall
with occasional calcifications can readily be seen by
axial CT (Valentini et al. 1998; Wang et al. 1997).
 706 M. M. Madkour

Fig. 39.3a, b. A 28-year-old female presented with recurrent dysuria, fever, and constitutional symptoms for more than 3 years. a
Abdominal radiograph showing lobar and curvilinear calcifications in the left kidney (arrowheads). There is also irregular cal-
cification at the distal end of the left ureter due to calcified granulomas (curved arrow). This is an infrequent finding in ureteric
tuberculosis. b Intravenous pyelogram showing partially functioning left kidney, narrow contracted renal pelvis due to fibrosis
(arrow), and abnormal pooling of contrast around the deformed upper pole calyces (black arrowheads) in the right kidney

a b

Fig. 39.4a, b. A 32-year-old male presented with recurrent episodes of right flank pain and dysuria for over 3 years. He had ultra-
sound of the abdomen 3 years before presentation and was reported as normal. Urine cultures showed E. coli on one occasion.
He had had many courses of antibiotics. a Abdominal radiograph showed no abnormal findings. b Intravenous pyelogram shows
nonfunctioning right kidney and normal left collecting system and ureter. Filling defect noted in the bladder at the site of the
ureterovesical junction caused by tuberculosis granulomas (arrowheads). Series of morning urine cultures for tuberculosis was
positive. Cytoscopy showed inflammation at the ureterovesical orifice with mass. Biopsy showed caseating granuloma
Genitourinary Tuberculosis 707

39.7.7.5 inflammation and stenosis. Biopsy from bladder

Magnetic Resonance Imaging (MRI) lesions may be obtained for histopathologic and
microbiologic investigations that may show granu-
MRI has the advantage over other imaging modalities lomata with caseation and necrosis and may grow
in differentiating large granulomatous caseating mass the bacilli. Piscioli et al. (1985) from Italy conducted
lesions from other mass lesions of the kidney due to a comparative study in 13 patients with tuberculosis
other causes. Caseating tuberculous lesions tend to of the bladder, diagnosed by cytologic findings of
be hypodense in T2 and fail to enhance. Other mass urinary sediment and positive histologic findings.
lesions of the kidney tend to be hypodense in Tl and Cytologic findings of urinary sediment were aimed
show high signal intensity on T2- and Tl-gadolinium at identifying epithelioid cells which were present in
contrast images. MRI is however at a disadvantage 38.5% and/or the identification of Langhans giant
in detecting calcification if compared with axial CT cells which were found in 84.6% of this series. His-
(Murata et al. 1996; Valentini et al. 1998). tologic diagnosis of the bladder tissue biopsy sample
obtained by cystoscopy showed typical granulomas
39.7.7.6 with Langhans giant cells in all 13 patients (100%)
Renal Scintigraphy and culture was positive in 11 (84.6%) as reported by
these authors.
Assessment of renal function can be done by isotope
renal scintigraphy. Renal perfusion, glomerular filtra-
tion rate, excretion, and drainage can be assessed by 39.1.8
this method. It can be done by using either: 200-mBq Polymerase Chain Reaction (peR)
Tc 99m DTPA, 100-mBq Tc 99m DMSA, or Cr 51 EDTA.
The size of the kidney, renal parenchymal uptake and Diagnosis of urinary tract tuberculosis depends on
perfusion, its shape, pelvicalyceal and ureteric shape, the gold standard of identifying the bacilli from the
obstruction, dilatations and drainage of the isotope urine or tissue biopsies. The standard methods of
through the ureter can be assessed. Cortical scarring urine sediments staining and culture remain unsat-
or mass lesions may be seen as photodense areas but isfactory. The use of PCR for the diagnosis of active
their nature cannot be differentiated by isotope scin- tuberculosis was mostly reported on sputum samples
tigraphy (Fig. 39.5). and only a few reports cite results on urine sediments
(Missirliu et al.1996; Moussa et al. 2000). Detection of
39.7.7.7 mycobacterial DNA by means of PCR requires only
Cystoscopy 1-10 organisms to yield a positive result while smear
examination requires at least IO,OOOorganisms/ml.
When frank hematuria is the sole presenting feature,
cystoscopy should be performed (Chattopadhyay et 39.7.8.7
al. 1997). Cystoscopy in such instances is done for the A Large Series of One Thousand Patients
purpose of diagnosing or excluding urinary bladder with Suspected GUTS
tumor. In one of our patients (from Buraidah, Saudi
Arabia) who presented with severe painless hema- In a prospective comparative study on 1,000 patients
turia and passing clots, cytoscopy showed a tumor with clinical suspicion of GUTB, the diagnosis was
mass in the urinary bladder. Histology was reported confirmed in 347 patients (34.7%), by parallel per-
as transitional cell carcinoma and in another cystos- formances of conventional urine smear and culture,
copy performed in another hospital it was reported radiometric liquid culture (BACTEC), and two differ-
as squamous cell carcinoma and he was referred to ent assays of PCR (IS6110-PCR and 16S rRNA-PCR)
the oncology department in our hospital for a third (Moussa et al. 2000). These patients attended the
opinion and was found to have tuberculous cystitis. Urology and Nephrology Center, Mansoura Univer-
Cystoscopy of urinary tuberculosis under general sity Hospital, in Egypt. Morning urine was obtained
anesthesia may show inflammation, mucosal ulcer- on 3 consecutive days from each patient. Urine sedi-
ation, or necrotic mass that simulates bladder tumor. ments were stained for AAFB, cultured simultane-
Edema, redness and inflammation of the trigone, ure- ously by using Lowenstein-Jensen (BACTEC), and
teric orifices inflammation and stenosis, contracted examined by two different PCR assays (IS611O-PCR
bladder with diminished capacity may be noted by and 16S rRNA-PCR). The results of these laboratory
cytoscopy. The urethra may also show features of investigations are shown in Table 39.1.
 f

h
 Genitourinary Tuberculosis 709

Fig. 39.5a-h. A 65-year-old man presented in October 1995 with left hemiparesis and brain CT pre- and postenhancement
showed (figure not included) multiple bilateral cortical infarcts in the centrum semiovale bilaterally, the left internal capsule,
and the subcortical region of the left parietal lobe. There was no evidence of enhancing foci. In December 1995 he presented
with upper abdominal pain. a Plain radiograph of the abdomen showed a dense irregular opacity in the left hypochondrium
(arrow) and 2 smaller punctuate opacities just inferomedial to the large opacity (small arrows). These were interpreted as rep-
resenting calcified hydatid cysts or old calcified hematomas. The lesions are probably located within the spleen and ultrasound
was suggested but not done as abdominal pain settled. Bilateral carotid Doppler sonography showed calcified atheromatous
plagues at left common carotid and right common carotid and right common carotid bifurcation. b On 24 June 1997 he presented
with I-week history of fever, left loin pain, and dysuria. Patient came in April 1996 with left foot pain. Plain X-ray (not shown)
showed calcification at the insertion of Achilles' tendon and vascular calcification. Chest X-ray demonstrates multiple calcified
granulomas in both lung fields (arrowheads). Blunting of the left costophrenic angle due to small effusion (curved arrow). c
On 25 June 1997 he had abdominal ultrasound. Ultrasound of the left hypochondriac region demonstrates irregular mixed
echogenic mass-like lesion (arrows), medial and inferior to the spleen (arrowhead) interpreted as possible abscess collection.
The left kidney could not be visualized clearly. d CT of the abdomen: axial section at the dome of the diaphragm demonstrates
multiple foci of pleural calcification (arrowheads) and pleural thickening with small effusion (arrow). e CT of abdomen: axial
section at a caudal level demonstrates irregular abscess collection (arrowheads) medial to the spleen (arrow). The perinephric
abscess was aspirated and 40 ml of thick yellowish pus was drained. CT-guided biopsy samples were also obtained. Culture grew
M. tuberculosis and histopathology showed granuloma. f At more caudal section the abscess is seen to extend to the posterior
abdominal wall (arrowheads). g More inferior section shows atrophic left kidney with multiple foci of calcification (white arrow
in c and d). The diagnosis of tuberculosis of the left kidney with autonephrectomy with atrophy and calcification noted first as
nonsymptomatic left hypochondrium calcification in October, 1995. Only in June 1997, did the patient present with fever, left
loin pain, and dysuria. h DMSA scintigraphy shows no evidence of tracer activity in the left kidney

..

Table 39.1. The results of AFB staining, PCR assays, relative to results of culture

Test Result Results of Culture Sensitivity Specificity PPV NPV

Positive Negative

AFB smear Positive 189 21 52.06% 96.7% 99% 77.9%

AFB smear Negative 174 616
IS6110-PCR Positive 347 12 95.59% 98.11% 96.66% 97.5%
IS6110-PCR Negative 16 625
16S rRNA-PCR Positive 316 7 87.05% 98.90% 97.83% 93.06%
16S rRNA-PCR Negative 47 630
ppv, Positive predictive value; NPV, Negative predictive value.
[Reproduced: Courtesy of Professor M. A. Ghoneim, director, Urology and Nephrology Center,
University of Mansoura, Egypt]

These authors recognized the factors that may alter 39.1.9

PCR assays of the urine. False-negative results are
known and found in 5% of the urine samples because
of the presence of urinary enzyme inhibitors and
contamination problems. These two obstacles were
overcome in this study by lysis of the mycobacterial
cell wall and by dUTP methods. The authors reported
that IS611O-PCR assay was more sensitive than 16S
rRNA-PCR. The detection limit of the former method
was 10 organisms by agarose gel electrophoresis and
only 1 cell by Southern blot hybridization. The sen-
sitivity and specificity of methods used in this study
are shown in Table 39.1. PCR is more sensitive than
culture and its specificity is close to 99%. Detection of
mycobacterial DNA can be achieved in 2-3 days.
Treatment of Renal Tuberculosis
Treatment for renal, male genital, and female genital
tuberculosis, and for other groups with renal diseases
who are at risk, will be discussed at the end of this
chapter.
39.1.10
Surgery of Urinary Tuberculosis
Reconstructive surgical intervention may play an
important role in the management of urinary tract
tuberculosis. Antituberculous antibiotics should be
710
initially given for a period of 4-6 weeks before any
surgical intervention.
Most surgical procedures for urinary tuberculosis
may include emergency operative management for
certain conditions (such as bilateral hydronephrosis,
hydronephrotic solitary kidney), if renal insufficiency
is getting worse and is occurring in association with
abscess formation such as perinephric tuberculous
abscess. Surgical procedures required may include
drainage of hydronephrosis, abscess drainage, partial
nephrectomy, reconstruction of upper urinary tract,
bladder augmentation, or urethral reconstruction
(Carl and Stark 1997; Gow and Barbosa 1984; Lazarus
and Peraino 1984; Psihramis and Donahoe 1986).
With advances in modern treatment in recent
years, nephroureterectomy is rarely needed. Every
effort should be made to preserve any remaining
functioning renal tissue. Restoration of renal func-
tion to what may appear as obstructive nonfunction-
ing kidney may occur after the removal of obstruc-
tion or constriction in the urinary collecting system.
Most reported series where nephroureterectomy was
performed involved patients seen in the 1950s, 1960s,
and 1970s (O'Flynn 1970; Gow 1979).
At present there are only rare occasions where
nephroureterectomy may still be indicated, despite
antituberculous treatment, such as recurrent urinary
tract infections, pain, persistent hematuria, failure of
chemotherapy, development of hypertension, or if
coexisting renal cancer is suspected.
The role of nephroureterectomy in the treatment
of a nonfunctioning kidney remains debated issue
among authors. Some believe it may allow the use of a
shorter course of chemotherapy, as viable bacilli were
found in resected kidneys of patients treated for about
3 months. However, such a short period of medication
would not eradicate the bacilli, but a longer course up
to 2 years would achieve that goal. Operative proce-
dures can be done by laparotomy or by laparoscopy.
Partial nephrectomy is rarely needed except during
fistulectomy with primary repair of the gut.
Reconstructive surgery is most commonly used
at present for pelviureteric, ureteric, or ureterovesi-
cal junction obstruction or stricture. Percutaneous
nephrostomy may initially be required to relieve
obstruction at the pelviureteric junction and to be
followed later by pyeloplasty. Cystoscopic dilatation
of strictures at the ureter or ureterovesical junction
by means of an expanding ureteric catheter has been
tried but commonly failed (Murphy et al. 1982).
Ureteric stinting may be tried and if it fails then
the stricture may be resected if it affects the short
segment of the ureter with reanastomosis. Uretero-
M. M. Madkour
vesical junction stricture may be resected, and the
ureter can be reimplanted in the urinary bladder by
using a reflux-preventing technique or by the use of
Boari flap. Longer segments of ureteric stricture may
require ureteroureterostomy, ureterocolic implant, or
ureteral replacement using ileal segment. But the last
has been found to be hazardous and has been associ-
ated with grossly dilated bowel segment, vesicoileal
reflux, bacteruria, excessive mucous secretion in
urine, and renal damage (Charghi 1979).
Bladder augmentation has rarely been indicated in
recent years. A bladder capacity of 100 ml or less may
be associated with troublesome frequency symptoms,
and enterocystoplasty, colocystoplasty, or cecocysto-
plasty may be indicated. Between 1959 and 1977, in a
review of 59 patients with bladder tuberculosis, aug-
mentation enterocystoplasty was performed on all of
them (Dounis and Gow 1979). These authors reported
operative mortality of 17% and postoperative compli-
cations in 50.8% of their patients.
The use of part of the stomach for bladder aug-
mentation (gastrocystoplasty) has been attempted.
It has advantages over enterocystoplasty because it
produces less mucous plaque, allows for easier ure-
teral reimplantation, and does not disturb the body
electrolytes (Shamsa 1998). Tuberculous urethral
strictures are extremely rare and may be managed by
dilatation, urethrotomy, or urethroplasty (Raghavaiah
1979; Indudhara et al.1992).
In conclusion, the need for surgery for urinary
tract tuberculosis has declined remarkably in recent
years due to the use of modern chemotherapy.
Reconstructive surgery of the collective system is
only required when stenotic lesions are not respond-
ing to antibiotics alone in the first 4 weeks.
39.2
Male Genital Tuberculosis
39.2.1
Epidemiology
Genital organ tuberculosis in general (male and
female) is more commonly reported than urinary
tract tuberculosis. Mokhtar and Salman (1983) from
Saudi Arabia reported genitourinary tuberculosis in 20
patients in their series of 125 patients with extrapul-
monary tuberculosis (16%). Genital involvement was
found in 16 patients (80%) (12 male and 4 female),
while the urinary tract was involved in only 4 patients
(20%). All 12 male patients had epididymo-orchitis.
Genitourinary Tuberculosis 711

Wechsler and colleagues (l960) reported on their series

mycobacterium was also found to cause male genital
tuberculosis particularly in homosexuals with mv.
of 127 male patients with early symptoms and signs of
These included M. kansasii, M. avium, M. celatum,
genitourinary tuberculosis, and found epididymitis in
and M. xenopi (Dahl et al. 1996).
36%. Other authors have noted that the epididymis is
the most common site of involvement in male genital The route of transmission of the bacilli to male
organs (Ferrie and Rundle 1983; Petersen et al. 1993).
genital organs is commonly via hematogenous spread,
While other authors reported that the epididymis and
direct spread from the urinary tract tuberculosis, or
the prostate were the most common male genital sitesexual intercourse (Sutherland 1982, 1985). Genital
of tuberculosis. In Spain, Garcia-Rodriguez et al. (l994)
tuberculosis in males may be the sole primary site of
reported on 81 patients with genitourinary tuberculo-
the disease, without urinary tract involvement. The
sis (51 male and 30 female). Epididymitis was foundurine culture may be positive in up to 50% of patients
in six patients (l1.8% of males) and the prostate was
with tuberculous epididymo-orchitis without any
infected in one (2%). In India, Chattopadhyay et al.
evidence of renal tract involvement (Ferrie and
(l997) reported nine children (five boys and four Rundle 1983). Carbal et al. (l985) reported two chil-
girls) with genitourinary tuberculosis. Two of the dren with tuberculous epididymo-orchitis that had
boys (40%) had epididymo~orchitis with no evidence no evidence of renal or other sites of involvement.
of renal involvement. Tuberculous epididymitis in aThese authors reviewed the literature for similar find-
group of 214 patients with scrotal inflammation whoings to their own and found 24 children reported by
were referred for ultrasonography was reported by others with tuberculous epididymo-orchitis without
Drudi et al. (l99l). They found 34 cases that were later
evidence of renal involvement. They also suggested
identified to be due to tuberculous epididymo-orchitis
that in children the transmission of infection to the
(l6%). Wechsler et al. (l960) reported a series of 127
epididymis is most likely via hematogenous route,
male patients with genitourinary tuberculosis and while in adults transmission by direct spread from
found tuberculous epididymo-orchitis in 36%. urinary tract disease is the most common. Early focal
Male genital tuberculosis occurs mostly in young
lesions may initially start at the tail or head, or may
men, with 60% of patients between 20 and 40 years of
affect the entire epididymis. Inflammatory cell infil-
age (Heaton et al.1989). Previous history of tuberculo-
trates, including epithelioid and polymorph nuclear
sis may be obtained in up to 70% of patients with epi-
cells with caseation and necrosis, may spread to the
didymitis (Chung et al. 1997). The incidence of tuber-
tunica vaginalis testis. Abscess formation may occur
culosis of the prostate in recent years with modern (Granadoz 1998) and may rupture through a single or
chemotherapy is not clearly determined but reportedmultiple scrotal sinuses. The scrotal involvement may
as low, while in the past, before the era of modernbe by direct spread of infection leading to swelling,
therapy, prostatic involvement ranged between 14% and the vas deferens may also be affected. Prostatic
and 95%. In India, Mondal et al. (l990) reported their
and seminal vesicle tuberculosis may be transmitted
findings on 126 cases of transurethral fine-needlevia hematogenous route or by direct extension of
aspiration cytology of prostatic nodules suspected infection from tuberculous cystitis (Kim et al. 1993a;
to be malignancies, and tuberculosis was found in Gow 1992; Hemal et al. 2000).
four samples (3%). The incidence of tuberculosis in Sporer and Auerbach (l978) reviewed 1,000
the vas deferens and seminal vesicles is not known,autopsies (l972-1976) from New Jersey and found
as they rarely occur in isolation. Penile tuberculosis is
tuberculous epididymitis in 26 cases. None of these
rare and the penis is the least commonly affected male
cases had evidence of urinary tract disease. They also
genital organ, yet there are many single case reports or
reported three cases from their own series. In one,
small series that have appeared in the literature. Penile
the kidney, bladder, and epididymitis were infected
tuberculosis may occur at any age from infancy to but not the prostate. In the second patient, the lungs
adulthood and among homosexuals with HIV (Lewis and the right kidney were involved but not the pros-
1946; Annobil et al.1990; Dahl et al.1996). tate. In the third patient, the prostate and seminal
vesicles were infected and urine culture was posi-
tive, yet there was no evidence of renal tuberculosis.
39.2.2 They also suggested that descending intracanalicular
Pathogenesis spread may be excluded by the absence of pathologic
changes of the bladder and ureter. They concluded
Mycobacterium tuberculosis is the most common that the lateral or peripheral parts of the prostate
cause of male genital tuberculosis. Nontuberculous were more often involved indicating the likelihood
712
of hematogenous spread. They concluded that if
descending infection was the route, then the disease
would begin near the prostatic urethra and that was
not the case in their findings. As the disease pro-
gresses and advances in the prostate, granulomatous
nodules, caseation, necrosis, and abscess formation
may occur. The prostatic tuberculous abscess may
break through the urethra or the prostatic capsule
(Kumar et al.1994). The vas deferens and the seminal
vesicles may be involved via direct spread of infec-
tion from epididymal or prostatic tuberculosis. The
vas deferens and the seminal vesicles may be affected
in patients with renal tuberculosis in up to 60% of
patients (Rains and Mann 1988).
Transmission of tuberculosis to the penis may
be either primary, when no tuberculosis elsewhere
in the urogenital tract is found, or secondary, when
other organs or systems are also affected. Primary
tuberculosis of the penis may occur as a result of
sexual intercourse with a partner having tubercu-
losis of the cervix or endometrium. Sengupta and
Mukherjee (1982) reported two patients with penile
tuberculosis, and sexual contact was definite in one
of them. Lewis (1946) reported five patients with
penile tuberculosis and reviewed 110 cases reported
by other authors. He found that primary penile tuber-
culosis was reported in 89 cases (77.4%), secondary
in 6.9%, hematogenous in 2.6%, while in 10 patients
the source of infection was not known. Among the
89 primary cases, 72 were due to ritual methods of
circumcision, 12 were definitely transmitted through
vaginal sexual intercourse, 2 from oral genital sex,
and 3 from different sources.
39.2.3
Clinical Features
In male genital tuberculosis, systemic constitutional
symptoms may be lacking. However, low grade fever,
weight loss, or past history of tuberculosis may be
present. Symptoms related to pulmonary or extrapul-
monary tuberculosis may also be present.
39.2.3.1
Clinical Features of Tuberculous Epididymo-orchitis
Tuberculous epididymitis or epididymo-orchitis has
no distinctive clinical features that differentiate it
from other causes of testicular mass. A high index of
suspicion by the treating clinician that tuberculosis
may be the cause is important to avoid unnecessary
orchiectomy. Epididymitis may be the sole presenting
M. M. Madkour
feature of the disease being insidious and progres-
sive in the absence of constitutional symptoms or
other features of tuberculosis in other organs in the
body. Testicular pain or swelling may be associated
with fever, sweating, loss of appetite, and weight
loss. Scrotal abscess or discharging sinus may be
the presenting feature. Unilateral involvement is the
most frequently reported, but bilateral epididymal
involvement may occur. The time interval between
testicular symptoms and presentation to clinician
may vary between 2 and 8 weeks (Izawa et al. 1999;
Carbal et al. 1985).
The epididymis is often swollen with nodularity,
and it may be adherent to the testis and difficult
to differentiate. Primary testicular tuberculosis
without epididymal disease is rare. The testicular
mass may be hard, similar to carcinoma. Tender-
ness may be slight and fluctuance of the mass may
be detected. The scrotum may be normal, inflamed,
indurated and adherent to the testicular mass.
Scrotal sinus with seropurulent discharge may be
noted at the time of presentation. The vas deferens
may be thickened and beaded with multiple small
nodules. Another rare presentation of tuberculous
epididymitis is obstructive azoospermia leading to
infertility (Abdel Razic and EI-Morsy 1990). Moon
et al. (1999) treated 44 patients who presented with
azoospermia. Seven of these cases were due to
tuberculosis of the epididymis (15.9%). They also
reported that sperm retrieval and intracytoplasmic
injection (ICSI) from these patients resulted in suc-
cessful pregnancies in a rate similar to that of non-
tuberculous patients.
39.2.3.2
Clinical Features of Tuberculous Prostatitis
and Seminal Vesiculitis
The clinical features of tuberculous prostatitis
include nocturia, dysuria, frequency, or urinary
retention. Irritative bladder symptoms and hematu-
ria may be the initial presenting manifestation. Rectal
examination may reveal enlarged nontender prostate
with firm or cystic swelling or with irregular hard
multiple nodules of variable sizes simulating other
pathologies such as carcinoma (Sporer and Auerbach
1978). Kumar et al. (1994) reported the clinical fea-
tures of two patients with tuberculosis of the prostate.
The first patient was a 29-year-old man with a 4-year
history of dysuria, urinary retention, and difficulty in
voiding of urine, and three episodes of hematuria.
Rectal examination revealed cystic prostatic mass.
Ultrasound showed bilateral hydronephrosis and
Genitourinary Tuberculosis
by transrectal imaging it confirmed the prostatic
cyst which was aspirated. Histopathology of biopsy
samples from kidney and prostatic cyst confirmed
the tuberculosis. He was given 9 months antituber-
culous treatment. The second patient presented with
rectal tenesmus, irritative voiding, and recurrent uri-
nary retention. This patient had similar investigative
findings and a prostatic tuberculous abscess that
communicated with the urethra. He was also treated
for 9 months with chemotherapy. The spermatic
cords may be thickened and beaded and the seminal
vesicles may be tender and craggy.
39.2.3.3
Clinical Features of Penile Tuberculosis
Symptoms of penile tuberculosis are not characteris-
tic and patients may present with pain and swelling
of the penis of gradual insidious onset. Constitu-
tional symptoms may be lacking. Dysuria, urethral
discharge, or tender masses in the groin may be the
presenting symptoms. Erectile dysfunction in adults
with penile tuberculosis may be a rare presenting
symptom (Pal 1997; Murali and Raja 1998, Nishigori
et al. 1986, Chatterjee et al. 1975).
In infants who have been circumcised, the parents
may note delayed healing of the wound, discharge,
and pus oozing at the site of the wound and asso-
ciated with fever (Lewis 1946; Annobil et al. 1990).
Lewis (1946) reviewed the literature on penile tuber-
culosis, found 110 reported cases, and added five of
his own. Most patients were Jewish infants who had
undergone ritual circumcision and sucking of their
penis by the operator to ensure hemostasis, a practice
which has been abandoned in recent times. Annobil
et al. (1990), reported a 4-month infant with penile
tuberculosis. Circumcised at the age of 6 weeks by a
"local barber" who spat several times on the razor
during its sharpening. The wound did not heal and
a penile ulcer and discharging sinus in both groins
developed 3 weeks later. Culture of the discharge
yielded M. tuberculosis. Family and contact screen-
ing were negative but the barber was not screened.
Physical examination may show lesions on any part
of the penis including ulcers, erythematous erup-
tion, a plague, papulonodular lesion, necrotic area,
subcutaneous nodules, chancre-like eruption, lupus
vulgaris, penile orificial swelling, fistulas or sinuses,
cold abscess, complete destruction of the glans penis,
or even gangrene (Lewis 1946; Jaisankar et al. 1994;
Annobil et al. 1990; Dahl et al. 1996; Nakamura et
al. 1989; Pal 1997; Murali and Raja 1998; Atalay and
Karaman 1997).
713
39.2.4
Diagnosis of Male Genital Tuberculosis
The hemogram is usually normal and ESR may be
raised. Biochemical parameters in the blood are
usually normal. Urine microscopy and culture may
yield M. tuberculosis. Screening for HIV should be
carried out, particularly in homosexuals with penile
tuberculosis (Dahl et al. 1996). Chest radiography
may show an associated active pulmonary disease.
Screening for associated renal tuberculosis should
be done if genital TB is discovered. The diagnosis
of male genital tuberculosis depends on a high
index of suspicion in order to avoid unnecessary
orchiectomy among patients with tuberculous epi-
didymo-orchitis. Imaging features using various
modalities are helpful in identifying the disease.
Ferrie and Rundle (1983) reviewed 20 cases with
tuberculous epididymo-orchitis and found that 50%
had evidence of tuberculous infection in the urine. In
this series IVU was done on 19 patients, and 14 had
abnormal findings in the kidneys, ureters, and uri-
nary bladder. Fine-needle aspiration and biopsies for
histopathologic and microbiologic investigations are
the most important diagnostic tools for male genital
tuberculosis. Polymerase chain reaction (PCR) may
also be helpful.
39.2.5
Imaging Features
39.2.5.7
Ultrasound and Color Doppler
Ultrasound of the scrotal mass may show epi-
didymal enlargement predominantly in the tail,
less frequently in the head, and perhaps the entire
organ (Figs. 39.6-39.9). Marked heterogeneity of the
echotexture with areas of calcification and small
hydrocele may be depicted (Drudi et al. 1991). The
testis may be enlarged, nonhomogeneous with cen-
tral hypoechoic areas of variable sizes, however, these
echographic appearances are impossible to differen-
tiate from those of neoplasm (Figs. 39.6-39.9). An
irregular margin between the testis and epididymis
may also be present (Heaton et al. 1989; Chung and
Harris 1991; Kim et al. 1993). The sinus tract may
also be detected by ultrasound. Chung et al. (1997)
reported the sonographic findings on 18 patients
with tuberculous epididymitis and epididymo-orchi-
tis. They reported abnormalities in 22 hemiscrotums
of 18 patients. Unilateral epididymal abnormalities
 714 M. M. Madkour

Fig. 39.6a, b. A 28-year-old man presented with fever, sweating, weight loss, and right testicular pain and swelling for a few
months. He denied drinking raw milk and brucella serology and culture were negative. a Doppler ultrasound of both testes
demonstrates enlargement of the right testis. There is a generalized reduction in its echogenicity and enhanced vascularity in
comparison with the left testis. b Ultrasound of the right testis shows ill-defined area ofreduced echogenicity (arrowheads), this
area demonstrates increased vascularity on Doppler ultrasound (arrow). Fine-needle aspirate showed caseating granuloma

Fig. 39.7. Tuberculous epididymo-orchitis. A 28-year-old

male with left testicular pain and swelling for 2 months. He
had constitutional symptoms with fever. Ultrasound of the
scrotum demonstrates marked enlargement of the left epi-
didymis (arrowheads), areas of necrosis (open arrow), and
foci of calcification (arrow). The left testis is also involved
(T), and there is moderate hydrocele (h). Fine-needle aspira-
tion showed caseating granulomata
 Genitourinary Tuberculosis 715

Fig. 39.8a, b. A 45-year-old man presented with left testicular dull aching pain and swelling for 7 months. a Ultrasound of the
left testis demonstrates echolucent lesion (open arrows), containing floating echogenic foci consistent with debris (arrowheads).
The wall of the lesion is thick and edematous (arrow). Testicular abscess was diagnosed. b Doppler ultrasound demonstrates
increased vascularity around the wall of the abscess (arrowheads). Tuberculosis was diagnosed by fine-needle aspirate and
positive culture

a b

Fig. 39.9a, b. A 36-year-old male presented with fever, weight loss, and left testicular painful swelling for 1 month duration. He
was given antibiotics but was not responding to treatment. Hard left testicular mass was felt and was slightly tender. a Initial
ultrasound of the left testis demonstrates ill-defined hypoechoic lesion (arrowheads) involving the epididymis and adjacent
testicular tissue. Hydrocele is also noted (arrow). He was planned for orchiectomy but the patient refused the operation. b
The patient came 1 month later with worsening of his left testicular symptoms. Ultrasound demonstrates localized abscess
formation with irregular thick wall (arrowheads). Fine-needle aspiration showed caseating granuloma and the culture yielded
Mycobacterium tuberculosis

were found in 14 patients and bilateral in 4 patients.
All of theepididymides involved were enlarged (dif-
fuse in 9 and focal in 13). Decreased echogenicity
was observed in 13, increased in 2, and mixed in 7.
Draining sinuses were found in 4 and hydrocele in 12.
The testis were involved in 8 cases and had diffuse
enlargement of a hypoechoic testis and an ill-defined
or well-demarcated hypoechoic lesion.
Colored Doppler may be helpful in differentiat-
ing tuberculous epididymitis from nontuberculous
epididymal disease (Yang et al. 2000). Yang and
colleagues (2000) from South Korea performed
colored Doppler on 11 consecutive patients with
histologically proven tuberculous epididymitis and
epididymo-orchitis. Color Doppler ultrasound find-
ings correlated well with histologic findings. It dem-
716
onstrated no blood flow in the epididymal lesions
except for focal linear or spotty flow signals in the
peripheral portion.
Transrectal ultrasound of the prostate may show
hypoechoic areas with an irregular pattern in the
peripheral zone of the prostate. The seminal vesicles
may appear as enlarged and dilated but these ultra-
sound changes are not specific for tuberculosis. Cal-
cifications, abscess formations, or cavities that are
communicating from the prostate to the urethra or
the seminal vesicles may be noted by ultrasound as
well as by CT and MRI (Valentini et al.1998).
39.2.5.2
CTandMRI
In a contrast-enhanced CT of prostatic tuberculosis,
areas of inflammation, caseation, or abscess forma-
tion will appear as hypoattenuation. However, these
features are not specific for tuberculosis, but may
be depicted in nontuberculous pyogenic prostatic
abscesses. CT may depict seminal vesicle involve-
ment better than ultrasound. A tuberculous pros-
tatic abscess demonstrates peripheral enhancement
(Engin et al. 2000). Low signal intensity in the pros-
tate may appear as diffuse, radiating, or streaking
("watermelon skin" sign). Wang et al. (1997) noted
that a CT of the prostate may show abscess and calci-
Fig. 39.10. a US of the left testis demonstrates ill-defined
area of reduced echogenicity (arrowheads). Small amount
of hydrocele noted (open arrow). band c MRls of the testis:
sagittal (b) and axial (c) Tl-weighted images demonstrate
irregular area of intensified signal (arrowheads) consistent
with abscess formation
M. M. Madkour
fication including in the seminal vesicles. If no calcifi-
cation is present, then a CT of the upper urinary tract
with contrast may show abnormalities, otherwise the
abscess may look like other pyogenic abscesses.
MRI of epididymal and testicular tuberculosis has
the advantage of showing more detailed structural
changes than ultrasound (Fig. 39.10), yet these are
not specific for tuberculosis.
39.2.6
Tissue Biopsy and Fine-Needle Aspiration
Tuberculous epididymitis or epididymo-orchitis is
frequently diagnosed after orchiectomy and histo-
logic examination for suspected malignancy (Ferrie
and Rundle 1983). However, in endemic areas, par-
ticularly when other evidence of tuberculosis in other
body organs is present, fine-needle aspiration under
ultrasound guidance of epididymal tissue may spare
the unnecessary orchiectomy as well as confirming
the diagnosis. Fine-needle aspiration is also useful
in the diagnosis of infertility caused by obstructive
azoospermia due to tuberculous epididymitis.
Transrectal fine-needle aspiration of the prostate
mass, transurethral tissue resection, or incidental
findings after prostatectomy for other indications are
the usual diagnostic tools of tuberculous prostatitis.
c
Genitourinary Tuberculosis
Other investigations should include urine culture,
biopsy and culture of penile lesions and sinus dis-
charge, chest radiography, and screening for renal
tuberculosis.
39.2.7
Obstructive Azoospermia and Tuberculosis
Male infertility due to obstructive azoospermia
caused by tuberculosis of the epididymis, the vas
deferens, seminal vesicle, or ejaculatory duct may
be a presenting sequela of the disease (Abdel Razic
and EI-Morsy 1990). Diagnostic procedures include
history of tuberculosis, physical examination, plain
chest radiograph, IVD, semen analyses and fructose
measurement, hormonal assays, testicular biopsy,
transrectal ultrasonography, and vasography (Wong
et al. 1973). The incidence of tuberculosis as a cause
of obstructive azoospermia and infertility due to
ejaculatory duct obstruction was reported in 34% of
a group of 50 infertile men from Korea (Paick et al.
2000). Microsurgical reconstruction ofthe epididymis
or vas deferens, sperm retrieval, and intracytoplasmic
sperm injection (leSI) can be successful in achieving
fertilization and normal pregnancy outcome. Moon
et al. (1999) from Seoul, Korea, reported 44 patients
with destructive azoospermia that were divided into
2 groups: 7 patients with tuberculous obstructive
azoospermia and 37 patients with nontuberculous
obstructive azoospermia. The rates of fertilization
and embryo cleavage, clinical pregnancy, embryo
quality, and pregnancy outcome were comparable in
these two groups. Infertility due to obstructive azo-
ospermia caused by tuberculous epididymitis is still
prevalent in Korea. It occurs in approximately 25% of
those with obstructive azoospermia (Lee 1987).
Table 39.2. The frequency of organ involvement in female genital tuberculosis, as cited by different authors using various
diagnostic methods
Authors (year) Diagnostic procedures Fallopian Uterus Ovaries Cervix
of FGTB and no. of patients tubes % %
Sutherland (1985) Surgical findings 710 100
Agarwal and Clinical and pathologic 94.7
Gupta (1993) specimens 501
Tripathy and Clinical and laparoscopy 37 100
Tripathy (1990)
Nogalez-Ortiz (1979) Pathologic specimens 1436 100
717
39.3
Female Genital Tuberculosis
39.3.1
Epidemiology of Female Genital Tuberculosis
The frequency of female genital tuberculosis (FGTB)
among the general population is difficult to deter-
mine. In many patients the disease is asymptomatic
and only discovered accidentally (Schaefer 1976).
Female genital tuberculosis is rare in the developed,
industrialized world, but the incidence is rapidly
rising (Miranda et al. 1996; Lueken et al. 1997). The
frequency of involvement of different organs in the
female genital tract has been reported by many
authors (See Table 39.2).
The incidence of female genital tuberculosis
is much higher than the literature has suggested.
Infertility is the most common presenting reason for
investigations to find out the possible cause, therefore
most of the reports on female genital tuberculosis
were gathered from those infertile women. Agarwal
and Gupta (1993) from India reported 501 females
with genital tuberculosis. Young women aged 20-
30 years were the most commonly affected (82.3%),
and sterility was the most common presenting
symptom (68.5%). Menstrual disturbances were the
second most common complaint in 22.7% of patients,
abdominal mass in 3.5%, leukorrhea in 3.2%, abdom-
inal pain in 1.2%, vaginal prolapse in 0.4%, and vulva
ulcers in 0.2%. The uterus was involved in 99.5%,
the tubes in 94%, the cervix in 81.5%, the ovaries in
62.5%, and the vulva in 0.2%.
In an attempt to find the incidence of female
genital tuberculosis among women admitted to a
medical ward with bacteriologically proven pulmo-
nary tuberculosis, Tripathy and Tripathy (1990) from
Other organs
% % or sites
90 20 1 NA
99.5 62.5 91.5 Vulva 0.2
92 32 NA Pouch of Douglas 22
79 11 24 Myometrium 20
718
India performed a pelvic laparoscopy on 62 patients.
Of this group, aged between 15 and 45 years, 11 were
nulliparous, 15 para 1, 15 para 2, 13 para 3, and 8
para 4 or more. Laparoscopic findings were normal
in 25 women (40.3%) and abnormal with evidence
of genital tuberculosis in 37 (59.7%). The uterus was
abnormal in 34 women (bands and adhesion, hyper-
emia, tubercles nodules on the surface, and variations
in the uterine size). Abnormalities in the tubes were
noted in all 37 women (peritoneal adhesions, tuber-
cles on the surface and/or blockage). Endometrial
curettage showed histologic evidence of tuberculosis
in 4 women. The authors however did not indicate if
the nulliparous were among those with normallapa-
roscopic findings.
Sutherland (1985) analyzed his own personal series
of 710 cases of female genital tuberculosis. He noted
that the fallopian tubes were involved in almost all
cases, the endometrium in about 90%, the ovaries in
about 20%, the cervix in about 1%, and there were
no reports of tuberculosis in the vulva or vagina. The
average age of FGTB patient was 31 years, the youngest
16 years and the oldest 72 years. Infertility was present
in 84%, and only 16% of the married women gave a
history of pregnancy. In Riyadh, Saudi Arabia, the inci-
dence of female genital tuberculosis was reported to
be 0.45% of all gynecologic admissions. Infertility with
genital tuberculosis as its cause accounted for 4.2% of
cases (Chattopadhyay et al.1986).
In Spain, Nogales-Ortiz (1979) reviewed 78,000
gynecologic specimens studied between 1946 to 1977
and found the diagnosis of female genital tuberculo-
sis in 1,436 cases (1.8%) of the total number of speci-
mens. All patients presented with infertility (primary
in 94% and secondary in 6%), fallopian tubes were
affected in 100%, endometrium in 79%, myometrium
in 20%, the cervix in 24%, and the ovaries in 11 %.
39.3.2
Pathogenesis of Female Genital Tuberculosis
Hematogenous spread remains the most common
route of transmission from primary tuberculous
lesions most likely in the lung, similar to that observed
in other forms of genitourinary tuberculosis. History
of extragenital tuberculosis is found in more than
80% of these patients (Sutherland 1985). The fallo-
pian tubes are the most favorable site, specifically on
the mucosa. It usually affects both tubes at the same
time. Direct spread from neighboring organs such as
the gastrointestinal tract, peritoneum, or urinary tract
is also common. Sexual intercourse was the primary
M. M. Madkour
source of transmission as reported by Sutherland
(1985). He noted that "In my own series, 5 out of 128
husbands examined by a urologist had active genito-
urinary tuberculosis and transmission to their wives
during sexual intercourse has occurred:'
Seeding tuberculomas form in the mucosa of the
fallopian tube bilaterally (Schaefer 1976; Saracoglu et
aI.1992).As these advance and progress,caseation and
necrosis with microabscess formation and blockage
of the tubes may occur. The lumen may be distended
and filled with purulent or serosanguinous fluid,
and it may be calcified (Fig. 39.11). Direct spread to
the uterus, cervix, and abdominal cavity may occur.
Uterine endometrium and even to a lesser extent the
myometrium may also be affected. This process is
always associated with extensive fibrosis, adhesion
to neighboring organs, calcification, and structural
changes, and may even lead to fistula formations in
the sigmoid colon, appendix, cecum, ileum, rectum,
and the urinary bladder (Sbihi et al. 1980; Silva et al.
1988). Direct spread of infection to the genital organs
may occur from contiguous abdominal tuberculous
lymphadenitis, peritonitis, or other viscera (Schaefer
1976). Zamberletti et al. (1981) tried to find out the
reasons for such poor prognosis as far as regaining
fertility is concerned after treatment. They exam-
ined five fallopian tubes by electron microscope to
determine the histomorphology. The authors noted
mucosal damage and that the architecture of the
folds and cellular population were compromised. The
folds were low in height and their directions were not
coaxial with the lumen of the tube. The number of
ciliated cells was very low, even scarce, and the cilia
were thin and disordered, with irregular heights.
None of these findings were restored by antituber-
culous chemotherapy. The uterus may appear bulky
or atrophic, calcified (Fig.39.11), hyperemic with
tubercles on the surface (Tripathy and Tripathy 1990;
Saracoglu et al. 1992). The ovaries are affected by adhe-
sions from fallopian tubes rather than from uterine
involvement. Cervix erosions and ulcers may occur
by direct spread of tuberculous infection from the
uterus. The cervix may also show a polyp, papillary
growth, or granulation tissue formation. The vagina
may be involved, although this is rare. Vesicovaginal
fistula may be the presenting feature of female genital
tuberculosis (Ba-Thike et al. 1992; Kutteh and Hatch
1992; Arora et al. 1994; Singh et al. 1988).
Vaginal tuberculosis is usually caused by direct
extension of the infection from uterine tuberculosis
(Coetzee 1972; Ma et al.1997). Vulval lesion is rare and
infection occurs as a result of direct spread from upper
genital tract tuberculosis or from sexual intercourse.
Genitourinary Tuberculosis 719

months. As pelvic pain becomes severe, it is aggravated

by sexual intercourse and during menstruation. Lower
abdominal pain may appear and episodic acute attacks
may occur. Fatigue, malaise, low grade fever, loss of
appetite, and weight loss may be noted. However, the
presenting features of the disease may be menstrual
disturbances such as amenorrhea, menorrhagia, oli-
gomenorrhea, irregular periods, or dysmenorrhea in
approximately 20% of patients (Agarwal and Gupta
1993). Postmenopausal bleeding may be the present-
ing feature in approximately 4%. Vaginal discharge
may be a rare presenting feature. Vaginal dribbling of
urine due to a vesicovaginal fistula may rarely be the
presenting symptom (Ba-Thike et al.1992).An ulcer in
the vagina is rarely a presenting feature. The frequency
of symptoms at the time of presentation in some series
is shown in Table 39.3.

Fig.39.11. A 50-year-old woman with past history of tuber-
culosis and infertility. Plain radiograph of the pelvis demon-
strates calcification of the fallopian tubes (arrowheads)
39.3.3
Clinical Features of Female Genital Tuberculosis
39.3.3.1
Presenting Symptoms
Female genital tuberculosis may pass unnoticed by
the patient for years and only infertility, occurring
in approximately 85% of cases, may be the first pre-
senting symptom (Schaefer 1976; Sutherland 1982;
Agarwal and Gupta 1993). Past history of tubercu-
losis mostly of the lung or less frequently of other
systems is noted in approximately 50% of patients.
Family history of tuberculosis may be found in 20%
of patients suffering from female genital tuberculosis.
Pelvic pain may occur in 50% of patients, but this
is usually mild and the time interval between this
symptom and seeking medical advice may be several
39.3.3.2
Physical Signs
Physical signs may be lacking or point to features
of pulmonary involvement, spinal tenderness, or
"doughy" sensation during abdominal palpation.
Lower abdominal tenderness may also be discov-
ered on deep palpation. Bimanual pelvic examination
may give the only positive signs if adnexal masses,
which are tender, are felt. Other rare physical signs
include leukorrhea and vulval ulcer. Bimanual pelvic
examination may show normal findings, though, if
the tubes are not enlarged. Tube ovarian mass may
be felt with variable sizes. The mass may be firm or
fluctuant and only slightly tender. Genital carcinoma
may coexist with genital tuberculosis.
39.3.3.3
Tuberculosis of the Fallopian Tube and Ovary
(Features and Investigations)
Tubal tuberculosis is the leading cause of infertility in
endemic areas of developing countries (Wang 1989,
Punnonen et al. 1983). In India, Parikh et al. (1997)

Table 39.3. Female genital tuberculosis: the frequency of presenting symptoms

Author (year) Total Infertility: Pelvic!abdominal Vaginal Amenorrhea: Vaginal Postmenopausal

number number (%) pain: number (%) bleeding: number (%) discharge: bleeding:
number (%) number (%) number (%)

Agarwal and 501 343 (68.5) 6 (1.2) 58 (U.S) 56 (11.2) 16 (3.2) 2 (0.4)
Gupta (1993)
Otieno (1983) 12 12 (100) 8 (66) 4 (33) 12 (100) 5 (40) NA
Sutherland (1985) 710 313 (44) 179 (25.2) 127 (17.9) 36 (5) 26 (3.6) 16 (2.2)
Saracoglu et al. 72 34 (47.2) 23 (31.9) 8 (1Ll) 1 (1.4) NA 2 (2.8)
(1992)
720
reported 300 women attending a specialized infertility
center, with tubal factors as the cause of their infertil-
ity. One hundred and seventeen women were found
to have tubal tuberculosis. In China, Yang et al. (1996)
reported 1,120 women with tubal factor infertility,
and tubal tuberculosis accounted for that infertility,
in 63.6% of these women. In Turkey, Saracoglu et al.
(1992) reported 72 women with pelvic tuberculosis.
Hysterosalpingography were performed on 34 patients
and tubal blockage was noted in 32 of them. Wang
(1989) reported 66 women with infertility due to tubal
obstruction as depicted by hysterosalpingography.
During operations, tubal patency was found in 28.8%
of women reported to have obstruction by hysterosal-
pingography. Tuberculosis as a cause of tubal obstruc-
tion was found in 31.3% of pathologic examinations
of 48 specimens. The coexistence of tubal tuberculosis
and primary carcinoma is rarely reported (Wiskind et
al. 1992). Plain radiography of the pelvis may depict
calcifications of the tubes and uterine (Fig. 39.11),
and may depict peritoneal involvement as well as of
the tuboovarian mass (Crowley et al. 1997; Tang et al.
1984; Svendsen et al. 1985). Hysterosalpingography
may depict tubal obstruction and tuboappendiceal,
tuboparietal, salpingosigmoid, or tubointestinal fistu-
las (Silva et al. 1988; Sbihi et al. 1980). Laparoscopy/
laparotomy is essential for the histopathologic analysis
(Tripathy and Tripathy 1990). Pregnancy after antitu-
berculous treatment of tubal destruction is very rare
and IVF should be attempted.
In vitro fertilization success or failure has been
assessed. Marcus et al. (1994) reported 10 patients
with genital tuberculosis with infertility. Clinical preg-
nancies after IVF were achieved in six patients and
resulted in three live births, one ectopic pregnancy
and one abortion. Patients with trophic endometrium
achieved pregnancy in 42.9% of cases, and those with
atrophic endometrium had a 0% success rate (no preg-
nancy). Similar findings were reported by Soussis et al.
(1998) in 13 women with genital tuberculosis who had
undergone IVE The ovaries are almost always involved
as part of tube and uterine tuberculosis. Agarwal and
Gupta (1993) reported two cases with purely ovar-
ian involvement with multiple large cysts 6-10 cm in
diameter filled with hemorrhagic and purulent fluid.
39.3.3.4
Tuberculosis of the Uterus
(Features and Investigations)
Tuberculous endometritis may occur in isolation
without other genital or extragenital involvement
(Shireman 1992). Pregnancy may still occur despite
M. M. Madkour
endometrial tuberculosis, and it may progress to full
term. Fetal congenital infection in the premature
infant may be a sequela (Myers et al. 1981; Balasub-
ramanian et al. 1999; Yip et al. 1999; Baumgartner et
al. 1980). Postmenopausal endometrial tuberculosis
is reported in up to 11 % of patients (Nogales-Ortiz
1979; Mantovani et al. 1998; Castelo-Branco et al.
1995; Dhillon et al. 1990; La Grange 1982).
The most common cause of uterine tuberculosis
is Mycobacterium tuberculosis but nontuberculous
mycobacterium may also cause endometritis with
similar clinical and pathologic features including
irregular menses and infertility. Rao et al. (1992)
reported 140 infertile women with endometrial
biopsies. Positive cultures of organisms were found
in 45%. Mycobacterium tuberculosis was identified in
eight patients and nontuberculous mycobacterium
organisms in 14 patients with M. scrofulaceum, M.
kansasii, M. fortuitum and other cultured organisms
also found in endometrial samples. Sterile endome-
trial specimens were noted in 76 patients.
An endometrial biopsy is usually done to investi-
gate either infertility or abnormal uterine bleeding
and is an essential tool to determine the cause. Histo-
pathologic examination and culture of endometrial
samples may yield evidence of tuberculosis which
may vary in its rate of positive yield from one series
to another depending on the endemicity of tuber-
culosis and the vigilance of the reporting medical
center. In Nigeria, Emembolu (1989) reported the
microbiologic cultures of endometrial biopsy sam-
ples in 114 infertile women. Various organisms were
grown in 42 patients and Mycobacterium tuberculosis
was found in seven (16.7%) and represented the most
common organism isolated as a cause of infertility.
Adewole et al. (1997) from Nigeria analyzed the
microbiologic and histopathologic findings of 7,211
endometrial specimens of women who presented
with infertility or abnormal vaginal bleeding due to
various causes, and found tuberculous endometritis
in 0.49%. In Turkey, Aka and Vural (1997) investi-
gated 57 women with active pulmonary tuberculosis,
attempting to identify evidence of genital tuberculo-
sis involvement. Menstrual blood culture, curettage,
and hysterosalpingography were performed. Evi-
dence of genital tuberculosis was found in 12.3% of
these women. Endometrial aspirates from cytologic
and histopathologic examinations were compared as
tools to diagnosing endometrial tuberculosis. Tripa-
thy and Mohanty (1990) found that cytology was reli-
able in 93% of cases in their series.
Polymerase chain reaction (peR) testing of
endometrial tissue and menstrual blood culture
Genitourinary Tuberculosis 721

were used by Hasimoto et al. (1994) and Missirliu et tal tuberculosis. Higher incidence was reported by
al. (1996) and were found to be positive in all cases Agarwal and Gupta who noted that the cervix was
they studied. PCR became negative 3 months after involved in 81.5% of patients in their series of 501
initiating treatment with antituberculous drugs. women with genital tuberculosis. Primary tubercu-
Difficulties in diagnosing endometrial sarcoidosis lous cervicitis without other genital involvement was
(which may rarely occur) is mainly due to the simi- noted by Chakraborty et al. (1992) in three women
larity of its clinical and pathologic features to those of his series of 91 patients with proven genital tuber-
of endometrial tuberculosis. Murphy et al. (1992) culosis living in the Darjeeling Hills, an area highly
reported a woman with irregular menses and pul- endemic with tuberculosis. Four other women had
monary sarcoidosis and a positive family history of cervicitis with uterine involvement. Other authors
sarcoidosis. Endometrial biopsy showed histopatho- have reported tuberculous cervicitis as an isolated
logic changes of granulomas. Endometrial sarcoid- primary site (Vuong et al. 1989).
osis and difficulties in the diagnosis were similarly Tuberculous cervicitis is commonly reported in
noted and reported by Skehan and McKenna (1985). young women but may also occur after menopause
The trophic or atrophic state of the endometrium (Muechler and Minkowitz 1971). Tuberculous cervi-
in patients with tuberculous endometritis will citis has no distinctive clinical or macroscopic fea-
determine the rate of success of pregnancy or IVF tures and may simulate malignancy or other granu-
as noted before. Hysteroscopic examination of the lomatous diseases such as bilharziasis, amebiasis,
uterus is a helpful tool of investigation among those syphilis, or granuloma inguinale. Koller (1975) from
presented with infertility, though it is rarely used. South Africa reported 46 women with granulomatous
cervicitis. Tuberculosis was found in 21, bilharziasis
39.3.3.5 in 21, and amoebic cervicitis in four patients. Tuber-
Synechia Uteri: Asherman's Syndrome culous cervicitis may initially be mistaken for carci-
(Features and Investigations) noma, particularly in nonendemic areas of developed
countries (Shobin and SaIl Pellman 1976; Koller 1975;
Intrauterine adhesions and fibrosis leading to infer- Vuong et al. 1989; Chahtane et al. 1992; El Mansouri
tility, premature labor, miscarriage, placenta previa, 1995). The coexistence of carcinoma and tuberculous
or menstrual disorder is also known as Asherman's cervicitis was reported by some authors, yet it is rare
syndrome. The most common cause in endemic (Yamabe et al. 1972; Hsu et al. 1985; Bhambhani et al.
areas of developing countries is uterine tubercu- 1985). The initial symptoms at presentation include
losis. Other causes include Bilharziasis, trauma to vaginal discharge, bleeding after sexual intercourse,
pregnant uterus or after curettage or due to other menorrhagia, and weight loss. A past history of
infections (Krolikowski et al. 1995). Plain radiogra- extragenital tuberculosis may also be found. Mac-
phy may show calcifications of the uterus (Fig. 39.11). roscopically the cervix is bulky, hypertrophied with
Hysterosalpingography may show the classic «glove ulcerations or erosions of the external OS. A necrotic
finger appearance." Laparoscopy/laparotomy and or fungating mass that bleeds easily with touch may
hysteroscopy, as well as the histopathologic findings be noted. Cytologic examination of cervicovaginal
of an endometrial biopsy, can confirm the diagno- smear may be helpful in diagnosing tuberculous
sis. Antituberculous treatment with hysteroscopic cervicitis. Granulomatous formation with epithelioid
synechialysis may help in restoring fertility on rare cells and Langhans giant cells may be seen (Misch
occasions. The prognosis with regard to fertility is et al. 1976; Miller and Herman 1979; Rivasi et al.
poor among these women (Bukulmez et al. 1999; 1980; Angrish and Verma 1981; Vuong et al. 1989).
Marcus et al. 1994). Histopathologic examination from cervical biopsy
sample and tissue culture will confirm the diagnosis.
39.3.3.6 Hysterosalpingography may be difficult to perform
Tuberculosis of the Uterine Cervix because of the size of the mass. Intravenous pyelog-
(Features and Investigations) raphy may depict an associated bilateral kidney and/
or ureteric involvement with calyceal deformities
Tuberculous cervicitis is commonly associated with (Shobin and Pellman 1976; Husermeyer 1977). Plain
other types of genital tuberculosis but may rarely radiography may show calcifications of the cervix,
occur in isolation as the primary site. The incidence uterus, and the tubes (Fig. 39.11). Chest radiography
of tuberculous cervicitis reported by most authors may show active pulmonary disease or fibrosis and
ranges between 0.4% and 16% of patients with geni- calcifications.
722
39.3.3.7
Tuberculosis of the Vagina
(Features and Investigations)
Vaginal tuberculosis is rare, and its incidence may
range between 0.07% and 0.2% (Nogales-Ortiz 1979;
Agarwal and Gupta 1993). Patients with HIV-posi-
tive serology may present with tuberculous vaginitis
(Arora et al. 1994). Dyspareunia, vaginal discharge,
history of urinary dribbling per vagina, or primary
infertility may be the presenting clinical features.
Pelvic examination may be difficult due to pain or
narrowing of the vagina by a circular band of fibrous
tissue, and examination under anesthesia may be
necessary. Urine may be found to be leaking from
the vagina and the orifice of the vesicovaginal fistula
may be felt in the anterior wall. Painful ulcer at the
posterior fourchette and lower vagina may be seen
(Coetzee 1972; Ma et al. 1997). Vesicovaginal fistula
may also be depicted by cystogram or intravenous
pyelography (Ba-Thike et al. 1992; Singh et al. 1988).
Confirmation of tuberculous vaginitis is done by
tissue biopsy of vaginal ulcer or other upper genital
specimens for histopathologic and microbiologic
examinations. Vaginal carcinoma and tuberculosis
may coexist and a high index of suspicion is essen-
tial (Kutteh and Hatch 1992) to avoid missing this
rare coexistence.
39.3.3.8
Tuberculosis of the Vulva
(Features and Investigations)
The vulva is the least common site of genital tuber-
culosis. It is usually associated with tuberculosis in
other sites including upper genital organs, chest, or
spine (Bhattacharya 1978; Pena et al. 1973; Schaefer
1970; Stewart 1968).
Rarely, vulval tuberculosis may be the only pri-
mary site, and transmission through intercourse
particularly in immunocompromised patients with
HIV may occur (Vani et al. 1993; Naika et al. 1987;
Millar et al. 1979).
The initial symptoms include vulval swelling,
ulceration, dyspareunia, or apareunia, vaginal dis-
charge, and weight loss. The most frequent signs are
ulcers with irregular outlines and undermined edges
affecting the labia. The ulcers are tender, friable, and
may bleed when touched. Inguinal lymphadenopa-
thy is a common finding. The second least common
signs are those of the swollen hypertrophied type
with thickened skin, nodular swellings, multiple
wart-like lesions, lupus vulgaris, with or without
M. M. Madkour
ulceration. Multiple discharging sinuses may also be
noted. These clinical features are not characteristic
and cannot be distinguished from other causes such
as carcinoma, bilharziasis, amebiasis, syphilis, or
elephantiasis (Heuvet et al. 1979; Gras et al. 1980).
Histopathologic examination and culture of biopsy
samples of vulval lesions are essential to confirm the
diagnosis. Investigations of other genital or respira-
tory tuberculosis may give associated important
clues to the diagnosis.
Medical treatment alone may be sufficient, par-
ticularly in those who present with ulcers. In the
hypertrophic type, surgical excision of the lesion
may be necessary.
39.3.3.9
Investigation of Female Genital Tuberculosis
The hemogram and blood biochemistry is usually
normal. Plain chest radiography may show fea-
tures of active tuberculosis or evidence of healed,
scarred, old lesions and plain pelvic radiography
may depict calcification of both tubes and uterus
(Fig. 39.11). A hysterosalpingography, although
relatively contraindicated in acute inflammatory
pelvic disease, is an essential tool for investigat-
ing women presenting with infertility (Siegler and
Koutopoulos 1979). Fallopian tube lumen may look
like saw teeth with ragged contours and multiple
strictures. The ampulla may appear tufted due to
occlusion. As fibrosis develops, the tube may appear
like a pipe stem in configuration. Filling and spillage
of contrast media may not appear in either or both
tubes due to obstruction. Tubal diverticulosis or fis-
tula formation in the appendix, cecum, ileum, or colon
may occasionally be seen in a hysterosalpingography
(Siegler and Koutopoulos 1979; Silva et al. 1988). The
uterine cavity may be deformed due to adhesions and
endometrial destruction. The endocervical canal
may be elongated and dilated. Lymphatic or venous
intravasation of contrast media may be noted in
endometrial and myometrial tuberculosis. Other
imaging modalities playa role in the diagnosis such
as transvaginal ultrasound and postenhanced pelvic
CT. Axial postenhanced CT of the pelvis may depict
cystic masses with ring enhancement in the fallopian
tubes, peritoneal thickening, and ascites (Fig. 39.12a,
b). MRI may show a tubo-ovarian cyst and peritoneal
effusion (Valentini et al. 1998).
 Genitourinary Tuberculosis 723

a b

Fig. 39.12a, b. A 23-year-old female married for 4 years presented with primary infertility. She denied any history of fever but
had weight loss, and abdominal and pelvic discomfort. On examination, she had a doughy abdomen with ascites but no organo-
megaly. a Axial postenhanced CT of the pelvis demonstrates peritoneal thickening particularly anteriorly (arrowheads) and
ascites (arrow). b Axial CT caudal to (a) demonstrates sizable ascites surrounding the uterus (curved arrows), complex cystic
masses with ring enhancement at the fibrillated end of the fallopian tubes (arrow heads). Tuberculous salpingo-oophoritis was
diagnosed by laparoscopic biopsy and positive culture

39.3.4 including salpingolysis and fibrinolysis, and tubal

Microbiology and Histopathology
Dilatation and curettage of the uterus for histopathol-
ogy and microbiological examination is important.
Positive histopathology and microbiology results
from endometrial specimens were present in 79%
to 99.5% of those with female genital tuberculosis as
reported by sources (Nogales-Ortiz 1979; Agarwal and
Gupta 1993). The best time for obtaining endometrial
biopsy samples is several days before the expected
menstrual period (Schaefer 1970, 1976) at which time
the tubercles reach their maximum growth.
39.3.5
Laparoscopy
Laparoscopic examination is an important diagnostic
procedure in investigating female infertility for those
suspected to have female genital tuberculosis without
symptoms but with abnormal hysterosalpingography.
Abnormal laparoscopic findings were mentioned
in "Features and investigations" for female genital
tuberculosis.
39.3.6
Treatment of FGTB
The chemotherapy regimen for FGTB is similar to
that for the genitourinary tract. The role of surgery,
anastomosis may be helpful in some patients for
restoring fertility. Salpingectomy or salpingo-oopho-
rectomy may rarely be required in some patients
(Saracoglu et al. 1992).
39.4
Tuberculosis in Patients with Chronic
Renal Failure, on Hemodialysis,
or with Renal Transplant
Patients with chronic renal failure, renal transplant,
and those on hemodialysis are considered as high-
risk groups. They have an increased susceptibility
to, and high prevalence of, infections with various
pathogens including Mycobacterium tuberculosis
(AI-Homrany 1997).
In these groups of patients, the cell-mediated
immune responses are impaired, the cytotoxic activ-
ity of natural killer lymphocytic cells CD4+, the
migration and responses of activated macrophages,
the production of cytokines including interleukins,
tumor necrosis factor a, and interferon-g are all sup-
pressed. The development of tuberculosis among
these high-risk groups of patients may be a major
cause of morbidity and mortality, if not suspected
early and treated promptly. The high mortality was
reported to range from 11% up to 75% in different
series (AI-Homrany 1997). These patients are immu-
nosuppressed by the underlying disease, dialysis,
724
or by use of the immunosuppressive agents used in
renal transplants.
39.4.1
Epidemiology
There are many case reports on tuberculosis among
this high-risk group of patients, yet prevalence for
only a few series has been reported. Some authors
have even indicated the rarity of tuberculosis among
these groups. The prevalence of tuberculosis among
these groups is closely related to the prevalence of
tuberculosis in the general population in any com-
munity. It tends to be several times higher compared
with the general population. In Japan, the incidence
was reported to be as much as 6 to 16 times higher.
In the United States, it was 10-15 times higher, and
in Saudi Arabia, 50 times higher than in the general
population (Murthy et al. 1997; Wajeh et al. 1990).
Wajeh et al. (990) from Saudi Arabia reported 403
renal transplant recipients and found tuberculosis
in 14 patients (3.5%). In these patients, tuberculosis
was disseminated in 64.3%, pulmonary in 28.6%, and
genitourinary in 7.1%. These authors noted nontu-
berculous mycobacterium in 29% of those with dis-
seminated disease. The mortality was higher among
those with disseminated tuberculosis than with
other forms (37% and 11%, respectively). The time
interval between transplantation and development of
tuberculosis ranged from 1 to 84 months, with male
predominance and with a mean age of 35.7 years. In
another large series from Jeddah, Saudi Arabia, AI-
Shohaib and colleagues (Al-Shohaib 2000) reported
137 patients with chronic renal failure waiting for
hemodialysis. Eighty were followed up for a period of
3 years and monitored for the development of tuber-
culosis. Eight patients did develop tuberculosis (10%)
before starting on hemodialysis. Pulmonary tubercu-
losis was the most common (50%), followed by renal
(20%), cervical lymphadenitis 00%), and tuberculous
meningitis in 10%. Female predominance was noted
in this series (62.5%) and ages ranged between 35 and
55 years. The same authors reported their experience
1 year earlier (Al-Shohaib et al. 1999), with regard to
the development of tuberculosis in their patients while
on hemodialysis. They found 17 of21O patients (8.1 %)
developed tuberculosis, among which they found
pulmonary tuberculosis in 10, and lymphadenitis in
8 (one had tuberculosis in the lungs and lymph node
sites, while in another patient, pulmonary and peri-
toneal sites were involved). Male predominance was
noted and the mean age was 48 years.
M. M. Madkour
In Riyadh, Saudi Arabia, Al-Homrany et al. (Al-
Homrany 1997) reported a large series of 250 patients
on maintenance hemodialysis seen during 1986 to
1993 inclusive. The authors reported tuberculosis
in 13 patients (4.8%), with female predominance
(77%) and a mean age of 51 years. The duration of
hemodialysis before diagnosis of tuberculosis was
19.23±14.43 months, and the mean duration of symp-
toms before diagnosis was 4.54±3.38 weeks. More
recently, Al-Jondeby et al. (2001) conducted a national
review of 5,706 patients on hemodialysis in different
regions of Saudi Arabia. These authors reported the
incidence of tuberculosis among these patients as
ranging from 8% to 10%, with male predominance at
a ratio of 2:1 and a mean age of 51.3±1.0 years.
39.4.2
Clinical Features onuberculosis in these Groups
The clinical presentations of tuberculosis among
these high-risk groups are usually insidious and
nonspecific. Symptoms such as fatigue or anorexia
are not unusual, but these are not clear indications
of the occurrence of tuberculosis. Therefore the diag-
nosis can be delayed for weeks unless the treating
nephrologist has a high index of suspicion that tuber-
culosis may be the cause of these nonspecific symp-
toms (AI-Homrany 1997; AI-Shohaib et al. 1999). The
most common symptoms are fever, anorexia, fatigue,
loss of appetite, weight loss, and cough. Other clini-
cal features at the time of presentation may include
hemoptysis, lymphadenopathy, erythema nodosum,
or abdominal distension due to tuberculosis perito-
nitis. Pulmonary tuberculosis is the most common
finding in 40% to 70% of patients, followed by dis-
seminated multisystem disease in 30% to 40%, fol-
lowed by extrapulmonary localizations (Wajeh et al.
1990; Cengiz 1996; AI-Homrany 1997; Al-Shohaib et
al. 1999; Al-Shohaib 2000).
39.4.3
Investigations
The hemogram may show normocytic normochromic
anemia with normal or accelerated sedimentation rate.
Biochemical parameters may show findings of chronic
renal failure in nondialysed or transplant patients: a
PPD test may not be of diagnostic help as it may be
positive in only 40% to 60% of patients, which does not
necessarily indicate active disease. Positive PPD may
be of significance, particularly with recent contact, in
Genitourinary Tuberculosis
countries with a low prevalence of tuberculosis among
the general population. Plain chest radiography should
be done when tuberculosis is suspected. Attempts to
identify the organisms from sputum, other body fluids,
or tissue should also be carried out.
Bacteriologic confirmation of diagnosis among
these high-risk groups was achieved at different
rates in different series and ranged from 50% to
100% (Wajeh et al. 1990; Mitwalli 1991; Cengiz 1996;
AI-Homrany 1997; Murthy and Periera 1997; Al
Shohaib 2000). Nontuberculous mycobacterium has
been reported among these high-risk groups. These
include M. kansasii, M. fortuitum, M. marinum, M.
avium, M. abscessus chelonae, M. haemophilum, and
M. scrofulaceum (Wajeh et al. 1990).
39.4.4
Treatment of Genitourinary Tuberculosis
The chemotherapy used for the treatment of genito-
urinary tuberculosis is similar to the regimen given for
pulmonary tuberculosis and 6 months duration is suf-
ficient to achieve cure (Garcia-Rodriguez et al. 1994).
Antituberculous agents for patients on hemodi-
alysis and fully functioning renal allografts should
be the same as in the standard regimens used for
immunocompetent patients. There has been no con-
trolled trial with regard to the optimum regimen in
these groups of patients. The Committee on Chemo-
therapy of Tuberculosis in the United States recom-
mended isoniazid and rifampicin, supplemented in
the initial phase by either ethambutol, streptomycin,
or pyrazinamide in immunosuppressed patients.
The duration of treatment is usually 12-24 months.
Shorter courses of treatment (6 months) with direct
observed therapy (DOT) is also effective in these
groups of patients. Streptomycin can be used twice to
thrice weekly and with serum level assays to avoid its
nephrotoxicity and nerve deafness. Isoniazid, rifam-
picin, and para-aminosalicyclic acid are generally
safe in this group of high-risk patients. AI-Homrany
(I997) treated 13 tuberculous patients on hemodialy-
sis using 3 or 4 drug regimens for 6 to 12 months with
full recovery and no recurrences on follow-up.
39.4.5
Dose Modification
The efficacy of steroids and cyclosporin used in
the management of renal allograft recipients may
be diminished in patients that are receiving rifam-
725
picin. Hepatic microsomal enzymes induction by
rifampicin may increase the clearance of steroids
and cyclosporin, so that the dosage may need to
be doubled and adjusted according to cyclosporin
serum levels. Streptomycin can be given to hemo-
dialysis patients 6-8 hours before each dialysis, and
serum trough level as well as creatinine clearance has
to be measured.
Chronic renal failure patients who are waiting
for hemodialysis may use pyrazinamide three times
weekly, however, some authors suggested avoiding its
use in those with severe chronic renal failure. Strep-
tomycin can be used in chronic renal failure patients
two to three times weekly with close monitoring of
serum trough levels as well as urea and creatinine.
Ethambutol is excreted unchanged in urine and
patients with chronic renal failure are at risk of suf-
fering from its adverse effects (loss of visual acuity,
color blindness). It is better avoided in these patients,
but if it has to be used, the dose should be reduced to
5 to 10 mg/kg daily.
References
Abdel Razic MM, EI-Morsy FE (1990) Genitourinary mycobac-
teria in infertile Egyptian men. Fertil Steril54:713-717
Adewole IF et al (1997) The value of routine endometrial
biopsy in gynaecological practice in Nigeria. West Afr J
Med 16:242-245
Agarwal J, Gupta JK (1993) Female genital tuberculosis: a
retrospective clinicopathologic study of 501 cases. Indian J
Pathol MicrobioI36:389-397
Aka N, Vural EZ (1997) Evaluation of patients with active
pulmonary tuberculosis for genital involvement. J Obstet
Gynaecol Res 23:336-340
AI-Homrany M (1997) Successful therapy of tuberculosis in
hemodialysis Patients. Am J NephroI17:32-35
Al-Jondeby MS et al (2001) Caring for hemodialysis patients
in Saudi Arabia, past, present and future. Saudi Med J 22:
199-204
AI-Shohaib S (2000) Tuberculosis in chronic renal failure in
Jeddah. J Infect 40:150-153
Al-Shohaib et al (1999) Tuberculosis in active dialysis patients
in Jeddah. Am J NephroI19:34-37
Alvarez S, McCabe WR (1984) Extrapulmonary tuberculosis
revisited: a review of experience at Boston City and other
hospitals. Medicine Anal Rev Gen Med Neurol Dermatol
Pediatr 63:25-55
Angrish K, Verma K(1981) Cytologic detection of tuberculosis
of the uterine cervix. Acta CytoI25:160-162
Annobil SH,AI-Hilfi Aet al (1990) Primary tuberculosis of the
penis in an infant. Tubercle 71:229-230
Arora VK et al (1994) Tuberculosis of the vagina in an HIV
seropositive case. Tuberc Lung Dis 75:239-240
Atalay AC, Karaman MI (1997) Distal penile gangrene in a
patient with chronic renal failure. Int J Urol 4:431-432
726
Balasubramanian S et al (1999) Congenital tuberculosis. Ind J
Pediatr 66:148-150
Balasubramanian V et al (1994) Pathogenesis of tuberculo-
sis: pathway to apical localization. Tuberc Lung Dis 75:
168-178
Ba-Thike K et al (1992) Tuberculosis vesico-vaginal fistula.1nt
J Gynaecol Obstet 37:126-130
Baumgartner W et al (1980) Congenital tuberculosis (author's
transation). Monatsschr Kinderheilkd 1280:563-566
Beck-Sague C et al (1992) Hospital outbreak of multidrug-
resistant mycobacterium tuberculosis infections. Factors
in transmission to staff and HIV-infected patients. JAMA
268:1280-1286
Berman LB et al (1960) Glomerular abnormalities in tubercu-
losis. Arch PathoI69:278-285
Bhambhani S et al (1985) Cervical tuberculosis with carci-
noma in situ: a cytodiagnosis. Acta CytoI29:87-88
Bhattacharya P (1978) Hypertrophic tuberculosis of the vulva.
Obstet Gynaecol1 [Suppl]:21s-22s
Birnbaum BA, Friedman JP, Lubat E et al (1990) Extra-renal
genitourinary tuberculosis: CT appearance of calcified
pipe-stem ureter and seminal vesicle abscess. J Comput
Assist Tomogr 14:653-655
Buchholz NP et al (2000) Genitourinary tuberculosis: a profile
of 55 in-patients. J Pak Med Assoc 50:265-269
Bukulmez 0 et al (1999) Total corporal synechiae due to tuber-
culosis carry a varied prognosis following hysteroscopic
synechialysis. Hum Reprod 14:1960-1961
Cabral DA, Johnson HW et al (1985) Tuberculous epididymitis
as a cause of testicular pseudomalignancy in two young
children. Pediatr Infect Dis 4:59-62
Cantwell MF et al (1994) Epidemiology of tuberculosis in the
United States, 1985 through 1992. JAMA 272:535-539
Carl P, Stark L (1997) Indications for surgical management of
genitourinary tuberculosis. World J Surg 21:505-510
<;:askurlu T et al (1998) Urinary tuberculosis in a two-year-old
boy. Int Urol Nephr 30:525-528
Castelo-Branco C et al (1995) Primary adenocarcinoma of
the endometrium associated genital tuberculosis: a case
report. J Reprod Med 40:673-675
Cengiz K (1996) Increased incidence of tuberculosis in
patients undergoing hemodialysis. Nephron 73:421-424
Centers for Disease Control (1985) Drug-resistant tuberculosis
among the homeless - Boston. MMWR Morb Mortal Wkly
Rep 34:429
Centers for Disease Control (1986) Tuberculosis and acquired
immunodeficiency syndrome - Florida. MMWR Morb
Mortal Wkly Rep 36:587-590
Centers for Disease Control (1987) Tuberculosis and acquired
immunodeficiency syndrome - New York. MMWR Morb
Mortal Wkly Rep 36:785-795
Centers for Disease Control (1995) Essential components of a
tuberculosis prevention and control program; and screen-
ing for tuberculosis and tuberculosis infection in high-risk
populations: recommendation from the advisory council
for the elimination of tuberculosis. MMWR Morb Mortal
Wkly Rep 44:1-34
Chahtane A et al (1992) Hypertrophic tuberculosis of the
cervix: three cases. J Gyneacol Obstet BioI Reprod (Paris)
21:424-427
Chakraborty P et al (1992) Tuberculosis cervicitis: a clinico-
pathological and bacteriological study. J Indian Med Assoc
93:167-168
M. M. Madkour
ChangAB et al (1998) Renal tuberculosis in an Australian-born
child. J Paediatr Child Health (Australian) 34:293-295
Charghi A (1979) Ureteral replacement using a new variation
of the tailored ileal segment. J UroI121:598-601
Chatterjee H et al (1975) Tuberculosis of the penis. J Ind Med
Assoc 65:232-234
Chattopadhyay A et al (1997) Genitourinary tuberculosis in pedi-
atric surgical practice. J Pediatr Surg (US) 32:1283-1286
Chattopadhyay SK et al (1986) The pattern of female genital
tuberculosis in Riyadh, Saudi Arabia. Br J Obstet Gynaecol
93:367-371
Christensen WI (1974) Genitourinary tuberculosis. Review of
102 cases. Medicine 53:377-390
Chung JJ et al (1997) Sonographic findings in tubreculo-
sis epididymitis and epididymo-orchitis. Eur Radiol 7:
1076-1078
Chung T, Harris RD (1991) Tuberculous epididymo-orchitis:
sonographic findings. J Clin Ultrasound 19:367-369
Coetzee LF (1972) Tuberculous vaginitis. S Afr Med 46:
1225-1226
Cohen MS (1986) Granulomatous nephritis. Urol Clin North
Am 13:647-657
Cosivi 0, Grange JM et al (1998) Zoonotic tuberculosis due to
Mycobacterium bovis in developing countries. Emerging
Infect Dis 4:59-70
Cremin BJ (1987) Radiological imaging of urogenital tuber-
culosis in children with emphasis on ultrasound. Paediatr
RadioI17:34-38
Crowley JJ et al (1997) Genital tract tuberculosis with peri-
toneal involvement: MR appearance. Abdom Imaging 22:
445-447
Dahl DM et al (1996) Penile mass caused by the newly
described organism mycobacterium celatum. Urology
(US) 47:266
Dankner WM, Waecker NJ et al (1993) Mycobacterium bovis
infections in San Diego: a clinicoepidemiologic study of 73
patients and a historical review of a forgotten pathogen.
Medicine (Baltimore) 72:11-37
Dhillon 55 et al (1990) Genital tuberculosis: case report and
literature review. Wis Med 89:14-17
Drudi FM et al (1991) Tubercular epididymitis and orchitis:
US patterns. Eur Radiol 7:1076-1078
Dwyer DE et al (1987) Extrapulmonary tuberculosis - a con-
tinuing problem in Australia. Aust NZ J Med 17:507-511
Ellner JJ (1997) Regulation of the human immune response
during tuberculosis. J Lab Clin Med 130:469-475
El Mansouri A (1995) Hypertrophic tuberculosis of the cervix.
Rev Fr Gynecol Obstet 90:91-93
Emembolu JO (1989) Endometrial flora of infertile women in
Zaire, northern Nigeria. Int J Gynaecol Obstet 30:155-159
Engin G et al (2000) Imaging of extrapulmonary tuberculosis.
Radiographics (US) 20:471-488
Feeney D, Quesada ET, Sirbasku DM et al (1994) Transitional
cell carcinoma in a tuberculous kidney: case report and
review of the literature. J UroI151:989-991
Ferrie BG, Rundle JS (1983) Genitourinary tuberculosis in
patients under twenty-five years of age. UroI25:576-578
Garcia-Rodriguez JA et al (1994) Genitourinary tuberculo-
sis in Spain: review of 81 cases. Clin Infect Dis (US) 18:
557-561
Goldman SM, Fushman EK, Hartman DS et al (1985) Com-
puted tomography of renal tuberculosis and its pathologi-
cal correlates. J Comput Assist Tomogr 9:771-776
Genitourinary Tuberculosis
Gow JG (1979a) Genitourinary tuberculosis. Br J Hosp Med
6:556-568
Gow JG (1979b) Genitourinary tuberculosis: a 7-year-old
review. Br J UroI51:239-244
Gow JG (1992) Genitourinary tuberculosis. In: Walsh PC et al
(eds) Campbell's urology, 6th edn. Saunders, Philadelphia,
p 961
Gow JG, Barbosa S (1984) Genitourinary tuberculosis. A study
of 117 cases over a period of 34 years. Br J Urol 56:449-455
Granados LEA (1998) Testicular abscess: a manifestation of
tuberculosis. Aetas Urol Esp 22:381-383
Grange JM et al (1995) Factors determining ethnic differences
in the incidence of bacteriologically confirmed genitouri-
nary tuberculosis in southeast England. J Infect 30:37-40
Gras C et al (1980) Schistosomiasis of vulvae (case report and
comments) (author's trans!.). Med Trop (Mars) 40:335-338
Hasimoto A et al (1994) A case of endometrial tuberculosis diag-
nosed by polymerase chain reaction. Kekkaku 69:27-30
Heaton ND, Hogan B et al (1989) Tuberculosis epididymis-
orchitis: Clinical and ultrasound observations. Br J Urol
64:305
Heaton ND, Hogan B, Michell M et al (1989) Tuberculous
epididymo-orchitis: clinical and ultrasound observations.
Br J UroI64:305-309
Hemal AK et al (1998) 'Autoprostatectomy': an unusual
manifestation in genitourinary tuberculosis. Br J Urol 82:
140-141
Heuvel NVD et al (1979) Lymphagiectases of the vulva in a
patient with lymph node tuberculosis. Int Soc Trop Der-
matol 18:65-66
Hsu CT et al (1985) The coexistence of carcinoma and tuber-
culosis in the uterine cervix, report of 2 cases. Asia Oceania
J Obstet Gynaecol11:363-369
Husermeyer RP (1977) Postmenopausal tuberculosis of the
cervix. Case report and review. Br J Obstet Gynaecol 42:
153-154
Indudhara R, Vaidyanathan S et al (1992) Urethral tuberculo-
sis. Urol Int 48:436-438
Izawa II, Armstrong CE et al (1999) Tuberculous epididymo-
orchitis: a case report. Can J UroI6:751-756
Jaisankar TJ et al (1994) Penile lupus vulgaris. Int J Dermatol
33:272-274
Kim S, Pollack M, Cho KS et al (1993a) Tuberculous epididy-
mitis and epididymo-orchitis: sonographic findings. J Urol
150:81-84
Kim S et al (1993b) Tuberculous stricture of the urinary tract:
antegrade balloon dilation and ureteral stenting. Abdom
Imaging 18:186-190
Koller AB (1975) Granulomatous lesions of the cervix uteri in
black patients. S Afr Med J 49:1228-1232
Krolikowski A et al (1995) Asherman syndrome caused by
schistosomiasis. Obstet Gynaecol 85:898-899
Kumar A et al (1994) Tubercular cavity behind the prostate
and bladder: an unusual presentation of genitourinary
tuberculosis. J UroI151:1351-1352
Kutteh WH, Hatch KD (1992) Primary vaginal tuberculosis
after vaginal carcinoma. Gynecol Oncol 44: 113-115
La Grange JJ (1982) Postmenopausal endometrial tuberculo-
sis: a report of 2 cases and a literature survey. S Afr Med
J 59:501-502
Lazarus L, Peraino RA (1984) Reversible uremia due to bilat-
eral urethral obstruction from tuberculosis. Am J Nephrol
4:322-327
727
Leder RA, Low VHS (1995) Tuberculosis of the abdomen.
Radiol Clin North Am 33:691-698
Lee HY (1987) A 20-year experience with epididymovasos-
tomy for pathologic epididymal obstruction. Fertil Steril
47:487-491
Lenk S, Schroeder J (2001) Genitourinary tuberculosis. Curr
Opin Urol 11:93-98
Lewis EL (1946) Tuberculosis of the penis: a report of 5 cases
and a complete review of the literature. J UroI56:737-745
Lueken RP et al (1997) Genital tuberculosis-increased inci-
dence or coincidence. Zentralbl GnakoI119:39-41
Ma YY et al (1997) Vulvovaginal tuberculosis: a case report.
Changgeng Yi Xue Za Zhi 20:66-70
Mantovani A et al (1998) Postmenopausal endometrial tuber-
culosis: a clinical case. Minerva Ginecol 50:93-96
Marcus SF et al (1994) Tuberculous infertility and in vitro
fertilization. Am J Obstet Gynecoll71:1593-1596
McGuinness F (2000) Tuberculosis of the genitourinary tract.
In: McGuinness F (ed) Clinical imaging of non-pulmonary
tuberculosis. Springer, Berlin Heidelberg New York, pp
139-150
McKenna MT et al (1995) The epidemiology of tuberculosis
among foreign-born persons in the United States, 1986-
1993. N Engl J Med 332:1071-1076
Medlar EM (1926) Cases of renal infection in pulmonary
tuberculosis. Am J PathoI2:401-414
Medlar EM et al (1949) Postmortem compared with clinical
diagnosis of genitourinary tuberculosis in adult males. J
UroI61:1078-1088
Millar JW et al (1979) Vulval tuberculosis. Tubercle 60:173-176
Miller RE, Herman CJ (1979) Automated cervical cytology
screening: performance requirements and instrument
evaluation. J Histochem Cytochem 27:512-519
Miranda P et al (1996) Pelvic tuberculosis presenting as an
asymptomatic pelvic mass with rising serum CA-125Ievels.
A case report. J Reprod Med 41:273-275
Misch KA et al (1976) Tuberculosis of the cervix: Cytology as
an aid to diagnosis. J Clin PathoI29:313-316
Missirliu A et al (1996) Genitourinary tuberculosis: rapid
diagnosis using the polymerase chain reaction. Eur Urol
30:523-525
Mitwalli A (1991) Tuberculosis in patients on maintenance
dialysis. Am J Kidney Dis 18:579-582
Mokhtar A, Salman D (1983) Extrapulmonary tuberculosis.
Saudi Med J 4:317-322
Mondal A, Ghosh E et al (1990) The role of transrectal fine
needle aspiration cytology in the diagnosis of prostatic
nodules suspicious of malignancy-study of 126 cases.
Indian J Pathol Microbiol 33:23-29
Moon SY, Kim SH et al (1999) The outcome of sperm retrieval
and intracytoplasmic sperm injection in patients with
obstructive azoospermia: impact of previous tuberculous
epididymitis. J Assist Reprod Genet 16:431-435
Mourad G, Soulillou JP et al (1985) Transmission of Myco-
bacterium tuberculosis with renal allografts. Nephron 41:
82-85
Moussa OM et al (2000) Rapid diagnosis of genitourinary
tuberculosis by polymerase chain reaction and non-radio-
active DNA hybridization. J UroI164:584-588
Muechler E, Minkowitz S (1971) Postmenopausal endometrial
tuberculosis. Obstet GynecoI38:768-770
Murali TR, Raja NS (1998) Cavernosal cold abscess: a rare
cause of impotence. J Urol 82:929-930
728
Murata Y, Yamada I, Sumiya Y et al (1996) Abdominal mac-
ronodular tuberculomas: MR findings. J Comput Assist
Tomogr 20:643-646
Murphy DM et al (1982) Tuberculous stricture of ureter. Urol-
ogy 10:382-384
Murphy 0 et al (1992) Endometrial and pulmonary sarcoid-
osis. Ir J Med Sci 161:14-15
Murthy BVR, Pereira BJG (1997) A 1990 s perspective of hepa-
titis C, human immunodeficiency virus, and tuberculosis
infections in dialysis patients. Semin Nephr 17:346-363
Myers JP et al (1981) Tuberculosis in pregnancy with fatal
congenital infection. Pediatrics 67:89-94
Naika RP et al (1987) Esthiomene resulting from cutaneous
tuberculosis of external genitalia. Genitourin Med 63:
133-134
Nakamura S, Aoki M et al (1989) Penis tuberculid (papulo-
necrotic tuberculid of the glans penis): treatment with a
combination of rifampicin and an extract from tubercle
bacilli (TB vaccine). J Dermatol 16:150-153
Nardell E et al (1986) Exogenous reinfection with tuberculosis
in a shelter for the homeless. N Engl J Med 315:1576
National Survey of Tuberculosis Notifications in England and
Wales (1983) Characteristics of disease. Report form the
Medical Research Council. Tuberculosis and chest disease
unit. Tubercle 198768:19-32
Nishigori C, Taniguchi S et al (1986) Penis tuberculosis: papu-
lonecrotic tuberculides on the glans penis. Dermatologica
172:93-97
Nogales-Ortiz F (1979) The pathology of female genital tuber-
culosis: a 31-year series of 1436 cases. Obstet Gynecol 53:
422-428
Nzerue C et al (2000) Genitourinary tuberculosis in patients
with HIV infection clinical features in an inner-city hospi-
tal population. Am J Med Sci 320:299-303
O'Flynn D (1970) Surgical treatment of genitourinary tuber-
culosis. A report of 762 cases. Br J UroI42:667-671
Ormerod LP (1993) Why does genitourinary tuberculosis
occur less often than might be expected in the ethnic
Indian subcontinent population living in the United King-
dom? J Infect 27:27-32
Otieno LS (1983) Genitourinary tuberculosis at Kenyatta
National Hospital 1973-1980. East African Med J 60:
232-236
Paick JS et al (2000) Ejaculator duct obstruction in infertile
men. BJU Int 85:720-724
Pal DK (1997) Erectile failure and destruction of glans penis
by tuberculosis. Trop Doct 27:178-179
Parikh FR, Nadkarni SG et al (1997) Genital tuberculosis - a
major pelvic factor causing infertility in Indian women.
Fertil Steril 67:497-500
Pearson et al (1992) Nosocomial transmission of multidrug-
resistant mycobacterium tuberculosis: a risk to patients
and health care workers. Ann Intern Med 117:191-196
Pena G et al (1973) Vulvar tuberculosis. lrev Chil Obstet
GynecoI38:67-75
Petersen L et al (1993) Male genitourinary tuberculosis. Report
of 12 cases and review of the literature. Scand J Urol Nephr
27:425-428
Pisciolo F, Pusiol T et al (1985) Urinary cytology of tubercu-
losis of the bladder. Acta CytoI29:125-131
Premkumar A, Lattimer J, Newhouse JH (1987) CT and sonog-
raphy of advanced urinary tract tuberculosis. AJR Am J
Roentgenol 148:65-69
M. M. Madkour
Psihramis KE, Donahoe PK (1986) Primary genitourinary
tuberculosis: rapid progression and tissue destruction
during treatment. J UroI135:1033-1036
Punnonen R et al (1983) Female genital tuberculosis and con-
sequent infertility. Int J Fertil28:235-238
Raghavaiah NV (1979) Tuberculosis of the male urethra. J
UroI122:417-418
Rao PS et al (1992) Bacterial flora of the endometrium in infertile
women from Manipal. Ind J Pathol MicrobioI35:357-361
Reider HL, Snider DE, Cauthen GM (1990) Extrapulmonary
tuberculosis in the United States. Am Rev Respir Dis 141:
347-351
Richards MJ,Angus D (1998) Possible sexual transmission ofgen-
itourinary tuberculosis (letter). Int J Tuberc Lung Dis 2:439
Rivasi F et al (1980) Diagnosis of cervical tuberculosis by
cytology. Report of cases. Quad Sclavo Diagn 16:239-247
Saracoglu OF et al (1992) Pelvic tuberculosis. Int J Gyncol
Obstet 37:115-120
Sbihi H et al (1980) Tubo-intestinal fistulae. A report of 4 cases.
Ann Radiol (Paris) 23:656-659
Schaefer G (1970) Tuberculosis of the female genital tract.
Gynecol Obstet 13:965-998
Schaefer G (1976) Female genital tuberculosis. Clin Obstet
GynecoI19:223-239
Selwyn PA et al (1989) A prospective study of the risk of tuber-
culosis among intravenous drug users with HIV infection.
N Engl J Med 320:545-550
Selwyn PA et al (1992) High risk of active tuberculosis in
HIV infected drug users with cutaneous anergy. JAMA
268:504-509
Sengupta P, Mukherjee SD (1982) Tuberculosis of the penis. J
Indian Med Assoc 78:47-49
Shafer RW et al (1991) Extrapulmonary tuberculosis in patients
with human immunodeficiency virus infection. Med 70:
384-397
Shamsa A (1998) Gastric neobladder for treatment of tuber-
culosis systitis. J Uro1159:202
Shireman PK (1992) Endometrial tuberculosis acquired by a
health Care Worker in Clinical laboratory. Arch Pathol Lab
Med 116:521-523
Shobin D, Pellman C (1976) Genitourinary tuberculosis simu-
lating cervical carcinoma. J Reprod Med 17:305-308
Siegler AM, Kontopoulos V (1979) Female genital tuberculosis
and the role of hysterosalpingography. Semin Roentgenol
14:295-304
Silva PD et al (1988) Diagnosis and management of a tuber-
culous, tuboappendiceal fistula. Am J Obstet Gynecol 159:
440-441
Simon HB et al (1977) Genitourinary tuberculosis. Clinical fea-
tures in a general hospital population. Am J Med 63:410
Singh A et al (1988) Tuberculous vesicovaginal fistula in a
child. Br J Uro162:615
Skehan M, McKenna P (1985) Sarcoidosis of the endometrium:
a case presentation. Ir J Med Sci 154: 114
Soussis I et al (1998) In vitro fertilization treatment in genital
tuberculosis. J Assist Reprod Genet 15:378-380
Sporer A, Auerbach 0 (1978) Tuberculosis of Prostate. Urol
11:362-365
Stewart DB (1968) Tropical gynaecology. Ulcerative and hyper-
trophic lesion the vulva. Proc R Soc Med 61:363-365
Stoller JK (1985) Late recurrence of mycobacterium bovis
genitourinary tuberculosis: case report and review of lit-
erature. J UroI134:565-566
Genitourinary Tuberculosis
Sutherland AM (1982) The changing pattern of tuberculosis of
the female genital tract. A thirty-year survey. Arch Gynecol
234:95-101
Sutherland AM (1985) Gynaecological tuberculosis: analysis
of a pesonal series 710 cases. Aust NZ J Obstet Gynaecol
25:203-207
Svendsen JH et al (1985) Peritonitis due to genital tuberculosis.
Ann Chir Gynecol 74:180-182
Tang LC et al (1984) Atypical presentation of female genital
tract tuberculosis. Eur J Obstet Reprod Bioi 17:355-363
Tripathy SN (1990) Place of aspiration cytology in dusfunc-
tional uterine bleeding. J Indian Med Assoc 88:247-248
Tripathy SN, Tripathy SN (1990) Laparoscopic observations
of pelvic organs in pulmonary tuberculosis. Int. J Gynecol
Obstet 32:129-131
Valentini AL et al (1998) Diagnostic imaging of genitourinary
tuberculosis. Rays 23:126-143
Vani R, Kuntal R, Pralhad K (1993) A rare location of extrapul-
monary tuberculosis: the vulva. Tuberc Lung Dis 74:64
Vuong PN et al (1989) Pseudotumoral tuberculosis of the uter-
ine cervix. Cytologic presentation. Iacta Cyto 33:305-308
Wajeh et al (1990) Mycobacterial infection after renal trans-
plantation - report of 14 cases and review of the literature.
Quart J Med New Ser 77:1039-1060
Wang LJ et al (1997) CT features of genitourinary tuberculosis.
J Comput Assisted Tomogr 21:254-258
Wang XY (1989) Female infertility due to tubal obstruction-
management clinicoopathologic study of 66 cases. Zhong-
hua Fu Chan Ke Za Zhi 24:201-202, 251
Webster A, Wright DJ (1985) Ziehl-Neelsen staining of urine
729
deposits in the diagnosis of genitourinary tuberculosis. J
Clin Pathol 38:236
Wechsler H et al (1960) The earliest signs and symptoms in 127
male patients with genitourinary tuberculosis. JUro183:801
Weir MR, Thornton GF (1985) Extra-pulmonary tuberculosis.
Experience of a community hospital and review of litera-
ture. Am J Med 79:467-478
Wiskind AK et al (1992) Primary fallopian tube carcinoma
with coexistent tuberculosis salpingitis: a case report. J
Med Assoc Ga 81:77-81
Wong TW et al (1973) Testicular biopsy in the study of male
infertility. Arch PathoI95:160-164
Yamabe T et al (1972) Coincidence of carcinoma and tubercu-
losis of the uterine cervix. Gan No Rinsho 18:151-153
Yang DM, Chang MS et al (2000) Chronic tuberculous epi-
didymitis: color Doppler US findings with histopathologic
correlation. Abdom Imaging 25:559-562
Yang Y et al (1996) Laparoscopic diagnosis of tubal infertility
and fallopian tube lesions. Zhonghua Fu Chan Ke Za Zhi
31:327-329
Yates MD, Pozniak A et al (1993) Isolation of mycobacteria from
patients seropositive for the human immunodeficiency virus
(HIV) in south east England. Thorax 48:990-995
Yip SK et al (1999) Unassisted conception with a normal
pregnancy outcome woman with active mycobacterium
tuberculosis infection the endometrium. A case report. J
Reprod Med 44:474-476
Zamberletti D et al (1981) Scanning electron microscopy
of the tubercular fallopian tube. Acta Eur Fertilitatis 12:
213-221
40 Ocular Manifestations of Tuberculosis
SHWU-JIUAN SHEU

CONTENTS
40.1 Historical Perspective 731
40.2 Pathogenesis 731
40.3 Clinical Features 732
40.3.1 External Diseases 732
40.3.1.1 Eyelid Disease 732
40.3.1.2 Conjunctival Tuberculosis 732
40.3.1.3 Phlyctenulosis 733
40.3.1.4 Corneal Disease 733
40.3.1.5 Scleral Tuberculos 733
40.3.2 Intraocular Disease~ 733
40.3.2.1 Uveitis 733
40.3.2.2 Retinal Tuberculosi~ 735
40.3.2.3 Endophthalmitis 'i)
40.3.2.4 Eales' Disease and Tuberculin
Hypersensitivity 735
40.3.2.5 Optic Neuropathy 735
40.3.3 Orbital Tuberculosis 736
40.4 Diagnosis 736
40.4.1 Clinical Manifestation 736
40.4.2 Smears and Cultures 737
40.4.3 Tuberculin Skin Test 738
40.4.4 Molecular Techniques 738
40.4.5 Isoniazid Diagnostic Trials 738
40.5 Treatment 738
40.5.1 Monitoring of the Response to Treatment
40.5.2 Toxicity of Anti-TB Chemotherapy 739
References 739
et al. 1982; Saini et al. 1986; Helm and Holland 1993).
Because the treatment of TB is relatively effective and
cost efficient (Raviglione and O'Brien 1998; Zumla et
al.1999),earlydiagnosis and prompt treatment are key
to saving the sight of patients with ocular TB.
40.1
Historical Perspective
In 1711, Maitre-Jan reported the earliest description
of ocular tuberculosis, an iris lesion that eventually
lead to corneal perforation (Maitre-Jan 1711). Gue-
neau de Mussy recognized the first choroidal tubercle
in miliary tuberculosis in 1830 (Wecker 1874) and
in 1867, Cohnheim showed that choroidal tubercles
seen clinically were identical to tubercles elsewhere
in the body. He was able to produce them experimen-
tally in guinea-pigs by injecting tuberculous material
(Cohnheim 1867). The tubercle bacillus was discov-
ered by Koch in 1882 (Koch 1882), and the diagnosis
738
of ocular tuberculosis was further proved when von
Michel identified tubercle bacilli in the eye one year
later (von Michel 1883).

Tuberculosis (TB) is the leading cause of death world-

wide due to an infectious agent, the problem being
aggravated by the human immunodeficiency virus
(HIV) pandemic and the recent increasing incidence
of microbial resistance to antibiotics (Brudney and
Gob"in 1991; Barnes et al.1991; Berenguer et al.1992).
The recurrence of TB as a major public health prob-
lem raises the possibility that ophthalmologists may
encounter an increasing number of ocular complica-
tions. Ocular TB may affect various regions of the eye
and cause severe visual loss if not treated properly (Ni
S.-]. SHEU, MD
Department of Ophthalmology, Kaohsiung Veterans General
Hospital, 386 Ta-Chung, 1st Road, Kaohsiung, Taiwan 813
40.2
Pathogenesis
Ocular Tuberculosis, caused by M. tuberculosis,
presents as a spectrum of ocular diseases. The patho-
genesis of ocular TB remains controversial. Ocular
manifestations associated with TB are either caused
by an active infection, or an immunological reaction
in the absence of any infectious agent, which is related
to delayed hypersensitivity and an aseptic reaction (Ni
et al. 1982; Helm and Holland 1993). Active infection
of the eye may be primary, or secondary, in nature.
In primary ocular TB, there is no other systemic
lesion, whereas secondary ocular TB is defined as an
infection resulting from contiguous spread from an

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
 732
adjacent structure or hematogenous spread. Almost all
primary ocular infections are limited to conjunctival
and corneal diseases that present as an ulceration, a
tumor mass, pWyctenulosis, or interstitial keratitis.
However, intraocular TB is usually associated with
systemic disease and is therefore considered to rep-
resent a secondary infection. The predominant route
by which tubercle bacilli reach the eye is through the
blood stream, after infecting the lung. The pulmonary
foci might not be evident clinically or radiographically.
The manner in which tuberculosis may present is
determined by several factors, including the amount
and virulence of the infecting bacteria, the degree of
tissue hypersensitivity already present or which may
result from the infection, and the amount of native and
acquired resistance possessed by the host.
40.3
Clinical Features
TB produces foci of granulomatous inflammation,
usually in the lungs, but can involve practically every
bodily organ. TB may involve any part of the eye and
appear in different clinical forms. The clinical mani-
festations of ocular TB are non-specific and protean.
The most common manifestation of ocular TB in
patients with pulmonary TB is choroiditis. Anterior
uveitis, choroid tubercles, retinal vasculitis, vascular
occlusion, dense vitritis, and papillitis may also occur
(Ni et al. 1982; Regillo et al. 1991; Helm and Holland
1993; Bouza et al. 1997).
The emergence and rapid development of the
HIVIAIDS pandemic has had a devastating impact
on the global burden of tuberculosis. The incidence
a
Fig. 40.1a, b. A 70-year-old man with a painless tumor in his left lower lid. a Pathology findings of the lid mass showed chronic
granulomatous inflammation with caseous necrosis (H&E, x1S0). (Reprinted from Sheu et al. 2001, with permission from Else-
vier Science) b Acid-fast stain showed acid-fast bacilli (Kinyoun's crystal Fuchsin acid-fast stain, x198)
S.-I. Sheu
of TB in patients with the acquired immunodeficiency
syndrome (AIDS) is almost 500 times the incidence in
the general population (Barnes et al. 1991). In this era
of the HIV pandemic, TB is becoming more prevalent,
and an increasing number of ocular TB cases may
eventually be seen. Unfortunately, early diagnosis of
TB in HIV-positive persons is not easy as the clini-
cal and radiological features are often atypical and
resemble those caused by other HIV-related infections
(Festenstein and Grange 1991; Daley 1995).
40.3.1
External Diseases
40.3.7.7
Eyelid Disease
Tuberculosis of the eyelid may present as a tumor or an
abscess. The latter may be mistaken for chalazion with
erythema and focal, or diffuse swelling of the eyelid.
Figure 40.1 shows the histopathology of a painless
eyelid mass in a 70-year-old man. Chronic granulo-
matous inflammation and caseous necrosis as well as
acid-fast bacilli proved the diagnosis of ocular TB.
40.3.7.2
Conjunctival Tuberculosis
The first case ofconjunctival tuberculosis was described
by Arlt in 1864 (Arlt 1864) and in the review by Eyre
in 1912, the conjunctival lesions were classified into
five groups based on clinical appearance: ulcerative,
miliary tubercle, hypertrophic granulation, lupus, and
pedunculated tumor (Eyre 1912). The majority were
unilateral and most commonly involved the upper
_ _ _ _.......... b
Ocular Manifestations of Tuberculosis
palpebral conjunctiva. Involvement of the lacrimal
sac can prevent lacrimal drainage and leads to fistula
draining to the skin.
40.3.1.3
Phlyctenulosis
Phlyctenulosis are small nodules which arise on the
conjunctiva or at the limbus. A small ulcer usually
develops at the apex of a phlyctenule several days
after its appearance. The lesions may 'wander' across
the cornea from their site of origin at the limbus,
leaving a narrow vascularized scar to mark its path.
An association between phlyctenulosis and tubercu-
loprotein hypersensitivity is suggested by both epide-
miological and experimental evidence (Gibson 1918;
Thygeson 1954).
40.3.1.4
Corneal Disease
Corneal manifestations of tuberculosis include infil-
trates, ulceration, and interstitial keratitis. Interstitial
keratitis is rare and it is usually unilateral and pain-
less (Donahue 1967; Thygeson 1974). Figure 40.2
demonstrates a corneal ulcer with descemetocele in a
58-year-old man. Polymerase chain reaction analysis
and culture from corneal scraping were both positive
for M. tuberculosis. The inflammation resolved after
anti-TB medication.
40.3.1.5
Scleral Tuberculosis
TB is a rare, but classic cause of scleritis (which is
usually anterior and necrotizing), and is associated
Fig.40.2a, b. A 58-year-old man presenting with a recurrent corneal ulcer in the right eye. a Slit-lamp biomicroscopy showed a
corneal ulcer with descemetocele. b Nested peR from corneal scraping showed a positive result for Mycobacterium tuberculosis
(Reprinted from Sheu et al. 2001, with permission from Elsevier Science)
733
with scleral ulceration (Saini et al. 1988; Nanda et
al. 1989).
40.3.2
Intraocular Diseases
40.3.2.1
Uveitis
The TB-associated uveitis is usually granulomatous,
with mutton-fat keratic precipitates, iris granulo-
mas, and posterior synechiae. Concomitant pos-
terior uveitis is a frequent occurrence, producing
focal or diffuse choroiditis. Choroiditis is the most
common ocular manifestation during dissemina-
tion of the bacillus via the blood stream because
of its high level of blood supply and oxygenation.
Choroidal tubercles are frequently unilateral, close
to the posterior pole and sometimes associated with
an inflammatory retinal detachment. The choroidal
tubercles range from 1/4 to several disc diameters
in size, usually with indistinct margins. The number
of lesions ranges from one to as many as 50 or 60,
with most eyes having fewer than five lesions. The
tubercles may be variously pigmented depending
upon the age of the lesion, with older lesions being
more hyperpigmented (Massaro et al. 1964; Olazabal
1967; Cangemi et al. 1980; Chung et al. 1989).
A case of bilateral choroidal tuberculosis is shown
in Fig. 40.3. A 54-year-old woman complained of
blurred vision, headache, fever, nausea, vomiting and
nuchal rigidity. TB meningitis was diagnosed from
her history of pulmonary TB (in her teenage years)
and from CSF analysis, (although the culture was
found to be negative). The fundus showed multiple
 734 S.-T. Sheu

a b

c d

e f

Fig.40.3a-f. A 54-year-old woman presenting with bilateral choroidal tuberculosis. a Right eye b Left eye. Fundus montage
showed 50 to 60 ill-defined yellowish-white elevated nodules mostly over the posterior pole. Some hemorrhage was noted over
the macular area and the optic disc margin was blurred. c Right eye. d Left eye. At the late phase of fluorescein angiography, all
the nodular lesions became hyperfluorescent, and mild dye leakage from the disc was also noted. e Right eye. f Left eye. After
anti- TB medication, the disc margin cleared and the nodular lesion became more well-defined. Choroidal neovascularization
developed in the macula of both eyes
 Ocular Manifestations of 'fuberculosis 735

ill-defined yellowish-white elevated nodules mostly who was treated for idiopathic uveitis and was operated
over the posterior pole. Some bleeding was noted for a complicated cataract. The ocular inflammation
over the macular area and the optic disc margin was execrated after cataract extraction. Pathologic findings
blurred. A fluorescein angiogram showed a block of the eyeball showed extensive caseous necrosis with
fluorescence in the early phase and leakage in the polymorphonuclear leukocyte infiltration and acid-fast
late phase. The meningitis and ocular lesion gradu- bacilli. Pulmonary TB was found in these two cases,
ally resolved after anti-TB therapy. Unfortunately, though it was not evident clinically or radiographically
choroidal neovascularization in the macula impaired in the initial stage of ocular inflammation. Therapy for
the visual outcome. Vitritis can be associated with TB successfully treated the pulmonary TB and ocular
choroidal tubercle and obscure the typical fundus inflammation, but it was not possible to save the eyes
picture as shown in Fig. 40.4. due to delayed diagnosis and treatment.

40.3.2.2 40.3.2.4
Retinal Tuberculosis Eales' Disease and Tuberculin Hypersensitivity

TB of the retina is very rare and usually occurs Eales' disease occurs predominantly in adult males in
following choroidal involvement. Retinal periphle- their third and fourth decades of life (Eales 1880). The
bitis, vascular occlusion and tumor mass have been disease presents as recurrent unilateral retinal and vit-
reported to be associated with TB (Clever 1980; Saini reous hemorrhaging with involvement of the other eye
et al. 1986; Sant 1994). after a few months. It is usually associated with retinal
vasculitis, ischemia and neovascularization with a high
40.3.2.3 risk of intraocular hemorrhage as well. Although an
Endophthalmitis immunologicalmechanism has been postulatedto playa
role in the pathogenesis of Eales' disease in patients with
Some cases of endophthalmitis as· a consequence of tuberculin hypersensitivity, it may be a clinical entity
TB infection have been reported (McMoli et al. 1978; with multiple causes. This is because not all patients
Grenzebach et al. 1996). Also, endogenous endophthal- have positive tuberculin skin tests (Vine 1992).
mitis following intravesicular bacille Calmette-Guerin
injection has been confirmed in cultures of the vitreous 40.3.2.5
fluid (Han et al. 1999). Figure 40.5 shows a case of TB Optic Neuropathy
panophthalmitis. The pathology of the transbronchial
biopsy and eyeball (from enucleation) showed the exis- Optic neuritis has been ascribed to TB. Optic atrophy
tence of acid-fast bacilli.Vitreous culture was also posi- can be seen after adhesive arachnoiditis. Optic disc
tive for M tuberculosis. Another case of TB endophthal- tubercle, papilledema, and tuberculous neuroretinitis
mitis is shown in Fig. 40.6.This shows a68-year-old man have also been reported (Mansour 1998; Stechschulte et

Fig.40.4a, b. A 47-year-old man presenting with a right temporal headache, fever and loss of vision in the left eye. TB men-
ingitis was suggested by his TB history and typical CSF findings, (although the culture was negative). a Fundus examination
showed remarkable vitreous opacity. b Fundus examination, one week later, showed several yellowish-white subretinal nodule
at posterior pole
 736 S.-J. Sheu

a b

c .........._ .....""""'"

Fig.40.5a-d. Panophthalmitis in a 75-year-old woman with pulmonary TB. a Slit-lamp biomicroscopy at presentation showed
ciliary congestion and dense anterior chamber reaction (Reprinted from Sheu et aI. 2001, with permission from Elsevier Science).
b Acid-fast stain of bronchial biopsy showed acid-fast bacilli (Kinyoun's crystal Fuchsin acid-fast stain, xI000). c Pathologic
findings of the eyeball specimen from enucleation showed extensive caseous necrosis of the retinal tissue (H&E, x60). d Pathol-
ogy findings of the eyeball specimen from an enucleation showed extensive infiltration of acid-fast bacilli (Kinyoun's crystal
Fuchsin acid-fast stain, x600) (Reprinted from Sheu et aI. 2001, with permission from Elsevier Science)

al.1999).A case ofpresumed tuberculous neuroretinitis 40.4

is shown in Figure 40.7. Fundus examination showed
disc hyperemia, and swelling, as well as multiple plac-
oid exudative retinal detachment in the posterior pole.
Tuberculin skin test resulted in intense reaction with
necrosis. The lesion resolved after therapy for TB.
40.3.3
Orbital Tuberculosis
Several cases of orbital tuberculosis have been
reported (Baghdassarian et al. 1972; Argrawal et al.
1977; Jain et al. 1979). Orbital tuberculosis occurs by
hematogenous spread, or by invasion from adjacent
structures (Khalil et al. 1985). The disease is usually
slow-growing, chronic, and unilateral. Orbital dis-
ease most commonly presents as either proptosis,
or as spontaneous fistulization of an orbital abscess.
Regional lymph node involvement is common.
Diagnosis
Ocular TB diagnosis is complicated by the difficul-
ties associated with ocular sampling. A high degree
suspicion is essential for the early diagnosis of TB.
The past medical history of patients and their fami-
lies is a major consideration. Confirmation of ocular
TB depends on the demonstration of tubercle bacilli
in the ocular specimen, but diagnosis remains fre-
quently presumptive for the time being.
40.4.1
Clinical Manifestation
The manifestations of ocular TB are nonspecific and
protean. The ocular findings can be seen with other
ocular and systemic diseases. The clinical features are
not pathognomonic for the diagnosis of ocular TB.
 Ocular Manifestations of Tuberculosis 737

Fig.40.6a-c. Endophthalmitis in a 68-year-old man with

miliary TB. a Slit-lamp biomicroscopy at presentation showed
pigmented keratic precipitates, corneal stroma edema, exudate
in the anterior chamber. b Pathology findings of the eyeball
specimen from an enucleation showed extensive caseous
necrosis with polymorphonuclear leukocytes infiltration
(H&E, x120). c Acid-fast stain showed acid-fast bacilli in the
area of caseous necrosis (Kinyoun's crystal Fuchsin acid-fast
stain, x1000) (Reprinted from Sheu et al. 2001, with permis-
c sion from Elsevier Science)

Fig. 40.7a, b. A 44-year-old woman presenting with presumed tuberculous neuroretinitis. a Fundus showed severe disc swelling
and hyperemia as well as multiple placoid exudative retinal detachment in the posterior pole. b Tuberculin skin test resulted
in an intense reaction with necrosis

40.4.2 ous smear. A definitive diagnosis is dependent on a

Smears and Cultures
A presumptive diagnosis is commonly based on
the finding of acid-fast bacilli during microscopic
examination (acid-fast bacilli microscopy) of a diag-
nostic specimen, such as a vitreous aspirate or aque-
positive culture of the organism from a diagnostic
specimen. Lowenstein-Jensen medium is widely used
for the isolation of M tuberculosis. Nevertheless, the
culture is difficult and time-consuming. Normally,
treatment should be started before the culture results
are available.
738
40.4.3
Tuberculin Skin Test
This reaction is an immunologic process based on
delayed hypersensitivity. A positive skin test result
is detectable 3- -8 weeks after the primary infection.
An intense skin reaction can become necrotic. There
is no specific amount of induration that confirms TB
and a negative test result does not exclude the diag-
nosis. Active disease may be associated with a weak
or negative reaction, especially in older patients, the
malnourished, immunosuppressed, or those patients
undergoing corticosteroid treatment. The tuberculin
skin test is of limited value because of its low sensitiv-
ity and specificity, especially in countries where the
BCG vaccine is used.
40.4.4
Molecular Techniques
In recent years, several methods have been developed
which shorten the time necessary for specific iden-
tification of mycobacteria (Good and Mastro 1989).
These methods include the polymerase chain reac-
tion (PCR), which uses a heat-stable DNA polymerase
to amplify mycobacterial DNA from clinical samples
(Musial et al. 1988). PCR has been used in ocular
specimens to diagnose M. tuberculosis infection
(Kotake et al. 1994; Gupta et al. 1998). PCR is a rapid
diagnostic test with high sensitivity and specificity.
Moreover, it is especially useful for diagnosing ocular
TB because only a small sample is needed and viable
cells are not required.
40.4.5
Isoniazid Diagnostic Trials
A two-week therapeutic trial of isoniazid had been
recommended in cases of presumed ocular tubercu-
losis (Abrams and Schlaegel 1982). The diagnostic
value of therapeutic trials is diminished by the
increasing prevalence of drug-resistant strains of M.
tuberculosis.
40.5
Treatment
Despite the difficulties in diagnosis, the treatment of
TB is relatively effective and cost-efficient. Systemic
S.-T. Sheu
therapy should always be given as the primary treat-
ment for ocular TB as pulmonary, or other, foci of
disease may coexist. Five major drugs are consid-
ered to be first-line agents for treating TB: isoniazid,
rifampin, pyrazinamide, ethambutol, and streptomy-
cin. They are recommended based on their bacteri-
cidal activity (ability to rapidly reduce the number
of viable organisms), their sterilizing activity (ability
to kill all bacilli and thus sterilize the affected organ,
measured in terms of the ability to prevent relapses),
and their low rate of induction of drug resistance.
A number of second-line drugs (kanamycin, ami-
kacin, capreomycin, ethionamide, cycloserine, PAS,
quinolone) are used only to treat patients with TB
who are resistant to the first-line drugs. Corticoste-
roids are sometimes required for the ocular inflam-
mation. Steroids should not be administered in the
absence of antimicrobial therapy. Ignoring this rule
exposes the patients to a risk of systemic dissemina-
tion of TB with a high mortality rate. A multidrug
regimen is recommended. Any patient with a disease
highly suggestive of ocular TB should be treated
from the outset with multiple antituberculous che-
motherapeutic drugs because of the increasing inci-
dence of resistance to isoniazid as well as compliance
problems. The guidelines of the American Thoracic
Society for treatment consist of a 2-month initial
phase of isoniazid, rifampin, and pyrazinamide fol-
lowed by a 4-month continuation phase of isoniazid
and rifampin, for a total of 6 months (American Tho-
racic Society 1992). According to the World Health
Organization (WHO) and the International Union
Against Tuberculosis and Lung Disease, the resis-
tance to one drug is higher than the combined form
(one drug, 9.1%; two or more drugs, 7.7%; multidrug
resistance, 4.3%) (World Health Organization 1997).
Treatment can be prolonged in immunocompro-
mised patients or those with a disseminated disease.
Second-line drugs or other adjuvants are considered
in cases of multiple drug resistance or contraindica-
tions. Immunotherapy to boost the efficiency of the
immune system in infected patients could be a valu-
able adjunct to anti-TB chemotherapy (Toossi 1998).
40.5.1
Monitoring of the Response to Treatment
Bacteriological evaluation is the preferred method
of monitoring the response to treatment for TB.
Patients with pulmonary disease should have their
sputum examined monthly until cultures are nega-
tive, whereas, bacteriological monitoring of patients
Ocular Manifestations of Tuberculosis
with ocular TB is more difficult and often not feasible.
The response to treatment must be assessed clini-
cally. An initial response usually occurs in two weeks.
During treatment, patients should be monitored for
drug toxicity.
40.5.2
Toxicity of Anti-TB Chemotherapy
The most common adverse reaction of significance is
hepatitis. Patients should be carefully educated about
the signs and symptoms of drug-induced hepatitis
and should be instructed to discontinue medica-
tion and seek medical assistance on developing any
signs of hepatitis. The drugs can be reintroduced
one at a time after liver function has returned to
normal. Hypersensitivity reactions usually require
discontinuation of all drugs and re-assessment to
determine which agent is responsible. Similarly, the
occurrence of optic neuritis with ethambutol (and
rarely isoniazid), as well as the development of eighth
nerve damage with streptomycin require permanent
discontinuation of these drugs. Studies that detect
subclinical evidence of toxicity before symptoms
occur have been sought. In a study by Yiannikas and
associates (Yiannikas et al. 1983), pattern-reversal
visual-evoked potentials were used to detect a revers-
ible prolongation oflatency and decreased amplitude
of the PlOD wave in asymptomatic patients receiving
ethambutol. Subclinical impairment of color dis-
crimination was found to be relatively common in
patients receiving ethambutol (Joubert et al. 1986). A
deuteran type of defect was found to occur early and
tritan defect was found to be a later manifestation
of toxicity in a group of patients with normal vision
who were treated with ethambutol for at least one
month. Chronic alcoholics, malnourished individu-
als, and persons with pre-existing cerebral damage
may be at increased risk of developing ocular toxicity.
Pyridoxine was advocated for both prophylaxis and
treatment of isoniazid-related optic neuritis (Kass et
al. 1957). Acute uveitis had also been reported fol-
lowing rifampin administration (Spindler et al.199l).
A complete baseline examination prior to treatment
and routine test of visual acuity during follow-up is
strongly recommended, especially in the unaffected
fellow eye. The assistance of an infectious disease
specialist, or pulmonologist, is helpful in treating
these patients.
739
References
Abrams J, Schlaegel TF Jr (1982) The role of isoniazid thera-
peutic test in tuberculous uveitis. Am J Ophthalmol 94:
511-515
American Thoracic Society (1992) Control of tuberculosis in
the United States. Am Rev Respir Dis 146:1623-1633
Argrawal PK, Nath J, Jain BS (1977) Orbital involvement in
tuberculosis. Am J OphthalmoI25:12-16
Arlt CF (1864) Ueber acne rosacea und lupus. Klin Monatsbl
Augenheilkd 2:329-330
Baghdassarian SA, Zakharia H, Asdourian KK (1972) Report
of a case of bilateral caseous tuberculous dacryoadenitis.
Am J Ophthalmol 74:744-746
Barnes PF et al (1991) Tuberculosis in patients with human
immunodeficiency virus infection. N Engl J Med 324:
1644-1650
Berenguer J et al (1992) Tuberculous meningitis in patients
infected with the human immunodeficiency virus. N Engl
J Med 326:668-672
Bouza E et al (1997) Ocular tuberculosis - a prospective study
in a general hospital. Medicine 76:53-61
Brudney K, Dobkin J (1991) Resurgent tuberculosis in New
York City. Am Rev Respir Dis 144:745-749
Cangemi FE, Friedman AH, Josephberg R (1980) Tuberculoma
of the choroid. Ophthalmology 87:252-258
Chung YM et al (1989) Macular subretinal neovascularization
in choroidal tuberculosis. Ann OphthalmoI21:225-229
Clever VG (1980) Presumed ocular tuberculosis. Ann Ophthal-
mol 12:424-426
Cohnheim J (1867) Uber Tuberculose der Choroidea. Virchows
Arch Pathol Anat 39:49-69
Daley CL (1995) The typically "atypical" radiographic pre-
sentation of tuberculosis in advanced HIV disease. Tuberc
Lung Dis 76:475-476
Donahue HC (1967) Ophthalmologic experience in a tubercu-
losis sanatorium. Am J OphthalmoI64:742-748
Duke-Elder S (1966) System of ophthalmology: disease of the
uveal tract, vol 9. Mosby, St Louis, p 248
Eales H (1880) Cases of retinal hemorrhage, associated with
epistaxis and constipation. Birmingham Med Rev 9:
262-273
Eyre JWH (1912) Tuberculosis of the conjunctiva: its etiology,
pathology, and diagnosis. Lancet 1:1319-1328
Festenstein F, Grange JM (1991) Tuberculosis and the acquired
immune deficiency syndrome. J Appl Bacteriol 71:19-30
Gibson WS (1918) The etiology of phlyctenular conjunctivitis.
Am J Dis Child 15:81-115
Good RC, Mastro TD (1989) The modern mycobacteriology
laboratory. How it can help the clinician. Clin Chest Med
10:315-322
Grenzebach UH et al (1996) Endophthalmitis induced by atyp-
ical mycobacterial infection. Ger J Ophthalmol 5:202-206
Gupta V et al (1998) Management of presumed intraocular
tuberculosis: possible role of the polymerase chain reac-
tion. Acta Ophthalmol Scand 106:290-294
Han DP et al (1999) Endophthalmitis from mycobacterium
bovis bacille Calmette-Guerin after intravesicular bacille
Calmette-Guerin injection for baldder carcinoma. Am J
OphthalmoI128:648-650
Helm CJ, Holland GN (1993) Ocular tuberculosis. Surv Oph-
thalmol 38:229-256
Jain MR, Chundawat HS, Batra V (1979) Tuberculosis of the
740
maxillary antrum and of the orbit. Ind J Ophthalmol 1:
18-29
Joubert H et al (1986) Subclinical impairment of colour
vision in patients receiving ethambutol. Br J Clin Pharm
21:213-216
Kass I et al (1957) Isoniazid as a cause of optic neuritis and
atrophy. JAMA 164:1740-1743
Khalil M, Lindley S, Matouk E (1985) Tuberculosis of the orbit.
Ophthalmology 92:1624-1627
Koch R (1882) Die Aetiologie der Tuberculose. Berliner Klin
Wochenschr 15:221-230
Kotake S et al (1994) Polymerase chain reaction for the detec-
tion of Mycobacterium tuberculosis in ocular tuberculosis.
Am J Ophthalmol 117:805-807
Mansour AM (1998) Optic disc tubercle. J Neuroophthalmol
18:201-203
Massaro D, Katz S, Sachs M (1964) Choroidal tubercles: a
clue to hematogenous tuberculosis. Ann Intern Med 60:
231-241
McMoli TE et al (1978) Tuberculous panophthalmitis.J Pediatr
Ophthalamol Strabis 15:383-385
Musial CE et al (1988) Identification of mycobacteria from
culture by using the GenProbe Rapid Diagnostic System
for Mycobacterium avium complex and Mycobacterium
tuberculosis complex. J Clin MicrobioI26:2120-2123
Nanda M, Pflugfelder SC, Holland S (1989) M tuberculosis
scleritis (letter). Am J Ophthalmoll08:736-737
Ni C et al (1982) Uveal tuberculosis. Int Ophthalmol Clin 22:
103-124
Olazabal F Jr (1967) Choroidal tubercles: a neglected sign.
JAMA 200:374-377
Raviglione MC, O'Brien RJ (1998) Tuberculosis. In: Frauci AS
(ed) Harrison internal medicine, part 7, 14th edn. McGraw-
Hill, New York
Regillo CD et al (1991) Ocular tuberculosis. JAMA 266:1490
Saini JS, Mukherjee AK, Nadkarni N (1986) Primary tubercu-
losis of the retina. Br J Ophthalmol 70:533-535
S.-J. Sheu
Saini JS, Sharma A, Pillai P(1988) Scleral tuberculosis. Trop
Geogr Med 40:350-352
Sant MS (1994) Ocular tuberculosis masquerading as retino-
blastoma. Indian J Pathol Microbiol 37:343-347
Sheu SJ et al (2001) Ocular manifestation of tuberculosis. Else-
vier Science, New York, p 1583 (Ophthalmology, vol 108)
Spindler E, Deplus S, Flageul B (1991) Acute uveitis in reversal
reactions. Acta Leprol 7:331-334
Stechschulte SU, Kim RY, Cunningham ET Jr (1999) Tubercu-
lous neuroretinitis. J NeuroophthalmoI19:201-204
Thygeson P (1954) Observation on nontuberculous phlycte-
nular keratoconjunctivitis. Trans Am Acad Ophthalmol
OtolaryngoI58:128-132
Thygeson P (1974) Nontuberculous phlyctenular keratocon-
junctivitis. In: Golden B (ed) Ocular inflammatory disease.
Thomas, Springfield, pp SO-52
Toossi Z (1998) Adjunctive immunotherapy of tuberculosis.
Cytokines Cell Mol Ther 4:105-112
Vine AK (1992) Severe periphlebitis, peripheral retinal isch-
emia, and preretinal neovascularization in patients with
multiple sclerosis. Am J OphthalmoI118:23-32
Von Michel J (1883) Ober Iris und Iritis. Albrecht v Grafes Arch
OphthalmoI27:171-282
Wecker LV (1874) Die Erkrankungen des Uvealtractus und
des Glaskorpers. Tuberkeln der Choroidea. Choroiditis
tuberculosa. In: Graefe A, Saemich T (eds) Handbuch der
gesammten Augenheikunde, vol 4. Engelmann, Leipzig, pp
642-648
World Health Organization (1997) Anti-tuberculosis drug
resistance in the world. The WHO/IUATLD project on anti-
tuberculosis drug resistance surveillance. World Health
Organization Global Tuberculosis Programme, Geneva
Yiannikas C, Walsh J, Mcleod JG (1983) Visual evoked poten-
tials in the detection of subclinical optic toxic effects sec-
ondary to ethambutol. Arch Neurol 40:645-648
Zumla A et al (1999) The tuberculosis pandemic - which way
now? J Infect 38:74-79
41 Otorhinolaryngeal Aspects of Tuberculosis
SUJATA MURANJAN

CONTENTS
now not confined to the third world countries alone.
41.1 Diagnosis 741 A 12 per cent increase was noted in the USA between
41.1.1 Otorhinolaryngeal Manifestations 742 1986 and 1991 and a greater than 30 per cent increase
41.2 Tubercular Cervical Lymphadenitis 742 was noted in Switzerland over the same period. Since
41.2.1 Anatomy of Cervical Lymph Nodes 742 1987, after a declining incidence for decades in Eng-
41.2.2 Clinical Presentation 742
41.3 Tuberculous Neck Abscesses 743
land and Wales, the number of cases of tuberculosis
41.3.1 Retropharyngeal Abscess 743 has shown a steady increase (Williams and Douglas-
41.3.1.1 Management 743 Jones 1995). The resurgence has been attributed to
41.3.2 Parapharyngeal Space Abscess 743 an increase in human migration which has caused
41.3.2.1 Investigations 744 rapid mingling of the infected and non-infected com-
41.3.2.2 Management 744
41.4 Tuberculous Laryngitis 745
munities (Agarwal and Bais 1998). These problems
41.4.1 Clinical Presentation 745 have been compounded by the explosive HIV/AIDS
41.5 Tubercular Infections of the Oral Cavity pandemic. The resultant immunosuppression has
and Pharynx 746 caused infection of this mycobacterium to spread
41.5.1 Clinical Features 746 globally with a vengeance. Thus to quote Williams
41.6 Tubercular Infection of the Nose 746
41.6.1 Management 747
and Douglas-Jones, the mycobacterium has "truly
41.7 Tuberculosis of the Salivary Glands 747 marched back" (Williams and Douglas-Jones 1995)
41.8 Tuberculosis of the Thyroid Gland 747 The bacterium has the propensity to affect virtu-
41.9 Tubercular Infections of the Ear 747 ally any organ in the body. The ear, nose and throat
41.9.1 Mode of Infection 747 are not excepted.
41.9.2 Signs and Symptoms 747
41.10 Medical Management 748
The general symptoms include low-grade fever
References 749 (with an evening rise), anorexia, weight loss and
lassitude. The disease however, can exist even in
the absence of these symptoms and is sometimes
detected accidentally.

In the early 18th century Jean Louis Petit accurately

described tuberculosis. Almost 100 years later, in
1882, Robert Koch discovered the tubercle bacillus
(Plester et al. 1980). Since then millions of deaths
have been attributed to tuberculosis. It now affects
more than a third of the world population (Sudre et
al. 1992). Approximately 1.5xl06 new cases are diag-
nosed annually in India alone (Mohanty 1999). In
the developing countries with poverty, malnutrition,
economic recession and overcrowding constantly on
the rise, the problem of tuberculosis is acute and con-
tinues to exist in spite of extensive tuberculosis eradi-
cation campaigns. This rising incidence is, however,
S. MURANJAN, MS (ENT), DNB, DORL
Consultant in ENT, Suman Appartment, 3rd Floor, 16B Naushir
Bharucha Road, Tardeo, Mumbai 400007, India
41.1
Diagnosis
The diagnosis may be suspected from the history and a
physical examination. A chest radiograph may show the
presence of a lesion suggestive of pulmonary tuberculo-
sis. A moderately raised ESR is corroborative. A defini-
tive diagnosis, however, has to be made prior to starting
anti-tubercular chemotherapy. This is possible by show-
ing the presence of acid fast bacilli in the sputum, secre-
tions or tissue. Cultures on classical Lowenstein-Jensen
slopes may take as long as six to eight weeks. The culture
process can be expedited by the use of new technology
such as the BACTEC H60 to provide results within three
to seven days (Evans et al. 1992). Culture is a necessary

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
742
pre-requisite for drug sensitivity testing of mycobac-
teria. Obtaining a drug sensitivity profile is becom-
ing more important as multidrug-resistant strains
are being increasingly isolated. Histopathology of the
biopsy tissue confirms the diagnosis. The pathognomic
features of a tubercular granuloma are central caseating
necrosis surrounded by epitheloid cells, giant cells and
lymphocytes. Application of modern molecular diag-
nostic technology provides a rapid and accurate means
for identification of M. tuberculosis from sputum and
other fresh samples. The technique involves extraction
of DNA and amplification of M. tuberculosis specific
DNA sequences using the polymerase chain reaction
(PCR). However it is important to realize that these
PCR techniques can also detect dead non-viable and
non-culturable organisms in sputum. This property can
therefore aid in retrospective diagnosis (Williams and
Douglas-Jones 1995).
41.1.1
Otorhinolaryngeal Manifestations
• Cervical Lymphadenitis
• Deep neck space tubercular abscesses
• Tubercular laryngitis
• Tubercular otitis media
• Tubercular nasal granulomas
• Lupus vulgaris
Rare cases of tubercular thyroid infection, tonsillo-
pharyngitis, tubercular otitic hydrocephalous, tuber-
culosis of the petrous apex and tubercular involve-
ment of the salivary glands have been reported.
A search has to be made for a primary focus of
infection in any diagnosed case of head and neck
tuberculosis.
41.2
Tubercular Cervical Lymphadenitis
41.2.1
Anatomy of Cervical Lymph Nodes
Cervical lymph nodes are classified into superficial
and deep groups. The superficial nodes include those
draining the scalp and face (i.e. the occipital, retro
auricular, parotid, facial, submandibular and the
submental group). The nodes comprising the deep
group are the retropharyngeal, jugulodigastric and
jugulo-omohyoid. Lymph nodes of the anterior or
S.Muranjan
central compartment of the neck are the infrahyoid,
prelaryngeal and the pretracheal group (Wright and
Kenyon 1987) (Fig. 41.1).
41.2.2
Clinical Presentation
Lymph node tuberculosis is one of the common-
est presentations of extrapulmonary tuberculosis.
Approximately five per cent of all patients with
tuberculosis develop tuberculous cervicallymphad-
enitis (Shikhani et al. 1989). Affected lymph nodes
may be discreetly enlarged. In 75 per cent of the cases
however, the lymph nodes are matted and clinically
indistinguishable from a lymphoma (Dandapat et
al. 1986). Perilymphadenitis is responsible for the
typical matted appearance which is clearly evident
on palpation. The initially discreet lymph nodes can
caseate subsequently to form abscesses. Tubercular
sinuses develop following rupture of these abscesses
(Raviglione and O'Brien 1998) (Fig. 41.2).
In a case with suspected inflammatory cervical
lymphadenopathy a trial of anti-inflammatory drugs
and antibiotics has to be given for a period of one
week to 10 days. If the lymph node does not regress
in size, it has to be subjected to fine needle aspiration
cytology (FNAC). When the FNAC shows the pathog-
nomonic appearance of a tubercular granuloma,
anti-tubercular therapy has to be promptly initiated.
Occasionally in cases of tuberculous lymph nodes,
with central caseation, aspiration of caseous material
Fig.41.1. Lymph nodes of the neck. 1 Submental; 2 subman-
dibular; 3, 4 jugulodigastric; 5 jugulo-omohyoid; 6 posterior
triangle nodes; 7 supraclavicular; 8 pre-laryngeal; 9 parotid;
10 occipital
Otorhinolaryngeal Aspects of Tuberculosis
Fig. 41.2. Tuberculous cervical lymphadenopathy showing
scars and sinuses
may be diagnostic. FNAC can accurately establish a
diagnosis in 83% of the cases (Dandapat et al. 1990).
If the FNAC is inconclusive and the lymph node does
not regress in size, an excision biopsy has to be per-
formed and submitted for histopathology.
The fate of the lymph nodes depends on the
immunity of the host, resistance of the organism
and effectiveness of the anti-tuberculous treatment.
Anti tuberculous treatment started at the earliest and
taken regularly results in the complete regression of
the disease. In up to a quarter of the patients the
nodes enlarge when treatment is commenced. This
phenomenon is thought to be due to a hypersensitiv-
ity reaction to tuberculin released from bacilli that
have been killed (Williams and Douglas-Jones 1995).
Multiple cervical lymph nodes which do not regress
in size despite adequate anti-tubercular medication
have to be excised by functional neck dissection. The
possibility of drug resistance also has to be consid-
ered in cases of cervical lymph nodes which do not
regress despite adequate anti-tuberculous treatment.
41.3
Tuberculous Neck Abscesses
Abscesses may originate from suppurating tubercu-
lar lymph nodes. These abscesses can dissect their
743
way through the various fascial planes in the neck
and considerably enlarge in size (Muranjan and
Kirtane 2001). An accurate knowledge of the anatomy
of the various fascial planes in the neck is essential
for drainage of these abscesses.
41.3.1
Retropharyngeal Abscess
A retropharyngeal abscess usually arises secondary
to tubercular infection of the cervical spine following
hematogenous spread. It can also develop subsequent
to a suppurated retropharyngeal tuberculous lymph
node.
Presenting manifestations are low grade fever,
odynophagia, drooling and painful neck movements.
A large abscess can cause airway obstruction leading
to stridor. As the disease progresses, patients often
need to support the head with their hands. Irritation
of the nerve roots leads to pain referred to the shoul-
ders or arms (Williams and DouglaS-Jones 1995).
Oropharyngeal examination generally reveals a
fluctuant bulge on the posterior pharyngeal wall.
A lateral radiograph of the cervical spine showing
destruction of the vertebrae along with an increase
in the pre-vertebral soft tissue space is a hallmark of
a retropharyngeal tubercular abscess. The normal
lordosis of the cervical spine is lost as a result of the
spasm of the cervical muscles.
41.3.1.1
Management
The treatment comprises gradual decompression
and drainage of the abscess intra-orally over its most
prominent part. This is accompanied by the institution
of anti-tubercular therapy. Immobilization with neck
splints is required in case of severe destruction of the
spine (Fig. 41.3). A sudden decompression and mobil-
ity can cause collapse of the vertebral column leading
to paraplegia. If the patient presents in stridor, an
emergency tracheostomy has to be performed prior to
embarking on drainage. Anti-tubercular drugs have to
be given for at least a year in cases of retropharyngeal
abscess due to tuberculosis of the cervical spine.
41.3.2
Parapharyngeal Space Abscess
A retropharyngeal abscess can track laterally in the
parapharyngeal space forming a parapharyngeal
744
Fig.41.3. Lateral plain radiograph of the neck showing a
splinted cervical spine following drainage of a tuberculous
retropharyngeal abscess
space abscess. This can also occur secondary to
suppuration of tubercular lymph nodes along the
internal jugular vein. These patients then present
with torticollis, tenderness and swelling in the lateral
aspect of the neck (Fig. 41.4).
41.3.2.1
Investigations
An ultrasound of the neck helps in identifying the
approximate size of the abscess. A computed tomogra-
phy (CT) scan can establish its exact size and extent.
41.3.2.2
Management
If the parapharyngeal abscess communicates with a
retropharyngeal component it can be drained intra-
orally (as mentioned above) so that drainage of the
retropharyngeal part itself will enable the parapha-
ryngeal component to drain. In case the abscess is
only restricted to the parapharyngeal space, it has
to be drained externally. Following a skin incision at
the level of the upper border of the thyroid cartilage,
the sternocleidomastoid muscle needs to be retracted
laterally. The carotid sheath has to be identified and
retracted following which the abscess cavity can be
approached and drained.
The pus obtained from any abscess drainage has to
be sent for smear and culture sensitivity tests. Granu-
S. Muranjan
FigA1.4. Left sided diffuse neck swelling due to an underlying
tuberculous parapharyngeal abscess
lations from walls of the abscess cavity have to be sent
for histopathological examination.
Complications encountered in cases of parapha-
ryngeal abscesses include encroachment on the
airway causing respiratory obstruction, sloughing
off of the carotid artery and internal jugular vein as a
result of the infection and spread to the mediastinum
(Fig. 41.5). Spread of the abscess superiorly towards
the base skull can lead to paralysis of the 9th, 10th,
lith and 12th cranial nerves.
It is not normal for a tubercular laryngeal abscess
to develop after suppuration of a pre-laryngeal
lymph node. In a previously reported case an abscess
had tracked along the anterior visceral space, inferior
Fig. 41.5. An axial CT scan showing a left tuberculous para-
pharyngeal space abscess tracking inferiorly into the superior
mediastinum
Otorhinolaryngeal Aspects of Tuberculosis 745

to the lamina of the thyroid cartilage, through the

cricothyroid membrane to extend intra-laryngeally.
The left true vocal fold was pushed medially, thus
encroaching upon the airway and leading to stri-
dor {Fig. 41.6). The abscess was drained externally.
Prompt institution of anti-tubercular treatment lead
to complete regression of the disease with restoration
of a normal airway (Muranjan and Kirtane 2001)
{Fig. 41.7).

41.4
Tuberculous Laryngitis
Fig. 41.6. An axial CT scan at the level of the vocal cords show-
At the turn of the century tuberculosis was the com- ing an abscess lateral to the left true vocal cord pushing it
monest condition to affect the larynx (Williams and medially and compromising the airway
Douglas-Jones 1995). Prior to the chemotherapeutic
era there was an associated morbidity of 70 per cent
with laryngeal disease (Ormerod 1951). The patho-
genesis of laryngeal tuberculosis is either primary or
secondary. Primary infection occurs in the absence
of pulmonary disease where the bacillus is transmit-
ted by aerosolized particles. The laryngeal mucosa
is thus directly infected. Secondary infection of the
larynx generally occurs in patients with pulmonary
tuberculosis. The infected sputum laden with the
acid-fast tubercular bacilli is expectorated and then
swallowed, thereby inoculating the larynx. Seeding
of the bacilli can also occur by the hematogenous
and lymphatic routes (Richter et al. 2001; Soda et al.
1989). The posterior part of the larynx was identified
as being the most frequently affected in early descrip-
Fig. 41.7. A post-operative CT scan of the same patient done
tions of laryngeal tuberculosis (Thompson 1924). two months after incision and drainage showing complete
This may have been due to its association with florid resolution of the disease and a normal airway
pulmonary disease in bed-ridden patients in whom
the infected sputum accumulated over the inter-
arytenoid region. Recent evidence indicates that the manifestations mimic those of a laryngeal carcinoma.
lesions are located in the anterior part of the larynx. A distinguishing feature is the presence of painful
This is thought to be related to the ambulatory treat- dysphagia which is a common symptom of tuber-
ment regimens now instituted which prevent pooling culous laryngitis (Bull 1966). Tuberculosis should
of saliva in the posterior larynx (Soda et al. 1989). therefore be considered in the differential diagnosis
of malignancy of the larynx and hypopharynx. This is
especially important in the perspective of resurgence
41.4.1 of tuberculosis in all parts of the world (Goyal et al.
Clinical Presentation 1998). The other differential diagnoses, though less
common but which should nevertheless be borne in
The patient presents with hoarseness of insidi- mind include leprosy, sarcoidosis, fungal infections,
ous onset, cough with occasional blood streaked Wegener's granulomatosis and cat-scratch disease.
sputum, odynophagia, dysphagia, dyspnea, and sore Systemic diseases like systemic lupus erythemato-
throat (Richter et al. 2001). Occasionally the patient sus, rheumatoid arthritis, relapsing polychondritis
may complain of referred otalgia. There may be and amyloidosis are also included in the differential
a history of pulmonary tuberculosis. The clinical diagnosis (Richter et al. 2001).
746
The vocal folds are the sites most commonly
affected by tuberculosis, closely followed by the
ventricular folds (Soda et al. 1989). The other sites
affected in approximately 10% to 15% of cases are
the aryepiglottic folds, arytenoids, posterior com-
missure, subglottis and epiglottis (Richter et al. 2001).
Contrary to the earlier descriptions the anterior half
of the larynx is affected twice as often as the posterior
half (Bailey and Windle-Taylor 1981).
A laryngoscopy will show the characteristic fea-
tures of granulations and irregularity of the vocal
cords termed as a "moth eaten appearance". The
posterior commissure and arytenoids may be cov-
ered with granulations. Infiltration of the lax mucosa
around the arytenoids produces a pale, fleshy swell-
ing that may appear tense and shiny as if full of fluid,
but is in fact solid. In the past this has been referred
to as pseudo-edema. When the lesions heal by fibro-
sis, the epiglottis assumes a typical 'turban shaped'
appearance and the arytenoids appear pyramidal.
Fibrosis at the crico-arytenoid joint can impair vocal
cord mobility. Stridor in tubercular laryngitis can
result from granulations which occlude the glottis in
the active phase of the disease, vocal fold paralysis
secondary to mediastinal disease (Shah and Ramak-
antan 1990) or stenosis or fibrosis of the larynx. A
tracheostomy may be required in such cases.
Smear and culture tests of the sputum may show
presence of the acid fast bacilli. Acid fast bacilli are
detected in the sputum in 70 per cent of patients (Bailey
and Windle-Taylor 1981). A microlaryngoscopy with
biopsy of the granular lesions will show the charac-
teristic tubercular granuloma on histopathology thus
distinguishing it from a malignancy. Laryngeal tuber-
culosis responds rapidly to anti-tubercular treatment.
Symptoms resolve within two weeks of instituting che-
motherapy and the sputum becomes negative for myco-
bacteria. The persistence of a suspicious lesion after
treatment may be due to non-compliance, a resistant
organism or a concomitant carcinoma which should be
excluded by a repeat biopsy (Hunter et al.1981).
41.5
Tubercular Infections of the Oral Cavity
and Pharynx
Though oral and pharyngeal involvement in this
disease is quite rare, there have been reports of the
pyriform fossa, tonsils and oropharynx being affected
by tuberculosis (Goyal et al. 1998; Gupta et al. 2001).
Tonsillar tuberculosis could be demonstrated in over
S.Muranjan
one third of patients with cervical lymphadenitis at
the beginning of the century. The tonsils were usually
infected with M. bovis by ingestion of infected cow's
milk. With pasteurization of milk, tonsillar tuberculo-
sis is now a rare entity. Primary infection of the oral
cavity can occur following mucosal damage but as a
rule the infection is secondary to pulmonary disease.
41.5.1
Clinical Features
The lesion manifests itself as a painless ulcer which
has an undermined edge. More often than not there
is an associated regional lymphadenopathy. The sites
affected are gums, tongue, palate and floor of mouth
(Prabhu et al. 1978; Haddad et al. 1987). The appear-
ance is indistinguishable from carcinoma and biopsy
is therefore mandatory.
The differential diagnosis includes ulcers due to
trauma, hematological disorders, actinomycosis,
syphilis, midline granulomas, Wegener's granuloma
and aphthous ulcers (Gupta et al. 2001). Sarcoidosis,
deep mycotic infections and malignancy must also be
borne in mind. Tubercular ulcers usually regress com-
pletely following institution of anti-tubercular drugs.
41.6
Tubercular Infection of the Nose
Tuberculosis of the nose is a rare entity which pres-
ents either in the nodular or the ulcerative form. It
generally occurs secondary to pulmonary tuber-
culosis. Lupus vulgaris is an indolent and chronic
form of tuberculosis affecting the skin of the face
and mucous membrane of the nose (Weir 1987). In
the nasal cavity the muco-cutaneous junction is the
commonest site of inoculation. The nasal lesions are
frequently associated with precursor nodules on
the face. The disease course is slow with periods of
remission and exacerbations. The early appearance
is typical. There is mucosal pallor with minute 'apple
jelly nodules. Application of topical adrenaline fails
to blanch these nodules which in turn are rendered
more obvious against the surrounding mucosa. As
these nodules coalesce the mucosa becomes granular
and the septum may perforate. The edge of the per-
foration is irregular and surrounded by pale mucosa
(Williams and Douglas-Jones 1995). Ulceration may
be followed by healing with fibrosis and contraction
with distortion of the alae nasi.
Otorhinolaryngeal Aspects of Tuberculosis
The disease may spread from the muco-cutaneous
junction to the floor of the nasal cavity, turbinates,
and posteriorly towards the choanae. The patient then
presents with mucopurulent and occasionally blood
stained nasal discharge and obstruction. Anterior rhi-
noscopy and diagnostic nasal endoscopy reveals a red-
dish nodule. Occasionally the disease may also present
as a nasal polyp (AI Serhani 2001). The diagnosis is
confirmed by visualization of the acid fast tubercular
bacilli in the nasal discharge and a biopsy of the nodule
will show the classical tubercular granuloma.
Complications like dacryocystitis, corneal ulcer-
ation and nasopharyngeal lupus can occur and atro-
phic rhinitis may be a sequel.
41.6.1
Management
Treatment comprises anti-tuberculous drugs which
are given for 6-9 months. When the disease is com-
pletely arrested, surgical repair of the nasal deformity
may be required.
41.7
Tuberculosis of the Salivary Glands
The parotid gland is the commonest gland to be
affected, followed by the submandibular and the
sublingual gland. Infection within the parotid most
often arises in the intraparotid lymph nodes and is
often the only site affected. The bacilli presumably
reach the gland either via the duct or through lym-
phatic channels. The focus of entry is either in the
pharynx or the tonsils (Batsakis 1974). Pain is a late
manifestation and facial palsy is a rare complication.
Fine needle aspiration is generally diagnostic and
conventional anti-tuberculous chemotherapy com-
menced early can avoid unnecessary surgery.
41.8
Tuberculosis of the Thyroid Gland
Tuberculous thyroiditis is an extremely rare condition.
It arises secondary to other tubercular foci of infec-
tion through hematogenous spread. The two types of
tuberculosis of the thyroid gland are the diffuse miliary
form and the caseous type. The miliary form is usually
less common and asymptomatic while the focal form
747
is more common. This usually presents as a painless
solitary nodular goiter in a euthyroid patient (Patankar
et al. 1999). Occasionally abscess formation may occur.
Fine needle aspiration cytology is a sensitive, specific
and reliable method for diagnosing tuberculous thy-
roiditis (Mondal and Patra 1995). Prompt institution of
anti-tubercular drugs is required to control the infec-
tion. In cases of multiple tubercular abscesses of the
gland, a thyroidectomy may be indicated.
41.9
Tubercular Infections of the Ear
Tuberculous otitis media was first reported over 100
years ago (OdetoyinboI988). The incidence of tuber-
culosis of the middle ear cleft is reportedly low in the
developed countries. However this varies from country
to country depending on a number of factors like the
socio-economic status, under-diagnosis and under-
reporting. Achieving an early diagnosis is difficult due
to several factors. Significant amongst these are low
index of suspicion, false negative cultures due to the
presence of other bacteria in the specimen which inter-
fere with the isolation of the bacillus and change in the
histopathology of the middle ear because ofwidespread
use of antibiotic ear drops such as neomycin and genta-
mycin with weak anti-tubercular activity (Yaniv 1987;
Grewal et al. 2000). In addition to the above factors, ear
swabs have a poor sensitivity for AFB on microscopy as
the yield of mycobacteria from clinical specimen other
than sputa is very low (Lee and Drysdale 1993).
41.9.1
Mode of Infection
Bacilli from the infected sputum in case of pulmonary
tuberculosis can gain access to the middle ear via the
Eustachian tube. In regions where unpasteurized milk
is consumed, spread of Mycobacterium bovis can occur
directly via the Eustachian tube. Hematogenous dis-
semination is also an important route of spread of
infection.
41.9.2
Signs and Symptoms
The disease is of an insidious onset. Classically the
patient presents with painless profuse otorrhoea and
profound hearing loss. The otorrhoea fails to respond
748
to standard antibiotics. Occasionally mild deep-seated
pain may be present possibly due to granulations within
the mastoid under pressure (Plester et al. 1980). On
otoscopic examination classical multiple perforations
can be seen on the tympanic membrane. These mul-
tiple perforations are caused by a number of tubercles
involving the tympanic membrane. Though these were
initially thought to be a hallmark of the disease, they are
rarely encountered today. A central, or a total, perfora-
tion is more commonly encountered (Farrugia 1997).
Occasionally pale granulations or polyps may be seen.
The disease can cause extensive destruction of the ossic-
ular chain. Osteomyelitis of the labyrinthine bone and
the lateral sinus lead to sequestra formation (Windle-
Taylor and Bailey 1980). The resulting deafness is gener-
ally out of proportion to the otoscopic findings and it
may be of a conductive nature. In case of labyrinthine
involvement deafness may be of a mixed or sensorineu-
ral type and is accompanied by symptoms of giddiness
and vomiting. Involvement of the fallopian canal by the
disease can give rise to a lower motor neuron type of
facial paralysis. Tuberculosis must be the prime suspect
in patients with facial paralysis in chronic middle ear
disease, without cholesteatoma, (Plester et al. 1980).
Spread of the disease can also occur along the petrous
apex where in addition to the otorrhoea, the patient
also presents with a lateral rectus palsy, diplopia and
pain along the distribution of the trigeminal nerve. This
symptom complex is classically termed as Gradenigo's
syndrome. A tympanomastoidectomy has to be per-
formed. All the granulations and bony sequestra have
to be cleared from the middle ear and mastoid cavity.
While exploring the ear, the possibilities of erosion of
the bony facial nerve canal and fistulae over the lateral
semi circular canal and promontory have to be borne
in mind. The granulations and bone have to be sent for
biopsy and antitubercular drugs must be started at the
earliest and administered for about 9-12 months.
In a previously operated mastoid cavity, tuber-
culous otitis media must be considered if there is a
recurrence of granulation tissue, slow wound healing,
persistent otorrhoea and formation of bone seques-
tra. Hence during any aural surgery if granulation
tissue is encountered, it should always be sent, with
the bone, for biopsy.
41.10
Medical Management
Discussion of the entire medical management of
tuberculosis would be beyond the scope of this
S.Muranjan
chapter. A combination of drugs is used for the
treatment of tuberculosis to reduce the probability
of emergence of drug-resistant mutants. Factors
such as monotherapy, poor compliance, omission of
some drugs, suboptimal dosage, poor drug absorp-
tion or an insufficient number of active agents in
a drug regimen contribute to the emergence of
resistance to multiple drugs within a matter of
months (Gostin 1993). The usual standard drug
regimen for head and neck tuberculosis include
a four drug therapy comprising rifampicin, iso-
niazid, pyrazinamide and ethambutol given over a
period of 2 months followed by a combination of
rifampicin and isoniazid for another 4-7 months.
Following administration of anti-tubercular drugs,
the patient's local as well systemic symptoms
along with the ESR have to be monitored. After
3-4 weeks of therapy in a completely compliant
patient, if the symptoms do not improve or if the
ESR remains persistently high, drug resistance has
to be suspected. Quinolones such as ciprofloxacin
or ofloxacin, aminoglycosides likes streptomycin or
kanamycin, or other drugs like ethionamide, PAS,
clofazimine, and cycloserine may have to be added
to the regimen. Culture-sensitivity tests have to be
performed to confirm the sensitivity of the bacilli
to these drugs. Any anti-tuberculous treatment has
to be preferably started following consultation with
a chest physician.
Poor compliance is thought to be the primary
cause for emergence of drug resistance. To increase
compliance many centers employ a combination of
strategies such as incentives and directly observed
therapy (DOT) in which a health worker actually wit-
nesses the patient taking the medicines. This ensures
complete compliance on part of the patient.
Thus maintaining a high index of suspicion, early
diagnosis and prompt treatment at an early stage can
lead to complete cure. However the dictum, "preven-
tion is better than cure" cannot be more adequately
stressed especially in case of tuberculosis. Strength-
ening the patient education and awareness programs,
easy accessibility to the health centers and prompt
and regular use of the antituberculous drugs are cru-
cial in bringing this dreaded disease under control.
Acknowledgements. I am thankful to Dr. Renuka
Bradoo, Professor and Head, Department of ENT, Lok-
manya Tilak Municipal Medical College and General
Hospital, Mumbai, India for the constant encourage-
ment given by her and Dr. Surendra Lele, Consultant
ENT Surgeon, Seven Hills Hospital, Visakhapatnam,
India for the valuable suggestions given by him.
Otorhinolaryngeal Aspects of Tuberculosis
References
Agarwal P, Bais AS (1998) A clinical and videostroboscopic
evaluation of laryngeal tuberculosis. J Laryngol Otol 112:
45-48
Al Serhani AM (2001) Mycobacterial infection of the head
and neck: presentation and diagnosis. Laryngoscope 111:
2012-2016
Bailey CM, Windle-Taylor PC (1981) Tuberculous laryngitis.
Laryngoscope 91 :93-100
Batsakis JG (1974) Non neoplastic diseases of the salivary
glands. In: Tumours of the head and neck. Williams and
Wilkins, Baltimore, pp 54-55
Bull TR (1966) Tuberculosis of the larynx. Br Med J 2:991-992
Dandapat M, Mishra B, Dash SP, Car PK (1990) Peripheral
lymph node tuberculosis: a review of 80 cases. Br J Surg
77:911-912
Dandapat M, Padhi N, Nanda BP (1986) Peripheral lymph
node tuberculosis - a comparison of various methods of
management. Indian J Tuberc 33:20-23
Evans KD, Nakasone AS, Sutherland PA, de la Maza LM, Peter-
son EM (1992) Identification of Mycobacterium tubercu-
losis and Mycobacterium intracellulare directly from pri-
mary BACTEC cultures by using acridinium-ester-Iabelled
DNA probes. J Clin MicrobioI30:2427-2431
Farrugia EJ (1997) Tuberculous otitis media- a case report. J
LaryngoIOtollll:58-59
Gostin LO (1993) Controlling the resurgent tuberculosis epi-
demic. J Am Med Assoc 269:255-260
Goyal A, Nagarkar NM, Uppal KS, Mohan H, Daas A (1998)
Tuberculosis of the pyriform fossa - a rare entity. J Laryn-
goIOtoI112:782-783
Grewal DS, Hathiram BT, Agarwal R, Dwivedi A, Walvekar R
(2000) Otitic hydrocephalus of tubercular origin: a rare
cause. J Laryngol OtoI114:874-877
Gupta K.B, Tandon S, Jaswal TS, Singh S (2001) Tuberculosis
of tonsil with unusual presentation. Indian J Tuberc 48:
223-224
Haddad N, Zaytoun G, Hadi U (1987) Tuberculosis of the soft
palate: an unusual presentation of oral tuberculosis. Oto-
laryngol Head Neck Surg 97:91-99
Hunter AM, Millar JW, Wightman AJA, Horne NW (1981) The
changing pattern oflaryngeal tuberculosis. J Laryngol Otol
95:393-398
Lee PYC, Drysdale AJ (1993) Tuberculous otitis media - a dif-
ficult diagnosis. J Laryngol Otoll07:339-341
Mohanty KC (1999) Tuberculosis and atypical mycobacteria.
In: Sainani GS (ed) API textbook of medicine, 6th edn.
Association of Physicians of India, Mumbai, pp SO-55
Mondal A, Patra DK (1995) Efficacy of fine needle aspiration
749
cytology in the diagnosis of tuberculosis of the thyroid
gland: a study of 18 cases. J Laryngol Otol109:36-38
Muranjan SN, Kirtane MV (2001) Tubercular laryngeal
abscess. J Laryngol OtoI115:660-662
Odetoyinbo 0 (1988) Early diagnosis of tuberculous otitis
media. J Laryngol Otol102:133-135
Ormerod F (1951) A review of tuberculosis of the upper air
passages during the past thirty years and its treatment by
streptomycin. J Laryngol OtoI65:461-471
Patankar T, Prasad S, Chowdhary S, James P (1999) Plunging
goitre due to tuberculous thyroiditis. J Assoc Physicians
India 47:347
Plester D, Pusalkar A, Steinbach E (1980) Middle ear tubercu-
losis. J Laryngol OtoI94:1415-1421
Prabhu S, Daftary D, Pholakia HM (1978) Tuberculosis ulcer of
the tongue: report of case. J Oral Surg 36:384-386
Raviglione MC, O'Brien RJ (1998) Tuberculosis. In: Fauci AS,
Braunwald E, Isselbacher KJ (eds) Harison's principles of
internal medicine, vol 1, 14th edn. McGraw-Hill, New York,
pp 1004-1014
Richter B, Fradis M, Kohler G, Ridder GJ (2001) Epiglottic
tuberculosis: differential diagnosis and treatment. Ann
Otol Rhinol Laryngolll0:197-201
Shah P, Ramakantan R (1990) Hoarseness of the voice due to
recurrent laryngeal nerve palsy in tuberculous mediasti-
nits (letter). Arch Otolaryngol Head Neck Surg 116(1):108
Shikhani AH, Hadi UM, Mufarrij AA, Zaytoun GM (1989)
Mycobacterial cervical lymphadenitis. Ear Nose Throat J
68:660-672
Soda A, Rubio H, Salazar M, Ganem J, Berlanga D, Sanchez
A (1989) Tuberculosis of the larynx: clinical aspects in 19
patients. Laryngoscope 99:1147-1150
Sudre P, Dam G, Kochi A (1992) Tuberculosis: a global over-
view of the situation today. Bull WHO 70:149-159
Thompson St Clair (1924) Tuberculosis of the larynx - its
significance to the physician. Br Med J 2:841-844
Weir N (1987) Acute and chronic inflammations of the nose.
In: Mackay IS, Bull TR (eds) Scott-Brown's otolaryngology,
vol 4. Rhinology. Jaypee, New Delhi, pp 115-141
Williams RG, Douglas-Jones T (1995) Mycobacterium marches
back. J Laryngol Otol 109:5-13
Windle-Taylor PC, Bailey CM (1980) Tuberculous otitis media
a- a series of 22 patients. Laryngoscope 90:1039-1044
Wright D, Kenyon G (1987) Cancer of the neck. In: Stell PM
(ed) Scott-Brown's otolaryngology, vol 5. Laryngology.
Jaypee, New Delhi, pp 315-339
Yaniv E (1987) Tuberculous otitis media: a clinical record.
Laryngoscope 97:1303-1306
42 Endocrine and Metabolic Manifestations
of Tuberculosis
SOHAIL INAM and MONA AL-SHAHED

CONTENTS in active tuberculosis results from either a direct

affliction of the organs with the disease (endocrine
42.1 Introduction 751
tuberculosis) or more commonly as an abnormal-
42.2 Endocrine Tuberculosis 751
42.2.1 Adrenal Tuberculosis 751 ity of circulating hormones as a consequence of
42.2.1.1 CT Features of Adrenals Tuberculosis 752 systemic disease. The latter results from the effects
42.2.1.2 Illustrative Case Report 754 of malnutrition, cytokines and other inflammatory
42.2.1.3 Subclinical Tuberculous Adrenalitis mediators on the hypothalamic-pituitary axis, hor-
and Related Controversies 755
mone binding proteins and peripheral metabolism
42.2.2 Thyroid Tuberculosis 756
42.2.2.1 Illustrative Case Report 757 of hormones.
42.2.3 Pituitary Tuberculosis 758
42.2.3.1 Illustrative Case Report 760
42.2.4 Parathyroid Tuberculosis 760
42.3 Endocrine Perturbations Related
42.2
to Systemic Tuberculosis 761
42.3.1 Hypercalcemia 761 Endocrine Tuberculosis
42.3.2 Water and Sodium Balance in Tuberculosis 762
42.3.3 Effect of Tuberculosis Endocrine tuberculosis is uncommon. It may affect
on Circulating Hormones 764 any endocrine organ and is the result of organ seed-
42.4 Tuberculosis and Diabetes Mellitus 765
ing following hematogenous spread or contiguous
42.5 Endocrine and Metabolic Effects
of Anti-Tuberculous Drugs 766 spread from a nearby-afflicted structure. Hematog-
References 766 enous dissemination of virulent tubercle bacilli can
occur during the course of primary infection or long
after the initial infection. Its clinical manifestation
may be part of the systemic disease or may be the
sole manifestation of the condition. This chapter
42.1 deals with tuberculosis affecting the adrenal, thyroid
Introduction and pituitary glands. Tuberculosis of the gonads is
covered elsewhere (chapter 39). Direct tuberculous
Tuberculosis (TB) is often termed 'a great mimicker', involvement of the parathyroid glands and the pan-
as it can affect almost any organ in the human body creatic islets is extremely rare.
and present in a variety of ways. The endocrine
system is no exception, though the incidence of
clinical disease is very low. The occurrence of extra- 42.2.1
pulmonary tuberculosis appears to be increasing in Adrenal Tuberculosis
the developed world and is probably a reflection of
the impact of mv infection. Endocrine dysfunction Primary adrenal insufficiency (PAl) secondary to
bilateral adrenal destruction by tuberculosis, was
originally described by Addison in 1855. In a review
S. INAM, MBBS, FRCP (Edin), FRCP of his eleven patients with adrenal insufficiency
Head of Endocrinology Division, Department of Medicine, seven had tuberculosis. In the past when the disease
Riyadh Armed Forces Hospital, P.O. Box 7897, Riyadh 11159, was rampant and untreatable, tuberculosis was the
Saudi Arabia
M. AL-SHAHED, MBBS, FRCR
commonest cause of PAl accounting for 70-80% of
Consultant Radiologist, Department of Radiology, Riyadh Armed the cases (Guttman 1930). With the development of
Forces Hospital, P.O. Box 7897, Riyadh 11159, Saudi Arabia effective chemotherapy and early treatment, tubercu-

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
752
losis has increasingly become less common as a cause
of PAl and presently contributes from 7-20% of the
cases (Nomura et al. 1994; Oelkers 1996). Tubercu-
lous PAl is still thought to be a major cause of PAl
in countries like Saudi Arabia where the disease is
common. However, in a series of 15 new adult patients
diagnosed with PAl since 1982 at the Riyadh Armed
Forces Hospital tuberculosis could only be estab-
lished in three as a cause.
Adrenal involvement occurs from hematogenous
seeding. The adrenal glands are especially vulnerable
because of their rich blood supply and immunosup-
pressive effects of high local production of cortico-
steroids. In an autopsy series of patients with late
generalized tuberculosis, adrenal seeding was seen in
53% of patients (Slavin et al. 1980). This is not neces-
sarily associated with clinical disease. Both adrenal
glands are involved, though they may not be equally
affected. Seeding is followed by an inflammatory
reaction which if unchecked causes gradual destruc-
tion of the glands. About 90% destruction of both
adrenal glands is required before the onset of clini-
cal adrenal insufficiency. Destruction of the adrenal
medulla is more common than that of the adrenal
cortex (Guttman 1930). The clinical manifestations,
however, are related to the deficiency of adrenal cor-
tical hormones. All the layers of the adrenal cortex
are affected but as the zona glomerulosa consists
of scattered cell groups rather than being a distinct
layer, it may be initially only partly destroyed, so that
the clinical manifestations may be solely those of
glucocorticoid deficiency.
The histology of adrenal TB is no different from
any other organ affliction (see Chapter 10 on Pathol-
ogy for details). Earlier in the course of disease there
is inflammatory cell infiltration followed by granu-
loma formation resulting in adrenal gland enlarge-
ment. As the disease progresses micro-abscess for-
mation and tissue necrosis occurs. Mass lesions may
develop secondary to cold abscesses. With progres-
sive destruction, the glands are gradually replaced by
caseous nodules and fibrosis so that after about two
years they become normal, or small, in size and ulti-
mately over time they become fibrosed and calcified
(Vita et al. 1985).
The clinical features of adrenal TB are similar to
those of any other cause of primary adrenal insuf-
ficiency. These include weakness, fatigue, weight loss,
abdominal pain, diarrhea, hyperpigmentation and
orthostatic hypotension. There may be associated
fever and abnormal mental activity. The laboratory
tests show hyponatremia, hyperkalemia, azotemia,
anemia and eosinophilia. The clinical syndrome may
S. Inam and M. Al-Shahed
accompany active systemic disease such as acute
pulmonary, miliary or other extra-pulmonary tuber-
culosis (Alvarez and McCabe 1984). More commonly
as adrenal destruction is a slow process the clinical
presentation occurs as an isolated entity years after
the initial infection (Nomura et al. 1994; Sanford and
Favour 1956; Vita et al. 1985). Rarely the presentation
is that of an adrenal mass (unilateral or bilateral) seen
on CT of the abdomen (Jayakar et al. 1998). Whereas
the presentation is insidous, patients may occasionally
present with an acute adrenal crisis following the use
of rifampicin (Elansary and Earis 1983).
The diagnosis of adrenal insufficiency is con-
firmed by demonstrating a decreased, or absent,
cortisol response to exogenous ACTH (synacthen
stimulation test). Elevated ACTH levels and associ-
ated mineralocorticoid deficiency confirm that the
adrenal insufficiency is due to adrenal destruction.
Mineralocorticoid deficiency is characterized by
hyponatremia and hyperkalemia and confirmed by
findings of high renin and low aldosterone levels.
Once a hormonal diagnosis of PAl is made the next
step is to determine its cause. A supportive CT scan,
or MRI of the adrenal glands, absence of adrenal
antibodies and other causes of this disease are
required to support TB as the cause of PAI.(Laureti et
al. 1998). The presence of extra-adrenal TB is highly
supportive. A final diagnosis requires either histo-
logical proof or positive tissue cultures. Although
not performed routinely a CT-guided needle biopsy
of the adrenals can provide diagnostic material and
is useful in doubtful cases. It is of great value during
investigation of unilateral or bilateral adrenal masses
(Yee et al. 1986).
42.2.1.1
CT Features of Adrenals Tuberculosis
CT plays a major role in diagnosis of established
tuberculous adrenal disease (Hauser and Gurret
1986; Villabona et al. 1993;Wang et al. 1998; Wilms
et al. 1983). The appearance of the adrenals on CT
will vary depending on the stage of disease. Adre-
nal enlargement has been commonly documented
during routine screening in pulmonary tuberculosis.
It is more likely to occur with acute pulmonary TB,
rather than chronic pulmonary disease (Kelestimur
et al. 1994). An enlargement of the adrenals seen in
patients with pulmonary tuberculosis during routine
CT probably represents stress related hyperplasia or
adrenalitis and not clinical disease (Gulmez et al.
1996; Kelestimur et al.1994). The following presenta-
tions are well recognized in tuberculous adrenalitis.
 Endocrine and Metabolic Manifestations of Tuberculosis
a. Bilateral adrenal enlargement: The enlargement is
smooth, has sharp margins and diffuse enhance-
ment post contrast. This appearance is non-specific
and may represent direct tuberculous involvement
of the glands or simply hyperplastic adrenals as
a part of stimulatory response to systemic illness
(Fig. 42.1).
b. A unilateral mass: Rarely the disease manifests as
a unilateral mass making it difficult to differenti-
ate it from an adrenal tumor (Fig. 42.2).
c. Irregularly enlarged adrenals with multiple ring
enhancements: This would be a classical appear-
ance and mimics that seen in other organs. The
appearances are due to multiple areas of caseation
necrosis or abscess formation (Fig. 42.3).
d. Small fibrotic glands with multiple calcifications.
This can also be seen on a plain abdominal radio-
graph (Fig. 42.4).
c both suprarenal glands (arrows)
753
An MRI of the adrenal glands shows features
similar to that of CT and offers no major advantage.
Ultrasound can also detect enlarged adrenals, adre-
nal masses and calcification. The appearances are
again not specific. Abdominal calcification can also
be seen on plain film radiographs of the abdomen
in the late stages of the disease. This occurs in about
50% of cases and is not specific since it can be seen
in other types of infectious adrenalitis and following
adrenal hemorrhage (Vita et al. 1985).
The treatment of adrenal tuberculosis consists
of standard anti-TB therapy in a similar manner
to other forms of extra-pulmonary TB, along with
adrenal hormone replacement. Duration of six
months of anti-TB therapy using first line medica-
tions is sufficient. Glucocorticoids are replaced with
hydrocortisone (20-30 mg per day in divided doses)
or prednisolone (4-7.5 mg/day) or dexamethasone
Fig.42.1. a Post-enhanced CT of the upper abdomen dem-
onstrating bilateral suprarenal gland enlargement showing
homogeneous enhancement (arrowheads). L. liver, S. spleen.
b Magnetic resonance image (MRI), of the same patient, Tl-
weighted image (TR 516, TE 14): The enlarged suprarenal
glands appear as low signal intensity with well defined mar-
gins (arrow). c MRI post-enhanced (gadolinium), TI (TR 516,
TE 14), image showing diffuse homogeneous enhancement of
754
Fig.42.2. Post-enhanced CT of the upper abdomen demon-
strates unilateral adrenal gland enlargement. The right adre-
nal gland is enlarged and is diffusely enhancing (large arrow).
Note multiple shotty lymph nodes in the para-aortic region
(small arrows). L Liver; S spleen
Fig.42.3. Post-enhanced CT of the upper abdomen show-
ing enlargement of both adrenal glands with multiple ring
enhancement consistent with abscess formation (arrows)
Fig. 42.4. Abdominal plain radiograph demonstrating bilateral
suprarenal gland calcification (arrowheads)
S. Inam and M. Al-Shahed
(0.25-0.75 mg/day). Mineralocorticoids are replaced
using fludrocortisone in a dose of 50-200 Jlg/day. As
rifampicin is an enzyme inducer and increases the
metabolism of steroid hormones the dose of gluco-
corticoids should be increased (Edwards et al. 1974).
Mineralocorticoids on the other hand are less likely
to be affected (Schulte et al. 1987). During rifampi-
cin therapy the dose of glucocorticoids is increased
to two to three times the replacement dose initially
and then slowly titrated down to the lowest possible
dose achieving adequate replacement. The treatment
is adjusted based on clinical response, serum electro-
lytes,ACTH levels and measurement of plasma renin.
ACTH levels are usually maintained between two to
three times the upper limits of normal. Initial pre-
sentation with an acute adrenal crisis is rare and may
follow a physical stress or the initiation of therapy
with rifampicin. This is a medical emergency. The
goal of therapy is treatment of hypotension, reversal
of electrolyte abnormalities and glucocorticoid defi-
ciency. Hypotension is treated with liberal infusion of
0.9% saline. Hydrocortisone 100 mg or dexametha-
sone 4 mg intravenously is started immediately after
drawing blood for cortisol and ACTH analysis. Dexa-
methasone therapy is preferred initially as it does
not interfere with the measurement of cortisol and
it does not interfere with the synacthen stimulation
test (for confirmation of the diagnosis). Intravenous
fluids are maintained for 24-48 hours and hydrocor-
tisone 100 mg is given every eight hours for the first
24 hours. The latter is then tapered to a replacement
dose over the next 24-48 hours and fludrocortisone
replacement started. Once adrenal damage manifests
as a clinically diagnosed adrenal insufficiency, a 90%
destruction of both adrenal glands has occurred and
this is usually irreversible (Bhatia et al. 1998). How-
ever case reports do document recovery, even after
a prolonged period indicating that, it may be appro-
priate to re-evaluate adrenal function after therapy
(Penrice and Nussey 1992). Although rifampicin
therapy is known to unmask underlying adrenal
insufficiency (Elansary and Earis 1983) it has no del-
eterious effects in the outcome of treatment (Barnes
et al. 1989).
42.2.7.2
Illustrative Case Report
A 50 year-old female presented with several months
history of productive cough, fever and weight loss.
A diagnosis of pulmonary IB was made based on a
representative chest radiograph, sputum smear posi-
tive for acid-fast bacilli (AFB) and positive culture.
Endocrine and Metabolic Manifestations of Tuberculosis
Her biochemistry at that time was unremarkable. She
was started on a four-drug regimen with isoniazid
(INH), rifampicin, pyrazinamide and ethambutol. A
few days later she developed a generalized skin rash.
All medications were stopped and later introduced
individually in low dosage. She redeveloped rash
with both pyrazinamide and ethambutol. She was
thus treated with INH and rifampicin. Six months
later she presented to the emergency room with
altered consciousness. She had been experiencing
poor appetite, increasing pigmentation, general
fatigue and dizzy spells for the past several weeks.
She was hypotensive and pigmented. Her glucose
on arrival was 1.6 mmolll, serum Na+ 116 mmolll,
K+ 6.7 mmolll. Her basal cortisol was 87 nmolll and
failed to respond to IV 250 flg of synacthen (30 and 60
minute samples 75 and 84 nmolll respectively). The
CT of her adrenals is shown in Fig. 42.3, which is clas-
sical of multiple abscesses. Adrenal auto-antibodies
were negative. She responded well to IV hydrocorti-
sone and hydration with 0.9% saline. She has since
been maintained on hydrocortisone and fludrocor-
tisone. Her follow-up CT scans showed progressive
shrinkage with no calcification. She has permanent
primary adrenal insufficiency.
42.2.1.3
Subclinical Tuberculous Adrenalitis and Related
Controversies
Whereas PAl secondary to tuberculosis is an estab-
lished entity, debate rages on the existence and signifi-
cance of subclinical adrenal disease during screening
of patients with active tuberculosis. Certain caveats,
however, need to be addressed to understand this
controversy. In systemic tuberculosis, as in other sys-
temic illnesses, there is adrenal stimulation as a result
of increased activation of the hypothalamic-pituitary
axis. Mycobacterium cell-wall components cause a
release of cytokines such as Ill, IL6 and TNF alpha, all
of which stimulate the release of CRH/ACTH resulting
in an increase in serum cortisol levels (Habib et al. 2001;
Fukata et al. 1993; Spangelo 1994; Zhang et al. 1993).
This stimulation can result in hyperplasia of the adre-
nal glands to the extent that they may appear enlarged
on CT scans (Gulmez et al. 1996). The levels of serum
cortisol and adrenal response to exogenous ACTH
vary depending upon the degree of stress, chronicity
of illness, malnutrition, drug therapy and adaptation to
stress. Basal cortisol levels are higher and often elevated
(Barnes et al.1989; Chan et al.1993). There may be loss
of the diurnal variation ofserum cortisol concentration
(Sarma et al. 1990; York et al. 1992) and the cortisol
755
levels are higher in acute and miliary disease than in
chronic TB (Barnes et al. 1989).
Adrenal inflammation in tuberculosis if signifi-
cant, is associated with abnormal hormonal reserve
with varying degrees of dysfunction. The most com-
monly applied test to assess adrenal reserve has been
the measurement of cortisol response to 250 flg of
synacthen (a synthetic form of ACTH). Application
of this test in screening for the presence of subclinical
disease in patients admitted with active pulmonary
or extrapulmonary tuberculosis is a major source of
this debate. The abnormal response rate has varied
from 0-55% (Barnes et al. 1989; Ellis and Tayoub
1986; Kelestimur et al. 1994; Keven et al. 1998; Post
et al. 1994; Prasad et al. 2000; Sarma et al. 1990; Sid-
diqi et al. 1997). The major reason for this difference
is the diagnostic criteria used in these studies. The
criteria used after synacthen stimulation have been
either a peak cortisol response of 200-300 nmolll
above baseline or an absolute value of 550 nmolll at
any time during the test. Using the first criterion the
frequency of abnormality has been between 30-55%
(Ellis and Tayoub 1986; Prasad et al. 2000; Sarma et al.
1990; Siddiqi et al. 1997). However if basal values are
already high then the absolute increase post ACTH is
likely to be lower because of an already near maximal
endogenous stimulation of the adrenals (Chan et al.
1993; Patel et al.1991). Thus studies using a threshold
value of 550 nmolll during the test, which is presently
favored, show a frequency of 0% (Chan et al. 1993;
Kelestimur et al. 1994; Post et al. 1994). Other con-
founding factors in these studies are the duration and
extent of disease, which may also have a bearing on
the adrenal response to ACTH stimulation (Barnes
et al. 1989; Sarma et al. 1990). Rifampicin an inducer
of cytochrome P450 microsomal enzymes increases
cortisol catabolism, some reports suggest a reduced
response to Synacthen with the use of rifampicin fur-
ther complicating the issue (Ellis and Tayoub 1986).
Alternatively ACTH levels have been used as a mea-
sure of detecting early adrenal dysfunction. Reduction
in adrenal reserve causes a decrease in circulating
free cortisol levels, which by interrupting the negative
feedback on the hypothalamic-pituitary unit results
in a rise in ACTH levels. Whereas this test has been
shown to be very sensitive in picking up both subclini-
cal and established adrenal insufficiency, its applica-
tion has been unable to clarify the controversy. This is
due to the fact that systemic illness itself alters ACTH
levels. The study by Post et al. (1994) clearly shows the
limitation of this test in systemic tuberculosis.
It is possible that the difference in the prevalence of
abnormal adrenal reserve following stimulation with
756
Synacthen, could be solely explained by the use of dif-
ferent diagnostic criteria, yet some patients do show
abnormal responses even when using more stringent
criteria (Barnes et al. 1989; Prasad et al. 2000). This
could represent a subtle adrenalitis or a functional
defect in cortisol secretion related to the effect of ill-
ness. The latter has been demonstrated in septic shock
(Rothwell et al. 1991). Improvement of the cortisol
response after treatment using the same criteria tends
to support these possibilities (Barnes et al. 1989; Chan
et al. 1993). The fact that subtle tuberculous adrenal-
itis does occur has been shown histologically. In the
autopsy series by Slavin adrenal seeding was seen in
53% of their cases whereas only one patient had clini-
cal adrenal insufficiency (Slavin et al. 1980). In a recent
report by Chan et al., none of three patients who died
of active pulmonary TB with bilateral adrenal granu-
lomata at autopsy (and who had also had a synacthen
stimulation test available), had cortisol deficiency
(Chan et al. 1993). Thus at least in acute pulmonary
TB, tuberculous adrenalitis may occur but is usually
not associated with PAL The blunted increase post-
stimulation with synacthen is related to an already
maximally stimulated adrenal gland with high basal
cortisol levels. Things may be different in chronic dis-
ease as PAl usually presents years later.
In order to address this controversy Kelestimur has
recently reported on the assessment of adrenal reserve
using the Imcg synacthen test in active pulmonary
tuberculosis (Kelestimur et al. 2000). This test uses a
low dose of ACTH, which causes a submaximal stimu-
lation and more suited to pick up subtle abnormalities
in adrenal hormonal reserve. They showed normal
adrenal reserve in all their patients and no advantage
over the standard 250 mcg synacthen test.
Thus the diagnosis of PAl due to tuberculous
requires a clearly abnormal response to synacthen.
Subtle adrenalitis does occur but it may not be of clin-
ical significance when treated early. Although a high
level of suspicion for this life threatening complication
should be maintained, routine screening for abnormal
adrenal reserve is not recommended. If such screening
is attempted it should be only done as a part of clinical
studies using more stringent and uniform criteria. The
implications of subtle abnormalities of adrenal reserve
in tuberculosis remain unclear.
42.2.2
Thyroid Tuberculosis
The first report of tuberculous thyroid involvement
was by Lebert in 1862. Clinical thyroid tuberculosis
S. Inam and M. Al-Shahed
(TT) is very rare. In thyroid surgical specimens Rankin
reported tuberculous thyroiditis in 0.1% of 20,758
specimens examined at the Mayo clinic (Rankin and
Graham 1932) and Levitt reported two cases amongst
2114 consecutive thyroidectomy specimens (Levitt
1952). Slavin reported a 14% seeding of the thyroid
gland in an autopsy series of selected patients in the
pre- and post-antibiotic eras (Slavin et al. 1980). In the
largest single clinical series of 18 cases diagnosed by
fine needle aspiration (FNA) with demonstrable AFB,
Mondal reported an incidence of 1.15% amongst 1565
thyroid FNAs (Mondal and Patra 1995).
Tuberculous involvement of the thyroid gland can
occur from hematogenous seeding or from conta-
gious spread from an adjacent afflicted lymph node.
TT secondary to congenital transmission in infants
of mothers with untreated tuberculosis has also been
reported (Kang and Chi 1990). The thyroid gland has
an extensive blood supply and hematogenous spread
would not be unusual, yet clinical tuberculosis of
the thyroid gland is rare. Tuberculosis causes the
formation of epitheloid granulomas with central
caseation necrosis, Langhans' giant cells and periph-
erallymphocyte cuffing (Kapoor et al. 1985). TT can
present in two forms. 1) As part of miliary tubercu-
losis, which rarely gives rise to clinical disease or 2)
As a goiter due to caseous tuberculosis (Barnes and
Weatherstone 1979).
The commonest clinical presentation is that of a
thyroid nodule or mass often mimicking a carcinoma
(Khan et al. 1993; Tan 1993; Takami and Kozakai
1994). Less commonly the patient presents as a cold
abscess (Surer et al. 2000) or a diffuse goiter (Barnes
and Weatherstone 1979). In a review of the litera-
ture of TT in Japan in 63.6% of 44 cases the clinical
diagnosis was that of a thyroid tumor (Takami and
Kozakai 1994). Pressure on the adjacent structures
may cause symptoms of dysphagia, dysphonia, dys-
pnea and at times laryngeal nerve palsy (Berger et
al. 1983). Patients may also present with symptoms
mimicking subacute thyroiditis (Berger et al. 1983;
Sachs et al. 1988). However in TT pain, tenderness
and fever are less common than in patients with bac-
terial thyroiditis and the presentation more indolent
(Berger et al. 1983). TT is usually associated with TB
elsewhere, occasionally though it may be the sole
manifestation of the disease (Mondal and Patra 1995)
and occasionally the disease is found incidentally in a
surgical specimen (Kapoor et al. 1985).
Most patients with TT are euthyroid. Infrequently
the patient may present because of abnormal thyroid
hormone function. Both primary hypothyroidism
secondary to extensive damage (Barnes and Weather-
Endocrine and Metabolic Manifestations of Tuberculosis
stone 1979) and thyrotoxicosis due to a subacute thy-
roiditis (Berger et al. 1983; Nieuwland et al.1992) have
been described. Abnormalities in circulating thyroid
hormones in TB are more commonly due to the effect
of systemic disease - the sick euthyroid syndrome.
This is characterized by a normal or low free T4 (FT4),
reduced levels of T3, increase in RT3 (reverse T3) and a
normal or low TSH. These abnormalities reverse with
treatment ofTB (Chow et al.1995).
The diagnosis ofTT can be made if there is a strong
clinical suspicion especially in the presence of extra-
thyroidal TB. Ultrasound of the thyroid is imprecise.
It may show a thyroid mass, heterogeneous thyroid
texture, abscess formation or fibrocalcific disease
(Kang et al. 2000). CT of the thyroid may be highly
suggestive if it shows nodules with centrallow-atten-
uation and peripheral ring enhancement (Moon et al.
1997; Kang et al. 2000). Thickening due to inflamma-
tion of the neighboring muscles, subcutaneous tissue
and skin may also be seen and is termed the 'dermal
sign'. Cervical lymphadenopathy is commonly pres-
ent. Radionuclide thyroid scans show focal, or diffuse,
diminished uptake in the affected tissue (Mondal and
Patra 1995). Histological diagnosis is required for
final confirmation. Recent reports document the effi-
cacy of FNA of the thyroid in the diagnosis of TT. The
presence of caseating granulomas is highly sugges-
tive and surgery can be avoided. Additionally thyroid
aspiration may also yield AFB on smear or culture
(Das et al. 1992; Mondal and Patra 1995). Mondal
reported positive AFB on smear in 12 cases and posi-
tive cultures in 14 of 18 cases diagnosed using FNA.
This high yield may relate to the rich blood supply
and consequent higher oxygen concentration avail-
able in thyroid tissue. Other granulomatous disor-
ders such as subacute granulomatous thyroiditis (de
Quervain's thyroiditis), sarcoidosis and syphilis need
to be considered in the differential diagnosis. The
type of granulomas and presence of caseation necro-
sis help in differentiation. Demonstration of AFB and
positive culture of course confirm the diagnosis.
TT is rare, however it should be considered in the
evaluation of a cold nodule in endemic areas espe-
cially in the presence of cervical lymphadenopathy.
Radiology is of limited value except where a CT scan
shows an area of central necrosis with ring enhance-
ment. FNA of the thyroid is the diagnostic tool of
choice. If the cytology shows caseation necrosis, or
pus, then smears should be stained for AFB and aspi-
rate sent for appropriate culture. FNA of associated
enlarged lymph nodes is likely to greatly increase the
diagnostic yield. A surgical biopsy should be limited
only to doubtful cases.
757
The optimal treatment of this rare condition
remains undetermined. Anti-TB therapy alone,
surgical excision or drainage with anti-TB therapy
have all been advocated. Although the literature
often recommends a combination of surgery and
anti-tuberculous chemotherapy (Berger et al. 1983),
this is because some form of surgery has often been
required for diagnosis (Khan et al. 1993; Sachs et al.
1988; Takami and Kozakai 1994; Tan 1993). In the
absence of an abscess anti-TB therapy alone should
suffice (Kang et al. 2000; Mondal and Patra 1995).
The duration of therapy should be six months using
standard drug regimens. The associated abnormality
of thyroid function when present needs additional
therapy. Hypothyroidism appears to be permanent
and requires long-term replacement therapy with
thyroxine, except when it occurs as a part of subacute
thyroiditis (Barnes and Weatherstone 1979). On the
other hand, the treatment of thyrotoxicosis is less
well established and would follow that of other causes
of subacute thyroiditis, i.e. beta-blockers, pain relief
and treating the underlying condition (Berger et al.
1983; Nieuwland et al. 1992). Corticosteroids could
be used in severe cases after starting anti-TB therapy.
Anti-tuberculous chemotherapy may also affect the
thyroid. Treatment with para-aminosalicylic acid
and ethionamide has been associated with goiter as
they have actions similar to propylthiouracil; these
drugs are rarely used in modern practice. Rifampicin
may enhance extrathyroidal metabolism of thyroid
hormones (Christensen et al. 1989), this does not
seem to be of clinical importance.
42.2.2.1
Illustrative Case Report
A 17 year-old male presented with a two month his-
tory of fever, weight loss, headache and a painful neck
swelling. He had clinical signs of thyrotoxicosis and
a diffuse non-tender firm goiter on examination.
Another smaller swelling was palpable in the left
supraclavicular fossa. His FT4 50 pmolll (N 11-23),
TSH <0.02 mUll, thyroglobulin 262 Ilgl1 (N 0-55),
ESR 60 mmllst hour and thyroid antibodies were
negative. An ultrasound of the neck revealed an
enlarged thyroid with non-homogeneous architec-
ture and several enlarged cervical lymph nodes with
necrotic centers. An isotopic thyroid scan showed a
very low thyroidal uptake suggestive of subacute thy-
roiditis. An FNA of the lymph node showed granu-
lomatous inflammation whereas the thyroid aspirate
was consistent with acute inflammation. AFB smears
were negative but the aspirates from both the lymph
758
node and thyroid grew Mycobacterium tuberculosis
on culture, fully sensitive to first line anti-tubercu-
lous drugs. The patient was commenced on rifampi-
cin, INH, pyrazinamide and pyridoxine. Thyrotoxi-
cosis was treated with beta-blockers. His sequential
thyroid function tests showed a classical picture of
recovery from thyroiditis with the TSH rising to 60
mUll before gradually returning to normal. He has
put on 4 Kg in weight and is doing well.
42.2.3
Pituitary Tuberculosis
Intracranial tuberculomas constitute 0.15-4% of
intracranial space occupying lesions (DeAngelis
1981; Mauruce-Williams 1972). In the developing
countries with poor socio-economic conditions they
may still account for 15-30% of neurosurgical cases
(Basaria et al. 2000; Esposito et al. 1987). Pituitary
tuberculosis is extremely rare. Coleman and Meredith
reported the first case of intrasellar tuberculoma in
the English literature in 1940. Since, there have been
36 cases reported in the English literature. The dis-
ease is more common in females, which account for
over 70% of the cases (Ashkan et al. 1977; Ranjan and
Chandy 1994; Sharma et al. 2000). An autopsy series
of 14,160 specimens in the pre-antibiotic era over 11
years yielded only two cases (Kirshbaum and Levy
1941) and Slavin reported a 4% incidence of pituitary
involvement in a select group of patients with late
generalized tuberculosis (Slavin et al.I980). The larg-
est clinical series is that by Sharma et al. of 18 cases
collected over a IS-year period (Sharma et al. 2000).
Intrasellar tuberculomas have also been reported
along with another pituitary tumor (Gazioglu et al.
1999; Sharma et al. 2000, 2001).
Pituitary involvement results from either seeding
during initial hematogenous dissemination with
subsequent activation or direct extension from adja-
cent structures such as the sphenoid sinus, brain and
meninges. Any part of the pituitary gland (anterior,
posterior, and stalk) or the hypothalamus may be
involved. The presentation often occurs in isolation
although it can occur in association with tuberculous
meningitis, or as part of miliary TB.
The clinical presentation is that of a non-functioning
pituitary tumor with headache being the most promi-
nent and commonest manifestation. Vomiting and
fever may be present. Compression of the pituitary and
its stalk leads to varying degrees of hypopituitarism,
hyperprolactinemia and diabetes insipidus. Reduced
growth hormone and gonadotropins are the common-
S. Inarn and M. Al-Shahed
est hormone deficiencies. Pressure on the optic chiasm
can lead to visual field disturbances and extension into
the cavernous sinus causes cranial nerve (oculomo-
tor and abducent) palsies. However there may be no
endocrine abnormality present at diagnosis (Sharma
et al. 2000). Sinha reported absence of endocrine dys-
function in 71.4% of the cases they reviewed (Sinha et
al. 2000). Systemic features are usually absent though
a raised ESR has been described in several reports
(Patankar et al. 2000; Petrossians et al. 1998).
The diagnosis is based on suggestive radiology,
surgical findings and demonstration of granulomas
consistent with TB at histology. Demonstration of
TB bacilli in pituitary tissue or culture is confirma-
tory. The presence of active TB at another site espe-
cially the CSF would be highly suggestive although,
abnormal CSF alone can be seen in other types of
granulomatous hypophysitis and even in lympho-
cytic hypophysitis (Cheung et al. 2001). Concomitant
active TB has been reported in only two cases (one
lymph node, one CSF) where as a history of previ-
ously treated TB is described in nine cases. Thus in
the vast majority intrasellar tuberculoma is the only
manifestation of TB.
The radiological features help in pointing to the
nature of the lesion (Ashkan et al. 1977; Patankar et
al. 2000; Ranjan and Chandy 1994; Sharma et al. 2000;
Sinha et al. 2000). This is important as differentiating
these lesions from those due to pituitary adenomas
can avoid surgery. Plain film skull radiographs may
show sellar or suprasellar calcification. Typically
CT scan shows an isointense lesion with diffuse,
intense enhancement with contrast. The latter can
be so marked that the appearance is similar to
that of an aneurysm. Less commonly the lesion is
hypointense with enhancement of the margins. MRI
is the modality of choice. MRI shows a mass which
is isointenselhypointense compared to the brain on
Tl weighted images (Tl WI), variably hyperintense
on T2 weighted images (T2WI) (which enhances
with gadolinium contrast). The enhancement pat-
tern may be diffuse or show areas of low intensity
within (Fig. 42.5). Thickening of the pituitary stalk
when present is highly suggestive (Pereira et al. 1995;
Sinha et al. 2000). However these features are shared
with other forms of granulomatous and lymphocytic
hypophysitis (Buxton and Robertson 2001; Cheung
et al. 2001). Recently a report by Shimizu et al. has
suggested that Gallium-67 scanning may help to dif-
ferentiate these inflammatory lesions from adenoma-
tous pituitary tumors (Shimizu et al. 1998).
The surgical findings always show dural thicken-
ing, intrapituitary fibrosis, and yellow-gray tissue
 Endocrine and Metabolic Manifestations of Tuberculosis
a b
that is difficult to remove. The tissue is usually avas-
cular (Ashkan et al. 1977; Ranjan and Chandy 1994;
Sharma et al. 2000; Sinha et al. 2000). Histological
findings are those of epitheloid granulomas with
central caseation necrosis, Langhans' giant cells and
peripheral lymphocyte cuffing. There is varying
degree of inflammation, destruction and fibrosis of
the normal pituitary tissue.
The definitive diagnosis requires demonstra-
tion of AFB in the specimen or on culture. This is
exceedingly rare and only one case has had posi-
tive TB culture from pituitary tissue (Patankar et al.
2000) and another two had demonstrable AFB in the
surgical specimen (Basaria et al. 2000; Sinha et al.
2000). Whereas Ziehl Neelsen stain for AFB has been
reported in most series the tissue has been rarely sent
for culture due to an initial diagnosis of pituitary
tumor. In a recently reported case a positive PCR for
TB from the pituitary tissue was used for diagnosis
(Petrossians et al. 1998).
The differential diagnosis is that of granuloma-
tous hypophysitis which is an increasingly described
entity (Buxton and Robertson 2001; Cheung et al.
2001). The granulomatous inflammation in this
condition can be primary (idiopathic/autoimmune)
or secondary to conditions such as TB, fungal infec-
tions, sarcoidosis, Wagner's granulomatosis, syphilis,
histiocytosis and Crohn's disease. Thus the presence
of granuloma alone on histology is not diagnostic of
TB. The presence of caseating necrosis along with
epitheloid cells and Langhans giant cells is highly
759
Fig. 42.5a, b. MRI of the pituitary gland. a T1
weighted coronal image demonstrates enlarge-
ment of the pituitary gland (straight arrow).
The enlarged gland appears hypointense to
the adjacent brain tissue. b Post-enhanced
(gadolinium) coronal T1 image shows diffuse
and intense enhancement of the pituitary gland
(curved arrow)
suggestive of TB in the absence of other causes of
a necrotising granulomatous inflammation. Of the
36 cases reported in the English literature only 18
have documented the presence of caseation necrosis.
Other reports have either documented no caseation
or failed to describe it casting some uncertainty
about the diagnosis.
The diagnosis of pituitary tuberculoma should be
suspected in young females presenting with a non-
functioning pituitary mass in which the radiology
is suggestive and confirmed by trans-sphenoidal
biopsy unless there is active TB elsewhere. The trans-
sphenoidal surgical approach is preferred as it avoids
opening the CSF space and bacterial seeding of the
meninges. Caseation necrosis with epitheloid granu-
lomas and Langhans giant cells are highly suggestive
of tuberculosis. Demonstration of TB bacilli on stain-
ing or culture would be diagnostic. The role of PCR in
the diagnosis of pituitary TB remains unclear.
The treatment consists of anti-tuberculous
therapy. It should be started soon after surgery to
avoid TB meningitis. The exact duration of therapy
remains unproven, based on the recommendation
for intracranial tuberculomas therapy duration of
9-18 months has been used in the cases reported
(Ashkan et al. 1977; Ranjan and Chandy 1994;
Sharma et al. 2000; Sinha et al. 2000). A 12 month
course of therapy based on the recommendations
for meningeal and disseminated TB however would
suffice (Horsburgh et al. 2000). Resolution of the
lesion is usually seen at six months. Use of addi-
760
tional corticosteroids does not seem to give any
extra advantage over anti-tuberculous therapy alone
in most cases. Primary granulomatous hypophysitis
results in resolution of the pituitary mass with time,
whether spontaneous or in response to steroids,
thus resolution with anti-TB therapy alone should
not be considered diagnostic.
Recovery of pituitary function in tuberculous
hypophysitis appears to be common (Sinha et al.
2000), this is in contrast to that of tuberculous adre-
nal involvement. Such recovery may be complete
or partial. Persistent hypopituitarism however may
result (Ranjan and Chandy 1994). Hypopituitarism
has also been described as a sequel to TB menin-
gitis in childhood (Haslam et al. 1969; Lam et al.
1993; Sherman et al. 1971). This is most probably
a result of hypothalamic damage. Lam et al. (1993)
reported in a subset of 49 patients with TB meningi-
tis in childhood that 20% had some abnormality of
pituitary function on stimulation testing. The study
was compromised by its retrospective nature and
by the fact that only 49 of 246 patients were avail-
able for evaluation. Growth hormone deficiency was
the most common sequel followed by gonadotropin
deficiency. MRI performed in 10 affected patients
showed a normal pituitary in seven. Pituitary atro-
phy, enhanceable tissue in the suprasellar cistern,
hypothalamus or pituitary stalk and dilatation of
the third ventricle were the abnormalities seen.
Central precocious puberty has also been described
in patients with a history of tuberculous meningitis.
Desai et al. have reported a history of tuberculous
meningitis in 37% of girls and 27% of boys with cen-
tral precocious puberty (Desai et al. 1993).
42.2.3.1
Illustrative Case Report
A 29 year mother of three presented with progres-
sive headache of six months duration. The headache
was like pressure deep and behind the eyes. Two
weeks prior to presentation the headaches became
worse and were associated with diplopia. She had
a long history of irregular periods. Her first two
pregnancies required bromocriptine therapy.
Following the delivery of her last child she had
taken four injections of medroxyprogesterone and
remained amenorrheic for 18 months. On presen-
tation she had a partial left third nerve palsy. FT4
6.1 pmolll (N 10-25), TSH 0.119 mUll, PRL 830, LH
0.4, Estradiol 76. An MRI showed a 1.6 cm pituitary
tumor extending into the left cavernous sinus and
the suprasellar cistern almost touching the optic
S. Inam and M. Al-Shahed
chiasm. The lesion was isointense on T1 WI, vari-
ably hyperintense on T2WI with diffuse enhance-
ment following gadolinium contrast (Fig. 42.5).
There was no other abnormality seen in the brain or
meninges. Trans-sphenoidal surgery was performed
which showed a very thickened dura, a firm fibrous
yellow-gray lesion that was difficult to excise. Histo-
pathology confirmed granulomatous inflammation
with caseation necrosis (Fig. 42.6). Stains for AFB
and fungi (PAS) were negative. No tissue was sent
for TB culture. CSF examination and CT scan of
the chest were normal. ESR 41, CRP 20 (0-6), ACE
39 (11-127). The patient has been treated with a
standard four-drug regimen for TB, with a plan to
continue for 12 months. Three months after com-
mencing treatment her FT4, TSH, PRL returned to
normal and a repeat MRI showed marked regression
of the pituitary lesion.
42.2.4
Parathyroid Tuberculosis
Involvement of the parathyroid glands can occur
through hematogenous spread or from an adjacent
affected structure. Clinical disease is extremely rare.
In a recent case report, Kar et al. documented the
presence of epitheloid granulomas in a parathyroid
adenoma. Whereas AFB smear and culture were
negative, a PCR for Mycobacterium tuberculosis from
the parathyroid tumor homogenate was positive. The
patient also had granulomatous lymphadenitis (Kar
et al. 2001).
Fig.42.6. Histological section showing necrotizing granulo-
matous hypophysitis. Note the frequent multinucleated giant
cells (arrowheads) and central necrosis (arrow). Residual acini
of adenohypophysis are seen on the far left. (Hematoxylin and
eosin, original magnification _40).
Endocrine and Metabolic Manifestations of Tuberculosis
42.3
Endocrine Perturbations Related
to Systemic Tuberculosis
42.3.1
Hypercalcemia
Active TB has been well documented as a cause
of hypercalcemia and hypercalciuria. It has been
reported both in adults and children. The incidence
of hypercalcemia varies from 0-50% in the literature
(Abbasi et al. 1979; Fuss et al. 1988; Kelestimur et al.
1996; Roussos et al. 2001; Sharma 1981; Shek et al.
1990). The wide variation in the incidence of hypercal-
cemia between countries results from the difference in
the intake of vitamin D and calcium and the variable
amount of sun exposure (Chan 1997). Countries like
the UK, Belgium and Saudi Arabia with low levels of
vitamin D have a very low incidence of hypercalcemia
(Anonymous 1981; Fuss et al. 1988; Taha et al. 1984).
Similarly a report from Turkey failed to show an
increase in calcium in patients with TB (Kelestimur
et al. 1996). On the other hand the prevalence may
be as high as 25% in Sweden and 50% in Australia
(Chan 1997). In Hong Kong tuberculosis is the second
commonest cause of hypercalcemia in hospitalized
patients after malignancy (Shek et al. 1990). Other
reasons for the differences in reporting relate to
analytical methods, cut-off values used and failure
to correct for protein binding, (hypoalbuminemia is
especially common in these patients at presentation)
(Morris et al. 1989; Roussos et al. 200!). The ionized
calcium (free calcium) is more likely to be elevated
since it is independent of protein binding. However
in a recent report measurement of ionized calcium
did not give added information over total calcium
corrected for serum albumin (Roussos et al. 200!).
Most patients with hypercalcemia have pulmonary
tuberculosis, it has been less frequently associated
with miliary infection, peritonitis and osteomyelitis
(Chan et al. 1994; Lim et al. 1994).
The mechanism of hypercalcemia is similar to that
of other granulomatous disorders and results from
increased production of 1,25 dihydroxycholecalcif-
erol (1,25 (OHh D3 or calcitriol), the active form of
vitamin D. Calcitriol levels are elevated in patients
with tuberculosis compared to controls (Epstein et
al. 1984; Peces and Alvarez 1987; Pouchot et al. 1993;
Sharma 2000; Yang et al. 2000). Normally calcitriol
is produced in the kidney and tightly regulated by
PTH and phosphate levels. In granulomatous dis-
orders there is an increased extra-renal production
by mononuclear cells. This has been documented
761
in vivo in anephric patients (Peces and Alvarez
1987; Pouchot et al. 1993; Gkonos et al. 1984) and in
vitro from macrophages and T cells obtained from
patients with tuberculosis (Cadranel et al. 1990). The
major sites of production are the lung and lymph
nodes. Elevated free calcitriol levels have also been
documented in pleural fluid in patients with tuber-
culosis (Barnes et al. 1989). Macrophages possess
the 1 alpha-hydroxylase enzyme which converts 25
OH D3 to its active form 1,25 (OHh D3. The local
production of this metabolite enhances the ability
of macrophages and monocytes to inhibit growth
of mycobacteria and influences the granulomatous
reaction (Cadranel et al. 1994; Crowle at al. 1989;
Rook 1988; Sharma 2000). The antibacterial efficacy
of pyrazinamide is also enhanced by vitamin D and
its active metabolites (Crowle et al. 1989). Normally
production of calcitriol by mononuclear cells is
locally regulated, increasing levels of 1,25 (OHh D3
inhibit further production and increases its own deg-
radation. If this process becomes unregulated then a
systemic spill over of the active vitamin D metabolite
occurs resulting in hypercalcemia. It has been shown
that gamma interferon may be responsible for both
induction of increased 1 alpha hydroxylase activity
and inhibition of the negative feed back response to
calcitriol. It may be one of the stimuli responsible
for allowing continued production and decreased
degradation of calcitriol (Dusso et al. 1997). Cal-
citriol causes hypercalcemia mainly from increased
intestinal absorption of calcium. It also causes
bone resorption. The parathyroid hormone levels
are suppressed. Other mechanisms contributing to
hypercalcemia in tuberculosis include an increased
production of parathyroid hormone related peptide
(PTHrP) at inflammatory sites and bone involve-
ment in extensive tuberculosis (Roussos et al. 2001;
Braman et al. 1973). The hypercalcemia may become
more evident on exposure to vitamin D supplements,
sunlight, increased calcium intake or if there is asso-
ciated renal impairment (Abbasi et al. 1979; Fuss et
al. 1988; Sharma 198!). The ability of the kidney to
excrete large amounts of calcium is protective and
may explain the mild degree of calcium elevation in
tuberculosis. Thus renal impairment can contribute
significantly to increase in serum calcium and result
in severe hypercalcemia (Carroll 1987). As in sarcoid-
osis, hypercalciuria is more common than hypercal-
cemia and being vitamin D-related is seen mainly in
the postabsorptive state (Martinez et al. 1993).
The degree of hypercalcemia is usually mild,
though severe hypercalcemia has been reported
especially in association with miliary disease (Isaacs
762
et al. 1987; Yang et al. 2000). Typically hypercalcemia
is seen at presentation. However this may be masked
by the associated hypoalbuminemia. Some reports
have first noticed the condition one week to several
months after start of therapy. This may relate to
improved albumin levels, antigen stimulation by
release of bacterial products, more sun exposure and
improved nutrition. The latter could increase vita-
min D and calcium in diet (Fuss et al. 1988; Sharma
1981). Renal impairment due to rifampicin-induced
interstitial nephritis was responsible in one patient
(Carroll 1987).
Hypercalcemia is usually mild and asymptomatic.
It is detected on routine biochemical tests. When
levels are over 2.9 mmolll the patient may experience
lethargy, anorexia, abdominal pain and vomiting.
It may also contribute to impaired mental status in
elderly individuals. The latter is especially likely to
occur with levels in excess of 3.2 mmolll. Labora-
tory findings in addition to hypercalcemia show an
elevated inorganic phosphate, suppressed PTH levels
and high normal or high 1,25 (OHh D3 levels. The
24-hour urine calcium is elevated.
Treatment of hypercalcemia usually consists of
adequate hydration, reduced intake of calcium and
vitamin D and anti-TB therapy. Vitamin D supple-
ments should be avoided in these patients. If calcium
intake is reduced then oxalate intake should also be
lowered, to avoid increased oxalate absorption and
renal stones. Anti-tuberculous therapy is associated
with a drop in levels of calcitriol and serum calcium
(Roussos et al. 2001; Yang et al. 2000). Rifampicin
induces hepatic microsomal enzyme induction
and increases catabolism of vitamin D, whereas
isoniazid (INH) inhibits 1 alpha hydroxylase activ-
ity and reduces synthesis of calcitriol (Brodie et al.
1982; Davies et al. 1987; Fuss et al. 1988). Both these
actions lead to a decrease in the levels of calcitriol
and calcium and a consequent rise in PTH. In coun-
tries with low stores of vitamin D, anti-tuberculous
therapy may not only prevent the development of
hypercalcemia but may induce osteomalacia (Chan
1996; Shah et al. 1981). The above measures usually
suffice in the management of this condition. Hyper-
calciuria also responds to anti-tuberculous therapy
(Martinez et al. 1996). The activation of vitamin D
is also inhibited by glucocorticoids, which can be
used in more symptomatic or severe cases. A dose of
10-30 mg of prednisolone is sufficient in inhibiting
this process and treat hypercalcemia (Sharma 2000).
Alternatively ketoconazole (a cytochrome P450
inhibitor), which affects the I-alpha hydroxylase
enzyme, can be effective in the management of this
S. lnam and M. Al-Shahed
condition (Saggese et al.1993).As in sarcoidosis both
chloroquine and hydroxychloroquine are also effec-
tive in lowering calcium levels by inhibiting vitamin
D activation (Sharma 2000). Bisphosphonates such as
pamidronate, which act predominantly by inhibiting
osteoclastic bone resorption are also effective as in
other causes of vitamin D toxicity, substantiating the
contribution of bone resorption in causing hyper-
calcemia (Isaacs et al. 1987). Severe hypercalcemia
is rare and when it happens the therapy consists of
aggressive fluid replacement, use of loop diuretics
after adequate hydration, intravenous hydrocorti-
sone and if required pamidronate or calcitonin may
be used. Dialysis using a low calcium solution may be
required if there is severe renal impairment or inabil-
ity to control hypercalcemia.
42.3.2
Water and Sodium Balance in Tuberculosis
Hyponatremia is a well-known manifestation of TB.
It occurs in 10-60% of patients with active pulmo-
nary TB (Chung and Hubbard 1969; Post et al. 1994;
Morris et al. 1989). The condition occurs with a
greater frequency in those with miliary disease and
TB meningitis (Karandanis and Shulman 1976; Singh
et al. 1994) and has also been described in tuber-
culous epididymo-orchitis (Motiwala et al. 1991). It
is usually mild and related to systemic illness. Less
frequently it is due to associated adrenal or pituitary
insufficiency.
Two processes are involved in its pathogenesis.
The first and major mechanism of hyponatremia
is altered excretion of water by the kidney due to
release of arginine vasopressin (AVP). The levels of
AVP when measured are usually elevated or inap-
propriately high to the degree of hypo-osmolality
(Cotton et al. 1991, 1993; Hill et al. 1990; Rose and
Post 2001; Usalan et al. 1998). Several mechanisms
may be responsible for this. Some patients are intra-
vascular volume depleted as a result of fever, poor
intake, increased diaphoresis and increased GI loss
causing appropriate secretion of AVP. The majority of
patients however are euvolemic and the levels of AVP
are inappropriate to the hypo-osmolality, consistent
with the syndrome of inappropriate secretion of anti-
diuretic hormone (SIADH).
SIADH in tuberculosis is due to the following
pathogenetic mechanisms: (i) Excessive production
of AVP or similar peptides from the hypothalamus or
inflammatory tissue in response to pulmonary infec-
tion. The usual source of AVP is the hypothalamus
Endocrine and Metabolic Manifestations of Tuberculosis 763

and is most probably cytokine-induced, although Urine osmolality is inappropriately high for the
other factors such as hypoxia and nausea may also serum osmolality and urine sodium is >40 mmol/l.
playa part (Dreyfuss et al. 1988; Hill et al. 1990). Pro- To fulfill the diagnostic criteria for SIADH the patient
duction in the involved tissue was shown by Vorherr should be euvolemic and there should be no adrenal,
et al. (1970). They isolated bio-assayable AVP from pituitary or renal disease. In states of renal salt wast-
the affected lung but not from the normal lung in ing there is volume depletion, urine sodium usually
a patient with fatal pulmonary tuberculosis. (ii) A exceeds 100 mmolll and the plasma urea level is high.
more common cause of the inappropriately elevated Elevated serum potassium should raise the possibil-
levels of AVP is altered osmoregulation because of ity of mineralocorticoid deficiency.
a reset osmostat. Normally as the serum osmolal- The treatment of hyponatremia will depend on
ity drops below 280-282 mosmollK AVP secretion its degree, symptoms and its cause. In the category
is turned off, thus hypo-osmolality should result associated with impaired water excretion the degree
in undetectable levels of AVP. In this condition the of hyponatremia is usually mild, self limiting and
osmostat is set to switch of at an even lower level of asymptomatic. The hyponatremia corrects with
osmolality, as a consequence AVP release continues treatment of the infection. Replacing volume in
in spite of hypo-osmolality. This secretion can be depleted patients or restricting fluid in euvolemic
turned off by further lowering the serum osmolality patients usually suffices. The aim being slow cor-
(such as by a water load), suggesting that osmo-regu- rection «10 mmol124 hours). In more severe cases
lation is intact but set at a lower point. This explains of symptomatic SIADH with CNS dysfunction the
the modest degree of hyponatremia in these patients, use of hypertonic saline or urea may be warranted
which is usually stable. The cause for the lower reset (Decaux et al. 1980). In adults 3% saline 500 ml is
of the osmostat is not well understood and is prob- given over 6-8 h. The aim is to raise the serum sodium
ably mediated through cytokines, vagal stimulation by 6-10 mmolll as this is enough to reduce cerebral
or other yet undefined mechanisms (Dreyfuss et al. edema and alleviate symptoms. This should be fol-
1988,1991; Hill et al.1990). lowed by fluid restriction. In this situation one should
The second mechanism of hyponatremia is renal avoid increasing serum sodium by >12-15 mmol124
salt wasting. This occurs in the following settings: h or reaching an absolute value> 125 mmolll to avoid
a) Mineralocorticoid deficiency secondary to tuber- osmotic demyelination.
culous destruction of the adrenals. This can cause In renal salt wasting replacement with fluids and
significant hyponatremia and should always be con- salt is necessary. Addition of mineralocorticoids
sidered as a possibility in these patients (O'Rahilly is necessary in tuberculous Addison's disease (see
1985). b) Cerebral salt wasting (CSW) which can be above section on adrenal tuberculosis). In tubercu-
seen in TB meningitis. CSW occurs due to the pro- lous meningitis the situation is more complex. The
duction of natriuretic substances and/or stimuli as a hyponatremia may be related to SIADH or CSw. Since
result of cerebral injury and can cause severe hypo- the treatment of these two conditions is diametrically
natremia. The natriuretic substances responsible are opposite extreme vigilance is required. In CSW there
ANP (atrial natriuretic peptide) and other similar is volume depletion related to renal salt wasting and
peptides of cerebral origin (Narotam et al. 1994; Ti et fluid restriction is deleterious. CSW should always
al. 1998). The rennin-angiotensin system is usually be suspected if there is evidence of volume deple-
suppressed. c) TB renal involvement can cause an tion, high urine output with a raised osmolality
interstitial nephritis with impaired tubular function and urinary sodium in excess of 100 mmol/l. This
and consequent renal salt wasting. is treated by aggressive fluid and salt replacement. A
Most patients are asymptomatic as the hyponatre- central line is recommended to gauge fluid replace-
mia is usually mild and slow in onset. Serum sodium ment. Hypertonic saline (1.8%) may be necessary to
levels <125 mmol/l may be associated with lethargy, replace the large amount of sodium loss. In severe
anorexia and vomiting. Severe hyponatremia with cases fludrocortisone, a mineralocorticoid may be
levels <120 mmolll especially if develops rapidly can required to reduce natriuresis (Sakarcan and Boc-
lead to an impaired state of consciousness. However chini 1998). Fludrocortisone is started at a dose of 50
CNS disturbance is more likely to be related to intra- Jlg/day and titrated up to 200Jlg/day based on 24-h
cranial infection. urine sodium and serum sodium values.
The laboratory finding will depend on the patho- Hypernatremia can also occur in tuberculosis
genesis. In SIADH there is hyponatremia, hypo- though it is uncommon. It is due to excessive free water
osmolality and normal to low urea in the plasma. loss from the kidneys secondary to diabetes insipidus
764
(DI). DI may be central or nephrogenic. Central DI
occurs in tuberculous meningitis due to hypothalamic
injury with consequent AVP deficiency. Nephrogenic
DI is secondary to tubulo-interstitial renal damage
and is characterized by high AVP levels. The treatment
for these conditions is similar to other causes of Dr.
The initial therapy is directed towards the volume defi-
cit. 0.9% saline is preferred initially in patients with
significant volume depletion. Once adequate volume
is achieved or in those who are not significantly
volume depleted 0.45% saline can be used to replenish
the remaining water deficit and slowly lower serum
sodium towards normal. In central DI the hormone
deficiency is replaced by DDAVP. In the nephrogenic
form a thiazide diuretic or non-steroidal anti-inflam-
matory drug such as indomethacin may be effective.
42.3.3
Effect ofTuberculosis on Circulating Hormones
The most common reason for an abnormality of
circulating hormones in tuberculosis is the effect of
systemic disease. These effects are mediated through
malnutrition, changes in hormone binding proteins,
altered peripheral metabolism of hormones and the
effects of cytokines and other inflammatory media-
tors on the hypothalamic-pituitary axis.
The adaptive response to illness is stereotypic and
is characterized by provision of energy to essential
organs, postponement of anabolism and activation
of the immune response while protecting the host
against the latter's deleterious biological effects. The
hypothalamus regulates endocrine function, auto-
nomic functions, temperature, appetite and weight to
name a few. It is in turn influenced by impulses from
higher centers, the autonomic nervous system, envi-
ronmental factors and altered homeostasis. It plays a
key role in the adaptive process to stress. It is thus no
surprise that almost any significant systemic disease
can alter its functions. These effects are mediated
through inflammatory mediators related to infection,
and injury, or neurotransmitters released by impulses
from the higher centers. Considered teleologically
the aim is to channel resources away from catabolic
and non-essential functions. Systemic illness causes
activation of the hypothalamic-pituitary-adrenal
axis, suppression of the thyroid and gonadal axis and
initially in increased growth hormone and prolactin
secretion (Chorousos and Gold 1992; van den Berghe
et al.1998). This results in minimal to marked altera-
tion in hormonal secretion and target organ effect
depending on the severity of illness.
S. Inam and M. Al-Shahed
Tuberculosis is associated with hypogonadism in
both males and females. Menstrual abnormalities
are not uncommon. Typically levels of testosterone
and estrogen are low and accompanied by inappro-
priate levels of LH, a picture of hypogonadotropic
hypogonadism. In the report by Post et al. (l994)
73% of their male patients with active pulmonary
tuberculosis had hypogonadotropic hypogonadism.
The decrease in circulating hormones is not due to
changes in sex hormone-binding globulin, as it is not
significantly affected by critical illness (Luppa et al.
1991; Woolf et al.1985). The inflammatory cytokines
release beta-endorphin, corticotropin releasing
hormone (CRH) and cortisol, which cause reduced
pulsatility and decreased secretion of gonadotro-
pins (Chrousos and Gold 1992; Elenkov et al. 1999;
Habib et al. 2001). Additionally there may also be
alteration in the biological activity of gonadotropins
(Semple et al. 1987). Similar abnormalities have been
reported in other systemic illnesses and relate to their
severity (Spratt et al. 1993). This type of functional
hypothalamic disorder is reversible and has to be
distinguished from that due to direct tuberculous
damage of the hypothalamus or pituitary. Elevated
levels of cortisol also interfere with the action of LH
on the gonads, this occurs early in illness and results
in reduced secretion of hormones from the gonads
(Chrousos and Gold 1992). Rarely tuberculosis can
cause direct damage to the gonads causing primary
gonadal failure, this is characterized by high levels of
gonadotropins.
Alteration of thyroid function tests in systemic ill-
ness is termed the 'sick euthyroid' syndrome. This is
characterized by normal to low levels of T4, reduced
levels ofT3 and increase in levels of reverse T3 (RT3).
Plasma levels of T4 are influenced by both a reduc-
tion in binding proteins (TBG, transthyretin and
albumin) and an alteration in their ability to bind
the hormone. The latter results from an alteration
in the structure of TBG and the presence of circulat-
ing inhibitors of thyroid hormone binding such as
free-fatty acids and bilirubin (Docter et al. 1993).
These alterations interfere with most commercially
available methods for FT4. In contrast measurement
of FT4 by equilibrium dialysis gives normal or even
elevated levels (Chopra et al. 1998). Levels of T3 are
more likely to be reduced since 80% of T3 is produced
peripherally from T4. Decreased peripheral produc-
tion of this hormone results from the inhibition of
the enzyme 5' -monodeiodonase. Inhibition of this
enzyme is caused by decreased caloric intake, the
effects of systemic illness through increased cortisol
levels, elevated non-esterified free fatty acids, cyto-
Endocrine and Metabolic Manifestations of Tuberculosis
kines (TNF-alpha, Ill, IL6) related to inflammation
and reduced transport of T4 across cell membranes
(Utiger 1995). RT3 on the other hand is degraded
by the same enzyme, which explains its elevation.
In addition to peripheral events, another reason for
reduced thyroid hormones is a form of transient
central hypothyroidism secondary to hypothalamic
effects of systemic illness. There is reduced TSH
secretion and in severe illness there is loss of the
nocturnal rise. Additionally there may be a decrease
in the biological activity of TSH due to abnormal
glycosylation (van den Berghe et al. 1998). The result
is decreased stimulation of the thyroid and reduced
production of thyroid hormones. The TSH levels are
either normal or low and thus inappropriate for the
level of T4. This form of central hypothyroidism is
mediated by increased secretion of cortisol, CRH
and dopamine (Habib et al. 2001; van den Berghe
et al. 1998). The incidence of the sick euthyroid syn-
drome in tuberculosis ranges from 63-92% (Chow
et al. 1995; Hill et al. 1995; Post et al. 1994). These
alterations may serve as an indicator of severity of
disease and predict mortality. The hormone levels
return to normal in survivors within one month of
therapy (Chow et al. 1995). Anti-tuberculous therapy
is associated with an increase in TBG levels, an effect
probably related to the hepatic effects of rifampicin
(Hill et al. 1995).
The pituitary-adrenal axis as discussed earlier is
activated. Basal cortisol and 24-hour urine cortisol
levels may be elevated. Like other causes of stress,
nocturnal levels are high and there may be loss
of diurnal variation. This is often associated with
decreased suppression with dexamethasone. Myco-
bacterium tuberculosis cell-wall components cause
release of cytokines such as Ill, IL6 and TNF alpha,
which stimulate the release of CRH/ACTH resulting
in "the above changes (Fukata et al. 1993; Spangelo
1994). Elevation of CRH and cortisol may be respon-
sible for decreased secretion of gonadotropins, TSH
and growth hormone as well as reduced responsive-
ness of their target organs during illness (Habib et
a1.2001).
Acute illness causes hypersecretion of growth
hormone (GH) but in chronic illness GH levels are
inhibited. This is presumably through increase in
somatostatin, CRH and cortisol levels (Chrousos
and Gold 1992; van den Berghe et al. 1998). The IGFI
(Insulin-like growth factor 1) levels are reduced early
in illness. The latter is produced by peripheral tissue
in response to GH stimulation. Peripheral effects of
cytokines, altered nutrition, excess cortisol and other
effects of systemic infection are responsible for its
765
decreased production. In chronic illness inhibition of
GH levels further contribute to decreased production
ofIGFl. This accounts for reduced growth in children
with active tuberculosis and contributes to the nega-
tive nitrogen balance.
Prolactin is elevated in acute physical and psy-
chological stress (van den Berghe et al. 1998). The
elevations are usually modest and mediated by the
vasoactive intestinal polypeptide, oxytocin and
dopaminergic pathways (Reichlin 1993). There is
likely to be an adaptation in chronic illness with a
reduction in levels back to normal and a decrease
in the pulsatile fraction. There has been no systemic
study of prolactin in human tuberculosis in the
English literature. Elevated prolactin levels are often
seen in pituitary and hypothalamic tuberculosis and
return to normal after adequate therapy (Sharma et
al. 2000).
42.4
Tuberculosis and Diabetes Mellitus
See chapter 7 on tuberculosis in special groups for
more details. In brief the prevalence of tuberculosis
amongst diabetics is 2-5 times higher than in the
non-diabetic population (Kim et al. 1995; Opsald et
al. 1961). The increase in risk is related to abnor-
malities of the immune cell function, impaired
cytokine production, pulmonary microangiopathy
and effects of non-enzymatic protein glycation on
alveolar membrane and the pulmonary basal lamina
(Koziel and Koziel 1995). These abnormalities are
more pronounced in uncontrolled patients making
them more susceptible.
There is evidence in the literature to suggest that
patients with TB have a higher prevalence of diabetes
(Mugusi et al.1990). This may be partly related to the
fact that in diabetic screening programs almost 50%
of patients are found to be undiagnosed. There also
appears to be an increased occurrence of impaired
glucose tolerance in patients with TB based on the
systematic use of the oral glucose tolerance test
(Jawad et al. 1995; Mugusi et al. 1990). This has been
shown to improve with therapy suggesting that it
may represent infection-related stress though other
factors such as poor carbohydrate intake and malnu-
trition may playa role.
The clinical presentation is similar to non-diabetics.
Involvement of the lower lobes, cavitary disease, mul-
tilobar involvement and pleural effusions are thought
to be more common (Morris et al.1992; Perez-Guzman
766
et al.2000). This has not been borne out by more recent
controlled studies (Bacakoglu et al. 2001).
The treatment is similar to non-diabetics and the
relapse rates are comparable. There is no major dis-
similarity amongst different types of diabetics and
the degree of control as regards treatment outcome.
There is a report of an increased likelihood of drug-
resistant strains during relapse in these patients
(Kameda et al. 1990).
42.5
Endocrine and Metabolic Effects
of Anti-Tuberculous Drugs
Rifampicin is a potent inducer of cytochrome P450
hepatic microsomal enzymes. This has a significant
effect on steroid hormone metabolism. Cortisol metab-
olism is enhanced resulting in increased requirements,
which has major clinical implications. In patients with
subclinical adrenal insufficiency, the introduction
of rifampicin can result in an acute adrenal crisis as
the damaged adrenal is unable to meet the increased
demand (Elansary and Earis 1983; Wilkins et al. 1989).
This is seen within 2 weeks after starting therapy
(Wilkins et al. 1989). In patients with established
adrenal insufficiency an adrenal crisis can be precipi-
tated unless the dose of glucocorticoids is increased
(Edwards et al. 1974). (See section 42.2.1.1 above on
the treatment of adrenal tuberculosis for details.) The
effect on mineralocorticoids is less pronounced and
there is no need to alter their dose (Schulte et al. 1987).
It may also impair cortisol response to synacthen (Ellis
and Tayoub 1986). Rifampicin increases catabolism of
vitamin D to inactive metabolites, decreasing levels of
both 25(OH)D3 and 1,25(OHhD3 (Brodie et al. 1982;
Davies et al. 1987). This can cause a significant rise in
PTH levels (Brodie et al. 1982) and in the presence of
low vitamin D reserves lead to hypocalcaemia and even
osteomalacia (Chan 1996; Shah et al. 1981). Rifampicin
does increase extrathyroidal metabolism of thyroid
hormones but this seems to be of no clinical signifi-
cance (Christensen et al. 1989). It has also been reported
to increase insulin requirements in type 1 diabetes mel-
litus. This was shown to be related to augmented intes-
tinal absorption of glucose (Atkin et al. 1993).
The major effect ofINH is on vitamin D metabolism.
It inhibits 1 alpha hydroxylase activity and reduces
synthesis of calcitriol (Brodie et al. 1982; Davies et al.
1987). In combination with rifampicin there are sub-
stantial decreases in the levels of both 25(OH)D3 and
1,25(OHhD3. PTH levels may rise by 57% (Brodie et al.
S. Inam and M. AI-Shahed
1982). Hypocalcaemia and osteomalacia may develop
in the presence of marginal vitamin D stores (Chan
1996; Shah et al.198l).As vitamin D deficiency reduces
the effect of PTH, anti-TB therapy can mask primary
hyperparathyroidism (Kovacs et al.1994).
Para-aminosalicylic acid (PAS) and ethionamide
have effects similar to propylthiouracil on the thy-
roid. They inhibit synthesis of thyroid hormones and
can lead to goiter and hypothyroidism (Munkner
1969). PAS has also been implicated as causing hypo-
glycemia (Dandona et al. 1980). Both these drugs are
seldom used in modern day practice.
Acknowledgements. We would like to express our
gratitude to Emeritus Professor David Price Evans
for reviewing the manuscript and his valuable com-
ments. We are also very grateful to Drs Fahd AI-
Sabaan and Fahd AI-Rabiah for allowing us to use
their cases as illustrative case reports.
References
Abbasi AA, Chemplavil JK, Farah S et al (1979) Hypercalcemia
in active tuberculosis. Ann Intern Med 90:324-328
Addison T (1855) On the constitutional and local effects of
disease of the supra-renal capsules. Highley, London
Alvarez S, McCabe WR (1984) Extrapulmonary tuberculosis
revisited: a review of experience at Boston City and other
hospitals. Medicine 63:25-55
Anonymous (1981) Subcommittee of the Research Committee
of the British Thoracic Association: a controlled trial of six
months chemotherapy in pulmonary tuberculosis. Br J Dis
Chest 75:141-153
Ashkan K, Papadopoulos MC, Casey AT et al (1977) Sellar
tuberculoma: report of two cases. Acta Neurochir (Wien)
139:523-525
Atkin SL, Masson EA, Bodmer CW et al (1993) Increased
insulin requirement in a patient with type 1 diabetes on
rifampicin. Diabet Med 10:392
Bacakoglu F, Basoglu OK, Cok G et al (2001) Pulmonary tuber-
culosis in patients with diabetes mellitus. Respiration 68:
595-600
Barnes DJ, Naraqi S, Temu P et al (1989) Adrenal function in
patients with active tuberculosis. Thorax 44:422-424
Barnes P, Weatherstone R (1979) Tuberculosis of the thyroid:
two case reports. Br J Dis Chest 73:187-191
Barnes PF, Modlin RL, Bikle DD et al (1989) Transpleural gra-
dient of 1,25-dihydroxyvitamin D in tuberculous pleuritis.
J Clin Invest 83:1527-1532
Basaria S, Ayala AR, Guerin C et al (2000) A rare pituitary
lesion. J Endocrinol Invest 23:189-192
Berger SA, Zonszein J, Villamena P et al (1983) Infectious dis-
eases of the thyroid gland. Rev Infect Dis 5: I08-122
Bhatia E,Jain SK, Gupta RK et al (1998) Tuberculous Addison's
disease: lack of normalization of adrenocortical function
after anti-tuberculous chemotherapy. Clin Endocrinol 48:
355-359
Endocrine and Metabolic Manifestations of Tuberculosis
Braman SS, Goldman AL, Schwarz MI (1973) Steroid-respon-
sive hypercalcemia in disseminated bone tuberculosis.
Arch Intern Med 132:269-271
Brodie MJ, Boobis AR, Hillyard CJ et aI (1982) Effect of
rifampicin and isoniazid on vitamin D metabolism. Clin
Pharmacol Ther 32:525-530
Buxton N, Robertson I (2001) Lymphocytic and granuloma-
tous hypophysitis: a single center experience. Br J Neuro-
surg 15:242-246
Cadranel J, Garabedian M, Milleron B et aI (1990) 1,25(OH)2
D3 production by T lymphocytes and alveolar macro-
phages recovered by lavage from normocalcemic patients
with tuberculosis. J Clin Invest 85:1588-1593
Cadranel J, Garabedian M, Milleron B et aI (1994) Vitamin D
metabolism by alveolar immune cells in tuberculosis. Cor-
relation with clinical metabolism and clinical manifesta-
tions. Eur Res J 7:1103-1110
Carroll N (1987) Tuberculous hypercalcemia with renal failure.
Br J Dis Chest 81:296-299
Chan CHS, Arnold M, Mak TWL et al (1993) Adrenocortical
function and involvement in high risk cases of pulmonary
tuberculosis. Tuberc Lung Dis 74:395-398
Chan TY (1996) Osteomalacia during rifampicin and isoniazid
therapy is rare in Hong Kong. Int J Clin Pharmacol Ther
34:533-534
Chan TY (1997) Differences in vitamin D status and calcium
intake: possible explanations for the regional variations
in the prevalence of hypercalcemia in active tuberculosis.
Calcif Tissue Int 60:91-93
Chan TY, Chan CHS, Shek CC (1994) The prevalence of
hypercalcemia in pulmonary and miliary tuberculosis: a
longitudinal study. Singapore Med J 35:613-615
Cheung CC, Ezzat S, Smyth HS et al (2001) The spectrum and
significance of primary hypophysitis. J Clin Endocrinol
Metab 86:1048-1053
Chopra II, Trong UT, Le A (1998) Simultaneous measurement of
free thyroxine and free 3,5,3' triiodothyronine in undiluted
serum by direct equilibrium dialysis/radioimmunoassay:
evidence that free triiodothyronine and free thyroxine
are normal in many patients with low triiodothyronine
syndrome. Thyroid 8:249-257
Chow CC, Mak TWL, Chan CHS et al (1995) Euthyroid sick
syndrome in pulmonary tuberculosis before and after
treatment. Ann Clin Biochem 32:385-391
Christensen HR, Simonsen K, Hegedus L et aI (1989) Influence
of rifampicin on thyroid gland volume, thyroid hormones
and antipyrine metabolism. Acta EndocrinoI121:406-410
Chrousos GP,Gold PW (1992) The concepts of stress and stress
system disorders: overview of physical and behavioural
homeostasis. JAmMed Assoc 267:1244-1252
Chung DK, Hubbard WW (1969) Hyponatremia in untreated
pulmonary tuberculosis. Am Rev Respir Dis 99:595-602
Coleman CC, Meredith JM (1940) Diffuse tuberculosis of the
pituitary gland stimulating tumour with postoperative
recovery. Arch Neurol Psychiatry 44:1076-1085
Cotton MF, Donald PR, Schoeman JF et aI (1991) Plasma
arginine vasopressin and the syndrome of inappropriate
antidiuretic hormone secretion in tuberculous meningitis.
Pediatr Infect Dis J 10:831-842
Cotton MF, Donald PR, Schoeman JF et al (1993) Raised intra-
cranial pressure, the syndrome of inappropriate antidi-
uretic hormone and arginine vasopressin in tuberculous
meningitis. Childs Nervous Syst 9:10-15
767
Crowle AJ, Ross EJ, May MH (1987) Inhibition by 1,25 (OH)2
vitamin D3 of the multiplication of virulent tubercle bacilli
in cultured human macrophage. Infect Immun 55:2945-2950
Crowle AJ, Salfinger M, May MH (1989) 1,25 (OH)2 vitamin
D3 synergises with pyrazinamide to kill tubercle bacilli
in cultured human macrophages. Am Rev Respir Dis 139:
549-52
Dandona P, Greenbury E, Beckett AG (1980) Para-amino-
salicylic acid-induced hypoglycaemia in a patient with
diabetic nephropathy. Postgrad Med J 56:135-136
Das DK, Pant CS, Chachra KL et aI (1992) Fine needle aspira-
tion cytology diagnosis of tuberculous thyroiditis: a report
of eight cases. Acta Cytol 36:517-522
Davies PD, Brown RC, Church HA et al (1987) The effect of
antituberculous chemotherapy on vitamin D and calcium
metabolism. Tubercle 38:261-266
DeAngelis LM (1981) Intracranial tuberculoma: case report
and review of the literature. Neurology 31:1133-1136
Decaux G, Unger J, Mockel J (1980) Urea therapy for inappro-
priate antidiuretic hormone secretion from tuberculous
meningitis. J Am Med Assoc 244:589-590
Desai M, Colaco MP, Choksi CS et al (1993) Isosexual precoc-
ity: the clinical and etiologic profile. Indian Pediatr 30:
607-623
Docter R, Krenning EP, de Jong M et aI (1993) The sick
euthyroid syndrome: change in thyroid hormone serum
parameters and hormone metabolism. Clin Endocrinol
39:499-518
Dreyfuss D, Leviel F, Paillard M et aI (1988) Acute infectious
pneumonia is accompanied by latent vasopressin-depen-
dent impairment of renal water excretion. Am Rev Respir
Dis 138:583-589
Dreyfuss D, Leviel F, Paillard M et aI (1991) Resetting of the
vasopressin osmostat during infectious pneumonia. Am J
Med 90:407-408
Dusso AS, Kamimura S, Gollieni M et aI (1997) Gamma
interferon induced resistance to 1,25 (OH)2 D3 in human
monocytes and macrophages: a mechanism for the hyper-
calcemia of various granulomatoses. J Clin Endocrinol
Metab 82:2222-2232
Edwards OM, Courtenay-Evans RJ, Galley JM et al (1974)
Changes in cortisol metabolism following rifampicin
therapy. Lancet 2:549-551
Elansary RH, Earis JE (1983) Rifampicin and adrenal crisis. Br
Med J 286:1861-1862
Elenkov IJ, Webster EL, Torpy DJ et aI (1999) Stress, corticotro-
pin-releasing hormone, glucocorticoids, and the immune/
inflammatory response: acute and chronic effects. Ann NY
Acad Sci 876:1-11
Ellis ME, Tayoub F (1986) Adrenal function in tuberculosis. Br
J Dis Chest 80:7-12
Epstein S, Stein PH, Bell NH et aI (1984) Evidence for abnormal
regulation of circulating 1 alpha, 25 dihydroxy vitamin D in
patients with pulmonary tuberculosis and normal serum
calcium metabolism. Calcif Tissue Int 36:541-544
Esposito V, Fraioli B, Ferrante L et aI (1987) Intrasellar tuber-
culoma: case report. Neurosurgery 21;721-723
Fukata J, Imura H, Nakao K (1993) Cytokines as mediators in
the regulation of the hypothalamic-pituitary-adrenocorti-
cal function. J Endocrinol Invest 16:141-155
Fuss M, Karmali R, Pepersack T et al (1988) Are tuberculous
patients at a great risk from hypercalcemia? Quart J Med
69:869-878
768
Gazioglu N, Ak H, Oz B et al (1999) Silent pituitary tubercu-
loma associated with pituitary adenoma. Acta Neurochir
(Wien) 141:785-786
Gkonos PJ, London R, Hendler ED (1984) Hypercalcemia and
elevated 1,25 (OH)2 D3levels in a patient with end stage renal
disease and active tuberculosis. N Eng! J Med 311:1683-1685
Gulmez I, Kelestimur F, Durak AC et al (1996) Changes in the
size of adrenal glands in acute pulmonary tuberculosis
with therapy. Endocr J 43:573-576
Guttman PH (1930) Addison's disease: a statistical analysis
of 566 cases and a study of pathology. Arch Pathol 10:
742-785,896-935
Habib KE, Gold PW, Chrousos GP (2001) Neuroendocrinology
of stress. Endocrinol Metab Clin North Am 30:695-728
Haslam RHA, Winternitz WW, Howieson J (1969) Selective
hypopituitarism following tuberculous meningitis. Am J
Dis Child 118:903-908
Hauser H, Gurret JP (1986) Miliary tuberculosis associated
with adrenal enlargement: CT appearance. J Comput Assist
Tomogr 10:254-256
Hill AR, Uribarri J, Mann J et al (1990) Altered water metabo-
lism in tuberculosis: Role of vasopressin. Am J Med 88:
357-364
Hill AR, Schmidt MF, Schussler GC (1995) Rapid changes in
thyroid function tests upon treatment of tuberculosis.
Tuberc Lung Dis 76:223-229
Horsburgh CR, Feldman S, RidlOn R (2000) Practice guide-
lines for the treatment of tuberculosis: guidelines from the
infectious diseases society of America. Clin Infect Dis 31:
633-639
Isaacs RD, Nicholson GI, Holdaway 1M (1987) Miliary tubercu-
losis with hypercalcemia and raised vitamin D concentra-
tions. Thorax 42:555-556
Jawad F, Shera AS, Memon R et al (1995) Glucose intolerance in
pulmonary tuberculosis. J Pakistan Med Assoc 45:237-238
Jayakar DV, Condemi G, DeSoto-LaPaix F et al (1998) Adrenal
tuberculosis - case report. Eur J Surg 164:975-978
Kameda K, Kawabata S, Masuda N (1990) Follow-up study
of short-course chemotherapy of pulmonary tuberculosis
complicated with diabetes mellitus. Kekkaku 65:791-803
Kang BC, Lee SW, Shim SS et al (2000) US and CT findings
of tuberculosis of the thyroid: three case reports. J Clin
Imaging 24:283-286
Kang GH, Chi JG (1990) Congenital tuberculosis- report of an
autopsy case. J Korean Med Sci 5:59-64
Kapoor VK, Subramani K, Das SK et al (1985) Tuberculosis of
the thyroid associated with thyrotoxicosis. Postgrad Med
J 61:339-340
Kar DK,Agarwal G, Mehta B et al (2001) Tuberculous granulo-
matous inflammation associated with adenoma of parathy-
roid gland manifesting as primary hyperparathyroidism.
Endocr Pathol 12:355-359
Karandanis D, Shulman JA (1976) Recent survey of infectious
meningitis in adults: review oflaboratory findings in bacte-
rial, tuberculous, and aseptic meningitis. South Med J 69:
449-457
Kelestimur F, UnluY, Ozesmi M et al (1994) A hormonal and
radiological evaluation of adrenal gland in patients with
acute and or chronic pulmonary tuberculosis. Clin Endo-
crinoI41:53-56
Kelestimur F, Guven M, Ozesmi M et al (1996) Does tubercu-
losis really cause hypercalcemia? J Endocrinol Invest 19:
678-681
S. Inam and M. Al-Shahed
Kelestimur F, Goktas Z, Gulmez I et al (2000) Low dose (1
mcg) adrenocorticotropin stimulation test in the evalu-
ation of hypothalamo-pituitary-adrenal axis in patients
with active pulmonary tuberculosis. J Endocrinol Invest
23:235-239
Keven K, Uysal AR, Erdogan G (1998) Adrenal function during
tuberculous infection and effects of antituberculous treat-
ment on endogenous and exogenous steroids. Int J Tuberc
Lung Dis 2:419-424
Khan EM, Haque I, Pandey R et al (1993) Tuberculosis of the
thyroid gland: a clinicopathological profile of four cases
and review of the literature. Austr NZ J Surg 63:807-810
Kim SJ, Hong YP, Lew WJ et al (1995) Incidence of pulmo-
nary tuberculosis among diabetics. Tuberc Lung Dis 76:
529-533
Kirshbaum JD, Levy HA (1941) Tuberculoma of hypophysis
with insufficiency of anterior lobe: a clinical and pathologi-
cal study of two cases. Arch Intern Med 68:1095-1104
Kovacs CS, Jones G, Yendt ER (1994) Primary hyperparathy-
roidism masked by antituberculous therapy-induced vita-
min D deficiency. Clin EndocrinoI41:831-836
Koziel H, Koziel MJ (1995) Pulmonary complications of dia-
betes mellitus: pneumonia. Infect Dis Clin North Am 9:
65-96
Lam KSL, Sham MMK, Tam SCF et al (1993) Hypopituitarism
after tuberculous meningitis in childhood. Ann Intern Med
118:701-706
Laureti S, Aubourg P, CalcinaroF et al (1998) Etiological diag-
nosis of primary adrenal insufficiency using an original
flowchart of immune and biochemical markers. J Clin
Endocrinol Metab 83:3163-3168
Levitt T (1952) The status oflymphadenoid goitre, Hashimoto's
and Riedel's diseases. Ann R CoIl Surgeons 10:369-404
Lim SM, Tsai SL, Chan CS (1994) Hypercalcemia in tubercu-
lous peritonitis without active pulmonary tuberculosis. Am
J GastroenteroI89:2249-2250
Luppa P, Munker R, Nigel D et al (1991) Serum androgens in
intensive-care patients: correlations with clinical findings.
Clin EndocrinoI34:305-31O
Martinez ME, Gonzalez J, Sanchez-Ceabezudo MJ et al (1993)
Evidence of absorptive hypercalciuria in tuberculous
patients. Calcif Tissue Int 53:384-387
Martinez ME, Gonzalez J, Sanchez Cabezulo MJ et al (1996)
Remission of hypercalciuria in patients with tuberculosis
after treatment. Calcif Tissue Int 59:17- 20
Mauruce-Williams RS (1972) Tuberculomas of the brain in
Britain. J Postgrad Med 48:679-681
Mondal A, Patra MS (1995) Efficacy of fine needle aspiration
cytology in the diagnosis of tuberculosis of the thyroid
gland: a study of 18 cases. J Laryngol Otoll09:36-38
Moon WK, Han MH, Chang KH et al (1997) CT and MR
imaging of head and neck tuberculosis. Radiographics
17:391-402
Morris CDW, Bird AR, Nell H (1989) The haematological and
biochemical changes in severe pulmonary tuberculosis.
Quart J Med 73:1151-1159
Morris JT, Seaworth BJ, McAllister CK (1992) Pulmonary
tuberculosis in diabetics. Chest 102:539-541
Motiwala HG, Sanghvi NP, Barjatiya MK et al (1991) Syndrome
of inappropriate antidiuretic hormone following tubercu-
lous epididymo-orchitis in renal transplant recipient: case
report. J Urol 146: 1366-1367
Mugusi F, Swai AB, Alberti KG et al (1990) Increased preva-
Endocrine and Metabolic Manifestations of Tuberculosis
lence of diabetes mellitus in patients with pulmonary
tuberculosis in Tanzania. Tubercle 71:271-276
Munkner T (1969) Studies on goitre due to para-aminosali-
cylic acid. Scand J Respir Dis 50:212-226
Narotam PK, Kemp M, Buck R et al (1994) Hyponatremic
natriuretic syndrome in tuberculous meningitis: the prob-
able role of atrial natriuretic peptide. Neurosurgery 34:
982-988
Nieuwland Y, Tan KY, Elte JW (1992) Miliary tuberculosis pre-
senting with thyrotoxicosis. Postgrad Med J 68:677-679
Nomura K, Demura H, Saruta T (1994) Addison's disease in
Japan: characteristics and changes revealed in a nationwide
survey. Intern Med 33:602-606
O'Rahilly S (1985) Secretion of antidiuretic hormone in
hyponatremia: not always "inappropriate". Br Med J 290:
1803-1804
Oelkers W (1996) Adrenal insufficiency. N Engl J Med 335:
1206-1212
Opsald R, Riddervald HO, Aas TW (1961) Pulmonary tuber-
culosis in mitral stenosis and diabetes. Acta Tuberc Scand
4:290
Patankar T, Patkar D, Bunting T et al (2000) Imaging in pitu-
itary tuberculosis. J Clin Imaging 24:89-92
Patel SR, Selby C, Jeffcoate WJ (1991) The short synacthen test
in acute hospital admissions. Clin Endocrinol 35:259-261
Peces R, Alvarez J (1987) Hypercalcemia and elevated
1,25(OH)zD3 levels in a dialysis patient with disseminated
tuberculosis. Nephron 46:377-379
Penrice J, Nussey SS (1992) Recovery of adrenal function
following treatment of tuberculous Addisons's disease.
Postgrad Med J 68:204-205
Pereira J, Vaz R, Carvalho D et al (1995) Thickening of the
pituitary stalk: a finding suggestive of intrasellar tubercu-
loma? Case report. Neurosurgery 36:1015-1015; discussion
1015-1016
Perez-Guzman C, Torres-Cruz A, Villarreal-Velarde H et al
(2000) Progressive age-related changes in pulmonary
tuberculosis images and the effect of diabetes. Am J Respir
Crit Care Med 162:1738-1740
Petrossians P, Delvenne P, Flandroy P et al (1998) An unusual
pituitary pathology. J Clin Endocrinol Metab 83:3454-
3458
Post FA, Soule SG, Wilcox PA et al (1994) The spectrum of
endocrine dysfunction in active pulmonary tuberculosis.
Clin Endocrinol40:367-371
Pouchot J, Dreyfuss D, Gardin JP et al (1993) Ectopic produc-
tion of 1,25-dihydroxyvitamin D3 in tuberculosis. Nephrol
Dial Transplant 8:560-562
Prasad GA, Sharma SK, Mohan A et al (2000) Adrenocortical
reserve and morphology in tuberculosis. Indian J Chest Dis
Allied Sci 42:83-93
Ranjan A, Chandy M (1994) Intrasellar tuberculoma. Br J
Neurosurg 8:179-185
Rankin FW, Graham AS (1932) Tuberculosis of the thyroid
gland. Ann Surg 96:625-648
Rook GA (1988) The role of vitamin D in tuberculosis. Am Rev
Respir Dis 138:768-770
Reichlin S (1993) Neuroendocrine-immune interactions. N
Engl J Med 329:1246-1253
Rose BD, Post TW (2001) Clinical physiology of acid-base and
electrolyte disorders, 5th edn. McGraw-Hill, New York, pp
729-733
Rothwell PM, Udwadia ZF, Lawler PG (1991) Cortisol response
769
to corticotrophin and survival in septic shock. Lancet 337:
582-583
Roussos A, Lagogianni I, Gonis A et al (2001) Hypercalcemia
in Greek patients with tuberculosis before the initiation of
anti-tuberculosis treatment. Respir Med 95: 187-190
Sachs MK, Dickinson G, Amazon K (1988) Tuberculosis
adenitis of the thyroid mimicking subacute thyroiditis.
Am J Med 85:573-575
Saggese G, Bertelloni S, Baroncelli GI et al (1993) Ketoconazole
decreases the serum ionised calcium and 1,25-dihydroxyvi-
tamin D levels in tuberculous-associated hypercalcemia.
Am J Dis Child 147:270-273
Sakarcan A, Bocchini J (1998) The role of fiudrocortisone
in a child with cerebral salt wasting. Pediatr Nephrol 12:
769-771
Sanford JP, Favour CB (1956) The interrelationships between
Addison's disease and active tuberculosis: a review of 125
cases of Addison's disease. Ann Intern Med 45:56-72
Sarma GR, Immanuel C, Ramachandaran G et al (1990)
Adrenal function in patients with pulmonary tuberculosis.
Tubercle 71:277-282
Schulte HM, Monig H, Benker G et al (1987) Pharmacokinetics
of aldosterone in patients with Addison's disease: effect of
rifampicin treatment on glucocorticoid and mineralocor-
ticoid metabolism. Clin Endocrinol 27:655-662
Semple CG, Gray CE, Beastall GH (1987) Male hypogonad-
ism-a non-specific consequence of illness. Quart J Med
64:601-607
Shah SC, Sharma RK, Hemangini et al (1981) Rifampicin
induced osteomalacia. Tubercle 62:207-209
Sharma MC, Arora R, Mahapatra AK et al (2000) Intrasellar
tuberculoma - an enigmatic pituitary infection: a series of
18 cases. Clin Neurol Neurosurg 102:72-77
Sharma MC, Vaish S, Arora R et al (2001) Composite pituitary
adenoma and intrasellar tuberculoma: report of a rare case.
Pathol Oncol Res 7:74-76
Sharma OP (2000) Hypercalcemia in granulomatous disor-
ders: a clinical review. Curr Opin Pulmon Med 6:442-447
Sharma SC (1981) Serum calcium in pulmonary tuberculosis.
Postgrad Med J 57:694-696
Shek CC, Natkunam A, Tsang Vet al (1990) Incidence, causes
and mechanism of hypercalcemia in a hospital population
in Hong Kong. Quart J Med 77:1277-1285
Sherman BM, Gorden P, di Chiro G (1971) Postmeningitic
selective hypopituitarism with suprasellar calcification.
Arch Intern Med 128:600-604
Shimizu C, Kubo M, Kijima H et al (1998) Giant cell granulo-
matous hypophysitis with remarkable uptake on Gallium-
67 scintigraphy. Clin EndocrinoI49:131-134
Siddiqi S, Elahi HA, Hussain M et al (1997) Evaluation of adre-
nal function in long standing pulmonary tuberculosis: a
study of 100 cases. J Pakistan Med Assoc 47:132-134
Singh BS; Patwari AK, Deb M (1994) Serum sodium and
osmolal changes in tuberculous meningitis. Indian Pediatr
31:1345-1350
Sinha S, Singh AK, Tatke M (2000) Hypophyseal tuberculoma:
Direct radiosurgery is contraindicated for a lesion with a
thickened pituitary stalk: case report. Neurosurgery 46:
735-739
Slavin RE, Walsh TJ,PollackAD (1980) Late generalized tuber-
culosis: a clinical pathologic analysis and comparison of
100 cases in the preantibiotic and antibiotic eras. Medicine
59:352-366
770
Spangelo BL (1994) Cytokine regulation of the neuroendo-
crine system. Curr Opin Endocrinol Diabetes 1:60-64
Spratt OI, Cox P, Orav J et al (1993) Reproductive axis sup-
pression in acute illness is related to disease severity. J Clin
Endocrinol Metab 76:1548-1554
Surer I, Ozturk H, Cetinkursun S (2000) Unusual presentation
of tuberculosis reactivation in childhood: an anterior neck
mass. J Pediatr Surg 35:1263-1265
Taha SA, Dost SM, Sedrani SH (1984) 25-hydroxyvitamin D
and total calcium: extraordinary low plasma concentra-
tions in Saudi mothers and their neonates. Pediatr Res 18:
739-741
Takami H, Kozakai M (1994) Tuberculous thyroiditis: report
of a case with review of the literature. Endocr J (Jpn) 41:
743-747
Tan KK (1993) Tuberculosis of the thyroid gland - a review.
Ann Acad Med Singapore 22:580-582
Ti LK, Kang SC, Cheong KF (1998) Acute hyponatremia
secondary to cerebral salt wasting syndrome in patients
with tuberculous meningitis. Anaesth Intensive Care 26:
420-423
Usalan C, Nar A, Erdem Y et al (1998) Severe hyponatremia
probably resulting from inappropriate secretion of antidi-
uretic hormone. Nephron 80:237-238
Utiger RD (1995) Altered thyroid function in nonthyroidal
illness and surgery. To treat or not to treat? N Engl J Med
333:1562
Van den Berghe GV, de Zegher F, Bouillon R (1998) Acute and
prolonged critical illness as different neuroendocrine para-
digms. J Clin Endocrinol Metab 83:1827-1834
Villabona CM, Sahun M, Ricart W et al (1993) Tuberculous
S. Inam and M. AI-Shahed
Addison's disease. Utility of CT diagnosis and follow-up.
Eur J RadioI17:210-213
Vita J, Silverberg SJ, Goland RS et al (1985) Clinical clues to the
cause of Addison's disease. Am J Med 78:461-466
Vorherr H, Mass SG, Fallet R et al (1970) Antidiuretic principle
in tuberculous lung tissue ofa patient with pulmonary tuber-
culosis and hyponatremia. Ann Intern Med 72:383-387
Wang YX, Chen CR, He GX et al (1998) CT findings of adrenal
glands in patients with tuberculous Addison's disease. J
Beige De RadioI81:226-228
Wilkins EGL, Hnizdo E, Cope A (1989) Addisonian crisis
induced by treatment with rifampicin. Tubercle 70:69-73
Wilms GE, Baert AL, Kint EJ et al (1983) Computed tomo-
graphic findings in bilateral adrenal tuberculosis. Radiol-
ogy 146:729-730
Woolf PD, Hamill RW, McDonald JV et al (1985) Transient
hypogonadotropic hypogonadism caused by critical ill-
ness. J Clin Endocrinol Metab 60:444-450
Yang WC, Chang IT, Tsai BL (2000) Hypercalcemia in renal
transplant patients with tuberculosis. Transplant Proc 32:
1882-1883
Yee CAN, Gopinath N, Ho CS et al (1986) Fine needle aspira-
tion biopsy of adrenal tuberculosis. J Can Assoc Radiol37:
287-289
York EL, Enarson DA, Nobert EJ et al (1992) Adrenocortical
function in patients investigated for active tuberculosis.
Chest 101:1338-1341
Zhang Y, Doerfler M, Lee TC et al (1993) Mechanisms of
stimulation of interleukin-1 beta and tumour necrosis
factor-alpha by mycobacterium tuberculosis components.
J Clin Invest 91:2076-2083
43 Tuberculous Vasculitis and Mycotic Aneurysms
M. MONIR MADKOUR

CONTENTS come, complications (including relapses and fatali-

ties) still occur (Atnip 1989; Chan 1989; Chiba et al.
43.1 Introduction 771 1996; Hollier 1993; Ikezawa et al. 1996; Strand et al.
43.2 Historical Aspects 771
43.3 Epidemiology 771
2001). The recent resurgence of tuberculosis that fol-
43.4 Pathogenesis 772 lowed the HIV epidemic was associated with increas-
43.5 Clinical Features 773 ing reports of patients with tuberculous aneurysms
43.5.1 Our Own Case Report 773 and co-existing HIV (Bojar et al. 1998; Boggian et al.
43.5.2 Surgical Findings (June 20, 1982) 774 1994; Gouny et al. 1992; Seelig et al. 1999). The use of
43.5.3 Successful Outcome of Surgery
and Follow-up for 20 Years 776 BCG in recent years for adjuvant immunotherapy of
43.6 Investigation 776 bladder carcinoma has been reported to cause tuber-
43.7 Treatment 776 culous aneurysms (Hakim et al. 1993; Seelig et al.
References 777 1999; Wolf et al.1995; Rozenblit et al. 1996).

43.1 43.2
Introduction Historical Aspects

Mycotic aneurysms (or more properly 'infected aneu-
rysms') are rare. They were first described by Sir Wil-
liam Osler, an Oxford physician born in 1849. The word
'mycotic' that he used came from their resemblance to
fungal growth. Osler described a patient with multiple
beadlike aneurysms caused by a supportive infection
of the vessel wall.
Despite its rarity, the greatest incidence of mycotic
aneurysms is found in patients with infective endo-
carditis due to Staphylococcus aureus and Streptococ-
cus viridans. Other organisms are also capable of
inducing mycotic aneurysms including M. tuberculo-
sis and Brucella melitensis (Madkour 2001).
Tuberculous mycotic aneurysms are rare but carry
a high mortality rate due to perforation, or rupture,
with massive hemorrhage and shock (Golzarian et al.
1999; de Kruijf et al. 2000; Ito et al. 1992; Yo et al. 1993;
Zhang et al. 1994). The diagnosis may only be found
when an autopsy is carried out. Despite the recent
advances in diagnostic techniques, potent anti-tuber-
culous agents and successful surgical-treatment out-
In the pre-antibiotic era, with limited imaging capabili-
ties, the diagnosis of tuberculous aneurysms was made
as an autopsy finding. Tuberculous aneurysm of the
aorta was first reported by Kamen in 1895 and femo-
ral artery tuberculous aneurysms were first reported
in 1933 by Baumgarten and Cantor (Baumgarten et al.
1933). Surgical attempts to repair tuberculous aneu-
rysms of the abdominal aorta was first reported by
Herndon and colleagues in 1952. It was not successful
and the patient died six days after the operation from
massive gastrointestinal bleeding. The first successful
excision of tuberculous aneurysms of the abdominal
aorta was reported by Rob and Eastcott in 1955. Four
years later successful resection of a tuberculous aneu-
rysm in the thoracic aorta was reported by de Prophe-
tis and colleagues in 1959. The association between
tuberculous aortitis and military tuberculosis was
first described by Weigert in 1882.
43.3
Epidemiology

M. M. MADKOUR, MD, DM, FRCP

Before antibiotics became available the incidence of
Consultant, Department of Medicine, Riyadh Armed Forces tuberculous mycotic aneurysms was estimated to
Hospital, P.O. Box 7897, C-119, Riyadh 11159, Saudi Arabia be 0.3% from more than 22,000 autopsies reported

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
772
by Parkhurst and Deckerin in 1955 (Parkhurst et
al. 1955). Slavin and colleagues (1990) reported a
series of 100 autopsies with confirmed late-general-
ized tuberculosis, only one instance of tuberculous
aortitis was found. Few authors have reviewed the
literature in an attempt to identify special clinical, or
diagnostic, patterns that could help in predicting its
occurrence in patients before its diagnosis at autopsy.
In 1913 and 1933 reviews concluded that tuberculous
mycotic aneurysms of the aorta had been recognized
but were rare. In a recent review from the English lit-
erature, only 39 cases were found over a period of 53
years (1945-1998) (Long et al.1999). The authors also
described two patients of their own and concluded
that the prevalence has remained relatively constant.
The recent resurgence of tuberculosis that followed
the HIV epidemic was expected to be associated with
more reports of tuberculous and mycotic aneurysms.
There are four reports of single cases each from USA,
Germany and France of young males in their early
thirties with HIV and tuberculous mycotic aneu-
rysms of the aorta (Bojar et al. 1998; Boggian et al.
1994; Gouny et al. 1992; Seelig et al. 1999).
In a review of the literature in just over 7 years
from 1990 to mid-2002 I found a total of 59 cases.
Males were more affected than females (55% vs.
45%, respectively. Tuberculous arteritis with mycotic
aneurysms were more frequent in the age group 30-
50 years (range 9-91 years). The aorta was affected
in 38 cases (75%) (20 in the thoracic aorta, 17 in
the abdominal aorta and one thoraco-abdominal).
Twenty-one reported cases involved other arteries
with mycotic aneurysms including; middle cerebral,
left posterior inferior cerebella, internal carotid,
bronchial, Rasmussen's aneurysms, internal mam-
mary, hepatic, renal, superior mesenteric, inferior
mesenteric, inferior gluteal, femoral and popliteal
(Cross et al. 1995; Ito et al. 1992; Hildebrandt et al.
1998; Jebara et al. 1998; Kao et al. 1999; Sanyik et
al. 1999; Patankar et al. 2000; Deshmukh et al. 2000;
Tsurutani et al. 2000; Oran et al. 2001; Beeresha et
al. 2000). Most reported cases in the literature came
from developed, industrialized countries particularly
from the USA and Japan. In developing countries
most of the reported cases came from India.
43.4
Pathogenesis
M. tuberculosis bacilli are the most common cause
of tuberculous aneurysms and followed by M. bovis
M. M. Madkour
(BCG) (Seelig et al.1999; Rozenblit et al.1996; Wolf et
al. 1995; Hakim et al. 1993; Woods et al. 1988).
At least three distinct mechanisms have been
implicated in the seeding of arteries.
1. The most common mode of transmission is direct
extension of infection from a contiguous focus in
the lung parenchyma, empyema, periaortic lymph-
adenitis, pericarditis, spondylitis or paravertebral
abscesses of the thoracic aorta (Ishibataku and
Onizuka 1998; Tsurutani et al. 2000; Long et al. 1999;
Yo et al. 1993; Wetteland et al. 1956). Direct spread
of infection to the abdominal aorta may occur from
a nearby retroperitoneal lymphadenitis, intestinal
tuberculosis, spondylitis, psoas abscess or the pros-
tate (Ghosh 1954; Hagino et al. 1996).
2. Hematogenous spread to the arterial intima is
difficult unless pre-existing damage by athero-
sclerotic plague is present.
3. Septic embolization of the arterial wall via the vasa
vasorum, or the lymphatics, will result in seeding
in the adventitia, or media, in patients with mili-
ary or disseminated tuberculosis (Ishibataku and
Onizuka 1998; Tsurutani et al. 2000). Iatrogenic
or exogenous introduction of live attenuated M.
bovis bacilli (BCG) may occur.
BCG, an attenuated live strain of M. bovis, has
been used to correct defects in cell-mediated immu-
nity that occur in patients with malignant tumors.
Intravesical administration of BCG to treat patients
with urinary bladder carcinoma has been reported.
Experimental evidence of systemic absorption of M.
bovis after intravesical administration of BCG has
been reported (Schellhammer et al. 1975).
In humans, acid-fast bacilli may persist in the
genitourinary tract up to 3 years or more after BCG
therapy for bladder carcinoma (Hurle et al. 1998).
In Italy, Hurle and Colleagues reported a 67-year-
old man with Benign Prostatic Hypertrophy (BPH)
and superficial urinary bladder cancer underwent
intravesical BCG therapy in 1992. Three years later
he developed hematuria. At cystoscopy there was no
tumor recurrence but bladder ulcers with inflamma-
tion were noted. Acid-fast bacilli were detected in the
urine by Ziehl-Neelsen staining. Histopathology of
bladder ulcers biopsy showed granulomatous necro-
tizing inflammation with numerous acid-fast bacilli.
Prostatic tissue biopsy also showed similar histo-
pathological changes with acid-fast bacilli.
At least six reports have appeared describing tuber-
culous mycotic aneurysms of the abdominal aorta
following intravesical BCG therapy (Woods et al.1988;
La Berge et al.1999; Hakim et al.1993; Wolf et al.1995;
Tuberculosis Vasculitis and Mycotic Aneurysms
Rozenblit et al.1996; Seelig et al.1999). The mechanism
of transmission of the bacilli from the bladder to the
aorta is not clear. After seeding of the bacilli into the
arterial wall, an inflammatory granulomatous process
with caseating necrosis and with fibrosis occurs. The
entire thickness of the arterial wall and surrounding
tissue are involved. Perforation may occur with con-
tained encapsulated leakage to form a perivascular
hematoma (Peyton 1965; Hatakeyama et al.1997; Girard
et al. 1997) (False or pseudo-aneurysm) or it may lead
to massive hemorrhage. Single saccular true aneuris-
mal dilatation is the most common but extensive mul-
tiple tuberculous aneurysms of the entire abdominal
aorta and its major branches may be affected (See our
own case illustration below). Infection may spread lon-
gitudinally leading to a true aneurysm and may involve
the entire abdominal aorta leading to an irregular
aortic lumen, stenosis due to fibrosis or even occlusion
(See below). Tuberculous pulmonary artery stenosis is
extremely rare and may occur secondary to adjacent
mediastinal tuberculous lymphadenitis (Beaconsfield
et al. 1998). Small, or medium size, arteries located in
the lung parenchymal cavity due to active tuberculosis,
may undergo local dilatation (Rasmussen aneurysm).
Intracranial carotid artery, or basilar arteries, tuber-
culous arteritis may occur. Rupture of the aneurysm
(pseudo or true) may occur into the lung with massive
hemoptysis, or into the esophagus (or bowel) leading
to massive bleeding.
43.5
Clinical Features
The clinical features of tuberculous aneurysm at
presentation may be dominated by disseminated
or miliary tuberculosis or other extrapulmonary
disease. However, presentation may be related to the
aneurysm with persistent pain mostly related to the
location of the aneurysm. Patients may present with
a palpable pulsating abdominal mass, a para-aortic
mass found on imaging, aneurismal leakage or rup-
ture (Ogawa et al. 1990; Takahashi et al. 1986; Quaini
et al.1985; Oran et al.2001; Muller-Wening et al.1982;
O'Leary et al. 1977).
Felson and colleagues (l997) reported three
patients with tuberculous aneurysms of thoracic
aorta. Two patients had constitutional symptoms and
miliary tuberculosis. The third patient presented with
hemoptysis without any physical or laboratory find-
ings. The diagnosis was initially suspected from chest
radiographs in these three patients. The operation was
773
successful in one patient who survived, while the other
two died post-operatively.
Constitutional symptoms, fever, sweating, loss of
appetite, weight loss and weakness may be present
in over 50% of patients. Long and colleagues (l999)
reviewed the clinical presentation of 39 patients
published by other authors in a literature review. Evi-
dence of pulmonary or extrapulmonary tuberculosis
were found in 63% of reported cases with tubercu-
lous aneurysms. They proposed a clinical presenta-
tion scenario as follows:
1. Persistent chest, abdominal or back pain in 64%.
2. Hypovolemic shock or other major bleeding in
38%.
3. Palpable or radiologically visible paraaortic mass
in 64%.
Fever was reported in 35% and hypertension in
only four patients (Long et al. 1999). Kao and col-
leagues (l999) reviewed the literature on tuberculous
pseudoaneurysm of the femoral artery and found 6
cases and add one of their own. Three patients had
pulmonary tuberculosis and two had spine and bursa
disease.
43.5.1
Our Own Case Report
This extremely rare case report of one of my patients
with multiple tuberculous mycotic aneurysms and
irregular arterial lumen including:
a. Abdominal aorta
b. Right hepatic artery
c. Left renal artery
d. Celiac artery
e. Inferior mesenteric artery
f. Both iliac arteries
g. Occlusion of the superior mesenteric artery
Detailed operative findings are including which
indicate the difficulties that the surgeon faced in
such complex case.
A 25-year-old female patient was referred from
the Arabian Oil Company in Al-Khafji (Saudi town
on the border with Kuwait) on April 1, 1981.
She presented with dizziness, weakness, headaches,
hypotension and bradycardia. She was on atenolol
prescribed for her in Kuwait two years earlier for
hypertension that was investigated with no appar-
ent cause. The atenolol was stopped and her blood
pressure ranged between 140/100 to 150/110 mmHg.
The hemogram showed hemoglobin 9.0 G/dl, normal
774 M. M. Madkour

white cell count but ESR was 92 mm/h. The biochemi- arteries were found. There was an occlusion of the
cal parameters and urine culture were normal. superior mesenteric artery 1 cm distal to its origin
She delivered in February 1981 and developed par- (Fig. 43.1a,b). The most likely diagnosis was tuber-
oxysmal attacks of bilateral loin pain that was radiat- culous aneurysms based on: clinical features, normal
ing to both inguinal areas. This was thought to be a and equal pulses, no discrepancies in blood pressure,
urinary tract infection and treated with antibiotics. chest radiography, strongly positive tuberculin test,
She had IVU, which did not show any abnormalities. predilection for abdominal aorta.
She was examined by a gynecologist who found no Anti-tuberculous treatment with rifampicin
abnormalities. 600 mg, isoniazid 300 mg, ethambutol 600 mg and
At presentation, April 1, 1981, she was ill-looking, pyridoxine 20 mg daily was given. Surgical manage-
weak, with low blood pressure 90/70 and pulse of ment was not possible at our hospital and arrange-
only 40/min. There were no other clinical findings. ments were made for her transfer to the USA.
She was referred to our hospital for the investigation The patient was admitted to Georgetown Uni-
of her recurrent abdominal pain, headaches, hyper- versity Hospital, Washington, D.C. She was fully
tension and raised ESR. She came to our hospital on assessed by clinicians from internal medicine, gas-
April 6, 1981 and gave a history of abdominal pain troenterology, rheumatology, infectious diseases
since February 1981. It was paroxysmal and associ- and vascular surgery to assist in the work up of this
ated with vomiting, night fever, sweating, loss of complex problem. She was admitted on May 25,1981.
appetite and weight loss. She gave a history of being Her anti-tuberculous medications were temporarily
investigated in Kuwait for headaches and found to stopped on admission. Detailed clinical history and
have hypertension in 1980. physical examination were carried out by these con-
No cause was found to explain her hypertension at sultants. Extensive laboratory tests were performed
this age and she was treated with atenololl 00 mg daily. in addition to bone marrow biopsy, lumbar puncture,
Other systems enquiries were non-contributory. On all were similar to ours. Aortic arteriography was not
examination, she looked ill, pale and in pain. The pulse repeated as the films and medical report were given
was 80/min regular and the peripheral pulse could be to the patient before leaving to USA.
felt and were normal. Her temperature was 38.1°C and Their initial pre-operative discharge diagnosis
blood pressure 200/140 and there were no discrep- was: Abnormal aneurysms? Tuberculous arteritis.
ancies in blood pressure. There were no lymphade- Anemia and hypertension.
nopathy or cardiac murmurs and the chest was clear.
Abdominal examination showed a soft abdomen and a
pulsatile central abdominal mass, slightly lateral to the 43.5.2
umbilicus measuring 3x3 cm was felt. There was no Surgical Findings (June 20, 1981)
organomegaly, no bruit over the mass and the bowel
sounds were normal. The hemogram showed hemo- The patient was assessed operatively on June 20, 1981.
globin 10.0 gldl, normal white cells and ESR 94 mm/h. The operative findings from a report include: "The
Biochemical parameters, tests for syphilis, brucella midline was incised and the abdomen somewhat
and salmonella all were normal. non-adhesive. Immediately upon opening of the
Chest radiography showed left pleural thicken- abdomen a very large marginal artery was prominent
ing with small fibrocalcific changes most likely coursing through the entire large colon mesentery. It
due to old tuberculosis. The Mantoux skin test was came out from an inferior portion of the aneurysm.
strongly positive with induration of 34 mm after 48 The superior mesenteric artery was obliterated.
hours. Morning urine for three days was sent for The aneurismal dilatation at the end of the aorta
cytology for AFB smear staining and cultures for M. was approximately 4x4 cm. The retro peritoneum
tuberculosis were negative. Abdominal ultrasound was then opened and investigation around the aorta
showed a tortuous, irregular and dilated abdominal revealed much lymphoid tissue, hyperplasia and reac-
aorta, celiac trunk and left iliac arteries. Complete tive nodes. The nodes were sent for frozen section and
aortic arteriography found no abnormalities in the came back as granuloma. It is significant to note that
thoracic aorta. The abdominal aorta was grossly there is much reaction over the entire aorta, particu-
irregular with multiple areas of aneurismal dilata- larly in the areas of the aneurismal dilatation. At this
tion and aneurysms. Also, aneurysmal dilatations point before resecting the aneurysm consideration
for the left common iliac, left renal, right hepatic, over to the left renal artery was delivered. There was
gastroduodenal, splenic and the inferior mesenteric a small rent in the renal vein, which was over sewn
 Tuberculosis Vasculitis and Mycotic Aneurysms 775

a b

Fig. 43.1. a TB vasculitis and aneurysmal formation. A 25-year-old female pre-

sented with fever, abdominal pain, headaches and hypertension. There is exten-
sive irregularity of the aortic wall and its branches (arrows). There are multiple
aneurysms of the hepatic, splenic and renal arteries (arrowheads). The superior
mesenteric artery was occluded. b Post-operative angiogram-the patient had
a graft (arrow) at the site of the aneurysm, but developed a large aneurysm of
her right renal artery (arrowheads). c Selective right renal angiogram showing
c the aneurysm to a better extent (arrowheads)

with a 5-0 Ethibond suture. The left renal artery as it
came off the aorta was aneurysmal and the dilatation
was for approximately 1.5 cm. It then followed a very
tortuous course with two further aneurysms, the last
one measured 3x2 cm and was deep within the pelvis
of the left kidney. This probably was a contraindication
to resection since nothing could be sewn into the renal
artery distally and it was wrapped with material to
prevent any further expansion. There was much fibro-
sis in this area and Surgicel was placed into the wound
around the area of the left renal artery aneurysm.
Next, the iliac vessels were surrounded and cleared
of the surrounding tissue. The aorta and iliac vessels
were cross-clamped and the aneurismal dilatation was
resected down to the area past the bifurcation of the
iliac vessels. A 16x8 bifurcation Dacron graft (Dacron
velour graft) was brought onto the field. The cross
clamp time for the large inferior mesenteric artery was
approximately 40 minutes. Woven Dacron wide graft
was inserted into the aorta and into both iliac vessels.
Frozen biopsy of the artery in the iliac region revealed
granulomatous inflammation. The graft was sewn
776
into place. The inferior mesenteric artery was anasto-
mosed into the aorta at the level above the area of the
bifurcation. There was a clot in the inferior mesenteric
artery, briefly anastomosed and the artery was incised
and the clot was removed. No blood oozing and the
abdomen was irrigated and was closed. The operative
blood loss was approximately 3,500 cc and the patient
was brought to the intensive care uniC'
43.S.3
Successful Outcome of Surgery
and Follow-up for 20 Years
Before the patient had returned to Saudi Arabia she
had been put back on her anti-tuberculous chemo-
therapy (post-operatively). As tuberculosis was con-
firmed by histopathological evidence of granuloma-
tous lesions found in the resected aneurysms as well
as the extensive tuberculous lymphadenitis all over the
abdomen. Her raised blood pressure remained high
and she was placed on atenolol in the USA. Her con-
stitutional symptoms, fever, weight loss and abdomi-
nal pain all resolved. The hemoglobin and ESR have
returned back to normal values. She has been followed
up at our hospital regularly every 3 months up until the
present time. She had no recurrences of her abdominal
aortic aneurysms, as we are performing post-opera-
tive complete aortic angiography once every years
for the initial five years and every four years after
that. Her blood pressure, however, failed to normalize
with atenolol two years post-operatively. Anti-hyper-
tensive medication was changed and she is currently
on amlodipine 10 mg once daily oral and is control-
ling her hypertension well. Her kidney functions are
regularly assessed with biochemical parameters and
creatinine clearance tests and remained normal. The
follow-up aortic angiography performed last year is
shown in Fig. 43.1b,c.
In conclusion, this young lady initially presented
with headaches and hypertension that was initially
investigated and no cause was found in Kuwait. Only
one year later, her constitutional symptoms were fever,
weight loss, anemia, abdominal pain and high ESR,
and she was referred to our hospital. The physical
finding of palpable pulsating aneurismal mass in the
abdomen, the presence of plain chest radiography of
lung parenchymal scarring and calcified hilar lymph-
adenopathy, high ESR and strongly positive tuberculin
test made us suspect tuberculosis as the cause of the
multiple aneurysms depicted on the ultrasound and
aortic angiography. She was placed on anti-tubercu-
lous chemotherapy and atenolol. Because of the mul-
M. M. Madkour
tiplicity of the aneurysms and the close proximity of
the renal artery aneurysm to the renal pelvis she was
referred to Washington, USA. During surgery, the diag-
nosis of tuberculosis was confirmed by tissue biopsies
obtained from the massively extensive abdominal
lymphadenopathy as well as from the tissues of the
resected aneurysms with granulomatous lesions seen
on histopathological examination.
The patient was already on anti-tuberculous che-
motherapy for two months before surgery and tissue
cultures were, as expected, negative. She resumed anti-
tuberculous treatment after surgery and completed a
18 months regimen. Her hypertension did not settle
and she required continuous anti-hypertensive treat-
ment till present. Her abdominal symptoms, anemia,
raised ESR and weight loss all resolved.
Follow-up for 20 years showed no evidence of
recurrence of the disease, or further expansion of the
aneurysms that could not be resected.
43.6
Investigation
Imaging modalities plays an important role in the
diagnosis of arterial mycotic aneurysms. Chest radi-
ography may depict miliary tuberculosis, severe lung
parenchymal damage due to tuberculosis. Sputum
smear, cultures for M . tuberculosis, serological tests
for brucella should be done in countries where these
two diseases are endemic. Serology testing for HIV
is required in individuals at high risk. Abdominal
ultrasounds are useful in depicting abdominal
aortic mycotic aneurysm or in its major branches.
It may also depict mesenteric or para-aortic lymph
adenopathy.
Complete aortic arteriography is essential in the
diagnosis for establishing the site of aneurysm, its
multiplicity in other arteries and the state of the
arterial wall irregularity and lumen narrowing or
occlusion. Magnetic resonance angiography is a
noninvasive diagnostic modality. It is more sensitive
than contrast arteriography for the detection of distal
runoff vessel patency (Kao et al. 1999).
43.7
Treatment
Patients may have surgery for the tuberculous aneu-
rysms before the true cause is discovered and receive
Tuberculosis Vasculitis and Mycotic Aneurysms
the anti-tuberculous chemotherapy after surgery.
The treatment of tuberculous aneurysms is the com-
bination of anti-tuberculous medication and surgery.
The choice of anti-tuberculous drugs is dependent
on local incidence of drug resistance as well as the
immune status of the patient. Treatment should be
extended for at least 9-12 months or longer (Allins et
al.1999; Penn et al.I996).Allins and colleagues (1999)
from California reviewed the English language litera-
ture and found 26 patients who were operated for
tuberculous aneurysms of the descending thoracic
or abdominal aorta. Six patients had emergency
operations for massive hemoptysis, aortoduodenal
fistula or abdominal rupture of the aneurysm with
a mortality rate at 30 days of 50%. Twenty patients
had elective or semi-elective aneurismal repair and
19 survived for more than 30 days. The outcome of
surgery for tuberculous aneurysm was also reported
by Long and colleagues (1999) of published reports
in the literature. A total of 22 patients had surgery for
tuberculous aneurysms of the thoracic and abdomi-
nal aorta and 19 patients (86%) survived one month
or more after surgery while the other three died
shortly after surgery.
In India, Choudhary et al. (2001) reported the sur-
gical outcome of five young patients with tuberculous
pseudo aneurysms seen over a 3-year period. Sites of
involvements were: ascending aorta, distal aortic arch,
proximal descending thoracic aorta, distal descending
thoracic aorta, infra-renal abdominal aorta. All had the
infection caused by a contiguous focus of tuberculous
tissue, or from an organ close to the location of the
tuberculous pseudo-aneurysm. All five patients were
known to have tuberculosis before the operations. The
reasons for the operations were rapid clinical deterio-
ration. Surgery was successful in all five patients, with
one patient developing recurrence at the same site of
the previous aneurysm 8 months later.
References
Allins AD et al (1999) Tuberculous infection of the descending
thoracic and abdominal aorta: Case report and literature
review. Am Vasc Surg 13:439-444
Atnip RG (1989) Mycotic aneurysms of the suprarenal
abdominal aorta: prolonged survival after in situ aortic
and viscera! reconstruction. J Vasc Surg 10:635-641
Baumgarten EC et al (1933) Tuberculous mesaarteritis with
aneurysm of the femoral artery: report of a case. JAMA
100:1918-1920
Beaconsfield T et al (1998) Case report: tuberculous pulmo-
nary arteritis - an unusual cause of light pulmonary artery
stenosis. Clin RadioI53:229-231
777
Beeresha GLH et al (2000) Hepatic artery mycotic aneurysm of
tubercular aetiology. J Assoc Physicians India 48:247-248
Boggian K et al (1994) True aneurysm of the ascending aorta
in HIV disease. Schweiz Med Wochenschr 124:2083-2087
Bojar RM, Turner MT, Valdez S et al (1998) Homograft repair
of a tuberculous pseudoaneurysm of the ascending order.
Chest 114:1774-1776
Chan FY (1989) In situ prosthetic graft replacement for mycotic
aneurysm of the aorta. Ann Thorac Surg 47:193-203
Chiba Yet al (1996) Surgical treatment of infected thoracic and
abdominal aortic aneurysms. Cardiovas Surg 4:476-479
Choudhary SK et a! (2001) Tubercular pseudoaneurysms of
aorta. Ann Thorac Surg 72:1239-1244
Cross DT et al (1995) Endovascular treatment of epistaxis in a
patient with tuberculosis and a giant petrous carotid pseu-
doaneurysm. AMJ NeuroradioI16:1084-1086
De KruijfEJF et al (2000) Tuberculous aortitis with aortodude-
nal fistula presenting as recurrent gastrointestinal bleed-
ing. Clin Infect Dis 31:841-842
De Prophetis N et al (1959) Rupture of tuberculous aortic
aneurysm into lung. Ann Surg 150:1046-1051
Deshmukh H et al (2000) Endovascular management of an
inferior gluteal artery pseudo aneurysm secondary to
tuberculous cold abscess. Cardiovasc Intervent Radiol 23:
80-82
Felson B et al (1977) Mycotic tuberculous aneurysm of the
thoracic aorta. JAMA 237:1104-1108
Ghosh H (1954) Tuberculous lymphadenitis: report of a case
with perforation of aorta into duodenum. Am J Clin Pathol
24:1044-1049
Girard P et al (1997) An unusual cause of aorto-bronchial fis-
tula: tuberculosis-aortitis. Rev Mal Respir 14:221-222
Glozarian J et al (1999) Tuberculous pseudoaneurysm of the
descending thoracic aorta. Tex Heart Inst J 26:232-235
Gouny P et al (1992) Human immunodeficiency virus and
infected aneurysm of the abdominal aorta: report of three
cases. Ann Vasc Surg 6:239-243
Hagino RT et al (1996) A case of Pott's disease of the spine
eroding into the paraspinal aorta. J Vasc Surg 24:482-486
Hakim S et al (1993) Psoas abscess following intravesical
bacillus calmette-Guerin for bladder cancer: a case report.
J UroI150:188-189
Hatakeyama S et al (1997) Massive hemoptysis in a patient
with a tuberculous thoracic aneurysm. Nihon Kyobu Shik-
kan Gakkhai Zasshi 35:106-110
Herndon JH et a! (1952) Ruptured tuberculous false aneurysm
of the abdominal aorta: report of a case with resection of
the aneurysm and survival for six days. Tex State J Med
48:336-338
Hildebrandt JW et al (1998) Diagnostic case study: Rasmussen
aneurysm. Semin Respir Infect 13:160-162
Hollier LH et al (1993) Direct replacement of mycotic thora-
coabdominal aneuryms. J Vasc Surg 18:477-485
Hude R et al (1998) Acid-fast bacilli in the genitourinary tract
3 years after intravesical Bacilli Calmette-Guerin therapy
for bladder carcinoma. Br J UroI81:496-497
Ikezawa T et al (1996) Tuberculous pseudoaneurysm of the
descending thoracic aorta: a case report and literature
review of surgically treated cases. J Vasc Surg 24:693-697
Ishibatake H, Onizuka R (1998) A successfully treated case
of miliary tuberculosis with adult respiratory distress
syndrome and tuberculous aneurysm of abdominal aorta.
Kakkakuy 73:403-411
778
Ito K et al (1992) A case of extra-anatomic bypass for ruptured
thoraco-abdominal aneurysm with tuberculosis. Kyobu
Geka 45:640-643
Jebara VA et al (1998) Mycotic aneurysm of the popliteal artery
secondary to tuberculosis. A case report and review of the
literature. Tex Heart Inst J 25: 136-139
Kamen L (1895) Aortenruptur auf tuberkuloser Grundlage.
Beitr Pathol Anat Allg Patho117:416
Kao CL et al (1999) Tuberculous pseudoaneurysm of the femo-
ral artery: a case report. J Vasc Surg 30:561-564
Long R et al (1999) Tuberculous mycotic aneurysm of the
Aorta. Review of published medical and surgical experi-
ence. Chest 115:522-531
Madkour MM (2001) Cardiovascular localizations - Brucella
Mycotic Aneurysms. In: Madkour MM (ed) Madkour's
Brucellosis, 2nd edn. Springer, Berlin Heidelberg New
York, pp 139-141
Muller-Wening D et al (1982) Tuberculous aneurysm of the
descending aorta with aorta-bronchial fistula (in German).
Prax Klin Pneumol 36:22-26
O'Leary M et al (1977) Rupture of a tuberculous pseudoa-
neurysm of the innominate artery into the trachea and
esophagus: report of a case and review of the literature.
Hum Pathol 8:458-467
Ogawa J et al (1990) A tuberculous pseudoaneurysm of the
thoracic aorta presenting as massive hemoptysis - a case of
successful surgical treatment. Jpn J Surg 20:107-110
Oran I et al (2001) Mesenteric artery aneurysm in intestinal
tuberculosis as a cause of lower gastrointestinal bleeding.
Abdom Imaging 26:131-133
Parkhurst GF et al (1955) Bacterial aortitis and mycotic aneu-
rysm of the aorta: a report of twelve cases. Am J Pathol
31:821-835
Patankar T et al (2000) Fatal hemoptysis caused by ruptured
giant Rasmussen's aneurysm. AJR Am J Roetgenol 174:
262-263
Penn RL et al (1996) Lower respiratory tract infections. In:
Reese RE, Betts RF (eds) A practical approach to infectious
diseases, 4th edn. Little Brown, Boston, pp 258-349
Peyton RW (1965) Surgical correction of tuberculous pseu-
doaneurysm of upper abdominal aorta: first reported case.
Ann Surg 162:1069-1074
Quaini E et al (1985) Mycotic tuberculous aneurysm of the
descending thoracic aorta. Tex Heart Inst J 12:257-260
Rob CG, Eastcott HHG (1955) Aortic aneurysm due to tuber-
culous lymphadenitis. BMJ 1:378-379
M. M. Madkour
Rozenblit A et al (1996) Infected aortic aneurysm and ver-
tebral osteomyelitis after intravesical bacillus calmette-
Guerin therapy. AJR Am J RoetgenoI167:711-713
Sanyika C et al (1999) Pulmonary angiography and emboli-
zation for severe hemoptysis due to cavitary pulmonary
tuberculosis. Cardiovasc Intervent RadioI22:457-460
Schellhammer PF et al (1975) Study of local and systemic
effects of intravesical BCG. Urology 6:562-567
Seelig MH et al (1999) Mycotic vascular infections of large
arteries with Mycobacterium bovis after introvesical bacil-
lus calmette-guerin therapy: Case report. J Vasc Surg 29:
377-381
Slavin RE et al (1990) Late generalized tuberculosis: a clinical
pathologic analysis and comparison of 100 cases in the
preantibiotic eras. Medicine 59:352-366
Strand BT et al (2001) Tuberculous aneurysm of the aorta
presenting with uncontrolled hypertension. J Vasc Interv
RadioI12:521-523
Takahashi et al (1986) A case of tuberculous abdominal aortic
aneurysm which ruptured and perforated the sigmoid (in
Japanese). Nippon Geka Gakkai Zasshi 87:99-104
Tsurutani H et al (2000) Hepatic artery pseudoaneurysms in
a patient treated for miliary tuberculosis. Intern Med 39:
994-998
Volini FI et al (1962) Tuberculosis of the aorta. JAMA 181:
78-83
Weigert C (1882) O'ber Venetuberkel und ihre Beziehungen
zur Tuberculosen Blutinfection. Vircow Arch Pathol Anat
88:307-379
Wetteland P et al (1956) Tuberculous aortic perforations.
Tubercle 37:177-182
WolfYG et al (1995) Infection of a ruptured aortic aneurysm
and an aortic graft with bacilli calmette-Guerin after intra-
vesical administration for bladder cancer. J Vasc Surg 22:
80-84
Woods JM et al (1988) Mycotic abdominal aortic aneurysm
induced by immunotherapy with bacilli Calmette-Guerin
vaccine for malignancy. J Vasc Surg 7:808-810
Yo Y et al (1993) Case report of miliary TB, with mediasti-
nal TB, involving the aorta, trachea and esophages with
death due to hemorrhage from perforation of aorta into
the esophagus during treatment. Nihon Koyobu Skikkan
Gakkai Zasshi 31:738-743
Zhang JS et al (1994) Bronchial arteriography and trans-
catheter embolization in the management of hemoptysis.
Cardiovasc Intervent RadioI17:276-279
Treatment
44 Treatment of Tuberculosis
DEAN E. SCHRAUFNAGEL

CONTENTS treatment. They are more authoritative than any

single authored work and should be consulted for
44.1 Introduction 781 the treatment and management of tuberculosis. This
44.2 Evolution of the Treatment of Tuberculosis 781
44.3 The Problem of Drug Resistance 782 chapter generally keeps within these standards, but
44.4 Latent and Paucibacillary Tuberculosis 783 occasionally goes beyond, or aside of, them especially
44.4.1 Which Therapy for Latent Tuberculosis when clinical trial data is unavailable or unreliable.
Is Best? 785 Anecdotal information, although less authoritative,
44.5 Extrapulmonary Tuberculosis 786
is often additive and helpful in dealing with clinical
44.6 Children 786
44.7 Pregnancy 787 problems and practice.
44.8 Liver Disease 787
44.8.1 Patients Who Start with Normal Livers 788
44.8.2 Treating Patients with Liver Disease 789
44.9 Renal Failure 789
44.2
44.10 Individual Drugs 790
44.10.1 Isoniazid 790 Evolution of the Treatment of Tuberculosis
44.10.2 Rifampin 791
44.10.3 Pyrazinamide 791 Koch's demonstration that tuberculosis was caused by
44.10.4 Ethambutol 792 a bacillus (Koch 1882) opened the way for treatment
44.10.5 Streptomycin and Other Aminoglycosides 792
and cure, but his early attempts at treatment through
44.10.6 Rifabutin 792
44.10.7 Rifapentine 793 immunomodulation failed (Roullion 1981). Between
44.10.8 Fluoroquinolones 793 the 1890s and 1940s claims for cures of tuberculo-
44.10.9 Capreomycin 794 sis included a variety of diets, drugs, and surgeries.
44.10.10 Ethionamide 794 Patients with tuberculosis often improve spontane-
44.1 0.11 Cycloserine 794 ously and this improvement confused both the patient
44.10.12 Para-aminosalicylic Acid (PAS) 794
44.10.13 Linezolid 794 and investigator. The problem for the investigator was
44.10.14 Other Treatments 795 that only about half of sputum positive patients would
References 795 die of tuberculosis in five years. About a fourth would
be 'cured' and about a fourth would become 'chronic'.
The cures were attributed to an intervention, but the
'cures' would often relapse. When 'galloping consump-
44.1 tion' set in with its 'hectic fevers' and wasting, progres-
Introduction sion and death were more certain.
The false hopes for cures lead to the first con-
Excellent statements of recommendation for the trolled clinical trial of streptomycin for the treat-
treatment of tuberculosis have been produced by ment of tuberculosis (Hill 1948). Even though this
the American Thoracic Society-Centers for Disease study showed that streptomycin 'cured' tuberculosis
Control and Prevention (American Thoracic Society (Hinshaw and Feldman 1945; Schatz and Waksman
et al., 2003), World Health Organization (1997), and 1944), after only a few months the disease often
other organizations. These statements are written returned and was then incurable as the organisms
by committees of experts and set the standard for had developed resistance to it.
It was not until Lehmann discovered that para-
D. E. SCHRAUFNAGEL, MD
aminosalicylic acid (PAS) had anti-tubercular
Department of Medicine M/C 719, University of Illinois at activity (Lehmann 1946) that a combination of this
Chicago, 840 S. Wood St, Chicago, IL 606127323, USA relatively weak drug and streptomycin could really

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
782
cure tuberculosis. The idea of combination therapy
was born. PAS was difficult to take and toxic. Fortu-
nately, isoniazid was soon discovered. During much
of the 1950s different combinations of doses and
durations of these three drugs were studied. Isonia-
zid was at first thought to be without side effects and
its advent heralded a new optimism for reducing the
burden and possibly eliminating tuberculosis. The
late 1950s and 1960s saw the introduction of pyrazin-
amide (Yeager et al. 1952), ethambutol, thiacetazone,
ethionamide, cycloserine, capreomycin, and amino-
glycosides. By the end of the 1960s, ethambutol was
deemed to be safe and the next most effective drug.
The standard treatment for tuberculosis was two
years of isoniazid and ethambutol, which was later
shortened to 18 months. PAS, streptomycin, and the
others were reserved for resistant disease, difficult
cases, and special situations such as meningitis.
Rifampin was discovered in 1966 (Maggi et al.
1966) and came into popular use in the late 1970s. A
series of well performed trials by the British Thoracic
Society established a lower dose of pyrazinamide
(-25 mg/kg) as a first line agent. Earlier studies
showed that higher doses (50 mg/kg) were too toxic
for use as a first line drug. Additional British Thoracic
Society trials, and subsequently others, showed that
pyrazinamide, rifampin and isoniazid for 6 months
was adequate (Aquinas 1978; Chan et al. 1994; Cohn
et al. 1990; Hong Kong Chest Service and British
Medical Research Council 1991; Hong Kong Chest
Service/British Medical Research Council 1987; Hong
Kong Chest Service/Tuberculosis Research Centre
and Madras/British Medical Research Council 1989;
Hong Kong Chest Service/Tuberculosis Research
Centre and Madras/British Medical Research Coun-
ci11991; Snider et al. 1984a,b). Experience and other
studies established these three drugs as the core of
the current treatment for tuberculosis. A fourth drug,
ethambutol, was added until the resistance patterns
were known. See Tables 44.1 and 44.2 for the doses of
the first and second line agents.
44.3
The Problem of Drug Resistance
The number of bacilli living in a person with
tuberculosis is great. A single cavity may have 109
organisms. Organisms found in nature or 'wild type'
organisms have rare, naturally occurring resistance
to the antibiotics in use at a frequency of about 10-
6
3 to 10- depending on the drug being considered
D. E. Schraufnagel
(David 1970). This means that a cavity may have 10 3
to 106 organisms that are resistant to a given antibi-
otic before the antibiotic is first given. Resistance to
different classes of antibiotics is independent of each
other so that the chances of a wild type organism
being resistant to two different antibiotics would be,
using our same range, (10- 3_10- 6 ) X (10- 3_10- 6) which
would be 10-6 _10- 12 • Using two to three antibiotics
would assure the killing of virtually all wild type
organisms.
When streptomycin was first given, most of the
organisms were killed. The progeny of the remain-
ing bacilli that were not killed, and thus resistant to
streptomycin, multiplied. It appeared to the clinician
that a recurrence had occurred about three months
later, which was an appropriate time for these slow
growing bacteria to repopulate. Combination therapy
prevents the development of resistance but only if at
least two effective drugs are given. Problems arise
with ineffective medication and poor compliance.
Giving anti-tuberculous agents one by one, with
adequate time for repopulation, promotes resistance.
Unfortunately drug resistance is common and
found in all countries. The main reasons for this are
patient non-adherence to a standard therapeutic
regimen, errors in prescribing and delivering medi-
cation, failure to teach and emphasize the importance
of taking the medication until completion, mono-
therapy for unrecognized tuberculosis, and unsus-
pected primary resistance. Patient non-adherence to
a regimen may result from drug side-effects.
Fortunately for populations where drug resistance
is high, good national tuberculosis control should
eventually reduce the resistance in the population.
This is because evolutionary principles decree that
wild type organisms are the fittest for survival under
the stable environmental conditions in which they
live. As selective pressures change, because incom-
plete anti-tuberculous therapy kills sensitive bac-
teria, native organisms are supplanted by resistant
strains. Using an overwhelming force of antibiotics
destroys all bacilli and does not allow the resistance
to develop. If the selective pressure from partial or
weak drug application is removed, there is a return to
the native state in which the wild type again becomes
the dominant strain in a region (Hong et al. 1998;
Schraufnagel and Abubaker 2000; Weiss et al. 1994).
To discover resistance early, health care workers
following patients on anti-tuberculous drugs must
check culture and sensitivities and follow the sputum,
although the latter is often difficult because most
patients stop producing sputum as they improve.
Radiographs should be taken until improvement is
Treatment of Tuberculosis 783

Table 44.1. Doses of standard (first line) anti-tuberculous Any of these features should prompt a search for a
drugs for adults pathological process.
Drug Daily Twice weekly Thrice weekly To counter problems of drug resistance and to
(Maximal) (Maximal) (Maximal) assure a continuous strong drug presence, directly
Isoniazid 5mg/kg 15 mg/kg 15 mg/kg
observed therapy (DOT) is now recommended
(300 mg) (900 mg) (900 mg) throughout the world. DOT means that a designated
Rifampin 10 mg/kg 10 mg/kg 10 mg/kg person watches the patient swallow each dose of the
(600 mg) (600 mg) (600 mg) prescribed drugs (Bayer and Wilkinson 1995). DOT
Ethambutol 15 mg/kg 45-50 mg/kg 25-35 mg/kg increases the cure rate and reduces the prevalence of
(1600 mg) (4000 mg) (2400 mg)
Pyrazinamide 25 mg/kg 40-50 mg/kg 35 mg/kg
tuberculosis (Weiss et al. 1994).
(2000 mg) (4000 mg) (3000 mg) Education is an important part therapy. Discussing
Rifabutin 5 mg/kg 5 mg/kg 5 mg/kg how patients obtained the disease and how to prevent
(300 mg) (300 mg) (300 mg) its spread may help in decreasing its transmission in
For further information consult the American Thoracic the community. Showing patients their chest radio-
Society statements (http://www.thoracic.org) or Centers for graph can engage them in their treatment and point
Disease Control website (http://www.cdc.govl) out the progress they are making. Praising them for
adherence to the regimen and being available for
questions are part of a good therapeutic plan.
Table 44.2. Doses of secondary anti-tuberculous drugs for adults When drug sensitivities are available the medi-
cation should be changed to take advantage of the
Drug Daily Twice weekly Thrice weekly
(Maximal) (Maximal) (Maximal)
sensitivities. With resistance to any single drug, the
chance for cure is nearly the same as for tuberculosis
Streptomycin" 15 mg/kg 15 mg/kg 15 mg/kg resulting from a fully sensitive organism, although
(1000 mg) (1000 mg) (1000 mg)
Levofloxacin b (500 mg)
the time of treatment may be longer. Even with using
Moxifloxacinb (400 mg) the same standard four-drug regimen, the survival
Gatifloxacin b (400mg) with isoniazid resistance is about the same (Perries
Capreomycin" 15 mg/kg 15 mg/kg 15 mg/kg et al. 1995). However, any drug resistance means a
(1000 mg) (1000 mg) (1000 mg) smaller margin of safety.
Rifapentine 10 mg/kg
Resistance in vitro may not always translate into
once or twice
weekly (600 mg) failure for that drug but susceptibility tests are the
Amikacin" 15 mg/kg 15 mg/kg 15 mg/kg best predictors of whether a drug will be valuable.
(1000 mg) (1000 mg) (1000 mg) Ethambutol resistance in vitro can be artifactual
Kanamycin" 15 mg/kg 15 mg/kg 15 mg/kg because it tends to degenerate in the media with time
(1000 mg) (1000 mg) (1000 mg)
Ethionamideb 15-20 mg/kg
and allow growth. Isolated ethambutol resistance
(500-750 mg) should be rechecked by another laboratory.
PAS b (8-12 g in 3 doses) Developing drug resistance is not important when
" Dose and interval depends on efficacy of companion drugs
dealing with latent tuberculosis because it is esti-
and mycobacterial burden. Reduce dose to 750 or 500 mg for mated that these patients have only a small number
the elderly. Start or move to twice or thrice weekly soon. of organisms, perhaps 10 2_10 4; this is readily treated
b Intermittent dose not established. by a single drug.

clear and the lesions are either resolved or stable.

Radiographic signs of active tuberculosis include
cavities, alveolar shadows, pleural effusions and
changing lesions with time. If radiographic signs of
activity are present when the course is scheduled for
completion, consideration should be given to con-
tinue treatment. This is especially true if the patient
has cavitary disease or prolonged sputum positivity.
On the follow-up visits, weight loss, new symp-
toms, and deterioration in general health may be
important clues to treatment failure or the appear-
ance of a new disease, such as coincident lung cancer.
44.4
Latent and Paucibacillary Tuberculosis
Latent tuberculosis is defined as a tuberculous infec-
tion in a person who has no symptom of disease and
the signs are limited to a positive tuberculin skin
test and a chronic stable chest radiograph with only
granulomata, adenopathy, apical capping, or pleural
thickening. Infected persons harbor bacilli and have
a rate activation that can be estimated in large part
784
from the comorbid factors affecting the immune
system and the duration of infection.
Almost a third of the world has latent tuberculosis
attesting to the prowess of M. tuberculosis as a human
pathogen. In deciding whether to treat these indi-
viduals, one must weigh the risk and benefit between
the chance of the tuberculous infection becoming
active versus the side effects of the required medica-
tion. Identifying the risk of activation is summarized
periodically by the American Thoracic Society and
other organizations (American Thoracic Society and
Center for Disease Control and Prevention 2000).
The recommendations to treat latent tuberculosis
were based on the risk benefit of using isoniazid for
a year, which has been well studied and had been the
standard for a least two decades (American Thoracic
Society and Center for Disease Control and Preven-
tion 2000). In 1994, the ATS recommendations were
to go for 6-12 months (Bass et al. 1994). In 2000, the
recommendations were changed to treat latent tuber-
culosis with isoniazid for nine months, rifampin for
four months, pyrazinamide and rifampin for two
months, or isoniazid for six months (American
Thoracic Society and Center for Disease Control
and Prevention 2000) but these recommendations
were based on much less data. If these, or other new,
regimens prove to be safer, the risk benefit of treating
more or all persons with latent tuberculosis should be
reconsidered. Latent syphilis, a disease with potential
for harm similar to tuberculosis, is virtually always
treated because of the safety and ease of using peni-
cillin. A comparable anti-tuberculous regimen may
dictate its use for all infected persons, but currently
this is not the case.
The treatment of latent tuberculosis has resulted
in a great reduction of the community prevalence
of tuberculosis (Comstock et al. 1967). Treating
tuberculosis prevents further transmission, which
means the treatment of latent tuberculosis has a
greater implication than merely aiding the infected
individual.
Paucibacillary tuberculosis is the presence of
symptoms and signs characteristic of tuberculosis
but an inability to find organisms on sputum smear.
For abacillary tuberculosis, sputum cultures are also
negative. The symptoms and signs improve with anti-
tuberculous treatment. The Hong Kong Chest Service
reported that 57% of patients who were thought to
have tuberculosis but had five negative sputa went
on to develop active tuberculosis within five years
(Hong Kong Chest Service/Tuberculosis Research
Centre and Madras/British Medical Research Coun-
cil 1989). A South African study reported similar
D. E. Schraufnagel
findings (Cowie et al.1985). These studies tell us that
patients presenting with symptoms and signs of typi-
cal tuberculosis should be treated even if the organ-
ism is not identified.
The failure to find bacilli should not be a deter-
rent from treating a person who presents with typi-
cal findings. In general the treatment should start
with four drugs, rifampin, isoniazid, pyrazinamide,
and ethambutol and continue for four months with
2-4 drugs depending on the resistance pattern in the
area. The two drug follow-up regimen, of isoniazid
and rifampin, is permissible if the case comes from
a community whose drug resistance is less than 4%.
The 4-month course has been shown to be sufficient
for paucibacillary disease (Dutt and Stead 1994;
Hong Kong Chest Service/Tuberculosis Research
Centre and Madras/British Medical Research Council
1989). If the person does not improve with treatment,
continuing the treatment and searching for another
diagnosis is often warranted.
'Treating only until cultures come back' is not
recommended in patients with characteristic symp-
toms, or signs, of tuberculosis and a positive skin
test, especially if they would qualify for treatment
of latent disease. A 4-month isoniazid and rifampin
treatment will treat abacillary and latent tuberculo-
sis. The treatment may help make another diagnosis.
For example, a patient has one or more lung nodules
and is at high risk for surgery. The differential diag-
nosis includes tuberculosis and cancer. If anti-tuber-
culous treatment is started and the lesion continues
to grow the diagnosis of tuberculosis is unlikely. If
the drugs were stopped with negative cultures, the
diagnosis of the growing lesion at a later date could
still be tuberculosis or cancer. Furthermore, the dif-
ference in side effects between treating active and
latent tuberculosis is small. Frequently, treatment for
latent disease is still required and the initial therapy
has already begun. The first days of any drug therapy
are the most common time for side effects to occur.
Of course, if latent tuberculosis is not a consideration
and another diagnosis has been made, it may be
appropriate to stop the anti-tuberculous therapy.
There are two fundamental rules for the treatment
of latent tuberculosis: First, rule out active disease.
This is especially true if only a single agent is used.
Ruling out active disease is often difficult when other
lung disease is present and the subtle symptoms of
tuberculosis may be overshadowed by chronic respi-
ratory symptoms and the radiograph is abnormal.
Subtle radiographic changes may be missed. It is pos-
sible that undetected active, or 'smoldering', tubercu-
losis may account for the failure of the 'prophylactic
Treatment of Tuberculosis
tuberculosis treatment' reported in silicosis. The
hazard of treating active disease as latent is causing
the development of drug resistance.
The second basic rule is: 'Do not harm the patient'.
The lifetime risk of developing active tuberculosis
for a tuberculin positive individual and no other
risk factor is probably only about 10%. In treating
latent tuberculosis, the harm is caused by side effects
of the medications. To reduce the chance of serious
harm, detailed education and careful follow-up must
be given. Patients should be told that this medicine,
like all drugs, can cause side effects. The warning
about hepatitis should be explicit. On each follow up
clinic visit the health care worker should specifically
ask about symptoms of hepatitis and compliance.
If a side effect occurs, the patient must stop taking
the medication and call the physician or nurse. The
patients should be warned that if they continue
taking the medication, the side effect may worsen and
could cause death. If the drugs are promptly stopped,
the problem will resolve. Patients should report any
new symptom to the health care worker who must
see beyond a patient's attribution of symptoms to the
cause. For example, a patient may attribute a lack of
appetite to a change in smoking or work schedule,
but the symptom may be the first sign of hepatitis.
Health care workers must be alert to clinic 'no-shows'
and be especially concerned if there is a language or
cultural difference.
The importance of treating latent tuberculosis
increases for a recent conversion and factors that
decrease the host defense against tuberculosis. A
person with a recent conversion has the greatest
chance of developing tuberculosis in the first two
years after the conversion. If a recent converter
has HIV, the risk of developing active disease may
increase to 41 % in 60 days (Di Perri et al. 1989).
44.4.1
Which Therapy for Latent Tuberculosis Is Best?
The ATS now recommends four choices of treat-
ments for latent tuberculosis: isoniazid for nine
months, rifampin for four months, pyrazinamide and
rifampin for two months, or isoniazid for six months
(American Thoracic Society and Center for Disease
Control and Prevention 2000). A study of the Inter-
national Union Against Tuberculosis showed that
taking isoniazid for 12 months reduced tuberculosis
in compliant patients by 93% compared to placebo
controls; six months of treatment reduced it by 69%
and three months by 31% (International Union
785
Against Tuberculosis and Lung Disease 1982). The
chance of developing a liver reaction from isoniazid
is about 0.1 % per month.
A physician choosing conservative treatment who
wants substantial data would probably select 9-12
months of treatment with isoniazid as the regimen
of choice to treat latent tuberculosis. The data on
12 months of isoniazid is substantial and convinc-
ing. That for nine months of isoniazid is scarce but
reasonable. It comes largely from post hoc analysis
of longer trials but has convinced many experts to
pick this regimen as their first choice. The regimen
of isoniazid alone may be best for patients on medi-
cine that might interact with rifampin. Its biggest
drawbacks are that non-adherence is correlated to
length of therapy and the additional cost of monthly
follow-up visits.
The course that most departments of health chose
after the 1994 ATS recommendation was six months
of therapy. This is still an ATS recommended choice,
but when I presented the six-month versus twelve-
month choice to more than 200 mostly health care
workers who were beginning treatment for latent
tuberculosis between 1994 and 1999, more than 95%
choose the 12-month course despite the inconve-
nience and greater risk of toxicity of longer therapy.
Choosing the rifampin-based treatments depends
on how comfortable one is with less data. After the
ATS-CDC recommendations appeared in the year
2000, many health departments switched to the
two-month course of rifampin and pyrazinamide,
which resulted in financial and manpower savings,
but additional reports of fatalities (American Tho-
racic Society and Center for Disease Control and
Prevention 2001) have caused the enthusiasm for
this regimen to diminish. The two-month rifampin
and pyrazinamide regimen is supported by an open,
controlled study (Gordin et al. 2000) that showed no
difference in efficacy and toxicity between it and iso-
niazid for 12 months. Other reports have found it is
more toxic (American Thoracic Society and Center
for Disease Control and Prevention 2001) which has
led to the retraction of this general recommentation.
Nevertheless, the 2-month rifampin-pyrazinamide
regimen might be applicable for those who cannot be
followed for a longer period, or do not wish to take
medication for the longer period. It may be appro-
priate for those who have already started four- drug
therapy and has the advantage that a two drug combi-
nation avoids monotherapy, if active disease is inad-
vertently treated. It has been shown to be effective in
patients with HIV and it may be useful to complete
the course quickly and resume antiretroviral therapy
786
that may have been stopped because of rifampin drug
interactions.
Rifampin alone for four months has advocates
who point out that it may be the single-most effec-
tive agent against tuberculosis. It is thought to be the
safest of the regimens, although comparative trials
have not found differences between these regimens.
It is convenient, but the data from clinical trials in
limited (Villarino et al. 1997). In Hong Kong, silicot-
ics with latent tuberculosis given three months of
rifampin did as well as three months of isoniazid
and rifampin and six months of isoniazid and much
better than placebo (Hong Kong Chest Service/
Tuberculosis Research Centre and Madras/British
Medical Research Council 1992).
44.5
Extrapulmonary Tuberculosis
and Corticosteroids
Although some authorities recommend that mili-
ary, meningeal, bone, and joint tuberculosis should
receive up to 12 months of treatment (American
Academy of Pediatrics 2000; American Thoracic
Society et al. 2003), clinical studies generally show
that the six-month treatment courses are effective for
treating extrapulmonary disease. However, antibiotic
therapy beyond six months should be considered for
any tuberculosis if the number of infecting organ-
isms is great, the antibiotics may fail to reach the
target in adequate strength, or there is a wish to
broaden the margins of safety.
Tuberculous cavities are teeming with bacilli
and cavitary disease is characteristic of pulmonary
tuberculosis. Extrapulmonary disease usually does
not have cavities, or has far fewer cavities, resulting
in a lower bacillary burden than pulmonary disease.
The argument for a longer therapy for meningeal,
bone, and joint tuberculosis is that tissue penetration
of the antibiotics is less.
Several studies have found that the standard six-
month therapy does not result in a worse outcome
than longer treatment for meningitis (Alarcon et al.
1990; Biddulph 1990; Chotmongkol 1991; Doganay
et al. 1995; Donald et al. 1998; Jacobs et al. 1992; van
Loenhout-Rooyackers et al. 2001). A British Thoracic
Society trial comparing six months versus nine months
of therapy for tuberculous adenopathy showed that
the six-month regimen was as good (British Thoracic
Society Research Committee 1992).
D. E. Schraufnagel
There have been many papers on the use of cortico-
steroids as an adjunct in the treatment of tuberculosis
(Dooley et al. 1997; Hosoglu et al. 1998,2002; Schoe-
man et al. 1997). The main damage done by tubercu-
losis results from the body's response rather than the
bacteria themselves. The shadows on the chest film do
not result from bacilli but rather the exudation elicited
by inflammatory cytokines and the leukocytes them-
selves. Advanced AIDS and neutropenic patients usu-
ally do not have cavitary disease and necrosis despite
lesions containing many bacilli. This concept also
explains the 'paradoxical' response seen with treating
tuberculosis. There is no paradox, if the damage is
attributed to the host's reaction, and bacterial prod-
ucts elicit the host's response (Afghani and Lieberman
1994; Dastur et al. 1995). A damage analogy might be
opening a fire hose in a finely furnished home. The
smoking toaster is extinguished but the water damage
may be great. Steroids control the water damage.
Many studies have shown no harm from corti-
costeroids if the tuberculosis is being treated with
appropriate antibiotics. Steroids reduce inflam-
mation that may cause scarring and its long-term
effects. Corticosteroids are recommended in menin-
gitis because small amounts of scarring can impede
the flow of cerebral spinal fluid. Steroids may reduce
the scarring seen with endobronchial tuberculosis
(Alzeer and FitzGerald 1993; Chan and Pang 1989;
Park et al. 1997), and may improve the outcome in
tuberculous pericarditis (Long et al. 1989; Trautner
and Darouiche 2001) and brain tuberculoma (Garg
1999; Gavazzi et aI.1998). They also reduce the inflam-
mation in tuberculous pleurisy (Dooley et al. 1997;
Galarza et al. 1995; Grewal et al. 1969; Habibullah
1968; Lee et al. 1988; Wyser et al.I996).
44.6
Children
The treatment of tuberculosis in children is essen-
tially the same as adults, although the Academy of
Pediatrics recommends not using ethambutol if
children cannot tell you if they have visual problems
(American Academy of Pediatrics 2000). At the dose
of 15 mg/kg and in the absence of renal disease, eth-
ambutol rarely causes eye or any other side effects,
but the benefit of ethambutol is also low in children
who usually have a low mycobacterial load. Etham-
butol does not shorten the course or add significantly
to the treatment of tuberculosis caused by fully sen-
sitive bacilli. The American Academy of Pediatrics
Treatment of Tuberculosis
also recommends that duration of anti-tuberculous
therapy should be at least nine months for children
and at least 12 months for children with miliary, bone,
joint and meningeal tuberculosis but there is no evi-
dence from clinical trials to support this.
The difficulty in diagnosing tuberculosis in chil-
dren may translate into difficulty in treating it. Chil-
dren usually do not have cavitary disease and do not
expectorate. The inability to isolate M. tuberculosis
makes drug susceptibility testing impossible. The
antibiotics given should cover the organism from the
presumed source. Quinolones are not recommended
in children (see below: Pregnancy).
44.7
Pregnancy
Work up and treatment of tuberculosis should not be
delayed in pregnant women. All pregnant women with
active tuberculosis should be treated. The risk to the
fetus is much higher from maternal tuberculosis than it
is from potential harm from the anti-tuberculous drugs.
No drug has been approved by the FDA for use in preg-
nancy but there is experience with almost all of them.
Isoniazid, rifampin, and ethambutol are routinely given
in pregnancy. The duration of treatment should be for
nine months if pyrazinamide is not given. The lack of
literature on the use of pyrazinamide in pregnancy is
a major drawback and the reason why it has not been
recommended (ATS et al. 2003). Streptomycin was used
extensively in pregnancy as a lifesaving measure when
it was first available; it has been associated with con-
genital deafness (Conway and Birt 1965; Robinson and
Cambon 1964) and is no longer used in pregnancy. The
other aminoglycosides and capreomycin have a simi-
lar side-effect profile and are also not used. Although
there is a report that seven out of 23 mothers taking
ethionamide had children with congenital anomalies
(Potworowska et al. 1966), careful examination of the
cases show that the link is not convincing; two received
ethionamide only in the last month of pregnancy and
two had finished their course months before the onset
of pregnancy. Of the three who took it throughout the
pregnancy, two had Down's Syndrome and one had
an ill-defined abnormality. Several other small series
found no abnormality with ethionamide. PAS has been
considered safe in pregnancy (Hamadeh and Glassroth
1992; Heinonen et al. 1977; Scheinhorn and Angelillo
1977; Wilson et al. 1973).
The quinolones have been associated with cartilage
and joint defects in animals and have interdicted for
787
use in pregnancy (ATS et al. 2003), but several good
studies have failed to uncover fetal harm (Anonymous
1999; Berkovitch et al.1994; Larsen et al. 2001; Loebstein
et al. 1998; Schaefer et al. 1996). As with several areas
in tuberculosis treatment, when the experience is
incomplete it is necessary to weigh the potential risks,
benefits, and choices. The argument for their use in
pediatrics is similar to their use in pregnancy.
The management of latent tuberculosis in preg-
nancy depends on several factors. As with all medi-
cal treatments, the benefits must be weighed against
the risks. The risk for someone with latent disease
converting to active disease varies with comorbid
factors and timing of the exposure. For persons with
long-standing latent tuberculosis and no additional
risk factors, the risk of developing active tuberculo-
sis is about 10% in a lifetime or about 0.1-0.2% per
year. The risk of maternal serious side effects from
the drugs is probably about 1%. The risk to the fetus
is unknown but low because no study has shown the
anti-tuberculous drugs used cause any adverse effect
on the fetus. A report suggesting that an increased
risk of isoniazid toxicity during pregnancy based
on cross-sectional data (Franks et al. 1989) is not
convincing.
Treating latent tuberculosis is favored for those with
an increased immediate risk, as seen with recent con-
verters, and those with high risk such as HIY. Waiting
until after the pregnancy, or at least until after the first
trimester, is favored in those with no other risk factor.
For those with other risk factors, the risk-benefit con-
siderations lie somewhere in between. A problem with
postponing therapy is that many times this is the only
contact with the health care system for these young
women. If they are not treated, it is a missed opportu-
nity and active tuberculosis may be discovered when
the children develop disease.
Breast feeding can be permitted despite the passage
of medications into the milk (Tran and Montakanti-
kuI1998). The concentrations in milk are small and
not sufficient to treat the infant. Information on the
use of drugs such as pyrazinamide, ethionamide, and
capreomycin during breast feeding is limited (Tran
and MontakantikulI998).
44.8
Liver Disease
Liver disease in the treatment of tuberculosis can be
perplexing because both tuberculosis and its treat-
ment can cause and exacerbate liver disease. Detect-
788
ing a liver reaction early may be challenging, but if the
treatment is not promptly stopped, the side effect can
be fatal. A fatality from hepatotoxicity is particularly
tragic when treating a patient for latent tuberculosis.
Virtually all patients with drug-induced liver disease
have abnormal transaminase levels. Yet, enzyme levels
are a poor predictor of significant toxicity. It is best to
consider this as two problems: 1) patients with normal
livers who develop liver disease from the anti-tubercu-
lous medications and 2) the treatment of tuberculosis
in patients with pre-existing liver disease. A third
problem is tuberculosis of the liver that is generally
treated by the standard regimen.
44.8.1
Patients Who Start with Normal Livers
Probably all anti-tuberculous agents can cause hepa-
titis in normal individuals, although it is rare with
the aminoglycosides, capreomycin, and ethambutol.
The first question to ask is what constitutes liver
disease? The definitions of liver reactions range
from an elevation of transaminases to fatal fulmi-
nate hepatitis. There is no controversy about two
issues: 1. Anti-tuberculous drugs can cause fatal
hepatitis. 2. When patients without previous liver
disease develop symptoms of liver toxicity and have
elevated liver enzymes, the drugs must be stopped
immediately. Restarting the offending drug after an
episode of clinical liver disease carries a substantial
risk of recurrent hepatotoxicity.
On the other hand, asymptomatic liver function
test elevations often revert to normal despite continu-
ing the same drug regimen. In the absence of symp-
toms, the offending drug, especially if it is isoniazid,
can be restarted. About 20% of patients who take iso-
niazid develop serologic liver-test abnormalities and
only about 1% of these patients have serious hepati-
tis. Most tests with such low specificity would not be
considered useful clinically, but the occurrence and
gravity of the delay in diagnosis of liver reactions, the
subtlety of the symptoms, and the lack of better tests
are the basis of the recommendations for serological
testing. Serological testing may be beneficial in get-
ting patients back to the clinic regularly to talk with
a health care worker about possible side effects. Sero-
logical testing should not be a substitute for careful
instruction to the patient about tuberculosis and its
treatment and careful clinical follow-up.
After the initial visit all patients should be
warned specifically with words such as, "At the
first sign of symptoms of hepatitis, which are loss
D. E. Schraufnagel
of appetite, fatigue, nausea, vomiting, abdominal
pain, dark urine and jaundice, you should stop
the medicine and call this number. You should not
restart the medicine until told to do so by a health
care professional."
At every visit, the patients should be asked ques-
tions about these symptoms. If patients have symp-
toms, even non-specific fatigue or decreased appetite,
liver enzymes should be drawn. If the enzyme levels
are normal, it is unlikely that the symptoms result
from hepatotoxicity and the drugs can be contin-
ued. If the liver enzymes are abnormal, especially if
symptoms are more than trivial, generally all drugs
should be stopped. Generally, the physician should
wait until the symptoms have disappeared and the
liver function tests have returned to normal before
restarting any medicine. One should judge the activ-
ity and severity of the tuberculosis and the likelihood
that the different drugs may be involved.
If hepatotoxicity occurs with treatment of latent
tuberculosis, especially late in the course of therapy,
it may appropriate to simply stop the medication
permanently. If a patient is sick with tuberculosis, or
the liver disease is severe, switching to ethambutol,
streptomycin, and a quinolone may be a reasonable
choice.
Although liver enzymes are not as useful as one
would like for following asymptomatic patients, they
are good indicators of liver damage for the symptom-
atic patient. Many patients with isoniazid hepatotox-
icity have a relatively abrupt onset of symptoms (a
few days) and improve in about the same time span,
but the enzyme levels do not return to normal for
several weeks.
If one is uncertain of the offending agent, espe-
cially if the reaction has not been life threatening,
it is common practice to reintroduce the drugs
starting with the least likely to be the cause. On re-
introducing the drugs, a rise in liver enzyme levels
without symptoms is usually sufficient to perma-
nently stop the drug. In patients with symptomatic
reactions, re-introducing a known offending drug
is likely to cause recurrence of the liver reaction
and is not recommended. Re-introducing the drugs
should be followed by frequent open communica-
tion and measurement of the enzyme levels on a
weekly basis.
A major problem is that the symptoms of hepatitis
may be subtle. Fatigue or anorexia may be the only
symptom for days to weeks. The patient may not
understand the seriousness of informing the medical
team about the symptoms. The health care team may
not be situated to respond to the patient's call.
Treatment of Tuberculosis
44.8.2
Treating Patients with Liver Disease
The rifampicins, isoniazid, pyrazinamide, ethion-
amide, and PAS are all metabolized in the liver and
liver disease affects their metabolism. Not only does
the liver disease affect the drugs but the drugs affect
the measurements ofliver function and can cause liver
dysfunction. The drugs can alter the metabolism of
other drugs that can cause other side effects. Finally,
patients with chronic liver disease have fluctuations
in their liver function as determined by the serologic
tests, which makes evaluation more difficult.
With severe liver disease, beginning therapy with
ethambutol, streptomycin, and a quinolone is usually
recommended, although these agents are clearly not
as good as the first line drugs. I will not forget treat-
ing a young woman with a liver transplantation, liver
failure, and tuberculosis with these drugs in our inten-
sive care unit for about six weeks. She died from and
had extensive disseminated tuberculosis at autopsy.
The extensive disease is almost never seen in patients
treated with rifampin, isoniazid, and pyrazinamide for
this time. As the liver failure improves, it is usually pos-
sible to add isoniazid and rifampin. This may be done
sequentially while monitoring the liver enzymes.
A more common problem is how to treat tubercu-
losis in an alcoholic or patient with viral hepatitis. If
the liver dysfunction is mild, the standard regimen
is usually tolerated. If the patient is symptomatic or
the liver enzymes are elevated, the first consideration
is to withhold pyrazinamide, which appears to be
the most hepatotoxic (Liu et al. 1991). If the treat-
ment does not include pyrazinamide, the rifampin
and isoniazid should be continued for nine months.
Adding or beginning with ethambutol, streptomycin,
and a quinolone if the mycobacteriological burden is
high, may also be helpful in moderate liver disease.
For patients with liver disease, transaminase levels
are used to follow the liver disease, but again there
are caveats, and serologic testing does not replace
close clinical follow-up. One alcoholic cirrhotic had a
tripling of his already abnormal enzymes for which I
was about to stop the drugs, when a medical resident
pointed out to me that the enzyme levels were sub-
stantially higher than the current level on other occa-
sions in the past, before the treatment was begun. We
continued the patient on the same course and his
liver enzyme levels reduced at the next clinic visit.
Elevated liver enzymes in patients with chronic
liver disease does not have the same meaning as those
with a normal liver at the start of therapy. For example,
if the transaminases rise 3-5 fold after starting treat-
789
ment for tuberculosis, most authorities recommend
stopping treatment, but many experts in tuberculosis
would begin treatment tuberculosis with isoniazid
and rifampin in a cirrhotic whose transaminases were
3-5 times normal. In the former, the assumption is that
the anti-tuberculous drugs are causing the problem;
in the latter it is that the patient's liver disease was the
cause. The reason for serological liver function testing
before beginning tuberculosis medications is to detect
occult pre-existing liver disease.
Liver transplantation and tuberculosis are an
increasing problem. The immunosuppression that
goes with liver transplantation substantially increases
the risk of developing active tuberculosis and cannot
be ignored. Liver failure, rejection and drug inter-
actions can make treatment difficult (Meyers et al.
2000). A reasonable approach is to test all transplant
candidates with tuberculin but to withhold treatment
for latent disease until the new liver is in place and
functioning well, usually about three months after
transplantation. The patients then should be treated
with isoniazid because rifampin interacts with the
immunosuppressive agents.
On the other end of the spectrum is Gilbert's
syndrome, which can be treated with the standard
four-drug regimen. Isoniazid and rifampin metabo-
lism appears to be unchanged (Adachi et al. 1975;
Pereira-Filho et al. 1985).
44.9
Renal Failure
Treating tuberculosis in patients with renal failure
can be a problem because the drug dosages are not
well established. For patients on hemodialysis, all
drugs should be given by directly observed therapy
immediately after dialysis. Thrice-weekly intermit-
tent therapy fits well with dialysis and medical per-
sonnel can administer the medications after dialysis
with ease. Dosing medication after dialysis avoids the
early inadvertent reduction of the levels by dialysis.
Ethionamide, clofazimine, rifampin and PAS are not
significantly dialyzed (Malone et al. 1999).
Treating patients with renal failure who are not on
dialysis is more difficult. The rifamycins can be given
at their usual doses. Isoniazid is generally given at its
usual dose, but its metabolism is changed and clear-
ance is decreased (Bowersox et al. 1973; Cheung et al.
1993; Gold et a1.1976; Kim et al. 1993). Pyrazinamide
is metabolized by the liver but its main metabolite,
pyrazinoic acid, is excreted by the kidneys. Etham-
790
butol can become dangerous in renal failure. Con-
sideration should be given to avoiding these agents
or giving them twice or thrice weekly at their daily
dose. Cycloserine and the aminoglycosides should
probably only used if no other agents are available
and blood levels are monitored. Regular checking of
vision in patients on ethambutol is important.
44.10
Individual Drugs
44.10.1
Isoniazid
Isoniazid has been the most used anti-tuberculous
drug through the years. It has excellent bactericidal
activity that is largely limited to M. tuberculosis and to
a lesser extent to other members of the mycobacterial
family. It appears to interfere with the mycolic acid
synthesis (Winder and Collins 1970). Resistance to
isoniazid was associated with less virulence in vitro and
decreased catalase activity (Cohn et al. 1954). This has
been attributed to mutations of the KatG gene, which
encodes a catalase-peroxidase required to activate iso-
niazid (Zhang et al. 1992). Mutations of the regulatory
region of the ahpC gene cause overproduction of alkyl
hydroperoxide reductase but do not confer virulence
(Heym et al.1993). Mutations in these two genes appear
to account for most of the isoniazid resistance.
The daily dose is 5 mg/kg for adults and 10-15 mgt
kg for children to 300 mg daily. For intermittent ther-
apy, 15 mg/kg up to 900 mg is given twice or thrice
weekly. See Table 44.1 for dosage. Suspension and
parenteral preparations are available.
The major problems with isoniazid are asymp-
tomatic elevation of liver enzymes and potentially
fatal hepatitis, which have been discussed, and addi-
tionally, rash, neuropathy, lupus reaction, fever, and a
variety of minor symptoms also can occur.
The rash, in my experience, usually has a small
macular or acneiform pattern on the chest, abdo-
men, and arms. It may be pruritic and may respond to
diphenhydramine and calamine lotion and diphen-
hydramine by mouth. The rash may abate despite
continuation of the medication, however, with attrac-
tive alternatives to isoniazid for the treatment of
latent tuberculosis, a rash should generally prompt
a change in medication. Stopping and restarting the
drug may not be associated with return of the rash
and the rash is not necessarily a contraindication to
using or restarting isoniazid.
D. E. Schraufnagel
Isoniazid neuropathy usually occurs in patients
with underlying nerve disease, diabetes, pregnancy,
alcoholism, and poor nutrition, but is occasionally
seen in otherwise healthy patients. The symptoms
are often distal paresthesia. The usual treatment,
when isoniazid must be continued, is pyridoxine
or increasing the pyridoxine if it is already being
given. There is no trial that shows the routine use
of pyridoxine is indicated in patients receiving iso-
niazid, but it is recommended for those at higher
risk to develop neural symptoms. It is given in doses
between 10 and 50 mg daily. With the onset of symp-
toms of neuropathy, 100-200 mg are usually given. If
the symptoms do not respond to the pyridoxine, the
isoniazid should be stopped.
Isoniazid-associated lupus occurs infrequently. It
usually presents with arthralgias. The serum anti-
nuclear antibodies are positive and the syndrome
usually progresses until the medication is stopped.
The lupus reaction may be preceded by a rash.
Isoniazid associated fever may be characteristic.
The fever may begin about six hours after ingestion of
the isoniazid, rise to a high level, be accompanied by a
chill, and last only a short time. Hours later the patient
may feel completely normal. The fever may return the
next day with clock-like regularity but it becomes
more intense and lasts longer if the medicine is contin-
ued. With each day, the fevers become more hectic, last
longer, and the timing breaks down. In other patients,
the course may be less characteristic but it is usually
debilitating and require that the medicine be stopped.
The fever usually recurs if the drug is restarted. The
fever may be preceded by a rash.
If patients are educated and told to report any
abnormal sensation, it is common to get mild and
unexpected complaints. Frequent ones include flush-
ing especially with alcohol, decreased energy, sleepi-
ness, headache, light-headedness, and an inability to
concentrate. Most often they respond to reassurance
and go away without changing the medication.
Isoniazid interacts with phenytoin and carbam-
azepine. The phenytoin concentrations predictably
go up and the patient becomes phenytoin toxic. The
phenytoin dose should be reduced and the serum
concentrations of it monitored.
The question of whether isoniazid should be used
despite low level resistance has never been solved
(Victor et al. 1997). Two types of drug resistance are
commonly seen with isoniazid. One is complete resis-
tance at all concentrations and the other is resistance
at low levels only. The latter is associated with resis-
tance to ethionamide. Older studies indicated that
isoniazid was not effective if any resistance was pres-
Treatment of Tuberculosis
ent (Bignall et al. 1969; Stewart and Crofton 1964).
The high-level resistance may be associated with
katG gene mutations, which are also associated with
lower virulence (Heym et al. 1995). The inhA gene
isolated from M. smegmatis was thought at first to be
associated with the low-level resistance but the effect
of the gene is less clear in M. tuberculosis (Mdluli et
al.1996). Isoniazid is not given when drug sensitivity
testing indicate complete resistance.
44.10.2
Rifampin
Rifampin may be the most important drug for the
treatment of all forms of tuberculosis. Short course
chemotherapy can not be accomplished without it.
The time to the reduction of almost all symptoms and
signs, radiographic improvement, and sputum nega-
tivity are lower when rifampin is used in tubercu-
losis and other mycobacterial disease (Schraufnagel
et al. 1984).
Rifampin inhibits a DNA-dependent RNA poly-
merase. This enzyme appears to be a complex oligo-
mer that is made up of different subunits encoded by
the rpo genes. Most rifampin resistance comes from
modification in the rpoB genes (Telenti et al. 1993).
Rifampin is usually given at a dose of 10 mglkg in
adults and 10-20 mglkg in children up to 600 mg daily.
See Table 44.1 for dosage. It is an integral part of inter-
mittent therapy but is given at the same dose twice or
thrice weekly. It can be given orally or parenterally.
The main problems with rifampin include the
expected reactions, such as the orange discoloration
of body fluids and drug interactions. Nausea, rash,
flu-like symptoms, and liver, renal, and hematologic
toxicity are uncommon.
Patients should be warned that rifampin and its
metabolites are red-orange and the urine will turn
that color as the drug is excreted. Oral contracep-
tives may no longer be effective after rifampin is
started. Antiretrovirals, cyclosporine, tacrolimus and
warfarin are some commonly used drugs that may
have markedly changed serum levels after rifampin
is started.
Rifampin may cause intractable nausea that resolves
only with stopping the medication. A flu-like syn-
drome can occur with intermittent therapy and is the
main reason why rifampin cannot easily be given once
weekly. The symptoms are sometimes subtle and can
also be seen in twice-weekly therapy. For self-adminis-
tered therapy, these symptoms suggest that the patient
is taking the medicine irregularly.
791
The hepatitis of rifampin may be more subtle and
slower to develop than that resulting from isoniazid,
but there is overlap and it is not easy to establish
which drug causes the hepatic reaction without stop-
ping and reintroducing them. Rifampin may cause an
isolated elevation of serum bilirubin.
As with isoniazid, the rash may be self-limited, but
switching to another regimen is recommended for
treatment of latent disease with other good alterna-
tives. Renal and hematological toxicity are uncommon
but can be fatal and are contraindications to restarting
the medication. Thrombocytopenia, hemolytic anemia,
and renal failure may occur separately or together.
44.10.3
Pyrazinamide
Pyrazinamide is the number three drug for the treat-
ment of tuberculosis. Its antituberculous activity was
not recognized until the 1950s and it was considered to
be too hepatotoxic at its earlier doses of up to 50 mgl
kg. The British Thoracic Society trials in the 1970s
established its place among the best agents for tuber-
culosis. Many studies have shown that it decreases the
time of culture positivity and allows a shorter course
of treatment. Reducing the dose to 25 mglkg allowed
its common use. However, it is still associated with
substantial hepatotoxicity and lower doses are being
studied. (See Table 44.1 for dosage.)
Pyrazinamide is interesting in that it only works
at low pH. Pyrazinamide may be converted to pyrazi-
noic acid in the secondary lysosome. In this location
it lowers the pH, which increases the killing of the
mycobacteria harbored there. The pyrazinoic acid
may be trapped in the macrophage and its accu-
mulation may add to its toxicity for the mycobac-
teria (Salfinger et al. 1990). However, sensitivity to
pyrazinamide is not correlated with pyrazinamidase
which metabolizes pyrazinamide to pyrazinoic acid.
Most pyrazinamide resistance occurs with mutated
pncA genes (Sreevatsan et al. 1997a).
The drug is now used in children and can be used
for intermittent therapy. The drug is usually given for
the first two months of therapy when the mycobacte-
rial burden is greatest.
In addition to hepatotoxicity, nausea, anorexia, and
gastric intolerance are common. The main metabolite
of pyrazinamide is pyrazinoic acid which competes
with uric acid metabolism and causes a rise in serum
uric acid. In persons with gout, the pyrazinamide
may trigger an attack. In the absence of pre-existing
gout, the arthralgias are different from gout and do
792
not respond to the treatment of gout. The symptoms
usually occur after about 6-8 weeks of therapy and
are usually mild. Fortunately, therapy with pyrazin-
amide usually goes for only 8 weeks, so patients can
usually be persuaded to continue therapy for the last
few days. Aspirin may give symptomatic relief.
Gastrointestinal symptoms can be a major deterrent
to taking pyrazinamide. The first approach to nausea
from any anti-tuberculous medications is to take them
with food, or before bed, and to split the dose.
44.10.4
Ethambutol
Ethambutol is the fourth major drug for the treatment
of tuberculosis. It can be given for all forms of tuber-
culosis and may be synergistic with most medications.
Soon after it was discovered ethambutol was shown
to decrease the development of resistance to isoniazid
and streptomycin. It is the least likely of the four major
drugs to cause side effects and this favorable side effect
profile also adds to its value. The dose is usually 15-
25 mg/kg daily. The dose for intermittent therapy is 25
to 50 mg/kg. (See Table 44.1 for dosage.)
Ethambutol inhibits mycobacterial arabinosyl
transferase and the synthesis of arabinogalactan, a
component of the polymer core of the mycobacterial
cell wall (Takayama and Kilburn 1989). Etharnbutol-
resistant mycobacteria have variant arabinosyl trans-
ferases and truncated lipoarabinomannans (Khoo et al.
1996). Most ethambutol resistance results from altera-
tion of embAB gene products (Belanger et al. 1996b;
Sreevatsan et al. 1997b) that cause an overproduction
of arabinosyl transferase (Belanger et al.1996a).
A rare but worrisome side effect of ethambutol is
optic neuritis that can cause blindness. It generally
occurs with doses higher than 15 mg/kg, or with renal
failure. Occasionally the renal failure is unknown or
develops after therapy is started. Patients at increased
risk for this side effect should have their visual acuity
and color vision tested. The 25 mg/kg dosage has not
been shown to be more effective than the 15 mg/kg. The
drug must be stopped and not restarted if visual toxicity
occurs. Rash and peripheral neuropathy are also rare.
44.10.5
Streptomycin and Other Aminoglycosides
Although streptomycin is still considered the fifth
drug for tuberculosis, it is being supplanted by the
quinolones. It was the first effective drug to be used
D. E. Schraufnagel
against tuberculosis so there is extensive experience
with it. It is effective but must be given intramuscularly
or intravenously. Giving it intravenously reduces the
muscular discomfort and gives a higher peak level. Its
broad use throughout the world has resulted in many
areas developing strains resistant to it.
The aminoglycosides penetrate the inner and
outer membranes probably through porin channels
and exert their influence by binding to bacterial
ribosomes. This reduces bacterial protein synthesis
and induces misreading of mRNA. There appears to
be several mechanisms of streptomycin resistance
(Douglass and Steyn 1993), but most arise from
altered genes, e.g. rpsL, which encodes the ribosomal
protein S12, and rrs, a 16S rRNA gene (Finken et al.
1993; Nair et al. 1993). These stabilize the structure
formed by the 16S rRNA and confer resistance.
Resistance is not uniform among the aminoglyco-
sides and resistance to one member of the family does
not necessarily mean that the organism is resistance
to another. Caution should be used in renal failure,
although giving after dialysis works satisfactorily.
The recommended dose of streptomycin is 15 mg/
kg up to 0.5 g to 1 g daily or intermittently. The higher
dose daily therapy is often restricted to the first 1-2
weeks when the mycobacterial burden is greatest.
Beginning with or making an early shift to intermit-
tent therapy reduces toxicity and may allow its use
for a longer duration. A lower dose should be used for
older patients especially if the companion drugs are
good and the mycobacterial burden is not high.
Its main problems are patients' non-acceptance
because of the painful injections and the eighth
nerve toxicity and to a lesser extent nephrotoxicity.
Other common problems include rash, eosinophilia,
and neurotoxicity.
Amikacin and kanamycin are more toxic and less
effective and should be used only for tuberculosis
caused by organisms that are resistant to the first line
anti-tuberculous drugs and sensitive to these drugs.
However, these agents are less commonly used and
are likely to have less resistance than streptomycin.
All patients on any aminoglycoside must be followed
closely with regular hearing tests. A report of bal-
ance disturbance should prompt pause or cessation
of therapy.
44.10.6
Rifabutin
Rifabutin is the second member of the rifamycin
family. It was discovered in about 1980 (Sanfilippo et
Treatment of Tuberculosis
al. 1980) and entered the United States formulary for
the treatment of M. avium in 1993. Most human stud-
ies show that it is equal to, or better, than rifampin
for tuberculosis (Chan et al. 1992; Grassi and Peona
1996; Hong Kong Chest Service/British Medical
Research Council 1992; McGregor et al. 1996). In a
Ugandan trial of patients with tuberculosis and AIDS,
patients on rifabutin had better sputum conversion at
two months than those in the rifampin group (81 %
compared to 48%) (Schwander et al. 1995). Rifabu-
tin is usually not effective against tuberculosis that
is resistant to rifampin. The side-effects of rifabutin
are different from rifampin and patients intolerant
to rifampin may take rifabutin or rifapentine. Uve-
itis (Tseng and Walmsley 1995) and leukopenia are
common with rifabutin. Patients on rifabutin who
develop new fever must have the medicine stopped at
once and have an urgent complete blood count. This
may be accompanied by gastrointestinal symptoms.
The symptoms usually respond within a few days to
withdrawal of the medication. I do not re-challenge
patients with this reaction to rifabutin.
The rifamycins are effective against M. avium
(Nightingale et al. 1993; Sullam et al. 1994) and they
should be used if it is unclear whether a patient has
disease from an environmental mycobacteria rather
than tuberculosis. Rifabutin is used to prevent myco-
bacteriosis avium in patients with AIDS (Gordin and
Masur 1994) and may be used for treatment of latent
tuberculosis (McGregor et al. 1996).
As discussed, the rifamycins enhance the metabo-
lism of the protease inhibitors, cyclosporine and
many other drugs. The drugs that interact with
the rifamycins must be carefully selected and the
patients must be monitored for side effects (Center
for Disease Control and Prevention 2000). Rifabutin
has a lower enzyme-inducing effect than rifampin.
The mechanism of action and resistance for the
rifamycins is similar but the dosing is different. The
standard dose of rifabutin is 300 mg daily but using
half of this has been successful (Gonzalez Montane
et al. 1994).
44.10.7
Rifapentine
Rifapentine is a newer rifamycin with a long half-life
of about 14 hours, which makes it a candidate for
twice and once weekly therapy. Trials have tested the
once weekly dosing after a two-month daily course
of the standard 4 drugs, isoniazid, rifampin, pyrazin-
amide and ethambutol. A Chinese study found daily
793
and twice weekly rifampin comparable to twice and
once weekly rifapentine along with isoniazid, eth-
ambutol, and pyrazinamide for eight months (Biya
et al. 1992).
A Hong Kong study found that the continuation
rifampin and isoniazid thrice weekly was better than
rifapentine and isoniazid once weekly, but the failure
of the rifapentine group may have been because of
decreased bioavailability of this compound that was
not manufactured by the original developer (Tam et
al.I997).A US Public Health Service study found con-
tinuation rifapentine and isoniazid once weekly fell
short of rifampin and isoniazid twice weekly for four
months in patients with AIDS. It is recognized that
spacing isoniazid to once weekly may make the regi-
men ineffective even though rifapentine is effective
for this time span. Coupling rifapentine with another
effective long-acting agent might improve its efficacy.
Rifapentine might be ideal for the treatment of latent
tuberculosis (Chapuis et al.1994).
44.10.8
Fluoroquinolones
The quinolone family has good activity against
mycobacteria. Sparfloxacin and levofloxacin appear
to be superior to the older agents, ciprofloxacin and
ofloxacin, in vitro (Berlin et al. 1987). Levofloxacin
is less likely to cause phototoxicity than sparfloxacin
but is slightly less effective in vitro and in mice Oi et
al. 1995; Lounis et al. 1997; Rastogi and Goh 1991).
Gatifloxacin (Alvirez-Freites et al. 2002), grepafloxa-
cin (Mizutani 1996) and moxifloxacin (Gillespie and
Billington 1999; Ii et al. 1998; Miyazaki et al. 1999;
Yoshimatsu et al. 2002) also have good antitubercu-
lous activity.
The quinolones act by interfering with a variety
of DNA-related processes primarily through their
actions on DNA gyrase, the enzyme that promotes
supercoiling in bacterial DNA. Many activities of
DNA including replication, transcription, and trans-
position are sensitive to supercoiling. DNA gyrase is
a topoisomerase composed of two A and two B sub-
units that are encoded by gyrA and gyrB. Mutations
of these genes has been shown to confer resistance in
many bacterial species (Musser 1995).
Although the data is limited, it appears that the
quinolones are at least equal to the second line anti-
tuberculous drugs and have much fewer side-effects
(Kennedy et al. 1993; Suo et al.I996).
794
44.10.9
Capreomycin
Capreomycin is a polypeptide antibiotic isolated
from s. capreolus that is made up of four microbio-
logically active components. It is closely allied with
the aminoglycosides in that it localized to ribosomes
(Yamada et al. 1976) and can have cross resistance
with kanamycin. Its dosage and side effects are simi-
lar to the aminoglycosides.
Capreomycin should be used only for tuberculosis
that is resistant to the first-line antituberculous drugs
and known to be sensitive to it. It usually is not cross
resistant with streptomycin.
44.10.10
Ethionamide
Ethionamide is a thioamide derivative of isonicotinic
acid and an analog of isoniazid. Similar to isoniazid,
it inhibits the mycolic acid biosynthetic pathway
(Schroeder et al. 2002). It is an effective agent against
tuberculosis (Lees 1963), but its side effects limit is
usefulness.
Ethionamide is metabolized to an active product
that is similar in structure to that formed by the
activation of isoniazid by the catalase-peroxidase
KatG. Overproduction of Rv3855 (EtaR), a regulatory
protein from M. tuberculosis confers resistance to
ethionamide (DeBarber et al. 2000).
Ethionamide is commonly associated with nausea,
vomiting, and weight loss. This may be profound.
It is common and dose related. It is generally used
against mycobacteria with resistance to the first-line
anti-tuberculous drugs and sensitivity to it. The dose
is usually 500-750 mg per day (15-20 mg/kg). There
may be cross resistance with isoniazid.
44.10.11
Cycloserine
Cycloserine is a structural analog of D-alanine and
exerts in influence by competitively blocking the
enzymes alanine racemase and D-alanyl-D-alanine
synthetase (Aranda 1996). The dipeptide D-alanyl-D-
alanine is essential for the synthesis of the mycobac-
terial wall. It is relatively weak agent with significant
side effects. Its problems include psychosis, which
can be subtle or severe. Paranoia, depression, delu-
sions, and seizures are common. Blood levels may be
helpful in avoiding side effects. The peak should be
D. E. Schraufnagel
between 20 and 35 f.Lglml (ATS et al. 2003). Pyridox-
ine (100-200 mg) should be given to reduce neural
toxicity. If pyridoxine is given at a dose greater than
300 mg it may reduce the potency of the drug. Serum
phenytoin level should be monitored.
Cycloserine should only be used against mycobac-
teria that are resistant to first-line drugs and sensitive
to it. It may be of value for patients with hepatitis
because like the ethambutol, arninoglycosides and
quinolones, it can be used in hepatic failure. The usual
dose is 250 mg twice or thrice daily (10-15 mglkg).
44.10.12
Para-aminosalicylic Acid (PAS)
PAS is a structural analog of para-aminobenzoic acid.
Its bacteriostatic affect on M. tuberculosis results from
its ability to block the conversion of para-aminoben-
zoic acid to folic acid. Folic acid, which is essential to
purine and DNA synthesis, is made in bacterial cells
from para-aminobenzoic acid (Aranda 1996).
PAS is a drug of great historical significance in
that it allowed the cure of most cases of tuberculosis
by being the second effective agent to be discovered
(Lehmann 1946). Now, it should be used only for
treating drug-resistant tuberculosis. The usual dose
is 8-12 g daily in divided doses. It comes as a granular
form in 4 g packets that is usually mixed with juice.
The unpleasant taste and nausea have caused at least
one of my patients to vow never to drink orange juice
again. Other side effects include bloating, anorexia,
vomiting, diarrhea, weight loss, fever, rash, fluid
retention, arthralgia, and eosinophilia with radio-
graphic lung shadows.
44.10.13
Linezolid
Linezolid is an oxazolidinone, a class of antibiotics
that inhibit early bacterial protein synthesis. It has
good anti-tuberculous activity in several models
(Barbachyn et al.1996; Cynamon et al.1999a,b). I have
used linezolid in a patient with multi-drug resistant
tuberculosis and multi-drug intolerance and found
it effective in eliminating his tuberculosis despite an
inadequate intake of his all other agents. The problem
with linezolid is its effect on peripheral blood cells.
The agent must be stopped if unexplained leukope-
nia, thrombocytopenia, or anemia develops. This side
effect has occurred with longer therapy, which makes
its use for tuberculosis problematic. This drug should
Treatment of Tuberculosis
only be used against tuberculous organisms that are
resistance to first and second-line anti-tuberculous
drugs and the bacilli are known to be sensitive to it.
Blood cell counts must be monitored carefully and the
drug should not be given for a prolonged period. The
dose is usually 400-600 mg twice daily.
44.10.14
Other Treatments
Many new drugs are on the horizon (Schraufnagel
1999), but until more information is available they
should be tried only when first and second line drugs
have been exhausted and sensitivity in vitro has been
demonstrated. Other approaches, such as immuno-
modulation, hold hope but are still insufficiently
studied to be recommended.
References
Adachi Y, Wakisaka G, Yamamoto T (1975) Salycylamide gluc-
uronide formation in liver disease and its change by drugs.
Gastroenterol Jpn 10:120-131
Afghani B, Lieberman JM (1994) Paradoxical enlargement or
development of intracranial tuberculomas during therapy:
case report and review. Clin Infect Dis 19:1092-1099
Alarcon F, Escalante L, Perez Y, Banda H, Chacon G, Duenas
G (1990) Tuberculous meningitis. Short course of chemo-
therapy. Arch NeuroI47:1313-1317
Alvirez-Freites EJ, Carter JL, Cynamon MH (2002) In vitro and
in vivo activities of gatifloxacin against Mycobacterium
tuberculosis. Antimicrob Agents Chernother 46:1022-1025
Alzeer AH, FitzGerald JM (1993) Corticosteroids and tuber-
culosis: risks and use as adjunct therapy. Tuberc Lung Dis
74:6-11
American Academy of Pediatrics (2000) Tuberculosis. In: Peter
G (ed) Red book: report of the committee on infectious
diseases. American Academy of Pediatrics, Elk Grove Vil-
lage,IL
American Thoracic Society, Center for Disease Control and
Prevention (2000) Targeted tuberculin testing and treat-
ment of latent tuberculosis infection. Am J Respir Crit
Care Med 161:S221-S247
American Thoracic Society, Center for Disease Control and
Prevention, Infectious Disease Society of America (2003)
Treatment of Tuberculosis. Am J Respir Crit Care Med
167:603-662
American Thoracic Society, Center for Disease Control and
Prevention (2001) Update: fatal and severe liver injuries
associated with rifampin and pyrazinamide for latent
tuberculosis infection, and revisions in American Tho-
racic Society/CDC recommendations - United States, 2001.
MMWR Morb Mortal Wkly Rep 50:733-735
Anonymous (1999) Quinolones and pregnancy: worrying
animal findings, few clinical data. Prescrire Int 8:29-31
Aquinas M (1978) Essai controle de regimes chimiothera-
795
piques de 6 mois et de 8 mois dans Ie traitement de la
tuberculose pulmonaire. Une etude des Hong Kong Chest
Service/British Medical Research Council. Bull Int Union
Tuberc 53:253
Aranda CP (1996) Second-line agents: p-aminosalicylic acid,
ethionamide, cycloserine, and thiacetazone. In: Rom WN,
Garay S (eds) Tuberculosis. Little Brown, Boston, pp
811-816
Barbachyn MR, Hutchinson DK, Brickner SJ, Cynamon MH,
Kilburn JO, Klemens SP, Glickman SE, Grega KC, Hendges
SK, Toops DS, Ford CW, Zurenko GE (1996) Identification
of a novel oxazolidinone (U-I00480) with potent antimy-
cobacterial activity. J Med Chern 39:680-685
Bass JB Jr, Farer LS, Hopewell PC, O'Brien R, Jacobs RF, Ruben
F, Snider DE Jr, Thornton G (1994) Treatment of tuber-
culosis and tuberculosis infection in adults and children.
American Thoracic Society and The Centers for Disease
Control and Prevention. Am J Respir Crit Care Med 149:
1359-1374
Bayer R, Wilkinson D (1995) Directly observed therapy for
tuberculosis: history of an idea. Lancet 345:1545-1548
Belanger AE, Besra GS, Ford ME, Mikusova K, Belisle JT, Bren-
nan PJ, Inamine JM (1996a) The embAB genes of Mycobac-
terium avium encode an arabinosyl transferase involved
in cell wall arabinan biosynthesis that is the target for the
antirnycobacterial drug ethambutol. Proc Natl Acad Sci
USA 93:11919-11924
Belanger AE, Besra GS, Ford ME, Mikusova K, Belisle JT, Bren-
nan PJ, Inamine JM (1996b) The embAB genes of Mycobac-
terium avium encode an arabinosyl transferase involved
in cell wall arabinan biosynthesis that is the target for the
antirnycobacterial drug ethambutol. Proc Natl Acad Sci
USA 93:11919-11924
Berkovitch M, Pastuszak A, Gazarian M, Lewis M, Koren G
(1994) Safety of the new quinolones in pregnancy. Obstet
GynecoI84:535-538
Berlin OG, Young LS, Bruckner DA (1987) In vitro activity of
six fluorinated quinolones against Mycobacterium tubercu-
losis. J Antimicrob Chemother 19:611-615
Biddulph J (1990) Short course chemotherapy for childhood
tuberculosis. Pediatr Infect Dis J 9:794-801
Bignall JR, Lind A, Rist N (1969) A comparison of regimens
of ethionamide, pyrazinamide and cycloserine in re-treat-
ment of patients with pulmonary tuberculosis. Bull Int
Union Tuberc 42:7-57
Biya Y, Lizhen Z, Fengying A, National Cooperative Group
on the Clinical Study of Rifapentine (1992) A clinical
comparative study of Chinese-made rifapentine on che-
motherapy of tuberculosis (followup results after 3 years).
Chin J Antibiot 7:313-322
Bowersox DW, Wintebauer RH, Stewart GL, Orme B, Barron
E (1973) Isoniazid dosage in patients with renal failure. N
Engl J Med 289:84-87
British Thoracic Society Research Committee (1992) Six-
months versus nine-months chemotherapy for tubercu-
losis of lymph nodes: preliminary results. Respir Med 86:
15-19
Center for Disease Control and Prevention (2000) Updated
guidelines for the use of rifabutin or rifampin for the
treatment and prevention of tuberculosis among HIV-
infected patients taking protease inhibitors or nonnucleo-
side reverse transcriptase inhibitors. MMWR Morb Mortal
Wkly Rep 49:185-189
796
Chan HS, Pang JA (1989) Effect of corticosteroids on dete-
rioration of endobronchial tuberculosis during chemo-
therapy. Chest 96:1195-1196
Chan SL, Yew WW, Ma WK, Girling DJ, Aber VR, Felmingham
D, Allen BW, Mitchison DA (1992) The early bactericidal
activity of rifabutin measured by sputum viable counts in
Hong Kong patients with pulmonary tuberculosis. Tuberc
Lung Dis 73:33-38
Chan SL, Wong PC, Tam CM (1994) 4-, 5- and 6-month regi-
mens containing isoniazid, rifampicin, pyrazinamide and
streptomycin for treatment of pulmonary tuberculosis
under program conditions in Hong Kong. Tuberc Lung
Dis 75:245-250
Chapuis L, Ji B, Truffot-Pernot C, O'Brien RJ, Raviglione MC,
Grosset JH (1994) Preventive therapy of tuberculosis with
rifapentine in immunocompetent and nude mice. Am J
Respir Crit Care Med 150:1355-1362
Cheung WC, Lo CY, Lo WK, Ip M, Cheng IK (1993) Isoniazid
induced encephalopathy in dialysis patients. Tuberc Lung
Dis 74:136-139
Chotmongkol V (1991) Treatment of tuberculous meningitis
with 6-month course of chemotherapy. Southeast Asian J
Trop Med Public Health 22:372-374
Cohn DL, Catlin BJ, Peterson KL, Judson FN, Sbarbaro JA
(1990) A 62-dose, 6-month therapy for pulmonary and
extrapulmonary tuberculosis. A twice-weekly, directly
observed, and cost- effective regimen. Ann Intern Med
112:407-415
Cohn ML, Kovitz C, Oda U, Middlebrook G (1954) Studies on
isoniazid and tubercle bacilli. II. The growth requirements,
catalase activities, and pathogenic properties of isoniazid
resistent mutants. Am Rev Tuberc 70:641-664
Comstock GW, Ferebee SH, Hammes LM (1967) A controlled
trial of community-wide isoniazid prophylaxis in Alaska.
Am Rev Respir Dis 95:935-943
Conway N, Birt BD (1965) Streptomycin in pregnancy: effect
on the foetal ear. BMJ 2:260-263
Cowie RL, Langton ME, Escreet BC (1985) Diagnosis of
sputum smear- and sputum culture-negative pulmonary
tuberculosis. S Afr Med J 68:878
Cynamon MH, Klemens SP, Sharpe CA, Chase S (1999a)
Activities of several novel oxazolidinones against Myco-
bacterium tuberculosis in a murine model. Antimicrob
Agents Chemother 43:1189-1191
Cynamon MH, Klemens SP, Sharpe CA, Chase S (1999b)
Activities of several novel oxazolidinones against Myco-
bacterium tuberculosis in a murine model. Antimicrob
Agents Chemother 43:1189-1191
Dastur DK, Manghani DK, Udani PM (1995) Pathology and
pathogenetic mechanisms in neurotuberculosis. Radiol
Clin North Am 33:733-752
David HL (1970) Probability distribution of drug-resistant
mutants in unselected populations of Mycobacterium
tuberculosis. Appl MicrobioI20:810-814
DeBarber AE, Mdluli K, Bosman M, Bekker LG, Barry CE III
(2000) Ethionamide activation and sensitivity in mul-
tidrug-resistant Mycobacterium tuberculosis. Proc Natl
Acad Sci USA 97:9677-9682
Di Perri G, Cruciani M, Danzi MC, Luzzati R, de Checchi G,
Malena M, Pizzighella S, Mazzi R, Solbiati M, Concia E
(1989) Nosocomial epidemic of active tuberculosis among
HIV-infected patients. Lancet 2:1502-1504
Doganay M, Calangu S, Turgut H, Bakir M, Aygen B (1995)
D. E. Schraufnagel
Treatment of tuberculous meningitis in Turkey. Scand J
Infect Dis 27:135-138
Donald PR, Schoeman JF, Van Zyl LE, de Villiers IN, Pretorius
M, Springer P (1998) Intensive short course chemotherapy
in the management of tuberculous meningitis. Int J Tuberc
Lung Dis 2:704-711
Dooley DP, Carpenter JL,Rademacher S (1997) Adjunctive cor-
ticosteroid therapy for tuberculosis: a critical reappraisal
of the literature. Clin Infect Dis 25:872-887
Douglass J, Steyn LM (1993) A ribosomal gene mutation in
streptomycin-resistant Mycobacterium tuberculosis iso-
lates. J Infect Dis 167:1505-1506
Dutt AK, Stead WW (1994) Smear-negative pulmonary tuber-
culosis. Semin Respir Infect 9:113-119
Finken M, Kirschner P, Meier A, Wrede A, Bottger EC (1993)
Molecular basis of streptomycin resistance in Mycobacte-
rium tuberculosis: alterations of the ribosomal S12 gene
and point mutation within a functional 16S ribosomal
RNA pseudoknot. Mol MicrobioI9:1239-1246
Franks AL, Binkin NJ, Snider DE Jr, Rokaw WM, Becker S
(1989) Isoniazid hepatitis among pregnant and postpar-
tum Hispanic patients. Publ Health Rep 104:151-155
Galarza I, Canete C, Granados A, Estopa R, Manresa F (1995)
Randomised trial of corticosteroids in the treatment of
tuberculous pleurisy. Thorax 50: 1305-1307
Garg RK (1999) Tuberculosis of the central nervous system.
Postgrad Med J 75:133-140
Gavazzi G, Bouchard 0, Bosseray A, Leclerq P, Micoud M
(1998) Miliaire tuberculeuse cerebrale et pulmonaire chez
un patient immunocompetent: aggravation sous traite-
ment adapte. Presse Med 27:958-962
Gillespie SH, Billington 0 (1999) Activity of moxifioxa-
cin against mycobacteria. J Antimicrob Chemother 44:
393-395
Gold CH, Buchanan N, Tringham V, Viljoen M, Strickwold B,
Moodley GP (1976) Isoniazid pharmacokinetics in patients
in chronic renal failure. Clin Nephrol 6:365-369
Gonzalez Montaner LJ, Natal S, Yongchaiyud P, Olliaro P, the
Rifabutin Study Group (1994) Rifabutin for the treatment
of newly-diagnosed pulmonary tuberculosis: a multina-
tional, randomized, comparative study versus rifampicin.
Tuberc Lung Dis 75:341-347
Gordin F, Masur H (1994) Prophylaxis of Mycobacterium
avium complex bacteremia in patients with AIDS. Clin
Infect Dis 18 [Suppl 3]:S223-S226
Gordin F, Chaisson RE, Matts JP, Miller C, de Lourdes GM,
Hafner R, Valdespino JL, Coberly J, Schechter M, Klukowicz
AJ, Barry MA, O'Brien RJ (2000) Rifampin and pyrazin-
amide vs isoniazid for prevention of tuberculosis in HIV-
infected persons: an international randomized trial. JAMA
283:1445-1450
Grassi C, Peona V (1996) Use of rifabutin in the treatment
of pulmonary tuberculosis. Clin Infect Dis 22 [Suppl 1]:
S50-S54
Grewal KS, Dixit RP, Sil DR (1969) A comparative study of
therapeutic regimens with and without corticosteroids in
the treatment of tuberculous pleural effusion. J Indian Med
Assoc 52:514-516
Habibullah CM (1968) A study of tuberculous pleural effusion
with special reference to corticosteroid therapy. Indian J
Chest Dis 10:36-39
Hamadeh MA, Glassroth J (1992) Tuberculosis and pregnancy.
Chest 101:1114-1120
Treatment of Tuberculosis
Heinonen OP, Slone D, Shapiro S (1977) Birth defects and
drugs in pregnancy. Publishing Science Group, Littleton,
MA
Heym B, Zhang Y, Poulet S, Young D, Cole ST (1993) Charac-
terization of the katG gene encoding a catalase-peroxidase
required for the isoniazid susceptibility of Mycobacterium
tuberculosis. J Bacteriol 175:4255-4259
Heym B,Alzari PM, Honore N,Cole ST (1995) Missense muta-
tions in the catalase-peroxidase gene, katG, are associated
with isoniazid resistance in Mycobacterium tuberculosis.
Mol MicrobioI15:235-245
Hill AB (1948) Streptoycin for the treatment of tuberclosis: a
Medical Research Council Investigation. BMJ 2:769-782
Hinshaw HC, Feldman WH (1945) Streptomycin in the treat-
ment of clinical tuberculosis: a preliminary report. Proc
Staff Meetings Mayo Clin 20:313-318
Hong Kong Chest Service/British Medical Research Council
(1987) Five-year follow-up of a controlled trial of five 6-
month regimens of chemotherapy for pulmonary tuber-
culosis. Am Rev Respir Dis 136:1339-1342
Hong Kong Chest Service, British Medical Research Council
(1991) Controlled trial of 2, 4, and 6 months of pyrazin-
amide in 6-month, three-times-weekly regimens for smear-
positive pulmonary tuberculosis, including an assessment
of a combined preparation of isoniazid, rifampin, and
pyrazinamide. Results at 30 months. Am Rev Respir Dis
143:700-706
Hong Kong Chest Service/British Medical Research Council
(1992) A controlled study of rifabutin and an uncontrolled
study of ofloxacin in the retreatment of patients with pul-
monary tuberculosis resistant to isoniazid, streptomycin
and rifampicin. Tubercle Lung Dis 73:59-67
Hong Kong Chest Service/Tuberculosis Research Centre,Madras/
British Medical Research Council (1989) A controlled trial of
3-month, 4-month, and 6-month regimens of chemotherapy
for sputum-smear-negative pulmonary tuberculosis. Results
at 5 years. Am Rev Respir Dis 139:871-876
Hong Kong Chest Service/Tuberculosis Research Centre,
Madras/British Medical Research Council (1991) A con-
trolled clinical comparison of 6 and 8 months of antitu-
berculosis chemotherapy in the treatment of patients with
silicotuberculosis in Hong Kong. Am Rev Respir Dis 143:
262-267
Hong Kong Chest Service/Tuberculosis Research Centre,
Madras/British Medical Research Council (1992) A double-
blind placebo-controlled clinical trial of three antitubercu-
losis chemoprophylaxis regimens in patients with silicosis
in Hong Kong. Am Rev Respir Dis 145:36-41
Hong YP, Kim SJ, Lew WJ, Lee EK, Han YC (1998) The seventh
nationwide tuberculosis prevalence survey in Korea, 1995.
Int J Tuberc Lung Dis 2:27-36
Hosoglu S, Ayaz C, Geyik MF, Kokoglu OF, Ceviz A (1998)
Tuberculous meningitis in adults: an eleven-year review.
Int J Tuberc Lung Dis 2:553-557
Hosoglu S, Geyik MF, Balik I, Aygen B, Erol S, Aygencel TG,
Mert A, Saltoglu N, Dokmetas I, Felek S, Sunbul M, Irmak
H, Aydin K, Kokoglu OF, Ucmak H, Altindis M, Loeb M
(2002) Predictors of outcome in patients with tuberculous
meningitis. Int J Tuberc Lung Dis 6:64-70
International Union Against Tuberculosis and Lung Disease
(1982) Efficacy of various durations of isoniazid preventa-
tive therapy for tuberculosis: five years of follow-up in the
IUAT trial. Bull WHO 60:555-564
797
Jacobs RF, Sunakorn P, Chotpitayasunonah T, Pope S, Kelleher
K (1992) Intensive short course chemotherapy for tubercu-
lous meningitis. Pediatr Infect Dis J 11:194-198
Ii B, Lounis N, Truffot-Pernot C, Grosset J (1995) In vitro and
in vivo activities of levofloxacin against Mycobacterium
tuberculosis. Antimicrob Agents Chemother 39:1341-1344
Ji B, Lounis N,Maslo C, Truffot-Pernot C, Bonnafous P, Grosset
J (1998) In vitro and in vivo activities of moxifloxacin and
clinafloxacin against Mycobacterium tuberculosis. Antimi-
crob Agents Chemother 42:2066-2069
Kennedy N, Fox R, Kisyombe GM, Saruni AO, Uiso LO, Ramsay
AR, Ngowi Fl, Gillespie SH (1993) Early bactericidal and
sterilizing activities of ciprofloxacin in pulmonary tuber-
culosis. Am Rev Respir Dis 148:1547-1551
Khoo KH, Douglas E, Azadi P, Inamine JM, Besra GS, Mikusova
K, Brennan PJ, Chatterjee D (1996) Truncated structural
variants of lipoarabinomannan in ethambutol drug-resis-
tant strains of Mycobacterium smegmatis. Inhibition of
arabinan biosynthesis by ethambutol. J Bioi Chern 271:
28682-28690
Kim YG, Shin JG, Shin SG, Jang II, Kim S, Lee JS, Han JS, Cha
YN (1993) Decreased acetylation of isoniazid in chronic
renal failure. Clin Pharmacol Ther 54:612-620
Koch R (1882) Die Aetiologie der Tuberculose. Berlin Klin
Wochenschr 19:221-230
Larsen H, Nielsen GL, Schonheyder HC, Olesen C, Sorensen
HT (2001) Birth outcome following maternal use of f1uoro-
quinolones. Int J Antimicrob Agents 18:259-262
Lee CH, Wang WJ, Lan RS, Tsai YH, Chiang YC (1988) Cor-
ticosteroids in the treatment of tuberculous pleurisy. A
double-blind, placebo-controlled, randomized study. Chest
94:1256-1259
Lees AW (1963) Ethionamide and streptomycin therapy in
previously untreated cases of pulmonary tuberculosis. Am
Rev Respir Dis 88:399-401
Lehmann J (1946) Para-aminosalicylic acid in the treatment of
tuberculosis: a preliminary communication. Lancet 1:15-16
Liu LW, Trotz M, Erdos EG, Said SI (1991) Vasoactive intesti-
nal peptide (VIP) and helodermin degradation by airway
enzymes (abstract). Am Rev Respir Dis 143:A618
Loebstein R, Addis A, Ho E, Andreou R, Sage S, Donnenfeld
AE, Schick B, Bonati M, Moretti M, Lalkin A, Pastuszak A,
Koren G (1998) Pregnancy outcome following gestational
exposure to f1uoroquinolones: a multicenter prospec-
tive controlled study. Antimicrob Agents Chemother 42:
1336-1339
Long R, Younes M, Patton N, Hershfield E (1989) Tuberculous
pericarditis: long-term outcome in patients who received
medical therapy alone. Am Heart J 117:1133-1139
Lounis N, Grosset J, Truffot-Pernot C, Ji B (1997) Which
aminoglycoside or f1uoroquinolone is more active against
Mycobacterium tuberculosis in mice. Antimicrob Agents
Chemother 41:607-610
Maggi N, Pasqualucci CR, Ballota R, Sensi P (1966) Rifampicin:
a new orally active rifamycin. Chemotherapy 11:285-292
Malone RS, Fish DN, Spiegel DM, Childs JM, Peloquin CA
(1999) The effect of hemodialysis on cycloserine, ethion-
amide, para-aminosalicylate, and clofazimine. Chest 116:
984-990
McGregor MM, Olliaro P, Wolmarans L, Mabuza B, Bredell M,
Felten MK, Fourie PB (1996) Efficacy and safety of rifabutin
in the treatment of patients with newly diagnosed tubercu-
losis. Am J Respir Crit Care Med 154:1462-1467
798
Mdluli K, Sherman DR, Hickey MJ, Kreiswirth BN, Morris S,
Stover CK, Barry CE III (1996) Biochemical and genetic
data suggest that InhA is not the primary target for acti-
vated isoniazid in Mycobacterium tuberculosis. J Infect Dis
174:1085-1090
Meyers BR, Papanicolaou GA, Sheiner P, Emre S, Miller
C (2000) Tuberculosis in orthotopic liver transplant
patients: increased toxicity of recommended agents; cure
of disseminated infection with nonconventional regimens.
Transplantation 69:64-69
Miyazaki E, Miyazaki M, Chen JM, Chaisson RE, Bishai WR
(1999) Moxifloxacin (BAY12-8039), a new 8-methoxyqui-
nolone, is active in a mouse model of tuberculosis. Anti-
microb Agents Chemother 43:85-89
Mizutani S (1996) Chemotherapy of pulmonary Mycobacte-
rium kansasii infection. Kekkaku 71:527-531
Musser JM (1995) Antimicrobial agent resistance in myco-
bacteria: molecular genetic insights. Clin Microbiol Rev
8:496-514
Nair J, Rouse DA, Bai GH, Morris SL (1993) The rpsL gene and
streptomycin resistance in single and multiple drug resis-
tant strains of Mycobacterium tuberculosis. Mol Microbiol
10:521-527
Nightingale SO, Cameron W, Gordin FM, Sullam PM, Cohn
DL, Chaisson RE, Eron LJ, Sparti PO, Bihari B, Kaufman
DL, Stern JJ, Pearce DD, Weinberg WG, LaMarta A, Siegal
FP (1993) Two controlled trials of rifabutin prophylaxis
against Mycobacterium avium complex infections in AIDS.
N Engl J Med 329:828-833
Park IW, Choi BW, Hue SH (1997) Prospective study of corti-
costeroid as an adjunct in the treatment of endobronchial
tuberculosis in adults. Respirology 2:275-281
Pereira-Filho RA, Seva-Pereira A, de Magalhaes AF (1985) A
acetilacao da isoniazida na sindrome de Gilbert. Arq Gas-
troenterol 22: 172-175
Perriens JH, St.Louis ME, Mukadi YB, Brown C, Prignot J,
Pouthier F, Portaels F, Willame JC, Mandala JK, Kaboto M,
Ryder RW, Roscigno G, Piot P (1995) Pulmonary tubercu-
losis in HIV-infected patients in Zaire. A controlled trial
of treatment for either 6 or 12 months. N Engl J Med 23:
779-784
Potworowska M, Sianozecka E, Szufladowicz R (1966)
Ethionamide treatment and pregnancy. Polish Med J 5:
1152-1158
Rastogi N, Goh KS (1991) In vitro activity of the new difluori-
nated quinolone sparfloxacin (AT-4140) against Mycobac-
terium tuberculosis compared with activities of ofloxacin
and ciprofloxacin. Antimicrob Agents Chemother 35:
1933-1936
Robinson GF, Cambon KG (1964) Hearing loss in infants of
tuberculous mothers treated with streptomycin during
pregnancy. N Engl J Med 271:949-951
Roullion A (1981) Centenary of the discovery of the tubercule
bacillus by Robert Koch 1882-1982. Introduction. Bull Int
Union Tuberc 56:83-85
Salfinger M, Crowle AJ, Reller LB (1990) Pyrazinamide and
pyrazinoic acid activity against tubercle bacilli in cultured
human macrophages and in the BACTEC system. J Infect
Dis 162:201-207
Sanfilippo A, Della Bruna C, Marsili L, Morvillo E, Pasqualucci
CR, Schioppacassi G, Ungheri 0 (1980) Biological activity
of a new class of rifamycins, spiropiperdyl-rifamycins. J
Antibiot 33:1193-1198
D. E. Schraufnagel
Schaefer C, Amoura-Elefant E, Vial T, Ornoy A, Garbis H,
Robert E, Rodriguez-Pinilla E, Pexieder T, Prapas N, Merlob
P (1996) Pregnancy outcome after prenatal quinolone
exposure. Evaluation of a case registry of the European
Network of Teratology Information Services (ENTIS). Eur
J Obstet Gynecol Reprod Bioi 69:83-89
Schatz A, Waksman SA (1944) Effect of streptomycin and other
antibiotic substances on Mycobacterium tuberculosis and
related organisms. Proc Soc Exp Bioi Med 57:244-248
Scheinhorn OJ, Angelillo VA (1977) Antituberculosis therapy
in pregnancy. Risk to the fetus. West J Med 127:195-198
Schoeman JF, van Zyl LE, Laubscher JA, Donald PR (1997) Effect
of corticosteroids on intracranial pressure, computed tomo-
graphic findings, and clinical outcome in young children
with tuberculous meningitis. Pediatrics 99:226-231
Schraufnagel DE (1999) Tuberculosis treatment for the begin-
ning of the next century. Int J Tuberc Lung Dis 3:651-662
Schraufnagel DE, Abubaker J (2000) Global action against
multidrug-resistant tuberculosis. JAMA 283:54-55
Schraufnagel DE, Leech JA, Schraufnagel MEN, Pollak B (1984)
Short-course chemotherapy for mycobacteriosis kansasii.
Can Med Assoc J 130:34-38
Schroeder EK, de Souza N, Santos OS, Blanchard JS, Basso
LA (2002) Drugs that inhibit mycolic acid biosynthesis
in Mycobacterium tuberculosis. Curr Pharm Biotechnol
3:197-225
Schwander S, Rusch-Gerdes S, Mateega A, Lutalo T, Tugume S,
Kityo C, Rubaramira R, Mugyenyi P, Okwere A, Mugerwa
R,Aisu T, Moser R, Ochen K, M'Bonye B, Dietrich M (1995)
A pilot study of antituberculosis combinations comparing
rifabutin with rifampicin in the treatment of HIV-l associ-
ated tuberculosis. Tuberc Lung Dis 76:210-218
Snider DE, Graczyk J, Bek E, Rogowski J (1984a) Supervised
six-months treatment of newly diagnosed pulmonary
tuberculosis using isoniazid, rifampin, and pyrazinamide
with and without streptomycin. Am Rev Respir Dis 130:
1091-1094
Snider DE, Jr., Long MW, Cross FS, Farer LS (1984b) Six-
months isoniazid-rifampin therapy for pulmonary tuber-
culosis. Report of a United States Public Health Service
Cooperative Trial. Am Rev Respir Dis 129:573-579
Sreevatsan S, Pan X, Zhang Y, Kreiswirth BN, Musser JM
(1997a) Mutations associated with pyrazinamide resis-
tance in pncA of Mycobacterium tuberculosis. Antimicrob
Agents Chemother 41 :636-640
Sreevatsan S, Stockbauer KE, Pan X, Kreiswirth BN, Moghazeh
SL, Jacobs WR Jr, Telenti A, Musser JM (1997b) Ethambu-
tol resistance in Mycobacterium tuberculosis: critical role
of embB mutations. Antimicrob Agents Chemother 41:
1677-1681
Stewart SM, Crofton JW (1964) The clinical significance oflow
degrees of drug resistance in pulmonary tuberculosis. Am
Rev Respir Dis 89:811-829
Sullam PM, Gordin FM, Wynne BA (1994) Efficacy of rifabutin
in the treatment of disseminated infection due to Myco-
bacterium avium complex. The Rifabutin Treatment Group.
Clin Infect Dis 19:84-86
SUO J, Lin TP, Chiang CY, Lee CN, Yu MC (1996) Treatment of
multidrug-resistant tuberculosis in Taiwan. Chemotherapy
42 [SuppI3):20-23
Takayama K, Kilburn JO (1989) Inhibition of synthesis of ara-
binogalactan by ethambutol in Mycobacterium smegmatis.
Antimicrob Agents Chemother 33: 1493-1499
Treatment of Tuberculosis
Tam CM, Chan SL, Lam CW, Dickinson JM, Mitchison DA
(1997) Bioavailability of Chinese rifapentine during a clini-
cal trial in Hong Kong. Int J Tuberc Lung Dis 1:411-416
Telenti A, Imboden P, Marchesi F, Lowrie D, Cole S, Colston
MJ, Matter L, Schopfer K, Bodmer T (1993) Detection of
rifampin-resistant mutations in Mycobacterium tuberculo-
sis. Lancet 341 :647-650
Tran JH, Montakantikul P (1998) The safety of antitubercu-
losis medications during breastfeeding. J Hum Lact 14:
337-340
Trautner BW, Darouiche RO (2001) Tuberculous pericarditis:
optimal diagnosis and management. Clin Infect Dis 33:
954-961
Tseng AL, Walmsley SL (1995) Rifabutin-associated uveitis.
Ann Pharmacother 29:1149-1155
Van Loenhout-Rooyackers JH, Keyser A, Laheij RJ, Verbeek
AL, van der Meer JW (2001) Tuberculous meningitis: is a
6-month treatment regimen sufficient? Int J Tuberc Lung
Dis 5:1028-1035
Victor TC, Warren R, Butt JL, Jordan AM, Felix JV, Venter A,
Sirgel FA, Schaaf HS, Donald PR, Richardson M, Cynamon
MH, van HeIden PD (1997) Genome and MIC stability in
Mycobacterium tuberculosis and indications for continua-
tion of use of isoniazid in multidrug- resistant tuberculo-
sis. J Med MicrobioI46:847-857
Villarino ME, Ridzon R, Weismuller PC, Elcock M, Maxwell
RM, Meador J, Smith PJ, Carson ML, Geiter LJ (1997)
Rifampin preventive therapy for tuberculosis infection:
experience with 157 adolescents. Am J Respir Crit Care
Med 155:1735-1738
799
Weiss SE, Slomm PC, Blais FY, King B, Nunn M, Matney
GB, Gomez E, Foresman BH (1994) The effect of directly
observed therapy on the rates of drug resistance on relapse
in tuberculosis. N Engl J Med 330:1179-1184
Wilson EA, Thelin TJ, Ditts PV (1973) Tuberculosis compli-
cated by pregnancy. Am J Obstet Gynecol 115:526-529
Winder FG, Collins PB (1970) Inhibition by isoniazid of syn-
thesis of mycolic acids in Mycobacteium tuberculosis. J Gen
MicrobioI63:41-48
World Health Organization (1997) Treatment of tuberculo-
sis: guidelines for national programmes, 2nd edn. World
Health Organization, Geneva
Wyser C, Walzl G, Smedema JP, Swart F, van Schalkwyk EM,
van de Wal BW (1996) Corticosteroids in the treatment of
tuberculous pleurisy. A double-blind, placebo-controlled,
randomized study. Chest 110:333-338
Yamada T, Mizuguchi Y, Suga K (1976) Localization of co-
resistance to streptomycin, kanamycin, capreomycin, and
tuberactinomycin in core particles derived from ribosomes
of viomycin-resistant Mycobacterium smegmatis. J Anti-
biot (Tokyo) 29:1124-1126
Yeager RL, Monroe WE, Dessau FI (1952) Pyrazinamide in the
treatment of tuberculosis. Am Rev Tuberc 65:523-546
Yoshimatsu T, Nuermberger E, Tyagi S, Chaisson R, Bishai W,
Grosset J (2002) Bactericidal activity of increasing daily
and weekly doses of moxiftoxacin in murine tuberculosis.
Antimicrob Agents Chemother 46:1875-1879
Zhang Y, Heym B, Allen B, Young D, Cole S (1992) The catalase-
peroxidase gene and isoniazid resistance of Mycobacterium
tuberculosis. Nature 358:591-593
4S Directly Observed Treatment for Tuberculosis Control
ZHANG LI-XING
CONTENTS
45.1 Introduction 801
45.2 Current Status of DOTS Implementation 801
45.3 Impact and Advantages of DOTS 802
45.4 Main Principles of Implementation of DOTS 805
45.5 Approaches to DOTS Implementation 806
45.6 Limitations of DOTS Strategy 808
References 808
45.1
Introduction
The goals of tuberculosis control are to control
infection from Mycobacterium tuberculosis, reduce
morbidity of the disease and finally to eliminate
tuberculosis in the community. To achieve this, the
first and foremost intervention is to break the chain
of transmission of M. tuberculosis by identifying
the sources of infection and rendering them non-
infectious by use of chemotherapy. Providing che-
motherapy to infectious cases of tuberculosis is the
most efficient way to interrupt transmission of M.
tuberculosis (Kamat et al. 1966; Enarson et al. 2001;
IUATLD 2000; Enarson 2000).
More than 50 years after the introduction of
chemotherapy for tuberculosis, many developing
countries and areas continue to have a low cure rate,
high relapse rate, development of multi-drug-resis-
tant tuberculosis, and chronic tuberculosis excretors.
Such poor treatment outcomes are mainly due to poor
adherence of new smear-positive cases to treatment,
which in turn is due to a poorly organized treatment
program. In addition, these failures have enhanced
the transmission of M. tuberculosis and worsened
the tuberculosis epidemic. Improving adherence to
treatment is a priority issue for tuberculosis con-
Z. LI-XING, MD
President,Chinese Anti-Tuberculosis Association, 5,Dong Gang
Hu-Tong, Beijing 100035, China
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
trol program. In the early 1960s, the British Medical
Research Council (BMRC) successfully researched,
and developed, fully supervised chemotherapy to
the patient adherence to the prescribed treatment
regimen (Fox 1963). In the 1980s, the International
Union Against Tuberculosis and Lung Disease
(IUATLD) gradually implemented this fully super-
vised chemotherapy in Tanzania and other African
countries (Enarson 1991). In the 1990s, building on
the pioneering experiences of BMRC and IUATLD,
the World Health Organization (WHO) developed
the directly observed treatment (DOTS) strategy
(IUATLD 2000; WHO 1993). As a package, the DOTS
strategy has five key components (WHO 1993):
ilII Strong government commitment to sustain tuber-
culosis control effort;
~ Case-detection using sputum smear microscopy
among symptomatic patients;
" Standardized short-course chemotherapy for at
least all confirmed sputum smear-positive cases,
with direct observation of drug intake for at least
the initial two months;
• A regular uninterrupted supply of all essential
anti-tuberculosis drugs; and
• Standardized recording and reporting system that
allows assessment of treatment results for each
patient and of the overall tuberculosis control
program.
45.2
Current Status of DOTS Implementation
Although the DOTS strategy is widely accepted and
its use has expanded from 10 countries in 1990 to 127
countries in 1999, only about 23% of the estimated
smear-positive tuberculosis cases in the world was
reported by programs using the DOTS strategy in
1999; this percentage is nearly the same as the 22%
in 1998 (WHO 2001a,b). lWenty-two countries, with
a combined population of 3.74 billion persons, have
80% of the global tuberculosis burden. In 1999, these
802 Z. Li-Xing

22 high tuberculosis burden countries had an esti-
mated 6,621,000 new tuberculosis cases, of which
2,936,000 were new smear-positive pulmonary
tuberculosis cases, and only about one-fourth cases
of smear positive treated with DOTS (WHO 2001a).
Today, DOTS is the most effective strategy available
for tuberculosis control at the global level (Enarson
2000; Fox 1963; WHO 1993; Zhang et al. 1985, 1989,
1995, 2000a,b; Zhang and Kan 1992; China Tuberculo-
sis Control Collaboration 1996). Expansion of DOTS
should be the top priority for tuberculosis control
in the world.
Cure Rate Greatly Increased: The most important ben-
efit from implementing DOTS is the greatly increased
cure rate, which is due to improved patient adherence
to treatment. The cure rate of new smear-positive
cases is about 50% in non-DOTS area, compared to
a cure rate of 85-96% in Beijing or other areas where
DOTS has been implemented (Table 45.1, Fig. 45.1)
(Enarson et al. 2001; IUATLD 2000; Enarson 2000; Fox
1963; Zhang et al. 2000a; Zhang and Kan 1992). The
two-year relapse rate for 4037 patients treated under
DOTS in China was only 3.2% (China Tuberculosis
Control Collaboration 1996).

Decline in Tuberculosis Mortality Accelerated: Prior

to the introduction of DOTS in Beijing, the annual
45.3 reduction in tuberculosis mortality averaged less
Impact and Advantages of DOTS than 3% between 1965 and 1975. Following the
introduction of DOTS in 1978, tuberculosis mortal-
The epidemiology of tuberculosis is influenced by a ity declined more rapidly, averaging more than 7%
number of factors. One is the relatively recent intro- per year (Table 45.2).
duction of the DOTS strategy. Another is the intro-
duction of the AIDS epidemic in some areas. In Bei-
jing, the DOTS strategy has successfully operated for
two decades; (Zhang et al. 1985, 1989, 1995, 2000a,b;
Zhang and Kan 1992) however, the prevalence of II e-cl~.s%
human immunodeficiency virus (HIV) infection is o.~_I.2llo
_ poolM o.n.
o OloclI.2llo
still very low. Therefore, the steady improvement in • DeIIuIoocl
C OIhoto.s-.
0."
tuberculosis epidemiology over the past two decades
can be used to illustrate the impact of DOTS on the
tuberculosis epidemic. Fig.45.1. Cohort analysis of results of new smear-positive
cases in China, 1991-2000

Table 45.1. Cohort analysis of treatment results among new smear-positive cases in
Beijing,1985-2000

Year Total Successfully Treatment Died % Defaulted % Transferred %

no. treated* % failure %

1985 1945 85.4 4.4 3.4 5.5 1.3

1986 1823 87.7 8.7 2.9 0.1 0.6
1987 1808 89.9 5.5 3.0 0.7 0.9
1988 1473 92.4 3.6 2.6 0.7 0.5
1989 1210 90.0 4.4 3.2 1.5 0.9
1990 1346 89.1 3.6 5.9 0.4 1.0
1991 1219 94.1 2.5 2.8 0.2 0.4
1992 923 93.2 2.7 3.7 0.1 0.3
1993 764 95.8 0.7 2.4 0.8 0.3
1994 785 92.7 1.9 3.6 1.0 0.8
1995 819 91.0 0.7 4.6 2.2 1.5
1996 746 96.6 1.6 0.0 0.7 1.1
1997 786 90.2 3.9 4.2 1.1 0.5
1998 802 90.0 2.9 5.7 0.9 0.6
1999 764 92.0 2.9 4.2 0.9 0.0
2000 733 91.7 3.8 2.7 1.0 0.8

z 17946 90.7 3.9 3.4 1.3 0.7

*Includes those completing treatment with or without final sputum smear examination
Directly Observed Treatment for Tuberculosis Control 803

Table 45.2. Trend in tuberculosis mortality in Beijing, 1965-2000 in 1988 in Shun-yi County, a rural county of Beijing
Mortality/ Trend Mean trend DOT coverage
Municipality. In 1995, all 14,420 first grade pupils
Year
100,000 (%) %/year % (aged 6-7 years old) in all 181 primary schools in
Shun-yi county were registered for standardized
1965 24.9
1975 17.6 (-29.3) -2.9
tuberculin testing with PPD RT23 + Tween 80. The
1980 11.2 (-36.4) -7.3 (45) prevalence of infection in this group of children was
1985 6.6 (-41.1) -8.2 (63) 1.4%. This translated to an estimated ART! of only
1990 4.2 (-36.4) -7.3 (93) 0.19% between 1988 and 1995 (Zhang et al. 2000b).
1995 2.6 (-38.1) -7.6 (96) The effective implementation and comprehensive
2000 1.8 (-30.0) -7.1 (92)
1965-2000 (-92.8) -7.2
coverage of DOTS appears to be the most likely expla-
nation for the low risk of tuberculous infection.

Additional evidence of reduced transmission of

Rapid Decline in Tuberculosis Prevalence: Following
the introduction of DOTS, the prevalence of smear-
positive pulmonary tuberculosis in Beijing fell from
127 cases/l00 000 persons in 1979 to 16 cases/lOO
000 persons in 1990 - a total reduction of 87.4% and
an annual reduction rate of 17.2% (Fig. 45.2) (Zhang
et al. 2000a). In the World Bank-supported DOTS
project in China, which covered 13 provinces half
of China's population from 1991-2002, the decline
in smear-positive tuberculosis prevalence was 5.7%
annually between 1990 and 2000. By contrast, over the
same period of time, the other half of China without
this project had an annual reduction of only 1.3% in
smear-positive tuberculosis prevalence (Table 45.3)
(National Technic Steering Group 2002).
M. tuberculosis following DOTS implementation
include a decline in the number of tuberculosis men-
ingitis among children from 0-4 years-old, especially
in an area without BCG vaccination for newborns
(Table 45.4, Fig. 45.2). The decline in rate of tubercu-
losis in children, which paralleled the decline in rate
of meningitis in young children, further confirms
a reduction in the transmission of M. tuberculosis,
especially to the younger generation (Fig. 45.3).
Reduction in Risk of Drug Resistant Tuberculosis:
The aim of the modern tuberculosis control pro-
gram is a rapid reduction in the transmission of M.
tuberculosis and permanent cure for infectious cases
through the effective application of chemotherapy.
1000

Reduction in Chronic Cases: The high prevalence of

chronic infectious cases in many developing countries
is one of the most difficult problems to solve in tuber- ,..
culosis control. The best way to begin is to start with
a well-organized and effective treatment program for
new smear-positive cases in order to halt the produc-
tion of new chronic cases. This is then followed by
,. ........-
--
-DOT

--ClvonIca
--McnIly
the re-treatment of previously treated cases (Zhang --loIoI*lgiIII

et al. 2000a). Following the introduction of DOTS and

re-treatment for chronic cases in Beijing, the preva-
lence of chronic smear positive cases notified in the •• 1'-'-----------~-=---------"==--
program decreased rapidly from 28.9 cases/lOOOOO 1'" 1110 un Uu uu un 11'0 1"2 1tN 19'"
persons in 1979 to 2.6 cases/IOOOOO persons in 1990
Fig.45.2. The association of tuberculosis indices in Beijing,
- an annual reduction rate of 19.7% (Fig. 45.2). 1978-1996

Reduction in Tuberculosis Infection Rate: The annual Table 45.3. National wide sampling survey of prevalence of
risk of tuberculous infection (ART!), estimated using smear positive pulmonary tuberculosis in China,1990-2000
on age-specific prevalence of tuberculous infection, is Region Standardized Reduction Annual
a reliable means to assess the overall epidemiological prevalence rate(%) reduction
situation and can be used to evaluate the impact of 1/100,000 rate(%)
DOTS. In countries with a high coverage of Bacillus 1990 2000
Calmette-Guerin (BCG) vaccination at birth, it is
DOTS project region 142 79 44.4 5.7
very difficult to monitor the ART!. To estimate the
Non-project region 130 114 12.3 1.3
ART! in Beijing, BCG vaccination was discontinued
 804 Z. Li-Xing

Table 45.4. Cases of tuberculosis meningitis in children aged 0-4 years, 1981-2000 in Beijing (other
than Shun-Yi County) and Shun-Yi County

Year Beijing (other than Shun-yi county) Shun-yi county

Meningitis cases BCG coverage(%) Meningitis cases BCG coverage (%)

BCG+ BCG- BCG+ BCG-

1981 2 8 98.5 0 1 99.0

1982 6 4 98.1 0 0 99.6
1983 3 6 98.6 0 0 99.2
1984 7 5 98.7 1 0 99.7
1985 2 4 99.3 1 0 99.9
1986 8 5 99.1 1 0 99.9
1987 0 6 99.1 3 0 99.9
1988* 2 2 98.9 1 0 BCG discontinued
1989 2 0 99.3 0 0 BCG discontinued
1990 3 0 99.7 0 1 BCG discontinued
1991 3 0 99.1 1 0 BCG discontinued
1992 2 1 99.5 0 1 BCG discontinued
1993 1 0 99.9 0 0 BCG discontinued
1994 0 0 99.9 0 0 BCG discontinued
1995 2 0 99.9 0 0 BCG discontinued
1996 0 0 99.2 0 0 BCG discontinued
1997 3 0 99.1 0 0 BCG discontinued
1998 0 0 98.9 0 0 BCG discontinued
1999 0 0 98.4 0 0 BCG discontinued
2000 0 0 98.5 0 0 BCG discontinued

Total 46 41 8 3

Abbreviations: BCG =bacillus Calmette-Guerin

*BCG discontinued in Shun-yi county on July 15, 1988

u
in Beijing, the levels of initial resistance to isoniazid
.0 .- and streptomycin gradually decreased from 13.9%
§2I and 12.3% in 1978-1979 to 4.2% and 5.8% in 1996.
Furthermore, the level of initial resistance to rifam-
@.o -------.ua picin and to both isoniazid and rifampicin (MDR)
..• '"0
1..
CD
-I"' remained low. (Table 45.5) (Zhang et al. 1995)
~10
Improvement in Case-finding: The aim of case-find-
oL--- --'
ing is to discover the most potent sources of infec-
o 15 '0 U 10
tion, namely those with smear-positive pulmonary
Age group tuberculosis. Passive case-finding now applied is the
Fig. 45.3. Trends in the notification of newly diagnosed sputum active search for patients with smear positive among
smear-positive tuberculosis in Beijing by age group,1980-1997 patients with suspicious symptoms who seek health
care. Case-finding is of no value by itself unless
At the same time, the program aims to minimize the a good treatment program is available to ensure
development of resistance to anti-tuberculosis drugs, that all new smear-positive cases discovered are
which is a consequence of inappropriate use of these appropriately treated. The DOTS strategy integrates
drugs in the community (National Technic Steering both case-finding and treatment into a single entity.
Group 2002). Drug-resistant tuberculosis, and multi- When well-organized and effectively implemented,
drug-resistant tuberculosis (MDR-TB) in particular, DOTS should increase the case-notification rate.
is clearly a man-made disease and remains one of Through the World Bank's Infectious and Endemic
the main obstacles to tuberculosis control. It can be Disease Control Project, DOTS expanded rapidly to
avoided by the effective implementation of the DOTS cover half of China's population in 13 municipalities,
strategy (Zhang et al. 1995; Kimerling 2000; Fujiwara provinces and autonomous regions. The case-finding
and Fine Sherman 1997). After implementing DOTS rate of new smear-positive cases notified in the proj-
 Directly Observed Treatment for Tuberculosis Control 805

Table 45.5. Trends of initially resistant anti-tuberculosis drugs Treatment costs under DOTS were a tenth to a twen-
(0/0) in Beijing, 1978-1996 tieth the cost with traditional (including in-patient)
Period INH SM RMP EMB INH+RMP care. According to one study, the rate of relapse or
acquired resistance with DOTS occurred in only 1.2%
1978-1979 13.9 12.3
1981-1982 7.8 10.8 0.4 0.4 0.4
of patients and accounted for 6.0% of cost. In compari-
1983-1984 11.6 11.3 0.3 0.7 0.3 son, relapse or acquired resistance occurred in 10.9%
1985-1986 10.0 12.7 1.2 0.4 0.0 of patients and accounted for 35.7% of cost with tra-
1987-1988 8.1 7.3 2.4 1.6 1.6 ditional care (Weis et al.1999). With DOTS, $5.5 could
1989-1990 3.9 5.8 1.9 0.6 0.0 save one disability-adjusted life year (DALY) whereas
1991-1992 6.8 4.2 1.7 1.7 0.0
1995 5.4 2.7 1.3 0.7
$57.5 is needed to save one DALY in a program without
6.0
1996 4.2 5.8 2.5 0.8 0.8 DOTS (Xu Qun et al. 2000).
INH=isoniazid; SM=streptomycin; RMP=rifampicin;
EMB=ethambutol

ect increased from 9.1 cases/100,000 persons in 1992
to 28.5 cases/lOO,OOO persons in 2000. In comparison,
the case rate in the other half of China that lacked
a comprehensive DOTS program increased from 7.1
cases/100 000 person in 1992 to 9.8 cases/100 000
persons in 2000 (Fig. 45.4).
There are a number of reasons for much higher
case-finding in the above DOTS program. First, free
diagnosis and treatment was provided to all patients
with smear-positive pulmonary tuberculosis. Second,
the project cured a high percentage of its patients.
Third, the program was easily accessible to patients
and good services attracted patients with symptoms
to seek health care in the project. Fourth, compre-
hensive training was provided to health personnel
working at various levels. Fifth, the project raised the
health care workers' awareness of tuberculosis and its
related symptoms. Sixth, health promotion on tuber-
culosis and its treatment was provided to the public.
Cost Effective Approach to Tuberculosis Control: Sev-
eral studies have clearly demonstrated that DOTS is
the most cost effective approach to control tuberculo-
sis (Murray et al. 1991; Xu Qun 2000; Weis et al. 1999).
•• 21.0'1 n." U.M 11.51
45.4
Main Principles of Implementation of DOTS
To achieve the impact of DOTS on tuberculosis con-
trol and to reach global tuberculosis control targets
- ensure at least 70% of estimated smear-positive
pulmonary cases are detected and 85% of smear-
positive pulmonary cases are cured - the following
key principles of DOTS implementation should be
implemented:
Government Commitment: Governments should give
high priority in order to intensify tuberculosis con-
trol effort; the national tuberculosis program (NTP)
should be provided with sufficient financial support
to sustain the implementation of DOTS; it should for-
mulate and implement laws and regulations related
to tuberculosis control, and build the capacity to
implement DOTS at different levels.
Nationwide Coverage: The DOTS strategy should
cover and be implemented in the entire country,
especially the rural area.
National Tuberculosis Program (NTP): Implemen-
tation of DOTS nationwide can only be conducted
successfully within the context of a NTP. This serves
to ensure a permanent and sustainable tuberculosis
control program.

.OOTS_
._OOTS_ Emphasis Place on Smear-positive Pulmonary Tuber-
culosis: The control of the most potent sources of
transmission of M. tuberculosis, namely smear-posi-
tive pulmonary tuberculosis, should be the highest
priority.
Year
Ambulatory Chemotherapy: The control program
Fig.45.4. The rate of new smear-positive cases registered by should fully utilize the powerful weapon of directly
the DOTs project in China,1992-2000 observed treatment with short-course ambulatory
806
chemotherapy. At the very least, this should be pro-
vided free of charge to every smear-positive case of
pulmonary tuberculosis.
Primary Health Care: It is essential to have a well-
functioning primary health care system to effectively
implement DOTS. DOTS can be integrated with pri-
mary health care or community health care.
45.5
Approaches to DOTS Implementation
Organizational Structure: A good nationwide system
of health services is essential for the effective imple-
mentation of DOTS. DOTS should be incorporated
into and should fully utilize the existing health
care system. DOTS implementation should enhance
health sector development, and vice-versa. The func-
tions of DOTS at different levels of the health sector
are shown in Table 45.6.
Demonstration Area: At the start of DOTS imple-
mentation, it is useful to establish a demonstration
area to pilot various aspects of DOTS. A country can
be chosen to obtain first-hand experience in DOTS
implementation. Thereafter, DOTS can be carefully
expanded. Although the main principles of DOTS
are the same everywhere, approaches in implemen-
tation can be adapted to local conditions. A carefully
selected program manager should be responsible for
the demonstration project. The objectives of this pilot
project include the following: establish a well-func-
tioning health section system for DOTS implemen-
tation; determine the acceptability, feasibility and
practicality of DOTS; and to determine the effective
of treatment under local conditions.
Training Course: The DOTS strategy is whole new
approach to tuberculosis control for many health
Table 45.6. DOTS function at different level of health sector
Level Function in DOTS
National Policymaking, NTP Planning, Budget, Manual of program, Training, Reference laboratory Quality control smear
Supervision Supplies of drugs and Materials, Center registration, microscopy, Culture and susceptibility for TB
Manual of DOTS
State or Training, supervision, Supplies of drugs and materials,
provinces Center registration, Quality control, Pilot study
District or Determine regimen, Implementing DOTS Maintaining supplies
county of drugs, Supervision, Training, Registration
PHC Directly observed treatment, Make up with treatment, Registration
Z. ii-Xing
care personnel. To implement DOTS, it is essential
for health care personnel at all levels to have a thor-
ough knowledge and a clear understanding of the
principles and elements of DOTS. This is especially
important to convince senior level doctors that DOTS
is an advance in tuberculosis control and has a sound
scientific basis. In addition, it is important to orient
health care workers to the public health approach to
tuberculosis control.
Proper training courses should contain the follow-
ing: Government commitment to tuberculosis con-
trol; smear-positive pulmonary tuberculosis cases
are the most infectious cases; passive case-finding
in tuberculosis; standardized treatment regimens for
tuberculosis; sputum examination for the diagnosis
of tuberculosis; and how to implement DOTS.
Patient Education Prior to the Start of Treatment:
Whenever a patient has symptoms suggestive of
pulmonary tuberculosis, sputum specimens should
be collected and examined by microscopy to look
for acid-fast bacilli (AFB). If AFB is detected, the
patient is confirmed to have smear-positive pulmo-
nary tuberculosis. Prior to the start of treatment, it
is important for the patient and his, or her, family
members to be educated on the following issues:
IJ Tuberculosis is an infectious disease and is spread
through the air
1II Purposes of chemotherapy for tuberculosis are to
cure the disease, to reduce relapse after treatment
completion, and to prevent transmission of M.
tuberculosis
llII Main reasons for treatment failure
Effectiveness of DOTS
~ How DOTS is performed in detail
(\I How to receive directly observed therapy and
cooperate with health personnel
At the same time, the health care personnel, fre-
quentlya nurse, should try to understand the follow-
ing issues:
Reference laboratory Quality control of smear
microscopy, culture and susceptibility for TB
Smear microscopy
Collect sputum specimen
Directly Observed Treatment for Tuberculosis Control
• Any obstacles for the patient to receive DOTS
• How to solve any problems that patient may have
during treatment
On average, about 20 minutes is needed to educate
the patient and family members. By spending the
time, the health care worker gains the trust of the
patient and treatment is more likely to be successful.
Standardized Regimen of Chemotherapy: Standard-
ized chemotherapeutic regimens, as recommended
by WHO and IUATLD, are based on rigorous, multi-
center, scientific studies conducted under various
settings. These regimens have high efficacy, low tox-
icity, good tolerability, and low cost. The following
standardized regimens of chemotherapy are used in
the DOTS program in China: For new smear positive
pulmonary tuberculosis, 2H3R3Z3E3/4H3R3: Isoniazid
(H): 600 mg,Rifampicin (R) 600 mg,Pyrazinamide(Z)
2000 mg and Ethambutol (E) 1250 mg or Streptomy-
cin (S) 750 mg given every other day for two months
and followed by four months of Hand R given every
other day. For smear-positive relapsing cases and
others re-treatment cases, 2H 3 R3 Z3 S3 E3/6H 3R3E3.
Supervised Chemotherapy: To ensure that DOTS is
provided conveniently to patients, a health care worker
should address the following issues with each patient:
• Chose the supervisor who will supervise the intake
of medication; this is usually a health care personnel
working in the primary health center.
• If the supervisor is not a health care personnel, spe-
cial training should be provided to this supervisor.
• Decide on the time and venue for supervision of
drug intake. This should be based on what is conve-
nient for the patient, e.g. at a primary health center,
home or work site.
The supervisor is responsible for observing the
intake of every dose of drugs. Every supervised dose
should be recorded on the treatment card. If a patient
fails to attend a supervised dose, the supervisor
should take immediate action, including calling the
patient or going to the patient's home to find out the
reason for the missed dose, then make up the treat-
ment on the same day. To avoid re-sale of drugs by
patients, drugs should be kept by the supervisor and
the amount of drugs remaining should be compared
with the treatment card.
Full DOTSfor Entire Duration ofShort-course Treatment:
During the initial phase of treatment, patient adherence
may be good because patients are still symptomatic.
807
However, by the end of the second month of treatment,
the rate of patient non-adherence often increases as
symptoms are resolving. Because the completion of the
entire 6-8 months of short-course chemotherapy will
significantly improve cure and reduce relapse rate, every
smear-positive case should be encouraged to complete
the full course of treatment.
In many areas, a patient's home may be far from
the primary health care center or the area may not
even have a health care center. This is especially
common in remote rural and mountainous areas. In
these areas, DOTS can only to given for the first two
months (intensive phase) of treatment and then the
patient can self-administer their TB drugs during the
remaining months of treatment.
Quality Control of DOTS: To ensure that patients
are really complying with DOTS and that they are
completing the full course of treatment, supervi-
sion at different levels should be carried out on a
regular basis. To maintain the quality of DOTS, a
team from the state or province level should makes
random visits to patients, to the village health posts,
and to the district tuberculosis centers to check the
treatment cards, the amount of remaining drugs and
collect urine specimen from patients to test for the
metabolites of anti-tuberculosis drugs.
Advocacy: Effective advocacy and social mobilization
is necessary for successful DOTS implementation.
Assessing Results ofDOTS: Evaluation of DOTS imple-
mentation is very important to maintain and improve
the program. The use of independent experts and the
collection and analysis of program data is the best
way to evaluate DOTS. The following data are impor-
tant to analysis:
• DOTS detection rate: The percentage of estimated
new smear-positive cases detected in a DOTS pro-
gram is an important indicator to evaluate the NTP.
• Sputum conversion: All initial sputum smear-
positive cases have to submit sputum specimens
for repeat smear examination at end of the second
month of treatment. The percentage of patients
with no AFB found on smear examination is
another important indicator.
• Cohort Analysis: A cohort analysis is conducted
at end of each year to evaluate treatment outcome
for all smear-positive cases, including those treated
under DOTS and on self-administered therapy. For
the cohort analysis, treatment outcomes include
cure, treatment completion, treatment failure, death,
transfers and treatment default. The percentage of
808
treatment failure, which should be less 5%, provides
an indication of the efficacy of the regimen and the
level of resistance to anti-tuberculosis drugs.
Logistic Supply: High quality anti-tuberculous drugs,
microscopes, and transportation tools should be
available in a timely manner.
45.6
Limitations of DOTS Strategy
In Beijing, results from the prevalence survey con-
ducted between 1979 and 1990 showed a consistent
decline in the prevalence of smear-positive tubercu-
losis (Fig. 45.1). However, the trend in notified rate of
new smear-positive pulmonary tuberculosis is differ-
ent. The notified rate showed a slight increase in the
early years after the introduction of DOTS. After 1987,
the rate began to decline but then leveled off and did
not change between 1993 and 2000. This likely reflects
that the fact that the majority of new cases are the result
of endogenous reactivation of old tuberculous infec-
tions. A consistent decline in tuberculosis incidence
will require a sustained effort to implement DOTS.
Over the long term, the implementation of DOTS
can be influenced by many factors including political
change, social-economic reform, health sector reform,
health personnel or program director change, and other
unpredictable events. It is not easy to implement DOTS
strategy for a sustained period. It is also difficult to cover
and reach all populations and areas in large countries
with a DOTS program because of socio-cultural, reli-
gious, geographical, and people-related barriers.
From a technical perspective, there are some
disadvantages and limitations to the current DOTS
strategy. For instance, it does not deal with those with
tuberculous infection, treatment duration is too long,
and it does not adequately address MDR-TB and the
HIV-TB co-epidemic. For these and other new chal-
lenges, new diagnostic tests, new drugs and new vac-
cines are needed in the fight against tuberculosis.
References
China Tuberculosis Control Collaboration (1996) Results of
directly observed short-course chemotherapy in 112 842
Chinese patients with smear-positive tuberculosis. Lancet
348:358-362
Enarson DA (1991) Principles of IUATLD collaborative tuberculo-
sis programmes. Bulllnt Union Tuberc Lung Dis 66:195-199
Enarson DA (2000) World tuberculosis control: how far have
we to go? Int J Tuberc Lung Dis 4:s219-s223
Z. Li-Xing
Enarson DA, Beyers N, Zhang LX (2001) The tuberculosis
pandemic today: routes of transmission and new target
groups. Scand J Infect Dis 33:9-12
Fox W (1963) Ambulatory chemotherapy in a developing
country: Clinical and epidemiological studies. Adv Tuberc
Res 12:28-149
Fujiwara PI, Fine Sherman L (1997) Multidrug-resistant tuber-
culosis: many paths, same truth (editorial). Int J Tuberc
Lung Dis 1:297-298
IUATLD (2000) Management of tuberculosis: a guide for low
income countries, 5th edn. International Against Tubercu-
losis and Lung Disease
Kamat SR et al (1966) A controlled study of the influence of
segregation of tuberculosis patients for one year on the
attack rate of tuberculosis in a 5-year period in close family
contacts in South India. Bull WHO 34:517-532
Kimerling ME (2000) The Russian equation : an evolving
paradigm in tuberculosis control. Int J Tuberc Lung Dis
4:S160-S167
Murray CJL et al (1991) Cost effectiveness of chemotherapy
for pulmonary tuberculosis in three sub-Saharan African
countries. Lancet 338: 1305-1308
National Technic Steering Group of the Epidemiological
Sampling Survey for Tuberculosis (2002) Report on fourth
national epidemiological sampling survey of tuberculosis.
Chin JTuberc Respir Dis 25:3-7
Weis SE, Foresman B, Matty KJ et al (1999) Treatment costs
of directly observed therapy and traditional therapy for
Mycobacterium tuberculosis: a comparative analysis. Int
J Tuberc Lung Dis 3:976-984
WHO (1993) Treatment of tuberculosis: guidelines for national
programmes. World Health Organization, Geneva
WHO (2001a) Global tuberculosis control, WHO report 2001
(WHO/CDS/TB/2001.287). World Health Organization,
Geneva
WHO (2001b) The global DOTS expansion plan. Communi-
cable Disease Information Resource Center. World Health
Organization, Geneva
World Health Organization! International Union Against tuber-
culosis and Lung Disease (1997) Anti-tuberculosis drug
resistance in the world. WHO/TB/97.229, Geneva, p 227
Xu Qun, Jin Shuigao, Zhang Lixing (2000) Cost-effectiveness of
DOTS and non-DOTS strategy for smear-positive pulmo-
nary tuberculosis. Bull Chin Antituberc Assoc 22:60-62
Zhang LX, Kan GQ (1992) Tuberculosis control program in
Beijing. Tubercle Lung Dis 73:162-166
Zhang LX et al (1985) Supervised intermittent chemotherapy
for pulmonary tuberculosis in a rural area of China.
Tubercle 66: 1-7
Zhang LX, Kan GQ, Liu CW (1989) A model of fully supervised
chemotherapy for pulmonary tuberculosis in the tubercu-
losis control program in a rural area of China. Bull Int
Union Tuberc Lung Dis 64:20-21
Zhang LX et al (1995) Trend of initial resistance of tubercle
bacilli isolated from new patients with pulmonary tuber-
culosis and its correlation with tuberculosis program in
Beijing. Tubercle Lung Dis 76:100-103
Zhang LX et al (2000a) Risk of tuberculosis infection and tuber-
culous meningitis after discontinuation of Bacillus Calmette-
Guerin in Beijing. Am J Respir Crit Care Med 162:1314-1317
Zhang LX, Tu DH, Enarson DA (2000b) The impact of directly
observed treatment on the epidemiology of tuberculosis in
Beijing. Int J Tuberc Lung Dis 4:904-910
46 Multi-Drug-Resistant Tuberculosis
PETER D. O. DAVIES

CONTENTS 46.6.11 The Individual Patient with MDR-TB 828

46.6.12 The Consequences for Others 829
46.1 The Epidemiology of Multi-Drug-Resistant 46.6.13 Implications for Public Health
Tuberculosis 809 and the Wider Economy 829
46.1.1 Definitions and Causation 809 46.6.14 Is There One World Message After All? 830
46.1.2 Global Epidemiology of Drug-resistant TB 810 46.6.15 Transmission of MDR-TB 831
46.1.3 Impact of SCC on the Treatment 46.7 Conclusions 832
of Drug-resistant TB 811 References 833
46.2 Examples of Important Areas of MDR-TB 812
46.2.1 India 812
46.2.1.1 DOTS and its Impact in India 813
46.2.1.2 Practical Problems DOTS Faces In India 813
46.2.2 MDR-TB in Africa 814 46.1
46.3 Drug Resistance in Prisons 816 The Epidemiology of Multi-Drug-Resistant
46.4 The Microbiology of Drug-Resistant Tuberculosis
Tuberculosis 816
46.4.1 Molecular Diagnostic Techniques 816
46.4.2 Molecular Testing for Drug Resistance 817 46.1.1
46.4.3 Rapid Tests in Routine Use 819 Definitions and Causation
46.4.4 How the Sequencing of the Mycobacterial Genome
Will Help the Development of New Drugs
Drug-resistant tuberculosis and, particularly, multi-
in the Treatment of Tuberculosis 819
46.4.4.1 Breaching the Wall
drug-resistant tuberculosis (MDR-TB; defined as
of "Fortress Mycobacterium" 819 strains of M. tuberculosis resistant to at least iso-
46.4.4.2 In Vivo Phenotypes: Targeting Persistent niazid and rifampicin, the two most powerful anti-TB
Organisms 820 drugs), have recently received heightened attention
46.4.4.3 Latent Tuberculosis 821 in the media. This has not only helped advancement
46.5 The Treatment of MDR-TB 821
46.5.1 Introduction 821 within the field of MDR-TB, but it has also brought TB
46.5.2 Management of Non-MDR Resistance 822 to the forefront of discussions in public health. Drug
46.5.2.1 Isolated Resistances 822 resistance in TB is primarily created by two man-made
46.5.3 Multi-Drug-Resistant Tuberculosis 822 mechanisms. The first is erroneous prescribing prac-
46.5.3.1 Infection Control 823 tices combined with the inability to implement systems
46.5.3.2 Clinical Management 823
46.6 Tuberculosis and HIV Infection 824 of patient adherence to treatment. This is often seen
46.6.1 Introduction 824 when private practitioners are not fully aware of TB
46.6.2 HIV and TB 824 control procedures (Mahmoudi and Iseman 1993). For
46.6.3 MDR-TB Outbreaks 824 instance, one study in India revealed 80 different treat-
46.6.4 Treatment 825 ment regimens were being recommended for difficult-
46.6.5 Alternative Therapies 825
46.6.6 Contacts 825 to-trace patients in a single slum in Bombay (Uplekar
46.6.7 Prognosis and Predictors 825 and Shepard 1991). The second is patient-related,
46.6.8 Sporadic MDR-TB in HIV 826 as they fail to follow the prescribed regimen, or are
46.6.9 The Financial Challenges of MDR-TB 827 affected by other diseases (such as co-infection with
46.6.10 Different Countries, Different Problems, HIV) that interfere with the mechanisms of anti-TB
Different Expectations 827
drugs (Kochi et al. 1993). These factors place selective
pressure upon the bacilli in a patient, thereby increas-
P. D. O. DAVIES, DM, FRCP
ing the growth of resistant bacilli (Pablos-Mendez
Director, Tuberculosis Research Unit, Cardiothoracic Centre, and Lessnau 2000). However, these two factors simply
Liverpool Ll4 3PE, UK reflect the principle that drug resistance tends to

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
810
arise when national TB programmes do not adhere to
international recommendations for TB control (World
Health Organization 1997).
46.1.2
Global Epidemiology of Drug-Resistant TB
Since 1994, the World Health Organization (WHO)
and the International Union Against Tuberculo-
sis and Lung Disease (IUATLD) have conducted a
global project on drug resistance surveillance. The
project builds upon international recommended
methods for drug resistance surveillance and drug
susceptibility testing to four first-line anti-TB drugs:
isoniazid, rifampicin, ethambutol, and streptomycin
(World Health Organization 1997a,c). The project
distinguishes between drug resistance among new
cases (the presence of resistant strains of M. tuber-
culosis in a newly diagnosed patient who has never
received TB drugs or has received them for less than
one month of treatment) and drug resistance among
previously treated cases (presence of resistant strains
of M. tuberculosis in a patient who has previously
received treatment for at least one month) (World
Health Organization 2000a).
Data from the two surveys conducted in 1994-1997
and 1997-2000, respectively, confirm the existence
of drug resistance among new cases in all countries
surveyed. Data also show that MDR-TB is present in
58 of 67 sites surveyed {Pablos-Mendez et al. 1998;
Espinal et al. 200l}. In the most recent survey (l997-
2000) the median prevalence of drug resistance
among previously treated cases was 10.7% (range
1.7-36.9%), and the median prevalence of MDR-
TB among new cases was 1.0% (range 0-14.1%).
The median prevalence of resistance to each drug
was as follows: isoniazid - 6.2%, rifampicin - 1.2%,
ethambutol - 0.6%, streptomycin - 5.2%. Surveys in
countries such as Botswana, Cuba, Chile, Slovenia,
Uruguay, Venezuela showed low prevalence of drug
resistance among new cases. This was expected, as
these countries are also known for the high quality
of their TB control programmes. However, surveys in
other countries demonstrated higher levels of drug
resistance among new cases. These countries include
those of the former Soviet Union (Russia, Estonia,
and Latvia) Bolivia, Thailand, Sierra Leone, and
some settings in China and India. The WHO/IUATLD
Global Project has revealed the existence of several
'hot spots' for MDR-TB (defined conventionally as
those where the prevalence of MDR-TB among new
cases was 3% or above) including Dominican Repub-
P. D. O. Davies
lie, Estonia, Iran, Ivory Coast, Latvia, Mozambique,
and selected areas in China, India, and the Russian
Federation {Table 46.l}.
Trends analysis from those countries that had
more than one survey done over the years clearly
outlines the importance of sound TB control prac-
tices in preventing the emergence of drug resistance.
For instance, Cuba and Chile have successfully imple-
mented the DOTS strategy for more than 20 years
(World Health Organization 2001). As a result, Cuba
had a prevalence of MDR-TB among all TB cases
varying from 0 to 1% between 1995-1998, and Chile
maintained a prevalence of MDR-TB of less than
1% from 1980-1998. However, Ivanovo Oblast in the
Russian Federation reported a prevalence of MDR-
TB of 6% in 1995, reaching 9% in 1998. In the case of
Estonia, an even more alarming situation has been
described where MDR-TB among new cases gradu-
ally increased from 10% in 1994 to 14% in 1998.
Given the recent trend of international migration
and the potential spread of MDR-TB across conti-
nents (Becerra et al. 1999) industrialised countries
have begun to examine national policies ofTB control
in foreign-born populations (Talbot et al. 2000). The
WHO/IUATLD project also attempted to determine
if foreign-born populations, generally immigrating
from high TB prevalence areas, have a higher or
lower level of MDR-TB than indigenous populations.
Not surprisingly, all industrialised countries exam-
ined showed a higher prevalence of drug-resistance
among foreign-born new cases. A similar trend was
observed with MDR-TB among foreign-born new
cases (Gilad et al. 2000).
Previously treated cases have a much higher likeli-
hood of being infected with drug- resistant strains,
mostly due to acquiring resistance during improper
treatment in the past (World Health Organization
1997d). In the most recent survey (l997-2000) the
median prevalence of resistance among previously
treated cases was 23.3% (range 0-93.8%), and that of
MDR-TB was 9.3% (range 0 - 48.2%) (Espinal et al.
2001). However, recognising that some of these cases
were infected with drug-resistant TB or MDR-TB at
the outset, the term 'acquired resistance' was aban-
doned after the first survey and replaced with 'resis-
tance among previously treated cases'. The use of
'acquired resistance' should be limited to settings that
possess the capability (technically and financially) to
perform drug-susceptibility testing and DNA fin-
gerprinting on all new TB cases. This is in order to
determine if a resistant strain, once isolated in a case
for re-treatment, is the same one that produced the
disease in the previous episode and acquired drug
Multi-Drug-Resistant Tuberculosis 811

Table 46.1. Drug resistance in patients with no history of prior treatment expressed as a percent-
age of cases

Country Year Sample Overall Resistance to poly-resistance

1 drug 2 drugs 3 drugs 4 drugs any MDR

Benin 94-97 333 8.4 6.0 2.1 0.3 0.0 2.4 0.3
Botswana 94-97 407 3.7 3.4 0.2 0.0 0.0 0.2 0.2
Botswana 1998 638 6.3 5.3 0.8 0.2 0.00 0.9 0.5
Cameroon 1995 31.8
Central African 1999 464 16.4 10.8 3.7 1.7 0.2 5.6 1.1
Republic
Ethiopia 1995 167 15.6
Ghana 54.5
Guinea 1998 539 14.7 9.8 4.3 0.6 0.0 4.8 0.6
Ivory Coast 94-97 320 13.4 5.3 6.3 1.6 0.3 8.1 5.3
Kenya 94-99 445 6.3 5.4 0.9 0.0 0.0 0.9 0.0
Lesotho 94-97 330 8.8 6.1 2.4 0.3 0.0 2.7 0.9
Mozambique 1999 1028 20.8 12.2 5.8 2.3 0.5 8.7 3.5
Sierra Leone 94-97 463 28.1 16.6 10.2 1.1 0.2 11.4 1.1
Sierra Leone 1997 117 24.8 17.9 6.0 0.9 0.0 6.8 0.9
South Africa 1996 8.9
(Hlabisa)
South Africa 1997 661 8.0 5.9 1.2 0.5 0.5 2.1 1.5
(Mpumalanga)
Swaziland 94-97 224 11.7
Uganda 1997 374 19.8 12.8 6.7 0.3 0.0 7.0 0.5
Zimbabwe 94-96 676 3.3 1.3 1.2 0.1 0.6 1.9 1.9

resistance during treatment, or whether it is a new
strain that re-infected the patient after the first epi-
sode was diagnosed and tested.
46.1.3
Impact of see on the Treatment
of Drug-ResistantTB
SCC recommended as part of the DOTS strategy can
generate treatment success rates as high as 98% in
new cases (Fox 1981). However, this high level of
cure was observed in settings with minimal levels
of primary drug resistance. Among cases resistant
to a single drug, SCC may achieve cure rates as
high as those in pan-susceptible cases. Evidence
also suggests that for new cases of TB resistant to
one or more anti-TB drugs, including isoniazid,
ethambutol, or streptomycin, rates of failure are
comparable to those of new drug-susceptible cases.
However, new cases of TB resistant to rifampicin
experience significantly higher failure rates than
drug-susceptible cases (Mitchison and Nunn 1986;
Espinal et al. 2000). For patients with MDR-TB,
SCC is not appropriate for treatment (Table 46.2)
(Espinal et al. 2000). In settings where effective
DOTS is implemented (Korea, Peru, Hong Kong),
the cure rate achieved with SCC is no greater than
59% in new MDR-TB cases as compared to 85-89%
among new drug-susceptible cases and 76-87%
among new isoniazid resistant (non-MDR) cases.
For regions and countries with generally poorer
TB control programmes (Ivanovo Oblast of the
Russian Federation and the Dominican Republic)
cure rates in MDR-TB cases range from 10 to 20%
(Espinal et al. 2000). As expected, increasing levels of
resistance correlate directly with decreasing rates of
treatment success (Coninx et al. 1999). For MDR-TB
cases, failure rates to SCC have reached up to 40% in
some settings (Espinal et al. 2000). Not only do these
cases continue to remain infectious, but also such
cases may undergo amplification of resistance to the
drugs to which they were exposed in the regimen on
which they failed (Farmer 1999). This 'amplification'
phenomenon is simply the well-described mecha-
nism of acquired drug-resistance. In conclusion,
these data indicate that alternative strategies for
the management of MDR-TB cases, including use
of regimens containing second-line anti-TB drugs,
should be considered in those settings where MDR-
TB is frequent and where there is need to treat cases
at a programmatic level (Espinal et al. 1999). WHO
and its partners are currently piloting a strategy
called 'DOTS-Plus' that aims to develop global
policy recommendations for the management of
MDR-TB with second-line anti-TB drugs (World
812 P. D. O. Davies

Table 46.2. Levels of drug resistance to each drug in selected African countries

Country Year Sample INH RMP EMB SM

size
mono any mono any mono any mono any

Benin 333 3.3 5.4 0.0 0.3 0.0 0.6 2.7 4.8
Botswana 407 1.2 1.5 0.7 1.0 0.0 0.0 1.5 4.5
Botswana 1999 638 3.6 4.4 0.2 0.6 0 0.2 1.6 2.2
Cameroon
CAR 1998 464 4.1 9.5 0.2 1.3 0.0 2.4 6.5 11.0
Ethiopia
Ghana
Guinea 1998 539 4.5 9.3 0.2 0.7 0.0 0.6 5.2 9.5
Ivory Coast 320 3.1 11.3 0.0 5.3 0.0 0.3 2.2 6.9
Kenya 445 5.4 6.3 0.0 0.0 0.0 0.0 0.0 0.9
Lesotho 330 5.2 7.9 0.0 0.9 0.0 0.0 0.9 0.3
Mozambique 1999 1028 7.9 16.5 1.8 5.3 0.0 0.5 2.5 10.5
Sierra Leone 463 2.6 13.0 0.2 1.3 0.6 2.4 13.0 24.0
Sierra Leone 1997 117 3.4 10.3 0.0 0.9 0.0 0.0 14.5 21.4
South Africa
(Hlabisa)
South Africa 1997 661 3.5 5.6 0.2 1.7 0.0 0.5 2.3 3.8
(Mpumalanga)
Swaziland 334 3.9 9.0 0.0 0.9 0.3 0.9 2.4 7.2
Uganda 1997 374 6.2 12.5 2.0 5.8 3.0 8.0 5.6 11.2
Zimbabwe 676 6.7 20.0 1.1 3.6 0.2 1.1 11.1 24.1

Health Organization 2000b). This strategy must be
based on the existence of an effective DOTS pro-
gramme. (See below)
46.2
Examples of Important Areas of MDR-T8
46.2.1
India
The epidemiology of tuberculosis in India is bedev-
illed by a number of confounding factors. Notifica-
tion is seldom if ever performed, few laboratories
in the country reliably perform sensitivity testing,
laboratory techniques are not standardised, and most
existing Indian studies fail to reliably distinguish pri-
mary from secondary drug resistance. Indeed, India's
TB data recording is woeful. In a study by Raviglione,
India was the only one of eight world regions with
less than 2% of its population covered (in 1997) by
WHO control strategy (Raviglione et al. 1997). It was
also the region with the worst kept data on case hold-
ing, cure, and outcome rates. These provisos must be
kept in mind when the epidemiology of MDR-TB is
discussed.
India fared poorly in the two global studies on
MDR-TB conducted by the WHO/ IUATLD Global
Surveillance Project. In the first conducted by Cohn
et aI., Indian MDR-TB rates in the four Indian studies
included in this Medline search, varied from 0.9% in
Jaipur to 33.8% in Gujarat (Cohn et al. 1978). These
rates were the second highest worldwide after Nepal.
However this study can be criticised for including
only 7939 isolates from four laboratories of widely
varying standards. Besides, primary resistance was
not reliably distinguished from secondary resistance
in any of these samples. The second study, conducted
by Pablos-Mendez was a major global co-ordinated
effort including data as it did from 35 countries
between 1994 and 1997. Here too, India emerged
with the second highest MDR rates globally after
Latvia (Pablos-Mendez et al. 1998). The combined
MDR-TB rate from India was 13.3%. This study too
can be criticised for including small numbers of iso-
lates (2240) from only one region in the country (New
Delhi). Again, as in Cohn's study, primary resistance
could not be reliably distinguished from acquired
resistance.
Both these global studies probably under-estimate
the extent of the MDR-TB problem in India. A closer
search reveals several other Indian MDR prevalence
studies, with primary prevalence rates varying from
0.9% to 36.6% and acquired resistance rates varying
from 6.3 to 56.6% (Gothi and Sen 1996). The highest
prevalence rates ever reported have been from centre
at the Hinduja hospital in Bombay (Udwadia et al.
Multi-Drug-Resistant Tuberculosis
1996). Here, Level 2 mycobacterial laboratory, which
serves as a reference lab for the teeming metropolis
of Bombay, currently reports around 60% of all iso-
lates it receives as MDR.
46.2.1.1
DOTS and its Impact in India
Most DOTS principles originated in Indian stud-
ies that emerged from research arising from the
Tuberculosis Research Centre in Madras and the
National Tuberculosis Institute in Bangalore (Bayer
and Wilkinson 1995). These two centres contributed
truly pioneering research in the 1950s and 1960s that
helped shape not only India's TB control programmes
but also those of many developed nations. It was in
1993 that DOTS began in India, starting with pilot
studies covering 2.35 million people in five states.
The numbers covered have steadily increased so
that 250 million were covered by the end of 2000,
and further cover of 100 million people each year is
planned, until the entire population is eventually cov-
ered. This makes the Indian DOTS programme the
second largest in the world after China. Case detec-
tion rates are higher than those reported from China
and Bangladesh and success rates are comparable to
those from South East Asia and Africa. Where DOTS
is used in India, TB patients are twice as likely to be
cured. Where DOTS is not used patients are seven
times more likely to die of TB. A recent report by
reveals that detection rates in all Indian states have
steadily increased over time with treatment outcomes
also steadily improving from 72% in 1993 to 90% in
2000. Overall failure rates run at an impressive 3.5%,
default rates at 9% and death rates at 4% (Khatri and
Frieden).
Thus the initial gains made by DOTS in India
have undoubtedly been impressive but it is doubt-
ful if the success of DOTS can be sustained to cover
the entire country. In the words of Mukund Uplekar,
noted TB health researcher; "DOTS will be as good
and as effective as the general health services in the
country:' Thomas Frieden, while WHO director S.E.
Asia, is cautiously optimistic about DOTS but admits;
"The key challenge is to balance the urgent need for
rapid expansion with the paramount importance of
ensuring quality of implementation:'
46.2.1.2
Practical Problems DOTS Faces In India
1. A verifiable address is mandatory before an Indian
patient can be registered at an Indian DOTS centre.
813
Where does that leave the homeless, the slum
dweller, the pavement dwellers and the migrants
that form up to 50% of the population of some big
Indian cities like Bombay?
2. Women are more hesitant to attend DOTS centres
in India finding them too obtrusive. Indeed tuber-
culosis in India robs Indian women of the little
respect and standing they have in society. Visit-
ing a DOTS clinic is something Indian women
are reluctant to do for fear of being stigmatised
further. It is less conspicuous for them to visit a
private doctor and as a consequence many Indian
DOTS clinics have skewed male to female ratios of
3:1.
3. 'Difficult' patients seem to be deliberately weeded
out in an attempt to make results appear more
impressive. Thus, the alcoholics, the drug addicts
and the marginalized whom DOTS centres should
be striving to incorporate find themselves being
excluded.
4. Patients complain that DOTS clinic hours are too
rigid and inflexible. A daily-wage labourer often
ends up going without pay, and his family there-
fore without food on the days they visit a DOTS
clinic. Travelling large distances across harsh ter-
rain in a country as vast as India is a real problem
in DOTS clinics located in more remote parts of
the country. This becomes particularly difficult
for the old, the sick, the invalid and the poor. The
incentives, which are an important component of
DOTS programmes in the developed world, are an
unaffordable luxury in India.
5. Private patients are outside the purview of DOTS.
Private practitioners rarely refer their patients to
DOTS centres, as this would mean loss of control.
Since at least 50% of TB patients would choose to
first visit a private practitioner over attending a
government clinic a large segment of the popula-
tion is excluded.
6. The extra workload imposed by DOTS may over-
whelm the system. The pre-DOTS system strug-
gled to cope with the large numbers of patients
it faced. Each patient registered at a DOTS clinic
needs 40 episodes of supervision over the course
of his or her six-month SCC. This may prove an
unmanageable additional load that under-staffed
and under-funded DOT centres may eventually
buckle under the strain. Successful implementa-
tion of DOTS would need much more in terms of
staff, infrastructure and funds than can be pres-
ently afforded. This is currently estimated to run
into an incremental recurrent cost of $100 million
a year.
814
46.2.2
MDR-T8 in Africa
A total of 80 % of all the reported incident cases were
said to have occurred in 22 countries and eight of the
ten countries with the highest incidence were found
in Africa. This increased burden of disease from
tuberculosis globally, in an era where there is effective
drug treatment, has been ascribed to factors such as
poor control in areas such as Southeast Asia, eastern
Europe and sub-Saharan Africa and co-infection with
HIV in some African countries (Dye et al. 1999).
Though Southeast Asia and the Western Pacific
regions had the highest number of reported cases of
TB, Africa had the highest incidence rate per capita
with an average of 259/100,000. This region also
has the highest rate of HIV infection in TB patients,
with an average of 32% (Dye et al. 1999) though
this figure varies from 17.8% in countries such as
Cameroon (Bercion and Kuaban 1997) to as high as
70% in countries such as Botswana, Zambia (Elliott
et al. 1995) and Zimbabwe. The case fatality rate for
tuberculosis in Africa is high, exceeding 50% in some
African countries, compared to a global figure of 23%
(World Health Organization 2000a). The high rate of
HIV co-infection has contributed to this high case
fatality rate.
The available data on the rates of drug resistance
in Africa is not extensive as many countries have
not conducted nationwide surveillance of the levels
of drug resistance. The available reports generally
cover a small area of the country or the sample size is
usually small and hence cannot be considered to be
representative of the country as a whole. In collabora-
tion with the International Union Against Tuberculo-
sis and Lung Disease (IUATLD), WHO has conducted
a major surveillance of the global rates of drug
resistance between 1994 and 1996 (World Health
Organization 1997a) and this exercise was repeated
between 1997 and 1999 (World Health Organization
2001). During the surveillance 16% of the countries
were sampled. In the first round eight countries from
the AFRO region (sub-Saharan Africa) were included
- these were Benin, Botswana, Ivory Coast, Kenya,
Lesotho, Sierra Leone, Swaziland and Zimbabwe.
In the second surveillance seven African countries
were included (Botswana, Central African Republic,
Guinea, Mozambique, Sierra Leone, South Africa and
Uganda). Thus only Botswana and Sierra Leone were
assessed in both surveys.
The levels of primary drug resistance in Africa vary
according to the country assessed. The presence of
drug resistance is an indicator of the treatment deliv-
P. D. O. Davies
ery process and resistance occurring in patients with
no prior history of treatment reflects poor treatment
in the past. The rate of drug resistance in patients
with a previous history of treatment for tuberculosis
is always much higher than in patients with no his-
tory of previous treatment for tuberculosis.
In general the levels of drug resistance in Africa
are low when compared to other parts of the world.
This is despite the HIV associated increase in TB
cases and political strife and wars. This is probably a
reflection of the presence of relatively well function-
ing control programmes, 61 % of the countries in the
African Region of WHO were covered by the DOTS
strategy compared to the global average of 42.6%.
The more recent introduction of rifampicin may also
contribute to this lower incidence of drug resistance.
In the first survey conducted by WHO and
IUATLD the overall level of resistance among new
cases for the African countries were as follows;
Zimbabwe 3.3%, Botswana 3.7%, Kenya 6.3%, Benin
8.4%, Lesotho 8.8%, Swaziland 11.7%, Ivory Coast
13.4% and Sierra Leone 28.1% (Table 46. 1)(World
Health Organization 2000a). In this survey, the
Dominican Republic had the highest level of resis-
tance to any drug, with a rate of 40.6%. Overall drug
resistance to any drug in the second survey con-
ducted between 1996 and 1999 in the African coun-
tries were as follows; Botswana 6.3%, South Africa
(Mpumalanga Province) 8.0%, Guinea 14.7%, Cen-
tral African Republic 16.4%, Uganda 19.8%,Mozam-
bique 20.8% and Sierra Leone 24.8%. Between the
two surveys, the level of resistance in Botswana
increased from 3.7% to 6.3%, while for Sierra Leone
the rate reduced from 28.1 % to 24.8%. The rate of
resistance in other countries not included in the
WHO survey vary widely from country to country.
The available figures are as follows; Malawi 11.8%,
Cameroon 31.8% (2), Ghana 54.5%, Ethiopia 15.6%
and Tanzania 2.8%, Senegal 37% (Demissie et al.
1997; Range et al. 2000).
The levels of resistance to isoniazid in new
patients or primary resistance varied from 1.2% in
Ivory Coast to 12.4% in Cameroon. Resistance to
streptomycin ranges from 0% in Kenya to 20.5% in
Cameroon. Levels of mono-resistance to rifampicin
are low, being less than 1% in most countries where
it has been reported (Botswana, Benin, Sierra Leone,
Central African Republic, Guinea, Cameroon) and
only Mozambique (1.8%) and Ethiopia (1.8%) had
higher rates. Only Sierra Leone (0.6%), Swaziland
(0.3%), Cameroon (0.4%) and Uganda (2.4%) have
reported any mono-resistance to Ethambutol. The
low levels of mono-resistance to rifampicin reflect
Multi-Drug-Resistant Tuberculosis 815

the recent introduction of rifampicin containing
regimens as well as the tendency to use rifampicin as
a combined tablet with isoniazid.
Available data on drug resistance indicates that
the rates of resistance are higher to one drug than
to two or more drugs. The level of multi-drug resis-
tance in Africa is relatively low with the highest level
reported for Ivory Coast (5.3%) (Dosso et al. 1999),
Mozambique (3.5%), Zimbabwe (1.9%), South Africa
Mpumalanga Province (1.5%), Sierra Leone (1.1 %)
(Range et al. 2000) and other countries reporting less
than 1% of MDR (Botswana, Benin, Lesotho, Swazi-
land, Guinea, Uganda and Tanzania)(Table 46.2). In
Kenya there was no multi-drug resistance reported in
the WHO survey conducted between 1994 and 1997,
though in the refugee population the level of MDR
was 2.9% (Githui et al.I999). It should be noted how-
ever, that the countries included in the WHO surveys
are countries, which have reasonably functioning
programmes. In countries where there is a lack of a
well-functioning programme the levels of MDR-TB
may be much higher. In a survey conducted in Cam-
eroon without a functioning control programme,
among patients with a previous history of treatment,
multi-drug resistance was observed in 27.6% of the
cases (Kuaban et al. 2000).
The level of drug resistance is substantially higher
in patients with a history of prior treatment with anti-
tuberculosis drugs than in patients with no previous
treatment for tuberculosis. This fact supports the
impression that the development of drug resistance
is primarily associated with irregular medication
and poor control of tuberculosis. In Durban, South
Africa, the strongest predictor of drug resistance was
a history of prior treatment with anti-tuberculosis
therapy (Table 46.3) (ATT.) (Anastasis et al. 1997). In
Hlabisa, Kwa Zulu-Natal, South Africa, resistance to
isoniazid was 6.4% in cases with no prior history of
treatment compared to 13.6% in cases with a history
of prior treatment (Davies et al.). In Uganda, resistance
to isoniazid in cases with no history of prior treatment
was 6.7% compared to 37.8% in those cases with prior
treatment. Rates of MDR are similarly higher in cases
with prior treatment (4.4%) compared to those with no
prior treatment (0.5%) (Bertzel et al.1999).
No significant association has been observed
between the level of resistance to TB drugs and HIV
serostatus in several studies (Bercion and Kuaban
1997; Anastasis et al. 1997; Churchyard et al. 2000;
Braun et al. 1992). However the presence of the
increased rates of infection due to co-infection with
HIV has placed a strain on the existing control mea-

Table 46.3. Levels of drug resistance on patients with a history of prior treatment

Country Sample Overall Resistance to poly-resistance

1 drug 2 drugs 3 drugs 4 drugs any MDR

Benin
Botswana 114 14.91 7.0 2.6 0.9 4.4 7.9 6.1
Botswana 145 22.8 12.4 6.2 4.1 0.0 10.3 9.0
Cameroon
CAR 33 36.4 12.1 6.1 15.2 3.0 24.2 18.2
Ethiopia
Ghana
Guinea 32 50.0 9.4 12.5 15.6 12.5 40.6 28.1
Ivory Coast
Kenya 46 37.0 30.4 6.5 0.0 0.0 6.5 0.0
Lesotho 53 34.0 20.8 5.7 5.7 1.9 13.2 5.7
Mozambique 122 45.1 22.1 21.3 0.8 0.8 23.0 3.3
Sierra Leone 172 52.9 16.3 24.4 5.2 7.0 36.6 12.8
Sierra Leone 13 61.5 30.8 7.7 23.1 0.0 30.8 23.1
South Africa
(Hlabisa)
South Africa 100 22.0 11.0 11.0 0.0 0.0 11.0 8.0
(Mpumalanga)
Swaziland 44 20.5 9.1 4.5 2.3 4.5 11.4 9.1
Uganda 45 51.1 28.9 20.0 2.2 0.0 22.2 4.4
Zimbabwe 36 13.9 5.6 5.6 2.8 0.0 8.3 8.3

N.B. Table 1-3 are adapted from the WHO/IUATLD Global Project on Anti-tuberculosis Drug
Resistance Surveillance and other available data
816
sures for tuberculosis and the ensuring falling quality
of control can lead to an increase in the overall levels
of drug resistance and in particular for MDR-TB. Due
to high levels of HIV infection, the health system has
been over-burdened. In some countries up to 75% of
the hospital admissions are due to HIV related infec-
tions, hospitals are over-crowded, bed-occupancy
rates are very high and isolation of infectious cases
may not be possible. Hence the situation where a
patient with open TB may be next to a patient with
HIV related complications. The possibility of noso-
comial infections are therefore high. Thus there is
the possibility that drug resistant strains will easily
spread within the community.
Though the levels of drug resistance in Africa are
relatively low compared to other countries such as
Russia and Estonia, it is vitally important that every
effort is made to maintain a low incidence. Currently
many African countries with high levels of tuberculo-
sis are unable to adequately fund tuberculosis control
efforts without external support from cooperating
partners. Though the cost of treatment for tuber-
culosis is one of the most cost effective strategy, in
some countries this figure is more than the per capita
expenditure available for the whole health budget.
It therefore follows that treatment of multi-drug
resistance tuberculosis will be beyond the reach of
most countries (Schlunger 2000). Thus prevention of
the development of resistance should be viewed as a
major priority for these countries.
46.3
Drug Resistance in Prisons
The world's prisons, jails, police cells and detention
centres are inherently unhealthy because people with a
number of diseases including HIV and tuberculosis are
concentrated. Rates of tuberculosis in prisons are very
high sometimes as much as 100 times as common as
in the community (World Health Organization 1997b).
The International Committee of the Red Cross (ICRC)
found that the rate of TB in prison was 60 times higher
than in the community and was a major cause of death
(ICRC 1998). In Kazakhstan in 1997, 1034 of the 1491
(63%) prisoner deaths that year were from tuberculosis
(Tuberculosis in Prisons Roundtable 2000).
If the treatment of TB in such an environment is
not well organised, MDR-TB can result. Information
on the extent of MDR in prisons is limited but recent
figures from WHO suggest that they may be as high
as 24% (Stern 1999).
P. D. O. Davies
Countries of the former Soviet Union are par-
ticularly at risk. An ICRe study found that 69% of
patients with tuberculosis had resistance to at least
one drug (Amnesty International 1999).
46.4
The Microbiology of Drug-Resistant
Tuberculosis
The principal aims of TB microbiology are, first, the
detection of mycobacteria, either by microscopy or
molecular techniques, secondly, the isolation and
identification by culture, using solid or liquid media
and thirdly, the susceptibility testing of cultures
and the identification of drug resistant, particularly
multiple drug resistant (MDR TB) isolates. The
laboratory is at the core of our understanding of the
epidemiology of TB transmission and surveillance of
drug resistance patterns.
Since the development of acid-fast staining tech-
niques in 1882 by Robert Koch, the principles of TB
microscopic diagnosis have changed little and the
acid-fast smear is still the most rapid method for the
detection of M. tuberculosis. The sensitivity is only
around 5_103 per ml and microscopy must be supple-
mented with culture, which while sensitive, is slow,
taking 4-6 weeks to produce a positive result on most
solid culture media, such as L- J slopes. The recent
development of automated liquid culture systems
such as the Mycobacterial Growth Indicator Tubes,
(MGIT Bactec 960, Becton Dickinson, Franklin Lakes,
USA) and the MB BacT Alert systems (Organon
Teknika, Boxtel, Netherlands) have improved time
to detection to 12-18 days and reduced hands-on
time compared to their semi-automated predecessor,
the Bactec 460 TB. For optimal results these systems
should be used in conjunction with solid-media
based culture. Safety of liquid culture systems is an
important issue, as handling of liquid cultures carries
an increased risk of potential microbial contamina-
tion of the laboratories from culture vials.
46.4.1
Molecular Diagnostic Techniques
The advent of molecular technology has led to the
development and implementation of tests, which
are both rapid and sensitive for the detection, and
identification of mycobacteria. They can be used in
several ways: for the detection of mycobacteria, for
Multi-Drug-Resistant Tuberculosis
the specific detection of M. tuberculosis complex, for
the differentiation of M. tuberculosis complex and
MOTT (mycobacteria other than tuberculosis), and
for the detection of drug resistance.
Molecular detection methods utilise a series of
common steps, starting with the amplification of
nucleic acid sequences, which can be achieved using
several techniques. The best known and most widely
employed is the polymerase chain reaction (PCR).
Other techniques include transcription mediated
amplification (Amplified MTD-2, Gen-probe, San-
Diego, USA) (Miller et al. 1994), strand displacement
amplification (Probe-Tec ET, Becton Dickinson,
Franklin Lakes, USA), (Bergmann et al. 2000) ligase
chain reaction (LCX MTB, Abbott, Illinois, USA)
(Piersimoni et al. 1998) and QB replicase Galileo
system (Gene-TraklVysis, Illinois, USA) (Smith et al.
1997; Della-Latta 1999). These alternative techniques
have shown comparable performance to PCR. While
some have been withdrawn at the clinical trial stage
(Vitros, Johnson & Johnson, Rochester, USA & QB
Replicase Galileo System, Gene-TraklVysis, Illinois,
USA) many are now on the market with sensitivity
and specificity consistently evaluated at over 90%.
Detection systems can be colorimetric, such as in the
Line probe assay, or LiPA, (Innogenetics, Zwijndre-
cht, Belgium) and Amplicor TB (Roche Diagnostics,
Indianapolis, USA) or through chemiluminescent,
or fluorescent, tagging of probes (Drobniewski et al.
2000a).
A major problem for these molecular techniques is
cost. While it is relatively easy to scale-up production
of kits of molecular detection methods, once they
have been developed, their execution requires skilled
laboratory personnel, the concurrent performance
of multiple controls, several dedicated laboratory
areas and the use of relatively expensive consumable
reagents. All of these factors make the costs prohibi-
tive for laboratories in developing nations. Further-
more, in developing nations, where perhaps 95% of
mycobacterial disease is due to MTB complex, the
use of molecular techniques to differentiate MOTT
is of little benefit as appropriate therapy for MOTT
disease is usually unattainable. Therefore kits such
as the Accuprobe system for the identification of a
small number of pathogenic mycobacteria, which
was introduced commercially on a wide-scale in 1987
(Gen-Probe, San-Diego) would be of questionable
value in these settings. A similar argument applies to
molecular amplification based methods.
817
46.4.2
Molecular Testing for Drug Resistance
Mycobacteria spontaneously develop drug resistance
and the rates of mutation are different for each drug.
For M. tuberculosis these have been shown to be
approximately 108 to 109 for isoniazid and streptomy-
cin, 10 10 for rifampicin, 107 for ethambutol and 109 for
cycloserine (Gangadharam 1984). It has been known
almost since the inception ofTB chemotherapy, when
the first relapses following streptomycin monother-
apy were seen, that therapy with any single TB drug
leads to the selection of a drug resistant population.
For example mono-therapy with streptomycin led to
a rise from 1 in 88,750 drug resistant mutant bacteria
in sputum to 1 in 367 after 15 weeks of treatment
(Herbert et al. 1999). Combination therapy must be
used, therefore, to reduce the probability of drug-
resistant strains emerging. Multiple drug resistant
TB (MDR TB) is defined as resistance to at least
isoniazid and rifampicin. In the UK, drug resistance
has been monitored by the laboratory-based system
MYCOBNET since 1994. MDR-TB prevalence in
initial isolates for the UK rose from 0.6% to 1.7%
between 1993 and 1996, and then fell to 1.1% in 1999.
In 1999,6.3% of initial TB cases were resistant to one
or more first-line TB drugs: 5.9% resistant to isonia-
zid, 0.9% resistant to rifampicin, 0.5% to ethambu-
tol, 0.4% to pyrazinamide, and 1.1% were MDR-TB
(Telenti et al. 1993). The traditional determination
of drug-resistance is by resistance ratio, minimum
inhibitory concentration (MIC) or the proportion
methods. These can be done in liquid media such as
Middlebrook 7H9, solid media such as Lowenstein-
Jensen slopes, or Middlebrook 7H10 agar. As with iso-
lation techniques, faster susceptibility testing meth-
ods have been developed using the semi-automated
Bactec system or automated systems such as MBBacT,
MGIT and Bactec. The first stage in the development
of molecular detection methods for drug resistance
is the identification of the relevant genes and muta-
tions involved in drug resistance. The simplest drug
to analyse has been found to be rifampicin, where
approximately 95% of all clinical isolates resistant to
rifampicin have a mutation in a 81 base pair region
(codons 507-533) of the rpoft gene encoding the
beta chain of the DNA dependent RNA polymerase
(Zhang et al. 1992). This has made the development
of genotypic methods for the detection of rifampi-
cin resistance relatively straightforward. As approxi-
mately 90% of all rifampicin resistant isolates in the
UK are also resistant to isoniazid, a positive result for
rifampicin resistance can be taken as a strong indica-
818
tor of MDR TE. However, isoniazid resistance is much
more complex, as at least four genes are known to be
important. These are the katG gene, which encodes
the catalase-peroxidase enzyme, inhAimabA, which
has a role in fatty acid elongation (Quemard et al.
1995) ahpC, which encodes the alkyl hydroperoxide
reductase C (Sreevatsan et al. 1997) and oxyR which
is the oxidative stress regulator (Alcaide et al. 1997).
Two genes have been identified which are important
in streptomycin resistance, the rrs gene, which codes
for the 16S rRNA, and rpsL which codes for ribo-
somal protein S12. Similarly, Ethambutol- resistance
involves mutations in the embA, B, and C, genes (Sre-
evatsan et al. 1997, ) encoding enzymes involved in
lipoarabinomannan and arabinogalactan synthesis.
Ciprofloxin resistance is conferred principally by
mutations in gyrA, gyrB and nor genes, encoding
DNA gyrase subunits A and B and an efflux protein,
respectively. Pyrazinamide resistance is conferred by
mutations in the pncA gene, which encodes pyrazin-
amidase. In a study by Sreevatsan et al. (1997; Telenti
et al. 1993). Although 72% of pyrazinamide resistant
isolates carried mutations in this region there were
many mutation sites throughout this gene, which
makes the development of tests for pyrazinamide
resistance far more complex than rifampicin.
Once genes such as these have been sequenced
and the specific mutations identified, there are
several options for the genotypic detection of drug
resistance. The gold standard is DNA sequencing, but
this is impractical in routine clinical use, due to the
expense, skill and time demands of the technique,
although systems have become increasingly user-
friendly. Alternative techniques which have been
employed include PCR single-strand conformation
polymorphism (PCR-SSCP) (Telenti et al. 1993c)
analysis, heteroduplex analysis, mutation specific
priming, restriction enzyme analysis (Rossau et
al. 1997) and solid-phase hybridisation methods
(Carrier et al.I997).
PCR-SSCP relies on the difference in tertiary
structure between two single strands of DNA,
which differ by single mutations or more. The con-
formational change of the tertiary structure can be
detected by a change in mobility when analysed on a
polyacrylamide gel.
Heteroduplex analysis involves mixing ampli-
fied DNA from the sample with that of reference
drug-sensitive strains. The DNA is then denatured
and cooled again to allow the DNA to hybridise into
hybrid double-stranded DNA. The DNA is then ana-
lysed on a denaturing electrophoresis gel. Where the
sample DNA carries a drug-resistant mutation there
P. D. O. Davies
will be a mismatch in the complimentary base-pair-
ing and the heteroduplex will have a different mobil-
ity to the homoduplex.
Solid-phase hybridisation analysis involves the
immobilisation of DNA probes complementary to
the most common mutations and to drug sensitive
patterns. Amplified DNA can then be hybridised to
the probes and detected through fluorescence, radio-
label or colorimetric reaction in the normal way. One
such assay, which is commercially available is the
line-probe assay (LiPA, Innogenetics, Zwijndrecht,
Belgium) for the detection of rifampicin resistance.
One possible disadvantage of genotypic detection
methods is that they give no indication of the suscep-
tible:resistant ratio of organisms in a population and
therefore may lead to the withdrawal of drugs from a
regime when their therapeutic value is still high.
Other rapid detection systems have recently been
developed based on phenotypic methods, which can
be adapted for use as susceptibility tests. One such
technique utilises mycobacteriophage such as phAE40
or phG518. Jacobs et al. (Jacobs 1993; Riska et al.1999)
inserted the lux gene, which codes for luciferase, into
the phage genome. This enzyme catalyses the reac-
tion of luciferin with ATP and emits light, which can
then be detected. If the mycobacteria are grown in the
presence of a drug, where the organism is susceptible
the mycobacteria will die and therefore the phage will
fail to replicate and no light will be emitted. While
this technique allows all types of resistant organism
to be detected, regardless of genotype, it requires
expensive equipment and therefore unsuitable for
the developing world where rapid detection sys-
tems for drug resistance are most urgently required.
The same group has addressed this problem by
developing a simpler microtitre plate format with a
film detection system (Wilson et al. 1997). A more
economical approach is the PhaB (phage amplified
biologically) assay (Eltringham et al. 1999). Myco-
bacteriophage are added to the test sample which
then infect any mycobacteria present where they are
protected from the subsequent addition of viricide,
which kills all external phage. The viricide is then
removed through wash steps and the sample incu-
bated until lysis of the bacteria releases the phage.
The phage progeny are detected by plating onto a
lawn of rapidly growing Mycobacterium smegmatis
in which, following overnight incubation, they form
plaques. The system has been adapted to allow sus-
ceptibility testing and showed 100% correlation with
resistance ratio testing for rifampicin (Eltringham
et al. 1999b), 94% for isoniazid, 96% for streptomy-
cin, 100% for ciprofloxacin, 88% for ethambutol and
Multi-Drug-Resistant Tuberculosis
87% for pyrazinamide (Kirk et al. 1998). This system
is also relatively simple, and cheap to perform.
However, stringent precautions must be taken in a
diagnostic or reference laboratory to avoid the con-
tamination of reference strains and archive samples
with mycobacteriophage. Other rapid phenotypic
methods which have been investigated include flow
cytometry (Cangelosi et al. 1996), which requires the
provision of prohibitively expensive equipment, and
reverse transcriptase PCR (RT-PCR).
One such RT-PCR assay which measured a reduc-
tion in inducible heat shock protein dnaK mRNA
levels in susceptible isolates exposed to rifampicin
showed a 96% (46/48) correlation in susceptible
isolates and 97% (35/36) correlation in resistant iso-
lates (Eltringham et aI.1999). Alternative RNA based
approaches have included measuring levels of pre-
16S rRNA stem sequences through hybridisation
to specific radiolabelled probes for rifampicin and
ciprofloxacin (Hellyer et al. 1999) and quantitative
analysis of 85B mRNA for rifampicin and isoniazid
(Watterson et al.).
46.4.3
Rapid Tests in Routine Use
The Public Health Laboratory Mycobacterium Refer-
ence Unit (PHLS MRU) offers a Fastrack diagnostic
service for the detection of rifampicin resistance,
using the LiPA (Innogenics, Zwijndrecht, UK) com-
mercially available kit (Drobniewski et al. 2000).
Correlation with standard methods has been over
90%. Where Fastrack was performed on cultures
the correlation was 33/36 (91.7%), on primary speci-
mens,55/61 (90.2%) and on smear positive sputum,
44/48 (91.7%). This rapid detection system allowed
drug resistance to be detected on average 27.6 days
earlier (n=56) for smear positive sputum specimens,
27.7 days earlier (n=61) for all primary specimens,
and 19.1 days (n=36) when used for cultures
(Raviglione et al. 1997).
46.4.4
How the Sequencing of the Mycobacterial
Genome Will Help the Development
of New Drugs in the Treatment of Tuberculosis
Elucidation of the complete genome sequence of M.
tuberculosis provides access to an immense reservoir
of fundamental information about the evolution and
constitution of the organism (Cole et al. 1998). Use
819
of this information to enhance understanding of
the biology of tuberculosis is the driving force for
research in the post-genomic era (Young 2001). In
principle, the genome sequence includes information
about all of the possible targets to which new anti-
mycobacterial agents might be directed. At the most
direct level, structural and functional information
about a particular protein target can be deduced from
the sequence of its encoding gene. Further analysis of
interactions between individual gene products pro-
vides information about the biosynthesis of more
complex molecular structures, and about the flux of
metabolites essential for bacterial viability. How can
we use this new information resource to accelerate
development of reagents for improved tuberculosis
control?
46.4.4.1
Breaching the Wall of "Fortress Mycobacterium"
The outer coat is commonly viewed as the defining
characteristic of a mycobacterium, determining its
microbiological staining properties and its relation-
ships with external reality (Brennan and Nikaido
1995). Many current anti-mycobacterials (isoniazid,
ethambutol, ethionamide and pyrazinamide) affect
the processes involved in cell-wall biosynthesis, and
there is every likelihood that novel reagents targeted
against similar processes will have comparable effi-
cacy. Many of these biosynthetic pathways are now
understood at a genetic level. Knowledge built up
through decades of traditional biochemistry allows
identification of the genes encoding enzymes that
make up the pathways, and recombinant DNA tech-
niques have opened a new era of detailed structure-
function analysis. One approach to exploiting this
information in drug discovery programmes is to
clone and express genes encoding selected biosyn-
thesis enzymes. The recombinant proteins can be
used in functional assays to screen for inhibitors,
and at the same time provide the opportunity to
generate structural information for optimised drug
design (Dessen et al. 1995). Following an alternative
approach, changes induced in the bacteria as a result
of the action of an existing drug - characteristic tran-
scriptional signals, for example (Slayden et al. 2000)
- can be used to screen for novel compounds that
exert an analogous inhibitory mechanism. Both of
these strategies are suitable for use in high-through-
put screen formats, allowing testing of millions of
novel compounds. Compounds for such screens are
typically derived by combinatorial chemistry tech-
niques, which use simple building block strategies
820
to facilitate rapid generation of extensive libraries of
structurally diverse small molecules. The principles
and practice of modern high-throughput screening,
and potential applications to tuberculosis, are com-
prehensively described in the Scientific Blueprint
recently developed by the Global Alliance for TB
Drug Development (Global Alliance for TB Drug
Development 2001).
Drug discovery programmes focused on cell-wall
targets have a strong scientific foundation and a
well-defined rationale; there are excellent prospects
for identification of novel inhibitory compounds.
This approach has the limitation, however, that new
compounds are likely to have properties broadly sim-
ilar to those of current drugs. It may be possible to
selectively improve the pharmacokinetic properties
and avoid established resistance mechanisms, but it
is hard to envisage new cell-wall inhibitors driving
major changes in the overall design of treatment regi-
mens. They would clearly be drugs of choice for treat-
ment of resistant disease, but cost factors are likely to
argue against their routine replacement of existing
drugs. In this particular case, would a pharmaceuti-
cal company be able to recoup the considerable costs
involved in taking a novel compound through the
complex process of developing a promising 'hit' into
a clinically useful drug? This would seem to represent
a significant economic risk, particularly in light of
strong public opposition to the practice of marketing
drugs in low-income countries at prices, which reflect
research and development costs. The economic risk
will have to be borne by the public sector; either by
funding a guaranteed market for new drugs to treat
multi-drug-resistant tuberculosis, or by carrying
out at least parts of the development process within
public or charitable institutions. The cost of develop-
ment of a new drug is generally estimated to be of the
order of half a billion dollars. Given the magnitude
of the effect of tuberculosis on the global health and
economy, this figure is well within the budget of insti-
tutions such as the World Bank, and mobilisation of
public-private financing to promote development of
new cell-wall inhibitors warrants high priority in the
battle to control tuberculosis.
46.4.4.2
In Vivo Phenotypes: Targeting Persistent Organisms
The economic case for drug development is much
stronger for a compound, which would significantly
improve current therapy. A drug which reduced
current treatment regimens from six months to six
weeks is likely to be adopted as standard therapy,
P. D. O. Davies
for example, and would have widespread application
even if unit costs were higher than existing drugs.
The biological factors dictating the requirement for
prolonged therapy in tuberculosis are the subject of
extensive speculation (McKinney 2000).A reasonable
hypothesis is that a proportion of the bacteria per-
sist in a form in which they are relatively resistant
to the action of the drug. For example, in the case of
a drug, which acts by inhibition of cell-wall biosyn-
thesis, bacteria will be immune to its effects if they
don't happen to be engaged in the construction of
new cell-wall components. The presence of the drug
must therefore be maintained until the persisting
bacteria enter a renewed phase of cell division. Can
this period be reduced by targeting some metabolic
process, which is essential even in non-dividing
bacteria?
It is widely recognised that all bacterial pathogens
adapt to the environmental conditions they experi-
ence within the infected host, and that the resulting
biological properties of the organism - the phenotype
- are therefore different from those seen in the micro-
biology laboratory. Many investigators are applying
post-genomic tools to study the in vivo phenotype
of M. tuberculosis; monitoring patterns of gene and
protein expression, and constructing mutants with
defects at individual genetic loci. An example of this
approach in the context of potential drug discovery is
the case of isocitrate lyase. McKinney and colleagues
(2000) demonstrated that this particular enzyme is
essential if M. tuberculosis is to set up a chronic infec-
tion in mice. In contrast to enzymes required for cell-
wall biosynthesis, isocitrate lyase is dispensable during
the initial phase of active replication; it is thought to
playa role in the lipid-based metabolism that prevails
during the persistent phase of the infection. Inhibition
of isocitrate lyase activity might therefore be a par-
ticularly effective route by which to attack persistent
bacteria. The isocitrate lyase gene has been cloned and
expressed, providing the basis for a novel drug discov-
ery programme (Sharma et al. 2000).
The targeting of in vivo phenotypes generates
scientific excitement, but the corresponding drug-
screens present important practical problems. It can be
anticipated that inhibitors of isocitrate lyase will have
little or no effect on mycobacterial growth in vitro, for
example, introducing a requirement for costly screen-
ing in animal models at an early stage in the drug dis-
covery programme. In an attempt to reproduce in vivo
phenotypes in a simpler experimental system, we have
developed an 'ex vivo' screen. This involves infecting
mice with luminescent reporter strains of M. tubercu-
losis (Snewin et al. 1999), sacrificing animals after sev-
Multi-Drug-Resistant Tuberculosis 821

eral days or weeks, and establishing cell cultures from described above - was defective in the chronic phase
spleen or lung cells. Mycobacteria maintained in these that develops subsequent to immune recognition. One
cultures resemble those in intact animals in being held possible explanation for these observations is that the
under the control of the host immune response, and higher level of antigen expression attracts a stronger
display a drug-susceptibility profile resembling that immune response, stimulating a greater effort to clear
seen in vivo. Establishment of cultures in 24- or 96- the infection. Induction of antigen expression during
well plate formats permits testing of dose-response latent infection might similarly trigger enhanced
relationships for multiple compounds from only a immune activity against the persisting organisms. An
small number of animals. intervention of this type - a drug acting in synergy
with the immune response - would be complementary
46.4.4.3 to the current control strategies based on antibacterial
Latent Tuberculosis therapy of active disease and prophylactic vaccination
prior to infection.
Elimination of latent tuberculosis presents a third
target for drug action. It is well-established that
conventional drug treatment can reduce the risk
of development of clinical disease in individuals 46.5
recently infected with M. tuberculosis; can we imag- The Treatment of MDR-T8
ine expanding this approach to prevent tuberculosis
in the one third of the global population estimated to 46.5.1
be harbouring latent infection? For 90% of infected Introduction
individuals the immune response provides a level
of protection sufficient to avoid the development of The specific management of drug-resistant patients
the active disease, and it is attractive to consider the is only possible where facilities exist for both myco-
development of drugs or vaccination protocols that bacterial culture and for drug susceptibility test-
might add to this protection. In contrast to the cell- ing, which excludes most of the developing world.
wall targets discussed above, the scientific platform Treatment guidelines in the United Kingdom, and
for a rational drug discovery programme in this elsewhere, are predicated on the drug-resistance
area has yet to be established. Latent infection is data prevailing in the circumstances of their use.
dependent on the presence of an effective immune In developed countries the inclusion of the fourth
response, but could be explained as an equilibrium drug (ethambutol but occasionally streptomycin)
reached by actively replicating mycobacteria together depends on the level of isoniazid resistance expected
with coincident immune killing, the existence of a or known in a given patient group.
non-replicating 'dormant' form of mycobacteria (per- In the United Kingdom, the British Thoracic
haps analogous to the persistent phenotype invoked Society's Guideline (Joint Tuberculosis Committee of
above), or some intermediate between these two the British Thoracic Society 1998), recommends the
extremes. Understanding how a pathogen can be so omission of ethambutol in the initial phase only if the
aggressive during active disease and yet can main- patient meets all the following criteria, based on the
tain a stable and apparently harmless interaction for drug resistance data held (Tuberculosis Update 1999;
years, or decades, within the infected host remains a Hayward et al. 1996):
fundamental challenge for tuberculosis research. • White ethnic origin
In a recent study, we have investigated the effect of • Previous untreated for tuberculosis
manipulation of the antigen expression on M. tubercu- • Known, or thought likely to be on risk assessment,
losis infection in a murine model (Stewart et al. 2001). HIV-negative
By interfering with regulatory circuits involved in the • Not a known contact of drug resistant disease.
control of gene expression, we constructed a mutant
strain of mycobacteria characterised by constitutive The European Respiratory Society (Migliori et al.
over-expression of a set of heat shock proteins. These 1999) recommend the inclusion of ethambutol in the
proteins protect cells during exposure to harsh envi- initial phase for those in WHO Treatment Group I,
ronments, but also provide an important signal by that is:- new sputum smear-positive tuberculosis,
which the immune system recognises the presence new smear-negative tuberculosis with extensive
of an infection. The mutant strain was able to initi- parenchymal involvement, new cases of severe extra-
ate an infection but - like the isocitrate lyase mutant pulmonary tuberculosis (not defined). A three drug
822
initial phase is recommended for those in WHO Cat-
egory III :- new smear-negative pulmonary tubercu-
losis (except in Category I) and new less severe forms
of extra-pulmonary tuberculosis. The American
Thoracic Society (American Thoracic Society 1994)
advise the inclusion of ethambutol or streptomycin
in the initial phase of a daily regimen at an isoniazid
resistance prevalence of 4% .
Most parts of the world do not have mycobacterial
culture and drug susceptibility testing capabilities
and in these regions the standard advised regimen
has to cover the possibility of the commoner drug
resistances. Studies in Hong Kong used regimens of
rifampicin, isoniazid and pyrazinamide, with strep-
tomycin or ethambutol, and both streptomycin and
Ethambutol (Hong Kong Chest Services 1981, 1982,
1987) and using rifampicin and isoniazid throughout
with pyrazinamide for 2, 4 or 6 months Hong Kong
Chest Service/British Medical Research Council.
Results at 30 months (Hong Kong Chest Service
1991) showed that all the regimens were effective in
patients with initial isoniazid and/or streptomycin
resistance. Treatment failures and relapse rates were
low. If routine drug susceptibility tests are not avail-
able, these regimens can be assumed to be highly
effective if such resistances are present. The results of
these studies (Hong Kong Chest Services 1981,1982,
1987) provide the rationale for using four drugs in
the initial phase, with rifampicin and isoniazid in
the continuation phase, in areas or in population
subgroups with significant incidences of isoniazid
and/or streptomycin resistance.
An analysis of the influence of initial drug resis-
tance on response to short-course regimens in the
Medical Research Council (MRC) collaborative trials
in Hong Kong, Singapore and Africa was reported
in 1986 (Mitchison and Nunn 1986). In those trials,
patients with initial isoniazid and/or streptomycin
resistance had a failure rate of 17% when given a
six-month rifampicin and isoniazid regimen, and a
failure rate of 12% in those given rifampicin in the
two-month initial phase. As the number of drugs
given in the regimen and the duration of rifampicin
treatment increased, the failure rate fell, reaching only
2% of those receiving 4-5 drugs including rifampicin
throughout a six-month regimen.
Relapse rates after chemotherapy were only
slightly increased with initially resistant organisms.
The key exception was that of rifampicin resistance,
where the outcome was much poorer.
Although this data applies to countries without
drug susceptibility testing, the view taken in the
United Kingdom is that where drug susceptibility
P. D. O. Davies
tests are available, they should be followed, and the
treatment modified (Joint Tuberculosis Committee
of the British Thoracic Society 1998). The strength
of the scientific evidence to support various recom-
mendations (Petrie et al. 1995) is also given in the
1998 BTS treatment guidelines (Joint Tuberculosis
Committee of the British Thoracic Society 1998).
46.5.2
Management of Non-MDR Resistance
46.5.2.1
Isolated Resistances
1) Streptomycin resistance. Some of the drug resistance
reported, particularly in ethnic minority groups, is
to streptomycin alone. This is not clinically impor-
tant since streptomycin is not often used as a first
line drug in developed countries, and the efficacy of
the regimen recommended for both respiratory and
non-respiratory tuberculosis is not affected.
2) Streptomycin and isoniazid resistance. Combined
streptomycin and isoniazid resistance is the com-
monest dual resistance in the United Kingdom.
Management should be the same as for isoniazid
resistance found after treatment is commenced
but with treatment fully supervised throughout.
3) Other combinations.
Other combinations of non-MDR-TB resistance
are uncommon. Treatment needs to be individualised
depending on the combination involved and is best
determined after discussion with a highly experi-
enced clinician and mycobacterial services. Patients
with drug resistance (excluding isolated streptomy-
cin resistance) should be followed up for 12 months
after cessation of therapy.
46.5.3
Multi-Drug-Resistant Tuberculosis
The risk factors for MDR-TB are those for ordinary
drug resistance but are even more exaggerated. The
United Kingdom data shows that the odds ratios for
risk factors were, previous treatment 11.7 (95% confi-
dence interval 5.5-20), HIV-positivity 8.9 (2.1-30.7),
birth in India 4.6 (1.2-6.1), residence in London 4.0
(1.7-9.3) and male sex 2.2 (1.1-4.3). In countries
without drug susceptibility testing a patient having
treatment failure should be considered to be at risk
of MDR-TB, and anyone failing a supervised Retreat-
Multi-Drug-Resistant Tuberculosis 823

ment with a WHO category II Retreatment regimen,
should be assumed to have MDR-TB.
46.5.3.1
Infection Control
Patients with suspected MDR-TB, should have
molecular testing of samples with Rifampicin resis-
tance probes where possible. Those with clinical or
microbiological suspicion/proof of MDR-TB should
be isolated in a negative pressure room, and have
infection control and HIV assessments made (The
Inter Departmental Working Group on Tuberculosis
1998; Joint Tuberculosis Committee of the British
Thoracic Society 2000). Criteria for the removal
from strict respiratory isolation are also given (Joint
Tuberculosis Committee of the British Thoracic Soci-
ety 1998, Tuberculosis Update 1999).
46.5.3.2
Clinical Management
The drug treatment of MDR-TB is time consuming
and demanding on both patient and physician. In the
United Kingdom the advice is that treatment should
only be carried out by:
A) Physicians with substantial experience in manag-
ing complex resistant cases, and
B) Only in hospitals with appropriate isolation facili-
ties, and
C) In very close liaison with Mycobacteriology Refer-
ence Centres.
This may require the transfer of patients to an
appropriate unit where both criteria (A) and (B)
above are met. Treatment of such patients has to be
planned on an individual basis (Goble et al. 1993;
Iseman 1993) and needs to include reserve drugs
(see Table 46.4). Such treatments must be closely
monitored because of the increased toxicity but,
more importantly, full compliance is essential to
prevent the emergence of further drug resistance,
so that all such treatment must be directly observed
throughout, both as an inpatient and as an outpa-
tient [B].
Treatment should start with five, or more, drugs to
which the organism is, or is likely to be, susceptible
and continued until sputum cultures become nega-
tive [B]. Drug treatment then has to be continued
with at least three drugs to which the organism is
susceptible on in vitro testing for a minimum of
nine further months, and perhaps up to, or beyond,
24 months, depending on the in vitro drug resistance
profile, the available drugs (Farmer 1999) and the
patients HIV status. Consideration may also have
to be given to resection of pulmonary lesions under
drug cover (Iseman 1993).
The outcome in MDR-TB depends on how rapidly
the diagnosis is made, what treatment and facilities
are available, and the patient's HIV status. Results
in patients who are HIV-positive have been poor
with high mortalities often because of late diagnosis
(Drobniewski 1997; Small et al. 1993), but the out-
come in those who are HIV-negative can be much
better where appropriate facilities exist 100 and where
the drug resistance profile is less extensive (Telzak

Table 46.4. Reserve drugs: dosages and side-effects

Drug Children Adults Main side-effects

Streptomycin 15 mg/kg 15 mglkg Tinnitus, ataxia, vertigo renal impairment

(max dose 1 gm)
Amikacin 15mg/kg 15mg/kg As for streptomycin
Kanamycin 15 mg/kg As for streptomycin
Capreomycin 15 mg/kg As for streptomycin
Ethionamide or 15-20mg/kg <50 kg 375 mg bd Gastrointestinal, hepatitis
Prothionamide >50 kg 500 mg bd avoid in pregnancy
Cycloserine 250-500 mgbd Depression: fits
Ofloxacin 400mgbd Abdominal distress, headache Tremulousness
Ciprofloxacin 750 mg bd As ofloxacin plus drug interactions
Azithromycin 500 mg od Gastrointestinal upset
Clarithromycin 500mgbd as azithromycin
Rifabutin 300-450mg as for rifampicin: uveitis can occur with drug interactions
e.g. macrolides. Often cross resistance with rifampicin.
Thiacetazone 4mg/kg 150 mg od Gastrointestinal, vertigo, rash Conjunctivitis. AVOID if HIV-Positive
(Stevens-Johnson syndrome)
Clofazimine 300 mg od Headache, diarrhoea, red skin discolouration
PAS sodium 300 mg/kg 10 gm od or 5 gm bd Gastrointestinal, hepatitis, rash, fever
824
et al. 1995). After treatment all MDR-TB patients
require long-term follow-up.
Michael Iseman has ,Ten Commandments' for the
treatment of tuberculosis. The first is ,Thou shalt not
add a single drug to a failing regimen'; the second
to tenth commandments repeat the first command-
ment nine times to make sure the message has been
received. It can therefore be argued that to give a
WHO re-treatment regimen (World Health Organi-
zation Tuberculosis Unit 1991). (Category II), which
adds only a single drug to the combination that has
failed (Category I) regimen, breaks the command-
ments and may actually be adding to the incidence of
MDR-TB, and this needs debating.
46.6
Tuberculosis and HIV Infection
46.6.1
Introduction
Outbreaks of multi-drug resistance tuberculosis
(MDR-TB) amongst HIV-positive patients in the
early 1990s especially in the USA highlighted the
need for improvement in public health and hospital
control of infection policy, laboratory and clinical
management, research, development and education
in tuberculosis. This new focus, together with an
increase in financial investment, has led to many
developed countries having a comprehensive tuber-
culosis control strategy for immunocompromised
individuals, which is more robust.
46.6.2
HIVandTB
The biggest single risk factor for developing tuberculo-
sis is HIV infection. In some countries in the world espe-
cially sub-Saharan Africa the co-infection rate of HIV
and TB is estimated to be over 1000 per 100,000 popula-
tion. In spite of some excellent TB control programmes,
many of these countries are still experiencing increases
in tuberculosis case rates because of HIV co-infection.
Few countries have universal HIV testing or compre-
hensive TB culture and drug-sensitivity reporting and
so the global epidemiology of MDR-TB in HIV positive
patients is as yet mostly unmeasured (Punnotok et al.
2000). In the US in 1998 it was estimated that 20% of all
patients with TB were HIV co-infected but the propor-
tion with MDR-TB was not known.
P. D. O. Davies
The reason that HIV is strongly associated with
MDR-TB is through outbreaks. These occur because
HIV positive patients have an increased risk of devel-
oping active tuberculous disease once infected with
Mycobacterium tuberculosis. In HIV non-infected
people for every ten people who are exposed and
infected with TB (whether drug sensitive or resis-
tant,) only one will develop the disease during their
lifetime. For those with HIV infection this risk of
around 10% in a whole lifetime is telescoped down
to only one to two years. Highly active antiretroviral
therapy does have an impact on rates of developing
tuberculosis however and may reduce the rates of
tuberculosis in countries where patients are offered
anti HIV treatment.
46.6.3
MDR-TB Outbreaks
These have occurred in hospitals in Europe
(Hannan et al. 2001; Moro et al. 2000; Breathnach
et al. 1998; Coronado et al. 1993; CDC 1993) and
clinics in the USA (Pitchenik et al. 1990) for HIV
positive patients, substance abusers (Conover et
al. 2001) in prisons (CDC 1992), and homeless
shelters. From 1990 to 1992 nine large outbreaks
of MDR-TB were reported from the USA, all the
organisms isolated were resistant to Isoniazid and
Rifampicin (as this is a definition of MDR-TB) most
had Streptomycin and Ethambutol resistance also.
(Table 46.5). The HIV infection rate amongst these
patients was from 20-100% and the mortality was
from 60-89%. The interval from TB diagnosis to
death was between 4-16 weeks. Various strains
of MDR-TB were circulating at this time, espe-
cially the notorious ,W' strain, which infected 199
patients in New York from 1991-1994 and involved
30 hospitals (Shafer et al. 1995). Other types the
named ,N2', ,WI' and ,AB' infected a large number
of patients in 10-16 hospitals. The majority of these
outbreaks were brought under control because of
public health administrative measures including
infection control policies for patients with HIV
who have a cough, segregation of potential infec-
tious patients, the use of negative pressure rooms
and submicron masks (Moro et al. 1998; Maloney
et al. 1995; Stroud et al. 1995). Procedures such as
nebulisation of pentamidine for PCP prophylaxis,
saline for induced sputa or even salbutamol for
those with obstructive airway disease were no
longer performed in open areas but confined to
negative pressure rooms.
Multi-Drug-Resistant Tuberculosis 825

46.6.4 46.6.6
Treatment Contacts

The drug treatment regimen used for outbreak
patients have been based on drug sensitivity patterns
(see Table 46.5). Most of them included an 8 Quino-
lone such as Ofloxacin together with Amikacin and if
sensitive, Pyrazinamide. Other drugs have been used
but the long-term efficacy is unknown. Recently it has
been shown that a combination of Amikacin, Spar-
floxacin and Ethionamide was useful in patients who
were sensitive to these drugs.
The optimum duration of treatment for MDR-TB
is still unknown and many cases are treated for two
years after cultures convert to negative.
For close contacts of patients with MDR-TB there
are no clear guidelines as to what to do. It is easier
to decide on a chemoprophylaxis regime if the drug
sensitivity patterns are known. However, it is still not
known whether or not this type of prophylaxis would
work. A 'Delphi symposium' decided that if infection
and disease progression was likely then Ethambutol
and Pyrazinamide or a Quinolone plus Pyrazinamide
might be used (Passannante et al. 1992). The role of
BCG in adults is still unknown. In New York it was
reported that physicians had BCG in the hope that
this might protect them if exposed.

46.6.5 46.6.7
Alternative Therapies Prognosis and Predictors

Some patients with localised disease and those with
good cardiac pulmonary reserve and low bacterial
burden can be considered for either partial, or total,
lung resection. Nebulised interferon has been used in
rendering sputum smears negative in such patients
prior to surgery. It has also been used as a method
to prevent spread of infection when all other drug
treatments have failed in non-HIV infected MDR-
TB cases. The use of the immunomodulation with
M. vaccae has not yet been used in a formal trial of
MDR-TB.
Prognosis in HIV patients who are HIV pOSItive
and have MDR-TB has been very poor with between
25-30% of people surviving to 6 months if they are
severely immune suppressed (Fischl et al. 1992).
Most of the factors responsible for MDR-TB out-
breaks have now been addressed by government, or
federal, policy but are still important to help as the
guiding principles for prevention. These include inad-
equate control programmes for TB; inadequate adher-
ence to medication; infection control procedure break-
down by putting all the immune suppressed patients
in one area such as an open hospital ward; having a
low index suspicion for tuberculosis leading to the
patients' infectiousness being prolonged and having
poor laboratory communication with clinicians.

Table 46.5. HIV associated multi-drug-resistant tuberculosis Outbreaks January 1990 to August 1992 in USA

Facility Total Resistance HIV Mortality Median

Cases Pattern infection % interval TB
Diagnosis
To death

HospA 65 H, R, (E, Eth) 93 72 7

Hosp B 35 H, S, (R,E) 100 89 16
Hosp C 70 H, R, S, (E,Eth,Ka,B) 94 82 4
HospD 29 H, R, (E,Eth) 91 83 4
Hosp E 7 H, R, S, (E,Eth,Ka,RB) 20 60 4
Hosp F 16 H, R, S, (Eth,Ka,RB) 82 82 4
HospG 13 HR (E) 100 85 4
Prison System 42 HR (S,E,Eth,Ka,RB) 91 74 4
826
The major risk factors for MDR-TB in HIV patients
are the same as in the general population. There is a
history of such as previous treatment, birth, travel or
work in an area endemic for MDR-TB, a history of
poor adherence, sputum positivity continuing after
two months of treatment or being culture positive
after three months. In addition some factors in the
HIV population have been recognised as predictors of
MDR-TB. Severely immune suppressed HIV positive
patients seem more likely to develop MDR-TB prob-
ably related to the high risk of progression to disease
once infected. A failure to become apyrexial by two
weeks of treatment and development of hilar lymph
nodes are other factors pointing towards possible
MDR-TB (Telzak et al.1999; Salomon et al.I995).
One of the best predictors that patients will sur-
vive is that they start on at least two drugs to which
the organism is susceptible within two weeks of diag-
nosis. This often means a patient is given multiple
drugs prior to the drug sensitivity patterns being
available (Park et al. 1996; Turett et al.I995).
In Europe there have been about six outbreaks
involving around 225 patients and these have major
implications for resource utilisation and public
health (see Tables 46.6 and 46.7) they use a large
number of staff in contact tracing and case finding. It
was interesting to note that in the second outbreak in
the UK very few people took up the prophylaxis and
of those who did, almost all of them had stopped by
two weeks.
One of the largest outbreaks of MDR-TB occurred
in Argentina over a fifteen-month period (Ritacco et
al.I997).A total of 101 patients had resistance to five
drugs and the survival rate was approximately 10%.
Most of these patients were in contact with an intra-
venous drug user who adhered poorly to treatment
and had developed resistance.
There have been outbreaks of multi-drug resistant
M bovis in HIV patients in Spain involving 19 cases
over 15 months. Resistance to 11 drugs was found and
all the patients died with a median survival of only 44
days. Their risk factor was severe immune suppres-
sion (Cobo et al. 2001).
It was thought that poor drug absorption might
increase the risk of HIV positive patients in devel-
oping MDR-TB but one study of drug absorption in
AIDS patients (Taylor and Smith 1998) showed there
was no difference in the T-max, C-max and median
area under the curve (AUC) for AIDS patients com-
pared with HIV-negative patients. In fact it appeared
that the AIDS patients, even those with GI problems,
were absorbing Rifampicin better than HIV negative
matched controls. It has also been suggested that
P. D. O. Davies
Table 46.6. Outbreak 1 UK Contact & Case Finding
- Index + 7 other all HIV + 2 alive now
- 187 HlV + contacts
- 60 Staff
- 57 Community contacts
Table 46.7. Outbreak 2 UK Contact & Case Finding
- Index HIV- + 6 other HlV +
- 1298 general medical patients exposed
- 169 recalled
- 898 staff
- 64 HlV patients
- 476 HIV outpatients
- R H,R,ANS,Clo,Cyc1o+/-Z,Cla,Cip
- S Eth,Cap,Strep,Ethio,Amik
- Prophyaxis PAS + ETH offered to 400
- 12 on prophylaxis 10 stopped most by 2 weeks
HIV-positive patients required longer treatment
for their tuberculosis and that inadequate length of
treatment might lead to MDR-TB. However there is
no evidence to support this as the duration of treat-
ment and relapse are the same whether HIV patients
are given six or nine months treatment for initial
fully drugs susceptible disease.
46.6.8
Sporadic MDR-TB in HIV
In South Africa individuals were examined to see
whether they were more likely to develop MDR-TB
outside an outbreak event. It was found that 12% of
the HIV-negative compared with 2% of HIV positive
patients developed MDR-TB during the follow-up
(Anastasis et al.I997). The reason for this is that good
public health control can prevent outbreaks even in
developing countries. Those HIV patients involved in
outbreaks may have fewer episodes of cavitary dis-
ease; more fevers and be more immunosuppressed
than those who develop sporadic MDR-TB (Sacks
et al. 1999). Data from the South African goldmines
in 1993-1997 where the potential for outbreaks was
great, help shed more light on the association of
HIV and MDR-TB. Miners' hostels accommodating
2000-3000 men showed an increase incidence of TB
from 1174 to 2476 per 100,000 over a 6 year period
from 1990-1996 with associated increase in HIV
Multi-Drug-Resistant Tuberculosis 827

rates. Despite an HIV positivity of 28% the MDR-TB Table 46.8. Cost of drugs used in the treatment of tuberculosis
in new patients was only 1% of total and even in re- Drug and daily dose Annual cost £ ($)
treatment cases was only 2.8%. There was no associa-
tion between MDR-TB and HIV status. The reason for Amikacin 1 g 6430 (9655)
the low incidence seen in re-treatment cases, was that Capreomycin 1 g 6580 (9870)
there was a very good Directly Observed Therapy, Ciprofioxacin 1 g 980 (1470)
(DOTS) strategy with a completion rate of treatment Clarithromycin 1 g 1040 (1560)
93% (Churchyard et al. 2000). Clofazimine 300 mg 76 (114)
A major concern for many is that HIV might cause Cycloserine 500 mg 1685 (2530)
global outbreaks of MDR-TB as more and more Ethambutol 1 g 356 (535)
patients receive anti-tuberculosis therapy in an envi- Isoniazid 300 mg 26 (39)
ronment where the political and financial situation PAS 7.4g 5190 (7785)
is unstable and DOT strategies are not implemented Prothionamide 750 mg 880 (1320)
in full. The direct impact of this would be outbreak Pyrazinamide 2 g 110 (165)
transmission to immunosuppressed patients and Rifabutin 900 mg 3140 (4710)
then onwards to immune competent individuals. Rifampicin 600 mg 85 (127)
As a result, the already fragile control programmes Streptomycin 1 g 2450 (3675)
would be swamped. There have been cases of immu-
nosuppressed patients being treated for fully drug-
sensitive TB and them being exposed and acquiring
drug-resistant strains whilst on therapy for the fully from becoming sick, countries have to face the prob-
sensitive strain (Drobniewski 1997). lem of the lost productivity of those individuals.
Long-term sick adults are a cost to the economy in
more than one way: as well as costing money to treat
46.6.9 and maintain, they are not contributing, in taxes or
The Financial Challenges of MDR-T8 productivity, to the national economy. To this must be
added yet a further cost if the responsibility of sup-
Drug-resistant tuberculosis gives rise to challenges, porting their dependants falls on the state rather than
which go far beyond the problems associated simply on the extended family - and where it does fall on the
with the clinical care of the patient. family, the impact can be devastating (Rajeswari et al.
At the level of the individual patient with MDR-TB, 1997; Kamolratanaku et al. 1999).
doctors and hospital managers are having to face the
fact that managing even a single case of this condi-
tion is immensely more expensive than dealing with 46.6.10
a case of sensitive disease. Just a few such additional Different Countries, Different Problems,
patients may cause considerable budgetary problems Different Expectations
in even a well-funded institution in the developed
world. TB is a global disease, and accordingly so is MDR-
At a higher level, it has then to be recognised TB. Self-evidently the financial problems, which
that as well as the costs associated with individual MDR-TB produces are different in different parts
patients, the occurrence of MDR-TB in some patients of the world: London has relatively few patients and
may have implications for the cost of dealing with relatively lavish resources. Sub-Saharan Africa has
all patients. Once it is acknowledged that drug resis- many more patients and far fewer resources. Facili-
tance is a possibility, the management of all cases of ties and finances are different, but it is often forgot-
TB, and in particular their initial management, must ten that so are expectations. Doctors in resource rich
be planned with that possibility in mind. At a higher countries can do some things beyond the dreams of
level still, spending on public health measures to pre- their colleagues working in resource-poor countries.
vent the spread of MDR-TB, or to control and reverse It must be kept in mind, though, that levels of public
the problem where spread has already occurred can health protection which are - quite rightly - regarded
have very large financial implications, even for the in some countries as magnificent achievements with
wealthiest countries. the resources available, could in some other countries
Finally, in addition to carrying the financial be regarded as being so far below what is expected as
burden of treating the sick and of preventing others to amount to medical malpractice.
828
The expectations of doctors and other healthcare
workers dealing with MDR-TB will thus differ in dif-
ferent parts of the world. At least, though, healthcare
professionals regard MDR-TB as one disease, who-
ever it is that has the disease, and wherever they are.
There may be different constraints upon what we can
deliver for our patients, but we all know what it is that
we ought to deliver in an ideal world. To the policy
makers, and the money providers, however, MDR-
TB is many different diseases depending upon your
viewpoint.
The cost of MDR-TB in many countries is perhaps
about using DOTS-plus - can it be afforded (probably
yes), and can the system possibly afford drug suscep-
tibility testing on all initial isolates of TB (almost
certainly not). By contrast, in many parts of the USA
and increasingly in some parts of Western Europe,
the political and societal imperatives regarding the
financing of MDR-TB are to identify enough money
to put into negative pressure isolation every single
patient coming through the door of the hospital who
might possibly have pulmonary TB, and keep them
there until it has been proven not to be MDR.
There is, accordingly, no "one world" message: the
costs of MDR-TB will depend upon these political
and societal imperatives, and the expectations that
they produce.
46.6.11
The Individual Patient with MDR-TB
In managing the individual patient with MDR-TB, the
first and most obvious expense is the cost of drugs.
Second line drugs are more expensive, and those
first line drugs which, may still be effective may be
used for longer periods - ethambutol, for instance, is
markedly more expensive than rifampicin or isonia-
zid. Table 46.8 shows the average costs in the UK of
some of the drugs used in MDR-TB.
A WHO/IUATLD survey suggests that the cost of
a one month course of streptomycin is about $38, for
amikacin the monthly cost is $640, and for ofloxacin
the price ranges between about $88 and $200 dollars
per month, with cycloserine and PAS being similarly
priced. As has been pointed out, in many countries
just one month's supply of just one of those drugs
exceeds the entire annual per capita health expendi-
ture (JCRC 1998).
In those countries where they can be (fairly) read-
ily afforded, the expectations for levels of care are
high. To the drug costs must accordingly be added
the costs of toxicity monitoring: plasma levels; audio-
P. D. O. Davies
grams; and assessments of renal impairment. But the
question of drug, and associated, costs is only just
the start. The cost of managing tuberculosis starts to
escalate when in-patient care of patients is necessary
(Salomon et al. 1995; Park et al. 1996).
Negative pressure rooms for isolation of infectious
MDR-TB incur both capital (building) and annual
running costs. These costs can be predicted for the
anticipated useful life of the facility, and costs per
room/day calculated. These calculations, however, are
themselves likely to underestimate the cost of man-
aging an individual case of MDR-TB unless predic-
tions of bed occupancy rates are robust. If a patient
is placed in a negative pressure room for two or three
weeks, the institution pays not just for the nights the
patient actually occupies the room, but also for when
the room is empty and available.
Moving beyond the physical facility, there is the
matter of nursing intensity. Staffing levels on the
negative pressure unit of 10 beds at St Bartholomew's
Hospital in London are 50% higher than on other
wards in the hospital. This probably reflects the pat-
tern elsewhere. Quite apart from medical time and
nursing staffing levels, there is the increased input of
other healthcare professionals. Our experience is that
there is a very significant psychological, and indeed
psychiatric, morbidity in individuals with long-term
isolation. Turning from the mental to the physical, the
assessment of costs needs to consider the additional
medical care which may be more likely to be necessary
in MDR-TB than drug-sensitive disease: the possibility
of thoracic surgical intervention is just one of them.
There have been a number of estimates produced
for the costs involved in the care of individual patients
with MDR-TB in resource-rich countries. Values for
individual, very expensive, patients running into
hundreds of thousands of pounds are probably well
known to many of us. It has been our experience that
those responsible for the allocation of resources are
unimpressed by the example of single very expensive
cases with many complications, so in our institution
we examined a small series of patients, and looked
at those only with straightforward pulmonary MDR-
TB and who were HIV negative (White and Moore-
Gillon 2000).
In this study, we attempted to take into account
all the factors mentioned above, and came up with a
minimum mean cost of £60,000 ($90,000) per patient.
We know that this underestimates the true sums,
because in making the calculations about facility
costs we have, for instance, assumed 100% bed occu-
pancy in our negative pressure suite over an assumed
15 year life for the facility, and we know that is not
Multi-Drug-Resistant Tuberculosis
the case. These values have been helpful in drawing
to the attention of our local health care planners and
government that even though the numbers of cases
are relatively small, MDR-TB is a very real financial
issue in our inner cities, which cannot just be picked
up in the standard respiratory medicine budget.
46.6.12
The Consequences for Others
Moving beyond the individual patient, the conse-
quences for others can be considered in at least two
ways. The first of these is the relatively superficial
analysis of the consequences if the patient with MDR
TB passes the disease on to somebody else. This next
person then becomes another individual with MDR-
TB, with all the costs that are discussed above. The
cost of one multi-institutional outbreak of MDR TB
in New York City was estimated as in excess of $(US)
25 million, when calculated on the number of days
in hospital alone, and the full cost is bound to be far
higher than that (Frieden et al. 1996).
Further, as more than one hospital in London has
discovered, the costs of transmission to others are
not purely medical; they have been sued when hos-
pital in-patients have been infected by patients with
MDR-TB, and the damages in some cases have been
very large.
At its very simplest, this has led to subtle changes
in initial drug regimens in many cases. In the UK,
although the national guidelines have superficially
not changed very much (Farmer et al. 1998; Kuaban
et al. 2000) clinical practice has changed. The way that
most clinicians used to think about TB was that an
initial three-drug regimen was used as routine, and a
four-drug regime was to be used if there was reason
to think there may be an increased risk of drug resis-
tance. The has been reversed: so that a four-drug
regime is routine, and three drugs are only used if
the patient is regarded as being at very low risk. This
increases the cost of the standard regimen by about
30%, since ethambutol - the usual fourth drug - is
significantly more expensive than the others.
It has had the desirable effect of making clinicians
keener on obtaining microbiological confirmation
of the disease, particularly non-pulmonary disease.
This is because the sensitivities are so important,
and data from the British Public Health Laboratory
Service year on year shows an increasing proportion
of microbiological confirmation (Espinal et al. 1999).
Again it must be kept in mind that this additional
invasive sampling and laboratory examination is an
829
increased expense brought about by increased fears
about drug resistance.
Additionally, supplemental second-line drugs may
be used initially if there is a strong clinical suspicion
of resistant disease, and their use is continued until it
is confirmed that the disease is, after all, fully sensi-
tive. There is thus more expense: for drugs, for their
monitoring and for toxicity testing, and there is the
increased risk to the patient. The doctor may well- in
pulmonary disease - use strict isolation procedures,
until it is known that the disease is sensitive and the
regimen is appropriate. All these consequences flow
not from the fact that the patient himself has MDR-
TB, but purely from the fact that others have it and so
he might have it.
Is there anything that can be done to reduce these
costs? It is absolutely clear that if the time needed
for sensitivity testing can be reduced then there is
at least the potential to save money. This should be
viewed from the right direction: not much money
is saved by proving somebody is drug resistant
more quickly. On the other hand, because most of
the patients are drug sensitive and are unnecessar-
ily expensive if they are initially managed as being
possibly drug resistant, a great deal of money may
be saved by proving more quickly that patients
are drug sensitive. There have been a number of
publications that have looked at various tests and
their cost effectiveness, but one of the best analyses
comes from Drobniewski and colleagues, clearly
demonstrating the overall economic benefits of
expenditure on molecular diagnosis of tuberculosis
and drug resistance (Iseman 2000).
46.6.13
Implications for Public Health and the Wider
Economy
These factors relating to the implications of MDR-TB
for others who may not have the disease relate very
closely to the wider public health costs. There is a
clear need to have measures in the community to con-
trol the spread of resistant disease - but even more
importantly to prevent the emergence of resistant
strains. Reversing the problems of New York, which
developed in the 1980s is estimated to have cost $1
billion (Kritski et al. 1996).
In London, a study commissioned by the Regional
Public Health Office estimated that MDR-TB, about
2.5% of culture confirmed cases, accounted for 20%
of spending on TB (Rullan et al. 1996). This was, as
the authors acknowledged, only a fairly crude analy-
830
sis since it was not the main focus of the paper, but it
does give an idea of the scale of the problem.
Beyond immediate public health issues like these
is the impact of the disease in terms of the economy
as a whole. Where the burden falls is dependent upon
national wealth and the national culture. In a country
like the United Kingdom, the burden that the state
places upon the family is essentially zero. Many fami-
lies do of course contribute to care, but if they don't
the state will more or less cover all the costs. The state
will pay for the fact that the individual is ill, and pay
for the fact that they cannot work. The state will bear
the burden of lost tax revenue and it will bear the
cost of supporting the dependants if the ill person
was the breadwinner. In other countries, there may
be no such support from the state at all.
In many, the situation is intermediate between
those two extremes. Although not specifically in
the context of resistant disease, there have been
numerous excellent studies from India. One such
paper, by Rajeswari and colleagues, investigated a
mixed population treated by government health-
care, non-governmental organisations and private
practitioners showed that the total costs, particularly
indirect costs, were very high (Rajeswari et al. 1997).
Expenditure due to TB accounted for as much as 40%
of the patients mean annual income. Also, strikingly,
almost 20% of school age children of infected parents
had to discontinue their school studies, either to care
for their parents or to start work to contribute to the
family income
An excellent paper from Thailand starkly dem-
onstrates the importance of interpreting economic
statistics with a full awareness of the issues which
underlie them: increased expenditure on food in
tuberculous households was not a reflection of an
awareness of the need for good nutrition, but simply
because one in six households had to sell part of their
property to pay their bills and could no longer rely on
home-grown produce (Kamolratanaku et al. 1999).
In industrialised nations, economic discussions
about healthcare interventions are now common-
place. For example, when a major pharmaceuti-
cal company applies to the licensing and national
healthcare bodies for funding for its new anti-influ-
enza drug, part of the data it submits is the saving to
the national economy which it says will result from
its Widespread use. There are similar arguments for
new asthma drugs and new interventions in coronary
artery disease. Are there similar arguments in devel-
oped countries for new tuberculosis interventions?
No, but then the concept of the cost to the economy
of lost productivity is harder to sell as an idea to the
P. D. O. Davies
UK Government since in this and many other coun-
tries the majority of patients with TB are not working
anyway.
Can countries afford not to treat MDR-TB? There
is at present an interesting debate on this issue. It
certainly is a debate or discussion, rather than an
argument, because the protagonists are all, essentially
on the same side. Greatly simplified, the arguments of
Farmer and colleagues might be summed up by stating
that we cannot afford not to treat MDR-TB with indi-
vidualised regimes, however poor the environment in
which we are trying to work (Farmer et al.I998).
On the other hand, Espinal et al.'s argument is that
resources are invariably finite, and often very limited,
and that there will be «an optimal allocation of funds
that minimises current and future illness and death"
(Espinal et al. 1999). Further, the case has not been
proved that expenditure on individualised regimens
rather than standardised third line regimens will be
that optimal allocation of funds.
Iseman has addressed this debate in a short but
thought-provoking editorial (Iseman 2000). The utili-
tarian philosophers Jeremy Bentham and John Stuart
Mill explored the concept of following policies, which
led to the greatest good or happiness to the greatest
number. As Iseman puts it: «To what extent can and
should we divert energies and monies to the care of
the relatively small number of patients with MDR-
TB? Will such efforts impede the implementation of
DOTS programmes, the current utilitarian strategy
favoured by the WHO". On the other hand, would a
true utilitarian approach really favour diverting the
funds to MDR-TB, because over the long term that
will result in the greatest happiness to the greatest
number, the utilitarian ideal.
46.6.14
Is There One World Message After All?
It's a tough job to persuade a Government in London
that putting money into Africa will eventually make
voters happier about their record on TB locally in the
UK - not least because it will take a few years to have
an effect and by then the next Government may get
the credit. The battle against TB has to be fought in
the corridors of power in Washington, London, Paris
and Berlin, as well as out in the field.
It has been argued that where some drug resis-
tance occurs introducing the standard four-drug
regimen of isoniazid, rifampicin, pyrazinamide and
ethambutol followed by isoniazid and rifampicin
may augment drug resistance. If the standard re-
Multi-Drug-Resistant Tuberculosis
treatment regimen including only the addition of
streptomycin is introduced then five-drug resistance
will emerge. This has been termed the amplifier effect
of short-course chemotherapy (Farmer et al. 1998).
The authors site evidence of this having occurred
in Peru during the 1980s. Furthermore MDR-TB
strains, created in Peru have now shown up on the
eastern seaboard of the USA. Transmission within
hospital settings is also problematic (Kritski et al.
1996; Rullan et al. 1996). This may be some evidence
for a much greater problem (World Health Organiza-
tion 1997d). A possible 50 million people across the
globe may be infected with an MDR-TB strain (Small
et al. 1993).
The authors define DOTS-plus as the provision of
drugs, second line if necessary, to provide appropri-
ate treatment for the MDR-TB patient. This may not
necessarily be based on sensitivity results, which
would not be obtainable in many settings where the
DOTS-plus regimen may be required. Rather it would
be based on failure of the standard regimens and
would provide second line drugs, which the patient
would not have been exposed to previously.
When MDR-TB occurs the only way of treating the
patient effectively is to provide expensive second line
drugs. But the problem arises as to how these can be
afforded.
Farmer and colleagues argue that unless the prob-
lem of treating MDR-TB is addressed, and funded,
the problem can only get bigger. They point out that
untreated MDR-TB causes a series of sub-epidemics
in a susceptible population: the so-called fast MDR-
TB of primary disease. But in addition there will be a
slow epidemic caused by reactivation of latent MDR
infection at any time between a year and a lifetime
from the initial infection.
If MDR-TB is ignored the relative contribution of
these resistant strains to the overall caseload can only
increase year by year. Though it is true that overall
rates of drug resistance have fallen when well-func-
tioning DOTS programmes are introduced, in no
setting in the world has already established MDR-TB
rates fallen when DOTS short course therapy (with
no other intervention such as second line drugs) has
been introduced.
The argument that second line drugs cannot be
afforded in developing countries, whereas they can
in the developed world implies a second-class citi-
zenship of poorer nations. Farmer and colleagues go
on to berate the complacency that refuses to fund the
treatment of MDR-TE. If we do not practice a form
of DOTS-plus now, providing specific treatment of
MDR-TB, we will be forced to spend a greater amount
831
later when the epidemic 'time bomb' explodes.
Farmer points to five signs for optimism in the battle
against MDR-TB.
First, we know that MDR-TB is not necessarily
an incurable disease. Our pharmacological arma-
mentarium is understocked but it is a failure to treat
rather than a treatment failure, which results in the
majority of MDR-TB deaths (World Health Organiza-
tion 1997b).
Second, patients and their families are ready to go
to any sacrifice to attempt cure. It is also the experi-
ence of most clinicians that they will go to any lengths
possible to achieve the best for their patients.
Third, we cannot hide the destitute sick of the
world from the rest of us. The world is increasingly
a global village and the drug-resistant patient is no
more than 24 hours fiying time from anywhere in the
world (Pablos-Mendez and Lessnau 2000).
The fourth reason derives from the third. If the
destitute MDR-TB sufferer turns up in New York, or
London, no rightful humanitarian is going to say they
cannot be treated. Thus the hypocrisy of saying they
can be treated in New York, but not Lima Peru, for
example is exposed.
The fifth reason is that we do have the resources
necessary to confront MDR-TB, especially if we act
promptly and in a co-ordinated fashion. Delaying
our response will only increase future outlays. As
WHO has warned" once MDR-TB is unleashed we
may never be able to stop it." There is compelling evi-
dence from Peru to show that DOTS-plus, a strategy
which targets specifically MDR-TB as well as drug
susceptible disease has had good results (Farmer
2001). This provides a new strategy that would target
for treatment all patients with active disease, includ-
ing those with MDR-TB (Farmer and Kim 1998).
Side effects were not apparently a great problem
(Furin et al. 2001). Ineffective treatment of MDR-TB
with short course chemotherapy may increase the
costs of tuberculosis treatment and prevention in the
long run (Heifets 1994).
There is certainly good evidence that public
governmental or non-governmental organisational
pressure is helping to steadily reduce the cost of pre-
viously expensive chemotherapy (Coghlan 2001).
46.6.15
Transmission of MDR-TB
A recent study from Los Angeles has re-opened the
issue of whether drug-resistant tuberculosis is less
infectious than drug-resistant disease {Nitta et al.
832
2002). Strain typing was done on isolates of 102 pulmo-
nary multidrug-resistant cases and a total of 94 (92%)
were sputum smear-positive for acid-fast bacilli and 71
(70%) had cavitary lesions. The great majority could
therefore be described as being infectious. Four molec-
ular clusters of two cases each and one closely related
pair were identified among the 102 cases. Among 946
contacts identified 6% had a positive tuberculin skin
test. The authors imply from their discussion that they
regarded this as a low rate of transmission. They attri-
bute this to good control measures. An accompanying
editorial points out that the paper is flawed by a lack
of controls (Daley 2002). The authors do not provide
a comparison with a matched group who had drug
susceptible tuberculosis by which true transmissibility
could be determined. An earlier study in a similar area
showed much the same incidence of TST conversion
in contacts of patients with drug-susceptible strains
(Marks et al. 2000).
The editorial also points out the reduced transmis-
sion may be due to reduced virulence of the drug-
resistant organism. Other evidence is referenced
showing that drug-resistant isolates are less likely to
be associated with molecular-based case clustering
compared with drug-susceptible controls (van Soolin-
gen et al. 1999). Also isoniazid-resistant strains cause
significantly less disease in guinea pigs than drug-sus-
ceptible strains, and mutations or deletions within the
katG gene result in a decrease in the pathogenically of
isoniazid-resistant strains of M. tuberculosis (Li et al.
1998). The editorial concludes that the multi-drug-
resistant organisms, being 'less fit' may have more lim-
ited transmissibility than drug-susceptible organisms.
In the practical management of tuberculosis control it
is prudent to treat the drug-resistant strains as being
as infectious as the drug-susceptible strains until more
evidence emerges.
46.7
Conclusions
The potential for mortality and morbidity as a result
of MDR-TB, especially in conjunction with HIV is
immense. The question is whether existing struc-
tures can prevent or contain the disease. The fact
that MDR-TB exists at all and is a major problem in
many parts of the world suggests that we are failing to
prevent and cure the disease. The extent of the prob-
lem is a manifestation of the medical incompetence,
which has caused it. A very recent publication gives
some grounds for optimism (Dye et al. 2002).
P. D. O. Davies
The authors believe that standard short-course
chemotherapy, based on a combination of cheap
and safe first-line drugs, can prevent the build up
of resistance in a population in which most are drug
sensitive. They believe they have evidence that using
first-line drugs correctly can actually reduce drug
resistance. Resistant strains may be less virulent.
Long-term trends indicate that MDR can be con-
tained without expensive second-line drugs. Also on
a broad geographical scale they do not believe that
HIV has exacerbated MDR spread. But, they admit
the evidence to support these claims is not incon-
trovertible.
The jury is still out but the situation may not be
as bad as was feared. Prevention of MDR by excellent
medical practice must be of greatest importance.
Acknowledgements. The author would like to thank
The Annals of the New York Academy of Science for
allowing some reproduction from the monograph
Drug-resistant tuberculosis. From Molecules to Macro-
economics. Edit: Peter Davies. Annals of the New York
Academy of Science 2001 ;953:87-253. Also to Mario
Raviglione, Zarir Udwadia, Alwyin Mwinga, Vivien
Stern, Francis Drobniewski, Douglas Young, Paul
Farmer, Peter Ormerod, Anton Pozniac and John
Moore-Gillon, who have given permission to use some
of their manuscripts in this chapter.
References
Alcaide F, Pfyffer G, Telenti A, Amalio I (1997) Role of embB in
natural and acquired resistance to ethambutol in mycobac-
teria. Antimicrob Agents Chemother 41:2270-2273
American Thoracic Society (1994) Treatment of tuberculosis
and tuberculosis infection in adults and children. Am J
Respir Crit Care Med 149:1359-1374
Amnesty International (1999) Human Rights violations
against detainees. Amnesty International Report
Anastasis D, Pillai G, Rambiritch V, Abdool Karim SS (1997a)
Related articles - a retrospective study of human immuno-
deficiency virus infection and drug-resistant tuberculosis
in Durban, South Africa. Int J Tuberc Lung Dis 3:220-224
Anastisis D, Pillai G. Rambiritch V, Abdool Karim SS (1997b)
A retrospective study of human immunodeficiency virus
infection and drug resistant tuberculosis in Durban, South
Africa. Int J Tuber Lung Dis 1:220-224
Bayer R, Wilkinson D (1995) Directly observed therapy for
tuberculosis: history of an idea. Lancet 345: 154-158
Becerra MC, Farmer PE, Kim JY (1999) The problem of drug-
resistant tuberculosis. In: The global impact of drug-resis-
tant tuberculosis. Boston, pp 1-38
Berdon R, Kuaban C (1997) Initial resistance to anti-tuber-
culosis drugs in Yaounde, Cameroon in 1995. Int J Tuberc
Lung Dis 1:110-114
Multi-Drug-Resistant Tuberculosis
Bergmann JS, Keating WE, Woods GL (2000) Clinical evalua-
tion of the BD Probe Tec ET system for rapid detection of
Mycobacterium tuberculosis. J Clin Microbiol 38:863-865
Bertzel G, Aziz M, Wendl-Richter U et al (1999) Anti-tubercu-
losis drug resistance surveillance in Uganda 1996-1997.1nt
J Tuberc Lung Dis 3:810-815
Braun M, Kilburn JO, Smithwick RW et al (1992) HIV infec-
tion and primary resistance to antituberculosis drugs in
Abidjan, Cote d'Ivoire. AIDS 6:1327-1330
Breathnach AS, de Ruiter A, Holdsworth GM, Bateman NT,
O'Sullivan DG, Rees PJ, Snashall D, Milburn HJ, Peter BS,
Watson J, Drobniewski FA, French GL (1998) An outbreak
of multi-drug resistant tuberculosis in a London teaching
hospital. J Hosp Infect 92:111-117
Brennan PJ, Nikaido H (1995) The envelope of mycobacteria.
Annu Rev Biochem 64:29-63
Cangelosi GA, Brabant WH, Britschgi TB, Wallis CK (1996)
Detection of rifampin and ciprofloxacin resistant Myco-
bacterium tuberculosis by using species-specific assays for
precursor rRNA. Antimicrob Agents Chem 40:1790-1795
Carriere C, Riska PF, Zimhony 0, Kriakov J, Bardarov S,
Burns J, Chan J, Jacobs WR Jr (1997) Conditionally rep-
licating luciferase reporter phages: improved sensitivity
for rapid detection and assessment of drug susceptibil-
ity of Mycobacterium tuberculosis. J Clin Microbiol 35:
3232-3239
CDC (1991) Transmission of multi-drug-resistant tuberculo-
sis among immunocompromised persons in correctional
system - New York. MMWR Morb Mortal Wkly Rep 4128:
507-509
CDC (1993) Outbreak of multidrug-resistant tuberculosis at a
hospital- New York City, 1991. MMWR Morb Mortal Wkly
Rep 4222:427,433-434
Churchyard GJ, Corbett EL, Kleinschmidt I, Mulder D, de Cock
KM (2000) Drug-resistant tuberculosis in South African
gold miners: incidence and associated factors. Int J Tuberc
Lung Dis 4:433-440
Cobo J, Asensio A, Moreno S, Navas E, Pintado V, Oliva J,
Gomez-Mampaso E, Guerrero A (2001) Risk factors for
nosocomial transmission of multidrug-resistant tubercu-
losis due to Mycobacterium bovis among HIV-infected
patients. Int J Tuberc Lung Dis 5:413-418
Coghlan A (2001) Lives before profit. Drug companies slash
the cost of treating resistant TB. New Scientist 2298:6
Cohn D, Bustreo F, Raviglione MC (1997) Drug resistant tuber-
culosis: review of the worldwide situation. Clin Infect Dis
24:s121-s130
Cole ST, Brosch R, Parkhill J et al (1998) Deciphering the
biology of Mycobacterium tuberculosis from the complete
genome sequence. Nature 393:537-544
Coninx R, Mathieu C, Debacker M, Mirzoev F, Ismaelov A,
de Haller R, Meddings DR (1999) First-line tuberculosis
therapy and drug-resistant Mycobacterium tuberculosis in
prisons. Lancet 353:969-973
Conover C, Ridzon R, Valway S, Schoenstadt L, McAuley J,
Onorato I, Paul W (2001) Outbreak of multi-drug-resis-
tant tuberculosis at a methadone treatment program. Int
J Tuberc Lung Dis 1:59-64
Coronado VG, Beck-Sague CM, Hutton MD, Davis BJ, Nicholas
P, Villareal C, Woodley CL, Kilburn JO, Crawford JT, Frie-
den TR et al (1993) Transmission of multidrug-resistant
Mycobacterium tuberculosis among persons with human
immunodeficiency virus infection in an urban hospital:
833
epidemiologic and restriction fragment length polymor-
phism analysis. J Infect Dis 1684:1052-1055
Daley CL Transmission of multidrug-resistant tuberculosis
limited by man or nature. Am J Respir Crit Care Med 165:
742-743
Davies G, Pillay M, Strum AW, Wilkinson D Emergence of
multi-drug resistant tuberculosis in a community based
daily-observed treatment program in South Africa. Int J
Tuberc Lung Dis 3:799-804
Della-Latta P (1999) The mycobacteriology milestones. Lab
Med 30:408-417
Demissie M, Gebeyehu M, Berhane Y (1997) Primary resis-
tance to anti-tuberculosis drugs in Addis Ababa, Ethiopia.
Int J Tuberc Lung Dis 1:64-67
Dessen A, Quemard A, Blanchard JS, Jacobs WR Jr, Sacchet-
tini JC (1995) Crystal structure and function of the iso-
niazid target of Mycobacterium tuberculosis. Science 267:
1638-1641
Dosso M, Bonard D, Msellati P et al (1999) Primary resistance
to antituberculosis drugs: a national survey conducted
in Cote d'Ivoire in 1995-1996. Int J Tuberc Lung Dis 3:
805-809
Drobniewski FA (1997) Is death inevitable with multi-resistant
TB plus HIV infection? Lancet 349:71-72
Drobniewski FA, More PG, Harris GS (2000a) Differentiation
of Mycobacterium tuberculosis complex and nontuber-
culous mycobacterial liquid cultures by using peptide
nucleic acid-fluorescence in situ hybridization probes. J
Clin Microbiol 38:444-447
Drobniewski FA, Watterson SA, Wilson SM, Harris GS (2000b)
A clinical, microbiological and economic analysis of a
national service for the rapid molecular diagnosis of
tuberculosis and rifampicin resistance in Mycobacterium
tuberculosis. J Med MicrobioI49:271-278
Dye C, Scheele S, Dolin P, Pathania V. Raviglione M (1999)
Global burden of tuberculosis estimated incidence, preva-
lence and mortality by country. JAMA 282:677-689
Dye C, Williams BG, Espinal MA, Raviglione M (2002) Erasing
the World's slow stain: strategies to beat multidrug-resis-
tant tuberculosis. Science 295:2042-2046
Elliott AM, Hclwindii B, Hayes RJ et al (1995) The impact of
human immunodeficiency virus on mortality of patients
treated for tuberculosis in a cohort study in Zambia. Trans
R Soc Trop Med Hyg 89:78-82
Eltringham IJ, Drobniewski FA, Magan JA, Butcher PD, Wilson
SM (1999a) Evaluation of reverse transcription - PCR and a
bacteriophage-based assay for rapid phenotypic detection
of rifampin resistance in clinical isolates of Mycobacterium
tuberculosis. J Clin Microbiol 37:3524-3527
<refEltringham IJ, Wilson SM, Drobniewski FA (1999b) Evalu-
ation of bacteriophage-based assay (Phage amplified bio-
logically assay) as a rapid screen for resistance to Isoniazid,
Ethambutol, streptomycin, pyrazinamide and ciprofloxacin
among clinical isolates of Mycobacterium tuberculosis. J
Clin Microbiol 37:3528-3532
Espinal MA, Dye C, Raviglione M, Kochi A (1999) Rational
,DOTS plus' for the control of MDRTB. Int J Tuberc Lung
Dis 3:561-563
Espinal MA, Kim SJ, Suarez PG et al (2000) Standard
short-course chemotherapy for drug-resistant tubercu-
losis: treatment outcome in six countries. JAMA 283:
2537-2545
Espinal MA, Laszlo A, Simonsen L et al (2001) Global trends
834
in resistance to antituberculosis drugs. N Engl J Med 344:
1294-1303
Farmer PE (1999) Managerial successes, clinical failures. Int J
Tuberc Lung Dis 3:365-367
Farmer PE (2001) DOTS and DOTS-plus: not the only answer.
In: Davies P (ed) Drug resistant tuberculosis: from mole-
cules to macro-economics. Ann NY Acad Sci 953:165-184
Farmer PE, Kim JY (1998) Community-based approaches to
the control of multidrug-resistant tuberculosis: introduc-
ing "DOTS-plus". Br Med J 317:671-674
Farmer P, Bayona J, Becerra M et al (1998) The dilemma
of MDR-TB in the global era. Int J Tuberc Lung Dis 2:
869-876
Fischl MA, Daikos GL, Uttamchandani RB, Poblete RB, Moreno
IN, Reyes RR, Boota AM, Thompson LM, Cleary TJ, Oldham
Set al (1992) Clinical presentation and outcome of patients
with HIV infection and tuberculosis caused by multiple-
drug-resistant bacilli. Ann Intern Med 117:184-190
Fox W (1981) Whither short-course chemotherapy. Br J Dis
Chest 75:331-357
Frieden TR, Sherman LF, Lay Maw K et al (1996) A multi-insti-
tutional outbreak of higWy drug-resistant tuberculosis.
JAMA 276:1229-1235
Furin JJ, Mitnick JJ, Shin SS et al (2001) Occurrence of seri-
ous adverse effects in patients receiving community based
therapy for multi-drug resistant tuberculosis. In J Tuberc
Lung Dis 5:648-656
Gangadharam PRJ (1984) Drug resistance in mycobacteria.
CRC Press, Boca Raton
Gilad J, Borer A, Riesenberg K, Peled N, ScWaeffer F (2000) Epi-
demiology and ethnic distribution of multidrug-resistant
tuberculosis in southern Israel, 1992-1997: the impact of
immigration. Chest 177:738-743
Githui WA, Hawken MP, Juma ES, Godfrey-Faussett P, Swai OB,
Kibuga DK, Porter J (1999) Drug resistance survey in refu-
gee and non-refugee populations in north-eastern Kenya.
Int J Tuberc Lung Dis 3:S118
Global alliance for TB drug development (2001) Scientific
blueprint for tuberculosis drug development. Tuberculosis
81 [Suppll):1-52
Goble M, Iseman M, Madsen LA et al (1993) Treatment of 171
patients with pulmonary tuberculosis resistant to isoniazid
and rifampin. N Engl J Med 328:527-532
Gothi GD, Sen PK (1996) Tuberculosis - an overview. Tubercu-
losis Association of India Oration, Bangalore
Hannan MM, Peres H, Maltez F, Hayward AC, Machado J, Mor-
gado A, Proenca R, Nelson MR, Bico J, Young DB, Gazzard
BS (2001) Investigation and control of a large outbreak of
multi-drug resistant tuberculosis at a central Lisbon hos-
pital. J Hosp Infect 72:91-97
Hayward AC, Bennett DE, Herbert J et al (1996) Risk factors
for drug resistance in patients with tuberculosis in England
and Wales 1993-1994. Thorax 51 [SuppI3]:S32
Heifets LB (1994) Anti-mycobacterial drugs. Semin Respir
Infect 9:84-103
Hellyer TJ, DesJardin LE, Hehman GL, Cave MD, Eisenach
KD (1999) Quantitative analysis of mRNA as a marker for
viability of Mycobacterium tuberculosis. J Clin Microbiol
37:290-295
Herbert J, Watson J, Drobniewski FA (1999) The UK Myco-
bacterial Resistance Network Annual Report. Mycobnet
Collaborating Group
Hong Kong Chest Service (1981) British Medical Research
P. D. O. Davies
Council. First report: Controlled trial of four thrice-weekly
regimens and a daily regimen all given for 6 months for
pulmonary tuberculosis. Lancet 1: 171-74
Hong Kong Chest Service (1982) British Medical Research
Council. Second report: Controlled trial of four thrice-
weekly regimens and a daily regimen all given for 6
months. The results up to 24 months. Tubercle 63:89-98
Hong Kong Chest Service (1987) British Medical Research
Council. Five-year follow up of a controlled trial of five
6-month regimens of chemotherapy for pulmonary tuber-
culosis. Am Rev Respir Dis 136:1339-1342
Hong Kong Chest Service (1991) British Medical Research
Council. Controlled trial of 2, 4 and 6 months of pyra-
zinamide in 6 month, three-times weekly regimens for
smear-positive tuberculosis, including an assessment
of a combined preparation of isoniazid, rifampicin and
pyrazinamide. Results at 30 months. Am R(~v Respir Dis
143:700-706
ICRC (1998) Georgia: ICRC TB programme in prisons is
underway. ICRC News 4 June 1998
Iseman M (1993) Treatment of multidrug resistant tuberculo-
sis. N Engl J Med 329:784-790
Iseman MD (2000) Tuberculosis control strategies and utili-
tarianism. Int J Tuberc Lung Dis 4:95
Jacobs WR (1993) Rapid assessment of drug susceptibilities
of Mycobacterium tuberculosis by means of luciferase
reporter phages. Science 260:819-822
Joint Tuberculosis Committee of the British Thoracic Society
(1998) Chemotherapy and management of tuberculosis
in the United Kingdom: recommendations. Thorax 53:
536-548
Joint Tuberculosis Committee of the British Thoracic Society
(2000) Control and prevention of tuberculosis in the United
Kingdom: code of practice 2000. Thorax 55:887-901
Kamolratanaku P, Sawert H, Kongsin S, Lertmaharit S, Sri-
wongsa J, Na-Songkhla S et al (1999) Economic impact of
tuberculosis at the household level. Int J Tuberc Lung Dis
3:596-602
Khatri GR, Frieden TR The status and prospects of tuberculo-
sis control in India. Int J Tuberc Lung Dis 4:193-200
Kirk SM, Mazurek GH, Callister SM, Moore AV (1998) Myco-
bacterium tuberculosis susceptibility results in 24 hours by
using flow cytometry. Clin Micro Boil Newslett 20:83-87
Kochi A, Vareldzis B, Styblo K (1993) Multidrug-resistant
tuberculosis and its control. Res Microb 144:104-110
Kristi A, Marques MJ, Rabahi MF et al (1996) Transmission of
tuberculosis to close contacts of patients with multi-drug
resistant tuberculosis. Am J Respir Crit Care Med 153:
331-335
Kuaban C, Bercion R, Jifon G, Cunin P, Ngu Blackett K (2000)
Acquired drug resistance in Yoaoude, Cameroon. Int J
Tuberc Lung Dis 4(5):427-432
Li Z, Kelly C, Collins F, Rouse D, Morris S (1998) Expression
of katG in Mycobacterium tuberculosis is associated with
its growth and persistence in mice and guinea pigs. J Infect
Dis 177:1030-1035
MacIntyre CR, Plant AJ, Hendrie D (2001) Shifting the bal-
ance between in-patient and out-patient care for tubercu-
losis results in economic savings. Int J Tuberc Lung Dis 5:
266-271
MahmoudiA, Iseman MD (1993) Pitfalls in the care of patients
with tuberculosis. JAMA 270:65-68
Maloney SA, Pearson ML, Gordon MT, del Castillo R, Boyle
Multi-Drug-Resistant Tuberculosis
JF, Jarvis WR (1995) Efficacy of control measures in pre-
venting nosocomial transmission of multidrug-resistant
tuberculosis to patients and health care workers. Ann
Intern Med 122:90-95
Marks SM, Taylor Z, Qualis NL, Shestha-Kuwahara RJ, Wilce
MA, Nguygn CH (2000) Outcomes of contact investigations
of infectious tuberculosis patients. Am J Respir Crit Care
Med 162:2033-2038
McKinney JD (2000) In vivo veritas: the search for TB drug
targets goes live. Nat Med 6:1330-1333
McKinney JD, Honer zu Bentrup K, Munoz-Elias EJ, Miczak A,
Chen B, Chan WT, Swenson D, Sacchettini JC, Jacobs WR
Jr, Russell DG (2000) Persistence of Mycobacterium tuber-
culosis in macrophages and mice requires the glyoxylate
shunt enzyme isocitrate lyase. Nature 406:735-738
Migliori GB, Raviglione MC, Schaberg T, Davies PDO, Zellwe-
ger JP et al (1999) Tuberculosis management in Europe.Eur
Respir J 14:978-992
Miller N, Hernandez SG, Cleary TJ (1994) Evaluation of gen-
probe amplified mycobacterium tuberculosis. Direct test
and PCR for direct detection of Mycobacterium tubercu-
losis in clinical specimens. J Clin MicrobioI32:393-397
Mitchison DA, Nunn AJ (1986) Influence of initial drug
resistance on the response to short-course chemo-
therapy of pulmonary tuberculosis. Am Rev Respir Dis
133:423-430
Moro ML, Gori A, Errante I, Infuso A, Franzetti F, Sodano L,
lemoli E (1998) An outbreak of multidrug-resistant tuber-
culosis involving HIV-infected patients of two hospitals
in Milan, Italy. Italian Multidrug-Resistant Tuberculosis
Outbreak Study Group. AIDS 129: 102
Moro ML, Errante I, Infuso A, Sodano L, Gori A, Orcese CA,
Salamina G, D'Amico C, Besozzi G, Caggese L (2000) Effec-
tiveness of infection control measures in controlling a noso-
comial outbreak of multidrug-resistant tuberculosis among
HIV patients in Italy. Int J Tuberc Lung Dis 41:61-68
Nitta AT, Knowles LS, Kim J, Lehnkering EL, Borenstein LS,
David PT, Harvey SM, de Koning ML (2002) Limited trans-
mission of Multi-drug-resistant tuberculosis despite a high
proportion of infectious cases in Los Angles County, Cali-
fornia. Am J Respir Crit Care Med 165:812-817
Pablos-Mendez A, Lessnau K (2000) Clinical mismanagement
and other factors producing antituberculosis drug resis-
tance. In: Bastian I, Portaels F (eds) Multidrug-resistant
tuberculosis. Dordrecht, pp 59-76
Pablos-Mendez A, Raviglione MC, Laszlo A et al (1998)
Global surveillance for antituberculosis-drug resistance,
1994-1997. N Engl J Med 338:1641-1649
Park MM, Davis AL, Schluger NW, Cohen H, Rom WN (1996)
Related articles outcome of MDR-TB patients, 1983-1993.
Prolonged survival with appropriate therapy. Am J Respir
Crit Care Med 531:317-324
Passannante MR, Gallagher CT, Reichman LB (1994) Preven-
tive therapy for contacts of multidrug-resistant tuberculo-
sis. A Delphi survey. Chest 1062:431-434
Petrie JG, Barnwell E, Grimshaw J, on behalf of the Scottish
Intercollegiate Guidelines Network (1995) Clinical guide-
lines: criteria for appraisal for national use. Royal College
of Physicians, Edinburgh
Piersimoni C, Callegaro A, Scarparo C, Penati V, Nista D,
Bornigia S, Lacchini C, Scagnelli M, Santini G, de Sio
G (1998) Comparative evaluation of the new gen-probe
mycobacterium tuberculosis amplified direct test and the
835
semi-automated abbott LCx mycobacterium tuberculosis
assay for direct detection of mycobacterium tuberculosis
complex in respiratory and extrapulmonary specimens. J
Clin MicrobioI36:3601-3604
Pitchenik AE, Burr J, Laufer M, Miller G, Cacciatore R, Bigler
WJ, Witte JJ, Cleary T (1990) Outbreaks of drug-resistant
tuberculosis at AIDS center. Lancet 336:440-441
Punnotok J, Shaffer N, Naiwatanakul T, Pumprueg U, Subhan-
nachart P, Ittiravivongs A, Chuchotthaworn C, Ponglert-
napagorn P, Chantharojwong H, Young NL, Limpakarnarat
K, Mastro TD (2000) Human immunodeficiency virus-
related tuberculosis and primary drug resistance in Bang-
kok, Thailand. Int J Tuberc Lung Dis 6:537-543
Quemard AJC, Sacchettini A, Dessen C, Vilcheze R, Bittman
WR, Jacobs WR Jr, Blanchard JS (1995) Enzymatic char-
acterization of the target for isoniazid in Mycobacterium
tuberculosis. Biochemistry 34:8235-8241
Rajeswari R, Balasubramanian R, Muniyandi M, Geetharamani
S, Thresa X, Venkatesan P (1997) Socio-economic impact of
tuberculosis on patients and family in India. Int J Tuberc
Lung Dis 3:869-877
Range NS, Mfinanga G, Chonde TM, Masanja H (2000) Anti-
TB resistance patterns and the relationship with HIV
prevalence among TB patients in Tanzania. Int J Tuberc
Lung Dis 3:S120
Raviglione MC, Dye C, Schmidt S, Kochi A (1997) An assessment
of worldwide tuberculosis control. Lancet 350:624-628
Riska PF, Su Y, Bardarov S, Freundlich L, Sarkis G, Hatfull
G, Carriere C, Kumar V, Chan J, Jacobs WR Jr (1999)
Film-based determination of antibiotic susceptibilities of
Mycobacterium tuberculosis strains by using a luciferase
reporter phage and the Bronx box. J Clin Microbiol 37:
1144-1149
Ritacco V, di Lonardo M, Reniero A, Ambroggi M, Barrera L,
Dambrosi A, Lopez B. Isola N, de Kantor IN (1997) Noso-
comial spread of human immunodeficiency virus-related
multidrug-resistant tuberculosis in Buenos Aires. J Infect
Dis 763:637-642
Rossau R, Traore H, de Bennhouwer H, Mijs W, Jannes G, de
Rijk P, Portabels F (1997) Evaluation of INNO-LiPA Rif.,
TB assay, a reverse hybridization assay for the simultane-
ous detection of Mycobacterium tuberculosis complex
and its resistance to rifampin. Antimicrob Agents Chern
41:2093-2098
Rullan J, Herrera D, Cano Ret al (1996) Nosocomial transmis-
sion of multidrug-resistant tuberculosis in Spain. Emerg
Infect Dis 2: 125-129
Sacks LV, Pendle S, Orlovic D, Blumberg L, Constantinou C
(1999) A comparison of outbreak-and nonoutbreak-related
multidrug-resistant tuberculosis among human immuno-
deficiency virus-infected patients in a South African hos-
pital. Clin Infect Dis 91:96-101
Salomon N, Perlman DC, Friedmann P, Buchstein S, Kreiswirth
BN, Mildvan D (1995) Predictors and outcome of multi-
drug-resistant tuberculosis. Clin Infect Dis 215:1245-1252
Schluger NW (2000) The impact of drug resistance on the
global tuberculosis epidemic. Int J Tuberc Lung Dis 4:
271-275
Shafer RW, Small PM, Larkin C, Singh SP, Kelly P, Sierra MF,
Schoolnik G, Chirgwin KD (1995) Temporal trends and
transmission patterns during the emergence of multi-
drug-resistant tuberculosis in New York City: a molecular
epidemiologic assessment: J Infect Dis 171:170-176
836
Sharma V, Sharma S, Hoener zu Bentrup K, McKinney JD,
Russell DG, Jacobs WR Jr, Sacchettini JC (2000) Structure
of isocitrate lyase, a persistence factor of Mycobacterium
tuberculosis. Nat Struct Bioi 7:663-668
Slayden RA, Lee RE, Barry CE 3rd (2000) Isoniazid affects
multiple components of the type II fatty acid synthase
system of Mycobacterium tuberculosis. Mol Microbiol 38:
514-525
Small PM, Shafer RW, Hopewell PC et al (1993) Exogenous
re-infection with multi-drug resistant Mycobacterium
tuberculosis in patients with advanced HIV infection. N
Engl J Med 328:1137-1144
Smith JH, Radcliffe G, Rigby S, Mahan DJ, Lane J, Klinger
JD (1997) Performance of an automated Q-beta replicate
amplification assay for Mycobacterium tuberculosis in a
clinical trial. J Clin MicrobioI35:1484-1491
Snewin VA, Gares MP, Gaora PO, Hasan Z, Brown IN, Young
DB (1999) Assessment of immunity to mycobacterial infec-
tion with luciferase reporter constructs. Infect Immun 67:
4586-4593
Sreevatsan S, Stockbauer K, Pan X, Kreiswirth BN, Moghazeh SL,
Jacobs WR Jr, Telenti A, Musser JM (1977) Ethambutol resis-
tance in Mycobacterium tuberculosis; critical role of embB
mutations. Antimicrob Agents Chemother 41:1677-1681
Sreevatsan S, Pan X, Musser J (1997) Analysis of the oxyR-ahpC
region in isoniazid-resistant and susceptible mycobacterium
tuberculosis complex organisms recovered from diseased
humans and animals in diverse localities. Antimicrob Agents
Chemother 41:600-606
Sreevatsan S, Pan X, Zhang Y, Kreiswirth BN, Musser J Muta-
tions associated with pyrazinamide resistance in pncA
of Mycobacterium tuberculosis complex organisms 41:
636-640
Stern V (1999) The problem of TB in Eastern Europe and
Central Asia. ICPS, London
Stewart GR, Snewin VA, Walzl G, Hussell T, Tormay P, O'Gaora
P, Goyal M, Betts J, Brown IN, Young BD (2001) Over-
expression of heat-shock proteins reduces survival of
Mycobacterium tuberculosis in the chronic phase of infec-
tion. Nat Med 7:732-737
Stroud LA, Tokars JI, Grieco MH, Crawford JT, Culver DH,
Edlin BR, Sordillo EM, Woodley CL, Gilligan ME, Sch-
neider N et al (1995) Infection Control Hosp. Evaluation
of infection control measures in preventing the nosoco-
mial transmission of multidrug-resistant Mycobacterium
tuberculosis in a New York City hospital. Epidemiology
3:141-147
Talbot EA, Moore M, McCray E, Binkin NJ (2000) Tubercu-
losis among foreign-born persons in the United States,
1993-1998. JAMA 284:2894-2900
Taylor B, Smith PJ (1998) Does AIDS impair the absorption of
antituberculosis agents? Int J Tuberc Lung Dis 2:670-675.
PID: 9712282 PubMed - indexed MEDLINE
Taylor Z, Marks SM, Rios Burrows NM, Weis SE, Stricot RL,
Miller B Causes and costs of hospitalization of tuberculo-
sis patients in the United States. Int J Tuberc Lung Dis 4:
931-939
Telenti A, Imboden P, Marchesi F (1993a) Detection of rifam-
picin-resistance mutations in Mycobacterium tuberculosis.
Lancet 341:647-649
Telenti A, Imboden P, Marchesi F (1993b) Direct, automated
detection of rifampin-resistant mycobacterium tuber-
culosis by polymerase chain reaction and single-strand
P. D. O. Davies
conformation polymorphism analysis. Antimicrob Agents
Chemother 10:2054-2058
Telenti A, Marchesi F, Balz M (1993c) Rapid identification of
mycobacteria to species level by polymerase chain reac-
tion and restriction enzyme analysis. J Clin Microbiol 31:
175-178
Telzak EE, Chirgwin KD, Nelson ET, Matts JP, Sepkkowitz KA,
Benson CA, Perlman DC, EI-Sadr WM (1999) Predictors
for multidrug-resistant tuberculosis among HIV-infected
patients and response to specific drug regimens. Terry
Beirn. Community Programs for Clinical Research on
AIDS CPCRA and the AIDS Clinical Trials Group ACTG,
National Institutes for Health. Int J Tuberc Lung Dis 4:
337-343
Telzak EE, Sepowitz K, Alpert P et al (1995) Multidrug-resis-
tant TB in patients without HIV infection. N Engl J Med
333: 907-911
The Inter Departmental Working Group on Tuberculosis (1998)
The prevention and Control of tuberculosis in the United
Kingdom: UK guidance on the prevention and control of
1) HIV-related tuberculosis and 2) Drug resistant, including
multiple drug resistant, tuberculosis. Department of Health,
The Scottish Office, The Welsh Office, Sept 1998
Tuberculosis in Prisons Roundtable (2000) Press releases. 1
December 2000. United States Embassy, Kazakstan
Tuberculosis in prisons (2000) World Health Organization,
March 2000
Tuberculosis update (1999) Public Health Laboratory Service
Communicable Disease Surveillance Centre. Dec 1999
Turett GS, Telzak EE, Torian LV, Blum S, Alland D, Weisfuse I,
Fazal BA (1995) Improved outcomes for patients with multi-
drug-resistant tuberculosis. Clin Infect Dis 215:1238-1244
Udwadia ZF, Hakiniyan A, Rodriguez C et al (1996) A profile of
drug-resistant tuberculosis in Bombay. Chest 110:228s
Uplekar MW, Shepard DS (1991) Treatment of tuberculosis by
private practitioners in India. Tubercle 72:284-290
Van Soolingen D, Borgdorff MW, de Haas PI, Sebek MMGG,
Veen J, Dessens M, Kremer K, Embden JDA (1999) Molec-
ular epidemiology of tuberculosis in the Netherlands: a
national wide study from 1993 through 1997. J Infect Dis
180:726-736
Watterson SA, Wilson SM, Yates YD, Drobniewski FA Com-
parison of three molecular assays for rapid detection of
rifampin resistance in Mycobacterium tuberculosis. J Clin
MicrobioI36:1959-1973
White VLC, Moore-Gillon J (2000) Resource implications of
multidrug resistant tuberculosis. Thorax 55:962-963
Wilson SM, Al-Suwaidi Z, McNerney R, Porter J, Drobniewski
FA (1997) Evaluation of a new rapid bacteriophage-based
method for the drug susceptibility testing of Mycobacte-
rium tuberculosis. Nature Med 3:465-468
World Health Organization Tuberculosis Unit (1991) Division
of Communicable Diseases. Guidelines for tuberculosis
treatment in adults and children in national Tuberculosis
Programmes. WHO, Geneva, WHO/TB/91, pp 1-61
World Health Organization (1997a) Guidelines for the man-
agement of drug-resistant tuberculosis. WHO/TB/96.210
(rev 2), Geneva
World Health Organization (1997b) TB treatment observer.
24 March 1997
World Health Organization (1997c) Treatment of tuberculo-
sis: guidelines for national programmes. WHO/TB/97.220,
Geneva
Multi-Drug-Resistant Tuberculosis
World Health Organization (1997d) International Union
Against Tuberculosis and Lung Disease. Anti-tuberculosis
drug resistance in the world: the WHO/IUATLD global
project on anti-tuberculosis drug resistance surveillance.
WHO/TB/97,229, Geneva
World Health Organization (1997e) International Union
Against Tuberculosis and Lung Disease. Guideline for
the surveillance in drug resistant tuberculosis. WHO/TB/
96.216, Geneva
World Health Organization (2000a) International Union
Against Tuberculosis and Lung Disease. Anti-tuberculosis
drug resistance in the world: the WHO/IUATLD global
837
project on anti-tuberculosis drug resistance surveillance,
report no 2. WHO/TBI200/278, Geneva
World Health Organization (2000b) Guidelines for establish-
ing DOTS-plus pilot projects for the management of multi-
drug-resistant tuberculosis (MDR-TB) (Gupta R, Arnadot-
tir T eds). WHO/CDS/B/2000.279, Geneva
World Health Organization (2001) Global tuberculosis control:
WHO report 2001. WHO.CDS. 2001.287, Geneva
Young DB (2001) A post-genomic perspective. Nat Med 7:11-13
Zhang Y, Heym B, Allen B, Young D, Cole S (1992) The cata-
lase-peroxidase gene and isoniazid resistance to Mycobac-
terium tuberculosis. Nature 358:591-593
47 Monitoring Treatment Efficacy
P. H. LAGRANGE, N. SIMONNEY, A. O. SOUSA, A. WARGNIER, J. L. HERRMANN

CONTENTS is difficult to accomplish. Sub-optimal TB treatment

is always associated with an increased risk of dis-
47.1 Introduction 839 semination into the general population and support
47.2 Clinical and Bacteriological Monitoring
of TB Patients 839 emergence of multiple-drug-resistant strains.
47.2.1 Monitoring Patients with TB Early identification and effective treatment of
Involves Three Main Objectives 839 active TB have been the key element in TB control
47.2.1.1 Susceptibility Testing 840 and prevention. Adequate therapy for patients with
47.2.1.2 Compliance of the Treated TB Patient 840
TB is the single most important strategy for the pre-
47.2.1.3 Early Follow-up of the Patient
and the Culture Conversion 840 vention of drug-resistant TB. Effective treatment of
47.2.2 Diagnostic Methods 841 infectious patients can also significantly reduce the
47.2.2.1 TB Infection 841 new transmission in preventing the occurrence of
47.2.2.2 TB Disease 841 disease. However, TB must be treated for a relatively
47.2.3 Immunological T- and B-Cell Responses
long duration compared with the treatment of many
in TB Patients 842
47.2.3.1 Ex Vivo T-Cell Immune Responses 842 other infectious diseases. Because of the potential
47.2.3.2 In Vitro B-Cell Immune Responses 842 emergence of multiple-drug-resistant TB and patient
47.3 Conclusions 849 non-adherence to treatment, it is essential to monitor
References 850 the treatment effectiveness.
Patients under treatment are presently monitored
by direct smear examination and culture of clinical
47.1 samples (Rieder 1996). However, the long time required
Introduction for M. tuberculosis culture is one of the major limita-
tions. Furthermore, effective monitoring of socially and
Tuberculosis (TB) remains the most frequent lethal economically deprived populations presents additional
infection observed in socio-economically deprived problems. There is therefore an urgent need for rapid,
populations in which effective long term treatment simple, and reliable methods for the evaluation of both
the efficiency and outcome of treatment. The develop-
P. H. LAGRANGE, MD ments of new immunological tests as surrogate markers
Service de Microbiologie, H6pital Saint Louis, Assistance Pub- for TB monitoring are the subject of this review.
lique-H6pitaux de Paris, Universite Denis Diderot, 1, avenue
Claude Vellefaux, 75475 Paris, France
N. SIMONNEY, MD
Service de Microbiologie, H6pital Saint Louis, Assistance Pub-
47.2
lique-H6pitaux de Paris, Universite Denis Diderot, 1, avenue Clinical and Bacteriological Monitoring
Claude Vellefaux, 75475 Paris, France of TB Patients
A. O. SOUSA, MD
Service de Microbiologie, H6pital Saint Louis, Assistance Pub- 47.2.1
lique-H6pitaux de Paris, Universite Denis Diderot, 1, avenue
Claude Vellefaux, 75475 Paris, France Monitoring Patients with TB
A. WARGNIER, MD Involves Three Main Objectives
Service de Microbiologie, H6pital Saint Louis, Assistance Pub-
lique-H6pitaux de Paris, Universite Denis Diderot, 1, avenue The first objective is the diagnosis: of TB infection
Claude Vellefaux, 75475 Paris, France by active or passive case finding. This is in order to
J. 1. HERRMANN, MD
Service de Microbiologie, H6pital Saint Louis, Assistance Pub-
follow the annual rate of transmission (at the popula-
lique-H6pitaux de Paris, Universite Denis Diderot, 1, avenue tion level) and to begin the preventive chemotherapy
Claude Vellefaux, 75475 Paris, France in order to prevent the occurrence of TB disease (at

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
840
the individual level). The diagnosis of TB disease
should be made in order to detect any active TB
cases, to treat them, cure the patients and decrease
the transmission to the community (WHO recom-
mendations = 70% detection and 85% treatment)
using the DOTS strategy (WHO 200l).
The second objective is to predict, as soon as
possible, the patient's compliance, effectiveness of
the specific antibiotics prescribed, with simple and
robust biological tests with the aim of discontinuing
the isolation of the 'hospitalised' TB patient (Small
and Fujiwara 200l).
The third objective is to detect pre-clinical and
bacteriological relapse by sequential follow-up of the
patient during, and after, therapy using for example
sputum culture conversion (Liu et al.1999).
In order to predict the potential effectiveness of
the treatment, several tools exist involving the TB
bacilli susceptibility testing, assessing patient com-
pliance, allowing early and sequential follow-up of
culture conversion.
47.2.1.1
Susceptibility Testing
In vitro susceptibility testing of the isolated TB bacilli
to the three or four drugs being usually administered
remains the most pertinent test. Nevertheless, such
susceptibility tests need locally available, and opera-
tional, culture facilities. Moreover, the results are
often delayed, depending on the number of viable
bacilli present in the tested specimen and the meth-
odologies performed.
New alternatives are genomic approaches such as
amplification (PCR-SSCP), sequencing, hybridisation
to detect known mutations associated with antibiotic
resistance (for instance for rifampin or isoniazid).
It is also possible to follow the content and the decay
ofmessenger RNA from the TB bacillus in the specimen
during therapy (Kennedy et al. 1994). However, such
approaches have not been yet validated and remain to
be shown practicable in routine laboratories.
47.2.1.2
Compliance of the Treated TB Patient
The compliance of the patient during the whole six
months of treatment is very important. The first
three to six weeks are particularly important since the
majority of treatment failures occur during this period.
Cessation of one, or more, drugs is usually associated
with the selection of mutants involved in secondary
resistance.
P. H. Lagrange et al.
Compliance can be improved by the administra-
tion of a single tablet containing three to four drugs
(such as RIFATER), and with follow-up of the excre-
tion of drugs in the urine.
It is most important for the patient's compliance
that comprehensive information is provided by
the physician, that the treatment needs to be taken
regularly and that the entire course is taken for the
defined period of time, without cessation.
The direct observation therapy (often referred
to as DOT) strategy is actually recommended in
patients with a high risk of non-adherence (Small
and Fujiwara 200l).
47.2.1.3
Early Follow-up of the Patient
and the Culture Conversion
The need for early follow-up (during the first 2-3
weeks) is usually based on clinical symptoms (cough,
appetite, weigh gain and fever) which are often asso-
ciated with a true effectiveness of the treatment. The
rapidity of symptom resolution, although helpful in
assessing an individual patient's response to therapy,
can be highly variable (Barnes et al. 1987).
Moreover, around 5-10% of patients do develop
some paradoxical reactions during this period
(Smith 1987). These clinical manifestations, together
with a visual worsening of the chest radiographs,
might confuse the expected evolution and give the
impression of a treatment failure.
The early bactericidal activity test (EBA) per-
formed during the first two or three days has been
used to demonstrate the efficacy of the drugs admin-
istrated (Jidani et al. 1980). However, its technical
realisation is demanding and has limitations such as
the calibration of the specimen. It is often used retro-
spectively to evaluate new drugs or combinations of
drugs (Sirgel et al. 2001).
In patients with positive sputum cultures, the
conversion to negative cultures provides the only
objective measure of successful treatment, and cul-
tures should be obtained monthly until conversion
is documented. More than 85% of patients who
received both isoniazid and rifampin have negative
sputum cultures within two months after initiation
of treatment (Small and Fujiwara 200l). Continued
monthly monitoring of sputum after culture conver-
sion identified a very small number of patients who
had culture reversion. Nevertheless, patients who
cannot tolerate, or adhere, to a standard regimen
may need continued monitoring to assess response
to treatment. For all patients a specimen should be
Monitoring Treatment Efficacy 841

collected at the end of treatment to document cure There are several tools to achieve the diagnosis of
(Sundaram et al. 2002). TB infection or TB disease.
The TB patients follow-up during, and after, the
treatment is very important, since it has been shown 47.2.2.1
that almost all failures have been observed when T8 Infection
patients were not followed-up regularly (Sevim et
al. 2002). The only available method to detect TB infection is
This follow-up is appropriate to verify the patient'sthe Mantoux test (tuberculin skin test, TST). This in
cure and the absence of relapse. The cure can be vivo skin test reaction has several drawbacks. The
declared at the end of the recommended duration first limitation is the poor diagnostic and predic-
of the treatment (being usually six months), but tive values: sensitivity and specificity are variable,
the absence of relapse is only confirmed after the influenced by environmental factors and individual
second year after finishing the treatment. Usually at responses. False-positive and false-negative responses
the end of the treatment both the clinical status and are common (Chaparas et al. 1985). The second limi-
radiological findings agree that the patient's health tation lies in the variability inherent in its application
has been restored(in these patients the bacteriologi- and reading: the strict intra-dermal inoculation, the
cal tests are negative without any isolation or smear measurement performed two days after inoculation
positive results). and the interpretation of the results.
It is only after the second year (after the end of the New tests which are more specific and less ambig-
treatment) that the complete cure can be definitely uous, need to be developed. Some are now undergo-
assessed if there have been no detected relapses. ing evaluation, either in vivo or in vitro in man and
However, this is certainly accurate for non-immu- in animals (Andersen et al. 2000).The use of specific
nocompromised TB patients. In fact, a recent study M. tuberculosis antigens, such as ESAT-6 or CFT-10,
demonstrated that in HIV-co-infected TB patients, antigens for the in vitro production of interferon-
relapses occur quite frequently and necessitate sec- gamma (IFN-y) by the patient's PBMC has been
ondary preventive chemotherapy for a minimum of demonstrated as being more specific than the in vivo
one year of isoniazid (Sonnenberg et al. 2001). PPD skin test (Ravn et al.1999). Such tests have been
used in order to detect high-risk contacts, but until
now, in contrast to the TST, no field-research has
47.2.2 been performed to evaluate its potential for disease
Diagnostic Methods prevention in prescribing chemotherapy in highly
positive subjects (Vekemans et al. 2001).
Apart from the gold standard of bacteriological dem- There is actually no immunological test avail-
onstration of the presence of M. tuberculosis there able to monitor the efficacy of preventive therapy
is a very well defined place for the immune-based in patients with a primary TB infection. Treatment
assays to diagnose TB in patients. These assays are efficacy is usually evaluated in terms of the absence
either direct or indirect. of occurrence of disease during or after the complete
The direct approach involves the detection of chemotherapy.
TB-specific antigens (Sada et al. 1992) and the best
results have been demonstrated in cerebrospinal fluid 47.2.2.2
(CSF) in TB meningitis cases(Araj et al.1993). Several T8 Disease
new approaches using monoclonal antibodies have
shown promising results with pulmonary specimens The two gold standards are the smear test for acid-
(Pereira Arias-Bouda et al. 2000). However, the diag- fast bacilli (AFB) and the isolation of the causative
nostic value was not higher than those obtained with bacteria from the specimens. However the smear test
classical bacteriological tests. has a variable sensitivity in diagnosing pulmonary
The indirect approach involves the detection of TB, as it is able to detect bacilli (limit of detection
any B- or T-cell mediated responses to any specific being 10 000 bacilli/ml) in almost 65% (at best) of
antigenic components of the tubercle bacillus. Such active TB cases and less than 30% of cases in some
approaches have been developed for more than a cen- recent African surveys (Shinnick 2000). The sensitiv-
tury, and one is still recommended in the normal med- ity of smear microscopy can be increased to 50-80%
ical practice (the skin test against tuberculin), others if specimens are liquefied and the bacilli concen-
are presently evaluated but none are yet validated. trated before AFB microscopy.
842
Culture and isolation of the TB bacillus are never
obtained in 100% of cases, and are usually positive
only in 70-80% of the patients with pulmonary TB
(Stone et al. 1997). The extent is even lower among
patients with extra-pulmonary TB, HIV-co-infected
patients and in childhood TB (Eamranond and Jara-
mill 2001). New tests using molecular and genetic
techniques have been developed, but their results are
not superior to those obtained with the culture from
smear positive patients (Soini and Musser 2001). In
some clinical settings, more invasive methods are
needed to obtain specimens contributing to the diag-
nosis of TB by pathological findings.
Thus, there is an urgent need for applying new
tests which are cheap, rapid, more specific, with an
higher sensitivity, and are capable of being easily per-
formed. Such tests have been sought for a long time,
and antibody detection has been proposed as good
complementary method.
47.2.3
Immunological T- and B-Cell Responses
in TB Patients
The peripheral immunological responses are classi-
fied according to the identification of specific cellular
(lymphocytes) or serum (antibody) components that
the immune system developed after TB infection and
during the TB disease.
The specific T-cell mediated immune responses are
detected in TB patients using the TST and more recently
with the use of ex vivo assays (Andersen et al. 2000).
The specific B-cell mediated immune responses
are detected in nearly all studies by the measurement
of circulating antibodies using ELISA tests with
native, or recombinant, purified proteins (Daniel and
Debanne 1987). There are also some studies demon-
strating the presence of antibodies and specific anti-
gens in circulating immune complexes (Simonney
et al. 1997). Some others have looked at the cellular
levels in body fluids, such as the CSF (Lu et al. 1990),
and here we will present recent data, concerning
the numeration, in the peripheral blood, of specific
secreting B-cells that recognise specific secreted anti-
gens from M. tuberculosis (Sousa et al. 2000).
47.2.3.7
Ex Vivo T-Cell Immune Responses
Several authors have used the in vitro production of
IFN-"{ by Peripheral Blood Mononuclear Cells (PBMC)
from patients with TB or from healthy donors. These
P. H. Lagrange et al.
cells are cultured for five days in the presence of
ESAT-6, CFPlO antigens or PPD. Production of IFN-"{
is measured in the supernatant (Ravn et al. 1999). Dif-
ferent cut-off points are used in order to evaluate the
diagnostic values. The results showed that, even if the
specific antigens ESAT-6 and CFP-I0 are associated
with a lower sensitivity compared with the tuberculin,
they are to a great extent more specific than the PPD
(van Pinxteren et al. 2000). For instance, when cut-off
values were either 300 or 1000 pg/ml, these tests were
more able to discriminate the patients not infected by
the TB bacillus.
Using the same ex vivo model, the diagnostic value
of ESAT-6 has been evaluated in our laboratory look-
ing at a group of 23 TB patients compared with 16
healthy controls, or hospitalised patients not having
TB. A statistically significant (p<O.OS) higher level of
INF-g production was observed for the TB patients
(290.9± 82.5 pg/ml) as compared to the control
values (3S.9±IS.9 pg/ml). However, if the specificity
was high (93.9%), as described above, the sensitivity
(43.9%) was much lower that those reported in the
current literature (Andersen et al. 2000).
From these results and those published by others,
there is evidence that the ex vivo test for the produc-
tion of INF-"{ by PBMC using the ESAT-6 is highly
specific and sensitive for TB infection (Lalvani et al.
2001). Although being also very specific for the TB
Disease, the sensitivity of the IFN-"{ test, still showed
a great variability depending upon the state and the
severity of the disease (Hirsch et al. 1999). Moreover,
it has been neither evaluated extensively in HIV-co-
infected nor in TB-treated patients.
47.2.3.2
In Vitro B-Cellimmune Responses
In almost all the reported experimental TB studies,
the B-cell immune response has never been associ-
ated with any protection against the TB bacillus,
neither using B-cell depleted mice (using anti-IgG
antibodies or in KO mice) nor after using passive
transfer of immune sera from infected, or surviv-
ing mice, or rats after infection with the TB bacillus
(Flynn and Chan 2001). However a puzzling observa-
tion published several years ago (Forget et al. 1976)
concerned the capacity of a rabbit immune serum
to increase the bacteria replication in target organs.
This increase was observed only when a low infec-
tive dose was used. No one has been able to provide
a pertinent explanation for this phenomenon. In B-
cell immuno-compromised human condition, such
as the agammaglobulinaemia, the frequency of TB is
Monitoring Treatment Efficacy
not greater than in the environmental population of
children with the same age.
The only interesting clinical observation, concern-
ing the potential antibody protection in TB children,
was those establishing a correlation between the
serum very low levels of anti-LAM antibodies and the
frequency of disseminated TB in children less than
four years of age (Costello et al. 1992). The interpre-
tation, given by the authors, was the well known lim-
ited capacity of these children to mount an antibody
response to the polysaccharide antigen at that age.
On the other hand, there are several published
reports (over several decades) describing the real
capacity of the infected TB hosts to produce specific
antibodies during TB (Hoffenbach et al. 1985; Daniel
1988). Also, many attempts have been made to trans-
late this into a diagnosis test, but without any success.
This was due mainly due to the use of non-purified
cross-reacting antigens and the non-validation of the
techniques proposed (Daniel and Debanne 1987).
Since the 1980s,newer approaches have been selected
using Enzyme-Linked Immunosorbent Assays (ELISA)
and more purified antigens produced by recombinant
technologies. The results demonstrated a great potential
for diagnosis of active pulmonary TB, but it has been
recently noted that the host presents an heterogeneous
production of antibodies to several antigens, requiring
a multiple antigens approach (Genarro 2000).
The real advantages of ELISA tests using purified
protein antigen are their rapidity (less than 24 hours)
and simplicity, with comparable, or higher, sensitiv-
ity and specificity with the AFB test. However, sev-
eral authors have reported that HIV-co-infected TB
patients were usually poor responders to these pro-
tein antigens (Verbon et al. 1993; Daniel et al. 1994).
Thus, other antigens and specifically non-protein
antigens need to be tested.
It is important to note that the absent, or very low,
specific antibody levels in serum could be due to the
presence of circulating immune complexes (CIC),
not being detected by classical ELISA tests. In fact,
several authors have demonstrated the occurrence
of specific CIC in TB patients, showing the presence
of specific antibodies and/or antigens in these CIC
(Raja et al.1995; Simonney et al.1997).
However, during the six months of therapy and
long after, several authors reported evidence that the
specific antibody levels either increased or remained
usually high; their decay began normally after sev-
eral months or years of chemotherapy. It was felt that
early TB-treatment monitoring using antibody level
follow-up was not helpful (Kaplan and Chase 1980;
Drowart et al. 1991).
843
By contrast, evidence was given recently that
the number of circulating specific B cells, able to
secrete antibody (Antibody-Secreting Cells - ASC), is
increased in TB patients before, and during, the first
week of their treatment (Sousa et al. 2000). It is worth
noting that these elevated numbers of Bcells began to
decrease during the second week after the onset of an
efficient treatment with an exponential decay, these
ASC being virtually absent after one month of effi-
cient chemotherapy. Two techniques have been used
to demonstrate the presence of ASC in the peripheral
blood, the so-called IVAP and the ELISPOT, as shown
below. The results of these new techniques were com-
pared to the classical serum antibody evaluation.
Prospective Antibodies Study of Patients with TB
The detection of specific antibodies in patients' serum
was first performed in patients before the beginning
of the treatment with an ELISA test using one of the
three glycolipid antigens that are specific for the M.
tuberculosis complex. Several sets of data have shown
that they are specific for the M. tuberculosis complex,
which are not present in all atypical mycobacteria
(Cruaud et al.1990).This work was undertaken follow-
ing the published results from Cruaud and colleagues
describing the reliable diagnostic values of an ELISA
test using two (the DAT and the PGLTb 1) of these anti-
gens (Cruaud et al. 1990). A third specific glycolipid
antigen, the lipo-oligosaccharide (LOS), which was
described as being only present in M. tuberculosis, and
not in M. bovis, M. africanum and M. microti (Papa et
al. 1993) has been evaluated concomitantly.
The three purified glycolipid antigens were kindly
provided first by Papa from the Pasteur Institute
(Paris), and later by Dafft (CNRS, Toulouse, France).
These three glycolipids have been purified after
chloroform/methanol extraction from M. tuberculo-
sis (H37Rv strain). The three antigens were the lipo-
oligosaccharide (LOS), the diacyl-trehalose (DAT)
and the phenolic glycolipid (PGL-Tb1). These three
antigens are located at the external surface of the M.
tuberculosis bacillus.
Initially, both IgG and IgM antibodies were evalu-
ated, but the IgM responses were found to be less dis-
criminating compared with the IgG antibodies (Dale-
ine and Lagrange 1995), thus all the data presented
here will concern only the antibodies IgG responses.
The population of tested patients consisted of
149 patients with documented TB (either after being
shown to be culture positive or after pathological
confirmation). The vast majority of these TB patients
were HIV tested, because such testing has been recom-
844
mended in France since 1994 for all new TB case. A total
of 134 healthy control individuals, mostly from the
Hospital Saint Louis Blood Bank, were also included
in this study in order to evaluate the specificity of our
ELISA test. The group ofTB patients was classified into
two main subgroups according to the HIV status (HIV
positive or negative) and further subdivided according
to location of the disease (pulmonary or extra-pulmo-
nary) and the results of the AFB test.
Nearly 75% of the patients were HIV negative
(25% positive) and the median age was not statisti-
cally different in these two groups. However, when
looking at the TB localisation, it was quite interesting
to note, as already reported in the current literature,
that HIV-positive TB patients presented a higher pro-
portion of extra-pulmonary disease (53%) compared
with HIV-negative TB patients (27%). Moreover, it
was also observed that the AFB tests were more often
positive in HIV-positive patients compared with
HIV-negative TB patients with extra-pulmonary
disease, since there were more invasive pathological
samples in the former.
The varying OD values did show a statistically
significant difference (p<O.OO1) between TB patients
compared to controls for each of the antigens tested
(Table 47.1). In order to calculate the diagnostic
values of our domestic ELISA test, the best cut off
value for each antigen was chosen after establishing
the ROC curves.
Using the best cut-off values, a specificity of 92.5%,
93.3% and 94.3% was obtained for the DAT, the
PGLTbl and the LOS antigens, respectively. Analo-
gous specificity (91.8%) was obtained in calculating
the cumulative positive results from the three tests.
There were some variability when the sensitiv-
ity was calculated according to the antigen tested:
67% for the DAT, 72% for the PGLTbl and near 78%
for the LOS antigen. As it was already observed with
other antigens (A60, 38 kDa), the patient's responses
are heterogeneous, some patients did respond to only
one antigen, some against two antigens, or to the three
antigens (Kaplan and Chase 1980; Simonney et al.
1996). The cumulative results using the three antigens
gave a much higher sensitivity than the individual
tests, reaching in this study a sensitivity of 85.2%.
The patient's characteristics were important in
analysing the diagnostic values of our test in compar-
ison to results from other ELISA tests using purified
proteins (which showed usually much lower sensi-
tivities in patients with extra-pulmonary TB), results
from those with smear AFB negative tests and more
dramatically the results in HIV-positive patients
independent of the TB site (Verbon et a1.1993).
P. H. Lagrange et al.
In contrast, our domestic ELISA tests using the
three glycolipid antigens always showed an higher
sensitivity for each antigen tested in HIV-positive TB
patients compared with HIV-negative TB patients
(Table 47.2). The cumulative sensitivity of the three
combined tests was 85.2% for all TB patients, but
95.2% in the HIV-positive group.
Similar sensitivities (95%) were observed in HIV-
positive patients with pulmonary, or with extra-pul-
monary, TB disease. The lowest sensitivity (68%) was
observed in HIV-negative patients with extra-pulmo-
nary TB. There was no significant difference between
ELISA sensitivity in smear-positive and smear-nega-
tive patients with pulmonary TB, for all HIV immune
status and sites of diseases. The main value of the
Elisa test using the glycolipids is thus to provide early
complementary information by antibody detection
in HIV-positive TB patients, particularly those with
a negative smear test.
Retrospective Study in a Cohort of TB Patients
The ELISA test has been then tested retrospectively
using the serum from several individuals belonging
to a cohort of HIV-positive and HIV-negative TB
patients, followed for several months at varying time-
points before and after the diagnosis of tuberculosis.
The sera have been conserved at -80 0 C, and all the
tests have been blinded.
In this cohort of 27 TB patients, 11 were HIV posi-
tive and 16 were HIV negative. For a specificity higher
than, or equal to, 93%, the sensitivity for each ELISA
test using individual antigen (DAT and PGL-Tb 1) was
found to be almost 80% in proven TB patients and
100% in culture-negative TB patients (clinical and
radiological assessments and treatment efficacy). As
described previously, the sensitivity was again higher
in HIV-positive patients compared with HIV-nega-
tive TB patients (100% and 81.3%, respectively).
In several TB patients, sequential sera have been
harvested at various times before TB diagnosis, during
and long after the 6 months treatment. The dynamics
of the IgG antibody response to the DAT and PGL Tbl
antigens was then evaluated using several sera from
each patient. The results of such an approach using the
DAT antigen are shown in Fig. 47.1.
The means of anti-DAT antibodies (OD values)
were higher than the cut-off level found in a control
group for all the period of time under study. In the
HIV-positive TB patients the values increased and
reached a peak at the time of the TB clinical and/or
bacteriological diagnosis. This is in sharp contrast
with the relative delayed occurrence of the anti-
Monitoring Treatment Efficacy 845

Table 47.1. IgG antibody levels against lipo-oligosaccharide (LOS), diacyl-trehalose

(OAT) and phenolic glycolipid (PGL-Tbl) measured by ELISA in the sera of 149
adults with documented tuberculosis and 134 healthy controls (mean±SEM)

Optical density (mean±SEM)

LOS (n)b OAT (n) PGL-TBI (n)

TB patients 134.3±1O.7 (116) 200.0±13.8 (149) 164.8±14.3 (149)

Controls 13.7±2.1 (105) 46.9±3.4 (134) 25.1±17 (134)
p Values p <0.0001 p <0.001 p <0.001
'Absorbance at 414 nmxlO-3
bNb of serum tested

Table 47.2. Comparative sensitivity of the ELISA tests using the three glycolipid
antigens (LOS, OAT and PGL-Tbl) and the cumulative results in the total tested
population, in 42 HIV-positive and 107 HIV-negative patients with documented
tuberculosis

Antigen Sensitivity'

HIV
Positive Negative Total
LOS 82.5 (33/40)b 75.0 (57/76) 77.6 (90/116)
OAT 73.8 (31/42) 64.5 (69/107) 67.1 (100/149)
PGLTbl 88.1 (37/42) 65.4 (70/107) 71.8 (107/149)
Cumulative results 95.2 (40/42) 81.3 (87/107) 85.2 (127/149)

'Cut-off values: LOS:0.043; DAT:0.110; PGLTbl:0.043

bNb of positive serum/Nb of serum tested

than those in HIV-negative patients. However, at 15

months in HIV-negative and 30 months in the HIV-
positive treated TB patients, the remaining antibody
levels against DAT remained still higher than the
threshold level . If this ELISA test is used system-
atically, it will not discriminate new TB patients from
non-active TB in already treated patients. Similar
results have been obtained using the PGL-Tbl anti-
It
gen (data not shown)
In conclusion, according to these data, two points
I Q" .....--+---,,--+,- >----1 are worth noting. First, the antibody follow- up do not
·15 ·10 ·5 o 10 15 20 25 30
seem to be very useful at monitoring early treatment
tim. (month)
efficacy, since both the anti-DAT and anti-PGLTbl
Fig. 47.1. Kinetics of free circulating IgG antibody levels antibody levels decline very slowly during the 6
against diacyl-trehalose (OAT) measured by ELISA in sequen- months period of treatment. Second, when a new
tial serum from 11 HIV-positive (black squares) and 16 HIV- patient is suspected of TB disease and for the inter-
negative (open squares) patients with tuberculosis during anti
tuberculosis treatment (mean ± SEM). The horizontal dotted
pretation of the anti-glycolipid antibodies results, it
line represents the cut-off value calculated with control sera is quite important to know if the patient has been
and the open arrow the onset of treatment treated in the recent past for a first episode of TB.

Prospective Study in TB Patients Under Treatment

body production in the HIV-negative group of TB
patients.
It is also important to note that, even if the mean
antibody titres showed a dramatic decline in this long
term follow-up study, the slope of the curve appears
to be less pronounced in HIV-positive TB patients
In a group of23 HIV-negative patients with documented
TB, 13 were followed sequentially from the onset of the
treatment to its completion. During this follow-up, one
tube ofserum was harvested every week during the first
month and at monthly intervals thereafter.
846
ELISA tests were performed using several indi-
vidual sera and the dynamics of the free circulating
IgG antibodies against a recombinant protein anti-
gen (r38 kDa) and a purified carbohydrate antigen
(LAM) were evaluated (Sousa et al. 2000).
In all sera, a small but consistent decrease in anti-
body levels over the first two weeks of chemotherapy
was observed. This decrease was followed by a grad-
ual increase in the case of both anti-38 kDa and anti-
LAM. Thereafter, the mean level of anti-38 kDa IgG
decreased from day 60 up to day 180. In all cases, this
decreased was observed after M. tuberculosis culture
conversion. In contrast, the mean level of anti-LAM
IgG remained elevated for the complete duration of
follow-up of patients. These results are in agreement
with our preceding findings concerning the antibod-
ies against the three glycolipids and also with those
against proteins reported in the literature (Kaplan
and Chase 1980; Drowart et al. 1991).
Circulating Immune Complexes (CIe)
Before and During Therapy
Detection of Antibodies in the Circulating Immune
Complexes
In order to understand the false-negative ELISA
tests in some TB patients, the presence of circulat-
ing immune complexes (CIC) in serum was first
evaluated. Then, after CIC dissociation, the patient's
anti-glycolipids IgG antibody contents were evalu-
ated. The presence of anti-glycolipids within CIC in
lepromatous leprosy sera has been already reported,
often masking the real production of antibodies
(Tomimori-Yamashita et al. 1999).
CIC were isolated in the serum from adults, and
children, with TB and in control sera using the 2.5%
PEG precipitation technique. This has already been
done with leprosy patient sera. Such PEG concentra-
tions did not precipitate aggregated IgG, and all the
CIC were precipitated after the first run. The quantity
of CIC was measured at 280 nM with a spectropho-
tometer. CIC levels were significantly higher (p<O.Ol)
in all the sera of both adults (Simonney et al. 1997)
and children (Simonney et al. 2000) with TB, as com-
pared with the control sera.
The precipitates were dissociated using an acidic
solution (pH 2.4) and antibodies against the LOS,
DAT and PGL-Tb1 antigens were immediately mea-
sured using an identical ELISA test. Free antibodies
to the same antigens were tested concomitantly using
separate aliquots of the non-precipitated serum from
the same patients.
P. H. Lagrange et aI.
As previously described, varying levels of the free
anti-glycolipids IgG antibodies were observed in the
sera of TB patients, and usually lower mean levels were
observed in HIV-negative compared with HIV-positive
TB patients. There was no direct correlation between
the levels of free circulating IgG antibodies and the
levels ofIgG antibodies detected within the CIe.
The ELISA sensitivities were calculated for the free
and the complexed antibodies present in the CIe. It
was observed that almost all of the complexed anti-
bodies were detected in false-negative HIV-negative
TB patients and not in the HIV-positive TB patients.
The number of ELISA-positive TB patients is higher
in the HIV-negative compared to the HIV-positive TB
patients. About 96% of the former had CIC contain-
ing specific antibodies to one, two or the three glyco-
lipids compared to 72% in the later.
Kinetics of Circulating Immune Complexes During
Treatment
Using the serum aliquots from this prospective study
the dynamics of the total CIC and their anti-DAT IgG
antibodies were evaluated during treatment. The
kinetics of the mean levels of the total CIC in the
serum according to the time is shown in Fig. 47.2.
It is interesting to note that after a sharp and short
increase during the first two weeks following the
treatment onset, the total CIC reached a plateau
for three months and gradually dropped to baseline
level by day 120. Similar findings have already been
described in the literature (Raja et al. 1995).
The level of circulating free IgG antibody against
the DAT antigen did not change significantly during
the treatment (Fig.47.3 lower line), except for a
slight drop during the first week. However, anti-DAT
(Fig. 47.3 upper line) and anti-ST-CF (data not shown)
IgG antibodies present in CIC, showed a marked
increase as soon as the treatment started which mir-
rored the kinetics of free, or newly, synthesised circu-
lating antibodies against STCF and DAT. Such findings
were also observed for free IgG antibodies against
LAM and the 38 kDa protein (data not shown).
In conclusion, the levels of free circulating IgG
antibody against several protein, carbohydrate or
glycolipid M. tuberculosis antigens (38 kDa, ST-CF,
LAM, DAT, PGL-Tbl), detected by ELISA in the
whole serum, varied depending on the antigen used.
The individual kinetics did not yield any information
on early treatment outcomes. These data confirm the
lack of predictive power of serological tests for the
treatment monitoring as demonstrated previously
(Drowart et al. 1991).
Monitoring Treatment Efficacy
Circulatin& luunYn cPUlplex
1 bodies were often associated with an increased of
1 complexed specific IgG antibodies to M. tuberculosis
o 50 "1.00
DAYS
Fig. 47.2. Kinetics of circulating total immune complex levels
measured by 2.5% polyethylene glycol (PEG) precipitation in
sequential plasma samples from 13 patients during anti-tuber-
culosis therapy (mean ± SEM) (from Sousa et al. 2000)
PAT ELISA
T I
1
CIC-TgG
T T 1 replicating micro-organisms. Therefore, the ELISPOT
1 1
"!~' r r _- " --1 _ ---__ -::r.
T
. T
1
Free I g e
DAYS
Fig.47.3. Kinetics of specific free IgG (lower line), immune
complex-associated IgG (upper line) antibody levels against
diacyl-trehalose (DAT) measured by ELISA in sequential
plasma samples from 13 patients during anti tuberculosis
therapy (mean ± SEM) (from Sousa et al. 2000)
Nevertheless, the early increase in total CIC and
complexed antibody to the DAT, PGLTbl and ST-CF
antigens, associated with a modest decrease in cir-
culating free antibodies to M. tuberculosis antigens,
might be the consequence of an increase of released
mycobacterial antigens, both at the infectious sites
of the disease and in the peripheral blood, reflecting
early bactericidal activity of anti-tuberculosis drugs.
Such results confirm previous studies, showing that
in CIC both specific antigens and antibodies were
significantly elevated in TB patients (Bhattacharya
et al. 1986; Dlugovitzzky et al. 1995). An increase in
CIC levels after two months of treatment followed by
a decrease at six months has already been reported by
Raja and colleagues (1995). However, as shown in our
study, no correlation was demonstrated with the early
resolution of clinical TB disease under treatment.
847
Ex Vivo Numeration of Specific Secreting B Cells
Having demonstrated that free circulating anti-
antigens during the first two weeks of the treatment,
this observation suggested a possible release of anti-
gens, and also a possible reactivation of the B-cell
immune system. In order to evaluate this event, the
"1.50
number of circulating specific B lymphocytes was
first evaluated.
Counting techniques for the circulating antibody
secreting B cells (ASC) using ELISPOT assays have
been developed for the diagnosis of several infectious
diseases due to cytomegalovirus (Segondy et al.1990),
HIV (Lee et al. 1989), Salmonella typhi (Kantele et al.
1986), Toxoplasma gondii (Vendrell et al.1993) Brucella
spp. (Vendrell et al. 1992) and Chlamydia trachomatis
(Ghaem-Maghani et al.1997). These studies have dem-
onstrated a correlation between the detection of cir-
culating ASC and the presence of antigens secreted by
assays in TB patients have been performed to reach
a diagnosis and also for their potential in monitoring
treatment efficacy in TB patients.
With the use of specific antigens (the short term
culture filtrate - ST-CF), corresponding to secreted
antigens released by replicating M. tuberculosis
(Andersen et al. 1991), we reported the presence of
specific ASC in a cohort of 13 patients with active
TB before and their regular decline during recom-
mended multiple-drug therapy (Sousa et al. 2000).
Fresh PBMC were first enriched for Blymphocytes
using positive selection, the isolated cell population
contained around 90-95% B cells measured by
cytometry analysis. ELISPOT plates were pre-coated
with either ST-CF or goat anti-human IgG F(ab')
antibody. Enriched B cells were added in complete
culture medium to the ELISPOT plates and were
incubated for overnight at 37°C. After washing the
B cells extensively in PBS supplemented with Tween
20 (0.05%), the specific antibodies bound to the solid
phase antigen were revealed by incubation with anti-
human IgG Fc specific peroxidase-labelled antibody.
Each spot on the nitrocellulose membrane
(Fig.47.4) indicated the presence of an ASC specific
to the coating antigens. The number of circulating
ASC reactive to ST-CF was evaluated by ELISPOT for
each TB patient before and at varying times during the
treatment. The cut-off value of the normal level ofASC
had been determined in a previous study using healthy
controls and was found to be less than 150 ASCIl0 6
circulating B cells corresponding to a specificity of
848
B cell ELiSP T
Controls T8 patients
Fig.47.4. Two ELISPOT wells showing specific B cell spots
against short term culture filtrate (ST-CF) on the nitro-cel-
lulose membrane stained with goat anti-human IgG from
PBMC of one control (left picture) and a tuberculosis patient
(right picture)
100%. Before initiation of therapy, 82% of TB patients
had detectable specific ASC with a wide range among
patients (200-5,460 ASC/I06 B cells).
The kinetics of the ASC during the treatment
are shown in the Fig. 47.5. From day 0 to day 8, the
median numbers of specific ASC increased markedly
and 100% of the patients had detectable numbers of
circulating ASC on day 8. This peak was followed by a
three-fold decrease of the median number at day 15,
and a lO-fold decrease at day 30. It is noteworthy that
one patient had persistently elevated numbers of ASC
after two months, accounting for the large standard
error in the mean at that time in the histogram. In all
of the remaining cases, the levels of ASC decreased
markedly and always before mycobacteria cultures
turned negative. Moreover, this ACS decrease was
detected either before (in eight cases) or at the same
time (two cases) as the sputum smear conversion.
The percentage of TB patients remaining positive
for sputum smear and cultures after the onset of the
treatment is shown in Fig. 47.6.
Eleven patients had pulmonary TB with moder-
ate to severe disease and two patients had pleurisy
(shown to be positive by biopsy). In all patients with
pulmonary TB, the sputum smears and cultures were
positive for M. tuberculosis before therapy. Only
patients with documented TB have been included in
this study, to obtain comparative data with previous
published studies of culture conversion time (Joloba
et al. 2000) and to demonstrate the value of ASC
numeration value for TB treatment monitoring.
All patients responded well to treatment with culture
conversion and clinical and radiological improvement.
However, one patient was lost to follow-up after the
second month and relapsed one year later. Among the
10 patients with pulmonary TB and a positive outcome,
P. H. Lagrange et al.
ELISPOT
1500
Cut-off: 150 ASC/1 0 6 B cells
o 8 '13 30 60 90 '120 '1.80 240
DAYS
•
Fig.47.S. Kinetics of the numbers of circulating antibody-
secreting cells (ASC) per one million of B cells at varying
time-point during anti-tuberculosis therapy in a group of 13
patients with pulmonary tuberculosis (median ± SEM). The
horizontal dotted line represents the cut-off value calculated
using control patients (from Sousa et al. 2000)
20 .10 1\0 80 100 120 IJO IW 110
Days
Fig.47.6. Kinetics of the percentage of patients remaining
bacteriologically positive for sputum smears (open squares)
and cultures (closed squares) at varying time-points during
anti-tuberculosis therapy (from Sousa et al. 2000)
the mean number of days from the initiation of appro-
priate therapy to obtain negative AFB smears was 27.4
± 7.6 days. Similar values for the first negative sputum
cultures were 45.0 ± 9.2 days with a range of 15-90
days. These results were almost identical with those
described in the current literature (Joloba et al. 2000).
In contrast to the exponential decrease of circulat-
ing specific ASC beginning after the 8 day of treat-
ment, the mean level of circulating free anti-ST-CF
IgG antibody tested by ELISA increased until day 60
and then declined slowly until the end of the treat-
ment, confirming the preceding results obtained with
other antigens.
Thus, for the first time, there is evidence for the
presence of M. tuberculosis specific circulating anti-
body secreting B cells (detected by an ELISPOT assay
in the peripheral blood of TB patients), that can be
Monitoring Treatment Efficacy
used during the treatment follow-up and showed a
predictive value for early and late treatment efficacy.
This is in contrast to other immune markers.
It is assumed that the kinetics of the specific B cell
response might reflect the bacterial dynamics under
anti-mycobacterial therapy. The possible reduction of
the secreted antigens production may account for the
decrease of anti ST-CF specific B cells. Similar findings
has been already demonstrated by other authors for
patients with chronic brucellosis (Vendrell et al. 1992).
Detection of ASC in PBMC suggests that in acute,
or chronic, infectious disease the host immune system
is stimulated continuously by replicating antigens for
several weeks, or months, after the onset of illness.
This has been already observed in patients with TB
meningitis by detecting anti-BCG secreting B cells in
cerebral fluid and to a lesser degree in PBMCs (Lu et
al. 1990; Baig 1995). A reduction of the antigenic load
occurred in TB patients B during effective therapy
either by detecting antigens in sputa (Wallis et al.
2001), in the peripheral blood (Sethna et al. 1998) or
in the urine (Hamasur et al. 2001).
At least, the great advantage of the ELISPOT tech-
nique is the ease of collecting peripheral blood, to quan-
tify the circulating specific ASC and to use the results
as a surrogate marker of the persistence of replicating
mycobacteria in tissues. Thus, stopping the mycobacte-
rial replication with an effective treatment is associated
with a drastic reduction of M. tuberculosis secreted pro-
teins, such as those contained in the ST-CF (Andersen
et al. 1991) and this explained the concomitant rapid
decrease in specific circulating B cells observed in the
peripheral blood of well-treated TB patients.
In Vitro Antibody Production
In order to simplify the ELISPOT assay (demonstrat-
ing the presence and the dynamics of the specific
secreting B cell response) a new approach, looking
at the In Vitro Antibody Production (IVAP), has been
evaluated first for the diagnosis of TB patients. This
has already used by several authors in other infec-
tious diseases (Vendrell et al. 1992, 1993).
It involves measuring the production of antibodies
derived from the culture of PBMC from TB patients.
The supernatants are harvested on the 6th day of the
culture without any antigen and the antibodies are
detected by ELISA using the same antigens as for the
ELISPOT. The IVAP mean levels of anti ST-CF IgG
antibody were significantly higher (p<O.OOI) in 35
patients with proven documented TB (OD at 490 nM:
0.735 ±0.136) as compared to the 19 healthy control
supernatants (OD at 490 nM: 0.027 ±0.007).
849
By using a threshold cut-off (mean + 3 SD), giving
a specificity of 100%, the sensitivity of this IVAP
assay was 70.6%. This was about the same order that
was obtained with the ELISPOT assay. Similar sensi-
tivity but lower specificity were obtained for antibod-
ies against ST-CF and PGLTbl tested in the plasma.
A prospective follow up of TB patients under treat-
ment is now being undertaken in our laboratory.
47.3
Conclusions
The results obtained with our in-house ELISPOT
and IVAP assays, using antigens present within the
ST-CF, demonstrate that B-cell specific antigens from
M. tuberculosis are useful and encouraging: firstly as
a diagnostic tool and secondly for the monitoring
of the TB disease in non-HIV patients. In contrast,
specific M. tuberculosis free- and CIC-associated
IgG antibodies against the glycolipid antigens were
shown to be helpful especially as complementary tests
for the diagnosis ofHIV-co-infected TB patients, but
without any help for monitoring treatment efficacy.
However, it is important firstly to mention the indi-
vidual polymorphism of the B-cell immune responses
and the need for using several antigens to detect posi-
tive TB patients. Secondly, the various types of antigen
used might give different types of information. For
example, glycolipid antigens, such as those used in our
domestic ELISA test, were more effective in detecting
TB in HIV-positive patients compared with the results
obtained in HIV-negative patients. Such results con-
trast with the published results of ELISA tests using
specific purified recombinant proteins (Daniel et al.
1994). It is possible that the results obtained using gly-
colipids might be due to the special restricted immune
response, mediated through the CD1 pathway recently
described for these lipid antigens (Park and Bendalac
2000), not affected by the HIV-induced CD4 depres-
sion. On the other hand, secreted antigens, such as
those present within the ST-CF (a surrogate marker
of bacterial multiplication) are of great interest for
monitoring TB treatment. In fact, the elevated num-
bers of circulating specific M. tuberculosis secreting B
cell occurring and persisting in a vast majority of TB
patients and their rapid decrease after the onset of the
treatment might be used as a strong surrogate marker
of treatment efficacy and effective patient compliance.
A secondary increase, or an absence of change in ASC
numbers, evaluated by regularly harvesting periph-
eral blood samples during the treatment, might then
850
indicate a poorly compliant patient, or the presence
of multiple-drug-resistant tuberculosis. Such ex vivo
ELISPOT and in vitro IVAP tests have been shown to
be as informative as the smear and culture conversion,
but their results were obtained much earlier than the
classical bacteriological gold standard tests.
References
Andersen P et al (1991) Proteins released from Mycobacterium
tuberculosis during growth. Infect Immun 99:1905-1910
Andersen P et al (2000) Review: specific immune-based diag-
nosis of tuberculosis. Lancet 356:1099-1104
Araj G et al (1993) Improved detection of mycobacterial antigens
in clinical specimens by combined enzyme-linked immuno-
sorbent assays. Diagn Microbiol Infect Dis 17:119-127
Baig SM (1995) Anti-purified protein derivative cell-enzyme
linked immunosorbent assay, a sensitive method for early
diagnosis of tuberculous meningitis. J Clin Microbiol 33:
3040-3041
Barnes PF, Chan LS, Wong SF (1987) The course of fever
during treatment of pulmonary tuberculosis. Tubercle 68:
255-260
Bhattacharya A et aL,(1986) Antibody-based enzyme-linked
immunosorbent assay for determinat!on of immune
complexes in clinical tuberculosis. Am Rev Respir Dis 134:
205-209
Chaparas SD et al (1985) Tuberculin test. Variability with the
Mantoux procedure. Am Rev Respir Dis 132:175-177
Costello AM et al (1992) Does antibody to mycobacterial
antigens, including lipoarabinomannan, limit dissemina-
tion in childhood tuberculosis? Trans R Soc Trop Med
Hyg 86:686-692
Cruaud P et al (1990) Evaluation of a novel 2-3-diacyl-
trehalose-2'sulfate (SL-IV) antigen for case finding and
diagnosis of leprosy and tuberculosis. Res Microbiol 141:
679-669
Daleine G, Lagrange PH (1995) Preliminary evaluation of
a Mycobacterium tuberculosis lipooligosaccharide (LOS)
antigen in the serological diagnosis of tuberculosis in
HIV seropositive and seronegative patients. Tubercle Lung
Disease 76:234-239
Daniel TM (1998) Antibody and antigen detection for immu-
nodiagnosis of tuberculosis: why not? What more needed?
Where we stand today? J Infect Dis 158:678-680
Daniel TM, Debanne SM (1987) The serodiagnosis of tuber-
culosis and other mycobacterial diseases by enzyme-linked
immunosorbent assay. Am Rev Respir Dis 135:1137-1151
Daniel TM et al (1994) Reduced sensitivity of tuberculosis
serodiagnosis in patients with AIDS in Uganda. Tubercle
Lung Dis 75:33-37
Dlugovitzky D et al (1995) Circulating immune complexes in
patients with advanced tuberculosis and their association
with auto-antibodies and reduced CD4+ lymphocytes.
Brazil J Med Bioi Res 28:331-335
Drowart A et al (1991) Antibody levels to whole culture filtrate
antigens and to purified P32 antigen during treatment of
smear positive tuberculosis. Chest 100:685-687
Eamranond P, Jaramill E (2001) Tuberculosis in children: reas-
P. H. Lagrange et al.
sessing the need for improved diagnosis in global control
strategies. Int J Tuberc Lung Dis 5:594-603
Flynn JL, Chan J (2001) Immunology of tuberculosis. Annu
Rev ImmunoI19:93-129
Forget A et al (1976) Enhancement activity of antimycobac-
terial sera in experimental Mycobacterium bovis (BCG)
infection in mice. Infect Immun 13:1301-1306
Gennaro ML (2000) Immunological diagnosis of tuberculosis.
Clin Infect Dis 30 [SuppI3]:S243-S246
Ghaem-Maghami S et al (1997) Characterisation of B-cell
responses to Chlamydia trachomatis antigens in humans
with trachoma. Infect Immun 65:4958-4964
Hamasur B et al (2001) Rapid diagnosis of tuberculosis by
detection of mycobacteriallipo-arabinomannan in urine.
J Microbiol Methods 45:41-52
Hirsch CS et al (1999) Depressed T-cell Interferon-g responses
in pulmonary tuberculosis: analysis of underlying mech-
anisms and modulation with therapy. J Infect Dis 180:
2069-2073
Hoffenbach A, Lagrange PH, Bach MA (1985) Strain variations
of lymphokine production and specific antibody secretion
in mice infected with Mycobacterium lepraemurium. Cell
ImmunoI91:1-11
Jidani A et al (1980) The early bactericidal activity of drugs
in patients with pulmonary tuberculosis. Am Rev Respir
Dis 121:939-949
Joloba ML et al (2000) Quantitative sputum bacillary load
during rifampin-containing. short-course. chemotherapy
in human 'immunodeficiency virus-infected and non-
infected adults with pulmonary tuberculosis. Int J Tuberc
Lung Dis 4:528-536
Kantele A, Arvilommi H, Jokinen I (1986) Specific immu-
noglobulin-secreting human blood cells after per-oral
vaccination against Salmonella typhi. J Infect Dis 153:
1126-1131
Kaplan MH, Chase MW (1980) Antibodies to mycobacteria in
human tuberculosis 1. Development of antibodies before
and after antimicrobial therapy. J Infect Dis 142:825-834
Kennedy Net al (1994) Polymerase chain reaction for assess-
ing treatment response in patients with pulmonary tuber-
culosis. J Infect Dis 170:713-716
Lalvani A et al (200l) Rapid detection of Mycobacterium tuber-
culosis infection by enumeration of antigen-specific T cells.
Am Rev Respir Crit Care Med 163:824-828
Lee FK et al (1989) ELlSPOT: a new approach to studying the
dynamics of virus-immune system interaction for diag-
nosis and monitoring of HIV infection. AIDS Res Hum
Retrovirus 5:517-523
Liu Z, Shilkret KL, Ellis HM (1999) Predictors of sputum
culture conversion among patients with tuberculosis in
the era of Tuberculosis resurgence. Arch Intern Med 159:
1110-1116
Lu CZ et al (1990) Early diagnosis of tuberculous meningitis
by detection of anti-BCG secreting cells in cerebral fluid.
Lancet 336:517-523
Papa F et al (1993) Isolation and characterisation of serology
reactive lipo-oligosaccharide from Mycobacterium tuber-
culosis. Res Microbioll44:91-99
Park SH, Bendalac A (2000) COl-restricted T-cell responses
and microbial infections. Lancet 406:788-792
Pereira Arias-Bouda LM et al (2000) Development of antigen
detection assay for diagnosis of tuberculosis using sputum
samples. J Clin Microbiol 38:2278-2283
Monitoring Treatment Efficacy
Raja A et al (1995) Characterisation of mycobacterial antigens
and antibodies in circulating immune complexes from pul-
monary tuberculosis. J Lab Clin Med 125:581-587
Ravn P et al (1999) Human T cell response to the ESAT-6 antigen
from Mycobacterium tuberculosis. J Infect Dis 179:637-645
Rieder HL (1996) Sputum smear conversion during directly
observed treatment of tuberculosis. Tuberc Lung Dis 77:
124-129
Sada E et al (1992) Detection of Lipoarabinomannan as a diag-
nostic test for tuberculosis. J Clin MicrobioI30:2415-2418
Segondy M et al (1990) Cytomegalovirus-specific Bcell activa-
tion as a potential marker for the diagnosis of cytomegalo-
virus infection. Eur J Clin Microbiol Infect 9:745-750
Sethna KB et al (1998) Longitudinal trends in serum levels
of mycobacterial secretory (30 kDa) and cytoplasmic
(65 kDa) antigens during chemotherapy of pulmonary
tuberculosis patients. Scand J Infect Dis 30:363-368
Sevim T et al (2002) Treatment outcome of relapse and defaulter
pulmonary tuberculosis patients. Int J Tuberc Lung 6:320-325
Shinnick TM (2000) Diagnostic test needs for evaluating
anti-tuberculosis vaccines. Clin Infect Dis 30 [Suppl 3):
S276-S278
Simonney N et al (1996) Comparison of A60 and three glyco-
lipid antigens in ELISA test for tuberculosis. Clin Microbiol
Infect 2:214-222
Simonney N et al (1997) Circulating immune complexes
in human tuberculosis sera: demonstration of specific
antibodies against Mycobacterium tuberculosis glyco-
lipid (DAT, PGLTbl, LOS) antigens in isolated circulating
immune complexes. Eur J Clin Invest 27:128-134
Simonney N, Bourrillon A, Lagrange PH (2000) Analysis of
circulating immune complexes (CICs) in childhood tuber-
culosis: levels of specific antibody to glycolipid antigens
and relationship with serum antibody. Int J Tuberc Lung
Dis 4:152-160
Sirgel F, Venter A, Mitchison DA (2001) Sources of variations in
studies of the early bactericidal activity of antituberculosis
drugs. J Antimicrob Chemother 47:177-182
Small P M, Fujiwara PI (200 I) Management of tuberculosis in
the United States. N Engl J Med 345:189-200
Smith H (1987) Paradoxical responses during chemotherapy
of tuberculosis. J Infect 15:1-3
851
Soini H, Musser JM (200 I) Molecular diagnosis of Mycobacte-
ria. Clin Chern 47:809-814
Sonnenberg P et al (2001) HIV-l and recurrence, relapse, and
re-infection of tuberculosis after cure: a cohort study in
South African mineworkers. Lancet 358:1687-1693
Sousa AO et al (2000) Kinetics of circulating antibodies,
immune complex and specific antibody-secreting cells in
tuberculosis patients during 6 months of antimicrobial
therapy. Tuberc Lung Dis 80:27-33
Stone BL et al (1997) The diagnostic yield of acid-fast-bacil-
Ius smear positive sputum specimens. J Clin Microbiol 35:
1030-1031
Sundaram V et al (2002) Yield of continued monthly sputum
evaluation among tuberculosis patients after culture con-
version. Int J Tuberc Lung Dis 6:238-245
Tomimori-Yamashita J et al (1999) Antibody-based ELISA for
determination of anti-PGLl specific circulating immune
complex in leprosy. Leprosy Rev 70:261-271
Van Pinxteren LAH et al (2000) Diagnosis of tuberculosis
based on the two specific antigens ESAT-6 and CFPI0.
Clin Diagn Lab Immunol 7:155-160
Vekemans J et al (2001) Tuberculosis contacts but not patients
have higher gamma Interferon responses to ESAT-6 than
do community controls in the Gambia. Infect Immun 69:
6554-6557
Vendrell JP et al (1992) In vitro antibody secretion by peripheral
blood mononuclear cells as an expression of the immune
response to Brucella spp in humans. J Clin Microbiol 30:
2200-2203
Vendrell JP et al (1993) In-vitro synthesis of antibodies to
Toxoplasma gondii by lymphocytes from HIV-1-infected
patients. Lancet 342:22-23
Verbon et al. (1993) Evaluation of different tests for the sero-
diagnosis of tuberculosis and the use of likelihood ratios
in serology. Am Rev Respir Dis 148:378-384
Wallis R et al (2001) Induction of isoniazid-induced
expression of Mycobacterium tuberculosis antigen 85
in sputum: potential surrogate marker in tuberculosis
chemotherapy trials. Antimicrob Agents Chemother 45:
1302-1304
WHO (2001) Global tuberculosis control. World Health
Organisation, Geneva, WHO/CDSI287
Control and Prevention
48 Tuberculosis Control in Developing
and Developed Countries
KEVIN SCHWARTZMAN

CONTENTS 48.7.4 Peru 873

48.7.5 United States 874
48.1 Introduction 855 48.7.6 England and Wales 876
48.2 Overview of Interventions 48.7.7 Russia 876
for Tuberculosis Control 856 48.8. Conclusion 877
48.2.1 Preventing Tuberculous Infection 856 References 877
48.2.2 Preventing Progression and Reactivation 856
48.3 Cost-Effectiveness of Tuberculosis Treatment 857
48.4 The World Health Organization's DOTS
Strategy 858
48.4.1 Description 858
48.4.1.1 Government Commitment 858 48.1
48.4.1.2 Standard Diagnostic Approach 858 Introduction
48.4.1.3 Supervised Provision of Standard Treatment
Regimens 859 In 1993, the World Health Organization declared
48.4.1.4 Documentation and Surveillance 859
48.4.2 Status of the DOTS Strategy 860 tuberculosis to be a global public health emergency.
48.4.3 Controversies 861 The vast worldwide burden of mortality and suffering
48.5 Other TB Control Interventions 862 attributed to tuberculosis has been reviewed elsewhere
48.5.1 Active Case-Finding in Resource-Poor Areas 862 (see Chapter 3): an estimated 9 million people develop
48.5.2 Screening Interventions in Resource-Rich
active TB each year, and 2 million die as a result. Yet
Areas 862
48.5.2.1 Screening of Migrants to Low-Incidence from a medical standpoint, the vast majority of tuber-
Countries 863 culosis patients can readily be treated and cured, using
48.5.2.2 Screening and Control in Specialized Settings 864 medications that have been available for decades. Why,
48.5.3 Bacille Calmette-Guerin (BCG) Vaccination 865 then, has it been so difficult to make sustained progress
48.5.4 Challenges to Global TB Control 866
toward reducing the burden imposed by tuberculosis?
48.5.4.1 Overcoming Neglect 866
48.5.4.2 The HIV Pandemic 866 Several factors contribute to this seeming paradox.
48.5.4.3 Multiple Drug-Resistant Tuberculosis 867 The explosion of tuberculosis in sub-Saharan Africa,
48.5.4.4 The Private Health Care Sector 868 in the context of the recent HIV pandemic, has been
48.5.4.5 Population Movement 869 widely publicized. So, too, has the growing threat of
48.6 Recent Global Initiatives in TB Control 869
multi-drug-resistant tuberculosis in a variety of set-
48.6.1 Global Partnership to Stop TB 869
48.6.2 The Global Fund to Fight AIDS, Tuberculosis tings- in both 'developed' and 'developing' countries.
and Malaria 870 Both these phenomena have highlighted, and further
48.6.3 Global Alliance for TB Drug Development 870 eroded, the limited infrastructure available for tuber-
48.6.4 Global Drug Facility 870 culosis control in many areas of the world.
48.6.5 TB-HIV 870 The World Health Organization (WHO), the
48.6.6 DOTS-Plus and the Green Light Committee 871
48.7 Recent Developments in Tuberculosis Control: International Union Against Tuberculosis and Lung
Examples from the Field 871 Disease (IUATLD), and other organizations have
48.7.1 India 871 developed and promoted systematic approaches to
48.7.2 China 873 TB control - based on sound medical, epidemiologic,
48.7.3 South Africa 873
and public health evidence. However, successful
implementation hinges on sustained political will as
K. SCHWARTZMAN, MD, MPH, FRCPC
well as the necessary material resources.
Assistant Professor, Departments of Medicine and Epide-
miology and Biostatistics, McGill University, Respiratory
Tuberculosis control refers to the comprehensive
Epidemiology Unit, Montreal Chest Institute, Room K1.23, interventions directed toward individuals, communi-
3650 St. Urbain, Montreal, Quebec, Canada H2X 2P4 ties and populations, with the goal of reducing and

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
856
ultimately eliminating the morbidity and mortality
related to tuberculosis. Optimal medical treatment of
individuals with tuberculosis is a cornerstone of TB
control, but is not sufficient in and of itself.
While more money and greater political will must
be mobilized to fight TB, they alone are likewise
insufficient to contain and conquer it. Instead, funds
and political commitment are needed to secure and
maintain a systematic approach to TB control in
every corner of the globe. Such an approach must
of course reflect fundamental knowledge about the
pathogenesis and transmission of M. tuberculosis
- but it must also reflect an understanding of public
health in every sense of the phrase.
Hence effective TB control involves not only
prompt diagnosis and suitable prescriptions, but also
such considerations as:
i\Il optimal delivery of care and of medications,
suited to local conditions
• recognition of important beliefs and practices
surrounding illness
It accurate surveillance, which in turn permits:
fI early identification of changes in epidemiologic
parameters
" proactive management of complex human and
health systems issues
In this chapter, the current context of TB control
activities is briefly reviewed. Core tenets of TB control
are described. Key challenges faced by developing and
developed countries are summarized, as are several
recent strategies and initiatives. Recent developments
in several countries are highlighted as examples- with
the caveat that the complexity of the problem, and
hence the diversity of local conditions and programs,
is vast, and clearly not amenable to simplification.
48.2
Overview of Interventions
for Tuberculosis Control
Fundamental medical interventions in tuberculosis
control are best understood with reference to the
pathogenesis of the disease (Fig. 48.1).
48.2.1
Preventing Tuberculous Infection
The incidence and prevalence of tuberculous infection
in any community reflects the level of exposure to per-
K. Schwartzman
sons with contagious tuberculosis- that is, pulmonary
or laryngeal disease. Styblo described a linear relation-
ship between the incidence ofsmear-positive active TB
and the annual risk of tuberculous infection, based on
observations in a variety of countries. He estimated
that for every 1% annual risk of infection, the corre-
sponding increase in annual incidence of smear-posi-
tive disease was 49 cases per 100,000 population (95%
confidence interval 39-59 per 100,000) (Styblo 1985).
The longer that persons with contagious TB
remain untreated (or ineffectively treated), the more
persons they will infect. Hence prompt diagnosis and
effective treatment of pulmonary and laryngeal TB
represent the highest priority and most cost-effec-
tive TB control activities - just as they represent the
surest means to reduce individual morbidity and
mortality (Borgdorff et al. 2002).
Active case-finding refers to interventions which
identify individuals with active tuberculosis who
have not yet sought medical attention for TB-related
symptoms. The goal is to begin treatment earlier in the
course of the disease, thereby reducing the numbers
of infected contacts. On the other hand, passive case-
finding occurs when the diagnosis of active tubercu-
losis is made only after patients have sought help for
symptoms of disease. Nonetheless, efficient passive
case-finding remains an important goal- delayed and
missed diagnoses must be minimized, once patients
seek help for symptoms.
Behavioral, administrative, and technical inter-
ventions also reduce infection risk. Patients are urged
to cover their mouths when they cough. In resource-
rich areas, a number of interventions reduce risks to
health care personnel and other patients when expo-
sure is anticipated, e.g. respiratory isolation protocols
for TB suspects, high-performance masks, negative-
pressure ventilation, ultraviolet light fixtures. How-
ever, even in these settings the most substantial risk
reductions stem from prompt diagnosis and manage-
ment of infectious patients.
48.2.2
Preventing Progression and Reactivation
Interventions which prevent progression and reacti-
vation aim to reduce tuberculosis incidence among
persons who have already acquired tuberculous
infection. To the extent that these interventions suc-
ceed in reducing the incidence of active TB, they also
prevent subsequent transmission in the community.
Once infected with M. tuberculosis, some patients
progress rapidly to active tuberculosis (notably the
Tuberculosis Control in Developing and Developed Countries 857

I Active Tuberculosis I
I Case Finding (Passive or Active) I
I Effective Drug Treatment I
I Patient Behaviour (e.g. Cover Mouth) I
I Airborne Droplets I
I Progression or I
I Respiratory Isolation I
Reactivation
I Ventilation and Air Filtration I
I BCG Vaccination I I Ultraviolet Light I

I Treatment
Diagnosis and
of Latent TB
I
I
I Inhalation by Others I
I Antiretroviral Therapy I Fig. 48.1. Overview of tuberculosis control interventions.
forHIV
I Prompt diagnosis and suitable treatment of persons with
active TB are the most important and cost-effective measures
I Latent Infection I to reduce community TB incidence

HIV-infected), some reactivate years later, and most
remain with lifelong latent infection, if untreated.
The primary goal of BeG vaccination is to prevent
progression to active TB following infection in
infancy or early childhood. Antibiotic treatment for
latent infection ('preventive therapy', 'chemopro-
phylaxis') is used primarily in resource-rich areas,
to prevent progression to active TB among recently
infected contacts. It is also used to prevent reactiva-
tion, particularly among persons at elevated risk such
as the HIV-infected, and those with radiographic evi-
dence of extensive scarring related to presumed prior
active TB. In resource-rich countries, certain screen-
ing activities target persons with latent infection
- e.g. among new immigrants. However, treatment
of latent infection is of lower priority; it is thus often
neither feasible, nor appropriate, in low-income,
high-burden areas.
Just as HIV infection represents the strongest
known risk factor for reactivation of latent infection,
effective management of HIV disease substantially
reduces the likelihood of reactivation. This point
was demonstrated by Badri and colleagues in South
Africa, who attributed an 81 % reduction in reactiva-
tion risk to the use of highly active antiretroviral
therapy (Badri et al. 2002). However, even among
HIV-seropositive individuals who received antiretro-
viral therapy, the incidence of active TB remained 2.4
per 100 person-years, 8-10 times higher than among
HIV-uninfected persons in the same area. Nonethe-
less, HIV treatment, where feasible, is an increasingly
important aspect of tuberculosis control in areas
where the two epidemics overlap.
48.3
Cost-Effectiveness of Tuberculosis
Treatment
Systematic treatment of TB within national control
programs is one of the most cost-effective public
health interventions worldwide. Murray and col-
leagues evaluated the cost-effectiveness of short-
course therapy for smear-positive TB under program
conditions in Malawi, Mozambique, and Tanzania
(Murray et al.I991). The treatment regimen consisted
of two months of isoniazid, rifampin, pyrazinamide,
and streptomycin, followed by six months of isonia-
zid and ethambutol. The analysis reflected program
costs for treated patients. It did not include program
costs for management and diagnosis of TB suspects,
or costs borne by patients and family members.
Even with hospitalization during the intensive
phase, the estimated total program cost per year of
life saved was $1 to $3 US, suggesting that short-
course TB therapy was even more cost-effective than
immunization for measles and neonatal tetanus, oral
rehydration therapy for diarrhea, and blood-bank
screening for HIV - all of which were estimated to
cost $5-10 US per year of life saved. The same analy-
sis suggested that treatment of HIV-infected persons
858 K. Schwartzman

is nearly as cost-effective, since much of the ben-
efit relates to reductions in subsequent TB incidence
- among both HIV-seronegative and HIV-seroposi-
tive individuals -through interruption of community
transmission.
Even when costs of establishing the public health
control framework are considered, TB treatment
remains higWy cost-effective. Expansion of the Indian
national tuberculosis program to cover 43% of the
country's population cost $50 million US over 8 years,
and saved an estimated 200,000 lives, for an average
cost of $250 US per life saved. When indirect benefits
(avoiding lost earnings) are considered, this $50 mil-
lion investment generated estimated societal benefits
of over $400 million (Khatri and Frieden 2002). A
computer simulation model of TB cases and costs
in Thailand predicted that expansion of the Thai TB
control program to meet WHO benchmarks (70% case
detection, 85% cures) would likely save US $8 million
in health care costs, and $2.5 billion in lost productiv-
ity, over a 20-year period (Sawert et al. 1997).
48.4
The World Health Organization's DOTS
Strategy
ally districts with populations of 100,000 to 150,000
- and was designed to be integrated into public
health activities at the district level. It was successfully
adopted in Tanzania in the early 1970s, followed by
other low-income, high-incidence countries.
48.4.7.7
Government Commitment
Each element of the DOTS strategy carries a number
of implications. Government commitment to TB
control activities means recognition of the serious
burden that TB represents, as well as concrete steps to
control it. Hence it means not only stable funding, but
also political and administrative support - including
the establishment and maintenance of regional, and
national, tuberculosis control programs. These pro-
vide supervision as well as technical and logistical
support to district-level activities, and are respon-
sible for broader surveillance.
Political commitment also entails working pro-
ductively with international partners (including
donor agencies, and non-governmental organiza-
tions) to foster improvements in TB treatment and
control. In resource-rich countries, political commit-
ment also entails support for TB control in poorer
regions of the world.

48.4.1 48.4.7.2
Description Standard Diagnostic Approach

DOTS is the 'brand name' for the tuberculosis control
strategy promoted by the World Health Organiza-
tion since the early 1990s. The acronym stands for
Directly Observed Therapy, Short-Course. In fact, it is
a comprehensive management approach that goes far
beyond the mechanics of administering short-course
TB therapy. The key elements of the DOTS strategy
are outlined in Table 48.1.
The DOTS strategy reflects the approach to TB con-
trol pioneered by Karel Styblo and the IUATLD. The
strategy takes advantage of existing health units - usu-
The DOTS strategy emphasizes the use of sputum
microscopy for diagnosis of active pulmonary tuber-
culosis, among persons with compatible symptoms
('TB suspects'). This means that individuals with
the most advanced and highly contagious disease
- smear-positive pulmonary TB -are diagnosed
(and then treated). While persons with smear-nega-
tive pulmonary TB can and do transmit infection,
they infect about 20% as many others as do those
with smear-positive disease (Behr et al. 1999a). Hence
they represent a lower priority for detection, but will

Table 48.1. The WHO DOTS strategy (Maher and Mikulencak 1999)

Government commitment to sustained TB control activities

Case detection by smear microscopy among symptomatic patients self-reporting
to health services
Standardized treatment regimen of six to eight months for at least all confirmed
sputum smear positive cases, with directly observed treatment (DOT) for at least
the initial two months, and direct observation whenever rifampin is used
A regular, uninterrupted supply of all essential anti-TB drugs
A standardized recording and reporting system that allows assessment of treatment
results for each patient and of the TB control program overall
Tuberculosis Control in Developing and Developed Countries
subsequently be diagnosed if smear-positive, under
the DOTS strategy.
Mycobacterial culture and sensitivity testing are
not part of the DOTS strategy, as their widespread
use is beyond the means of most resource-poor
countries. The most cost-effective means to reduce
transmission, morbidity, and mortality is provided
by diagnosis and empirical, standardized treatment
of smear-positive cases. However, the success of this
approach may be compromised when multi-drug
resistance is highly prevalent.
Although chest radiography is a standard diag-
nostic and screening tool in many resource-rich
countries, the chest radiograph alone cannot make
or exclude the diagnosis of active pulmonary TB.
Limitations with respect to sensitivity, specificity,
and consistency of interpretation have been widely
reported, and chest radiography is not feasible in
many remote districts.
On the other hand, the DOTS strategy requires the
availability of dependable sputum microscopy at the
local level - meaning that microscopes, supplies, and
trained technicians must be available, with suitable
quality control and maintenance of proficiency. The
International Union Against Tuberculosis and Lung
Disease has published detailed guidelines for sputum
microscopy (Enarson et al. 2000).
Beyond the procedural aspects of sputum micros-
copy, the success of the DOTS strategy depends on
prompt diagnosis among patients who seek help for
TB-related symptoms. Hence providers must con-
sistently and rapidly refer 'TB suspects' for sputum
analysis, ensure that microscopy has in fact been com-
pleted, and check and act on the results (optimization
of passive case-finding). In addition, authorities must
recognize and address obstacles to prompt diagnosis
such as ignorance ofTB signs and symptoms (among
both providers and patients), financial barriers, and
limited accessibility of health facilities.
48.4.7.3
Supervised Provision of Standard Treatment
Regimens
Once the diagnosis of active TB has been made by
smear microscopy, successful treatment hinges on
rapid institution of a standardized patient manage-
ment strategy. At a minimum, the WHO recommends
that for new cases, the intensive phase of short-course
therapy (the first two months) involve direct obser-
vation of every drug dose - meaning that the health
worker watches the patient ingest every dose of
medication. Direct observation is emphasized during
859
this period, as adherence ensures rapid killing of M.
tuberculosis organisms and hence a rapid reduction
in the degree of contagiousness.
During the continuation phase of treatment for
new cases (the subsequent 4-6 months), therapy
may be observed or self-administered. However, the
WHO has urged maintenance of directly observed
treatment when the continuation phase includes
rifampin. This reflects the major adverse conse-
quences of rifampin resistance which is more likely
to develop with poor adherence to self-administered
rifampin-based regimens (Espinal et al. 2000). For the
same reasons, the WHO also recommends the use of
fixed-drug combination pills, which reduce the likeli-
hood that partial adherence will result in acquired
drug resistance. Short-course treatment regimens
are listed in Table 48.2.
The WHO recommends an empirical retreatment
regimen for treatment failures and relapses, and
situations where treatment has been interrupted
(Table 48.2). Patients who remain sputum smear-
positive after completion of a directly observed
re-treatment regimen are likely to harbor multi-
drug-resistant organisms, and should be referred for
treatment accordingly. Approaches to the control of
multi-drug-resistant TB are discussed later in this
chapter.
Just as the successful diagnosis of TB suspects
requires suitable equipment and technicians, the
successful treatment of TB patients hinges on the
consistent availability of effective medications and
capable health workers. These must be available to
every district, and in turn to every patient. Hence,
to ensure treatment completion, health workers may
need to travel to patients' homes and workplaces.
48.4.7.4
Documentation and Surveillance
Often overlooked is the importance of documenta-
tion and surveillance, which is the final key compo-
nent of the DOTS strategy. At the individual patient
level, this means:
• practitioners and laboratory personnel must
ensure that three sputum samples are obtained
and then analyzed for every TB suspect
health workers must document drug dispensing
and ingestion
providers must document results of repeat sputum
microscopy during the continuation phase of
treatment (i.e. assessment of sputum conversion)
;II providers must maintain follow-up and document
clinical outcomes
860 K. Schwartzman

Table 48.2. Standardized empiric treatment regimens. Modified from Maher et al. (2002)

Treatment Patients Standard treatment regimens

category
Initial phase (daily or Continuation phase
3 times per week)

New smear-positive pulmonary TB 2 EHRZ or 2 SHRZ 6 HE or 4 HR or 4 H3R3

New smear-negative pulmonary TB with extensive parenchymal
involvement
New cases of severe forms of extra-pulmonary TB.
n Re-treatment: Sputum smear-positive, with: 2 SHRZ then 1 HRZE 5 H3R3E3 or 5 HRE
Relapse
Treatment failure
Treatment after interruption.
III New smear-negative pulmonary TB (other than in Category I) 2 HRZ 6 HE or 4 HR or 4 H3R3
New, less severe forms of extra-pulmonary TB.
IV Chronic cases Refer to WHO guidelines for use of second-line
(still sputum-positive after supervised re-treatment) drugs in specialized centers

Drug abbreviations: streptomycin (S), isoniazid (H), rifampicin (R), pyrazinamide (Z) and ethambutol (E).
Each regimen consists of 2 phases. The numbers before a phase is the duration of that phase in months. A number in subscript
(e.g. 3) after a letter is the number of doses of that drug per week. If there is no longer a subscript after a letter, then the treatment
with that drug is daily. An alternative drug (or drugs) appears as a letter (or letters) in brackets

Table 48.3. Treatment outcomes for standardized reporting among new and re-treatment patients with smear-positive pulmo-
nary tuberculosis (Uplekar 2002)
Cure, defined as a patient who is sputum smear negative during the last month of treatment, and at least once previously
Treatment completion, defined as a patient who has completed treatment but does not meet criteria for cure or failure
Treatment failure, defined as a patient who is sputum smear positive at five months or later during treatment
Death
Default, defined as treatment interruption for two or more consecutive months
Transfer out, defined as a patient who has moved to another reporting unit and for whom treatment outcome is unknown
Treatment success is defined as the sum of patients cured plus those who completed treatment

At the district level, this means
• the microbiology laboratory must keep a log of all
microscopy results, with positive smears flagged
and responsible personnel (health worker, nurse,
physician) informed
• the district TB control worker must maintain a
registry of all patients diagnosed with TB
11 district authorities must make accurate noti-
through ongoing evaluation and operations research.
Key treatment outcomes are listed in Table 48.3, as
defined by the WHO.
The WHO and the International Union Against
Tuberculosis and Lung Disease have developed stan-
dardized materials for documentation and surveil-
lance (Enarson et al. 2000).

fication (count) data made available to central

authorities, including categories of reported cases 48.4.2
(e.g. new versus re-treatment) Status of the DOTS Strategy
<I!I district authorities must ensure that treatment

completion and outcome are ascertained and
documented for every patient in the registry
;ll district authorities must conduct cohort analyses
of key parameters and outcomes by time period,
to identify problems and trends
Accurate documentation and surveillance permit
reliable incidence estimates, for smear-positive dis-
ease. Moreover, they are essential if health authorities
are to monitor and improve program performance,
The first step in the DOTS strategy is the establish-
ment of a centrally administered National Tubercu-
losis Program (NTP) in every country. This unit is
then responsible for development and implementa-
tion of a systematic national program of TB control
- a program which functions in, and is adapted to,
every district, and which reflects the key elements
of the DOTS package. Following implementation,
the WHO has set two key benchmarks (Uplekar
2002):
Tuberculosis Control in Developing and Developed Countries
1) In each country, at least 70% of estimated smear-
positive cases of pulmonary TB should be detected
within the DOTS program, meaning diagnosed
and reported by practitioners who work within
the DOTS framework.
2) In each country, at least 85% of smear-positive
cases should successfully complete short-course
treatment, meaning the proportions of definite
cures plus treatment completers. The target is
higher (95%) for resource-rich countries.
During 2001, the WHO surveyed 210 countries
with respect to TB control strategies and notifica-
tions in 2000 (Blanc et al. 2002). By the end of 2000,
148 countries had implemented or were implement-
ing the DOTS strategy (up from lOin 1990), and 55%
of the world's population lived in areas covered by
DOTS. However, the 1.02 million smear-positive cases
detected under DOTS represented just over a quarter
of the estimated total number. DOTS programs
successfully treated 80% of all new smear-positive
patients registered in 1999, but this corresponded to
only 19% of all estimated smear-positive cases that
year. Of the 22 highest-burden countries, only Viet-
nam had reached the targets for 70% case detection
and 85% successful treatment.
The target date for global attainment of these
benchmarks was originally set for 2000, and then
revised to 2005. However, at current rates of expan-
sion of the DOTS program, these targets will only be
reached in 2013. Hence the WHO and other organiza-
tions have urged substantial acceleration in the pace
of DOTS expansion.
48.4.3
Controversies
While there is no doubt as to the importance and
benefit of systematic TB control programs world-
wide, some elements of the DOTS strategy have
been questioned. In particular, some authors have
suggested that the direct observation of every treat- .
ment dose may not be necessary in all cases or set-
tings. Community-based directly observed treatment
after a brief hospitalization is certainly cheaper than
hospitalization throughout the two-month intensive
phase of treatment (Floyd et al. 1997; Murray et al.
1991). In Hlabisa Health District, South Africa, the
estimated total cost of community-based DOT was
one-third that of hospital-based treatment ($741 vs.
$2048 US per patient) (Floyd et al.1997). Community-
based DOT was at least as effective as hospital-based
861
treatment, and hence substantially more cost-effec-
tive ($891 vs. $2096-3700 per patient cured).
However, in some treatment settings the value
added by direct observation of medication ingestion
is less certain, as compared with self-administered
treatment within a well-structured program of treat-
ment and diagnosis. A randomized, controlled trial
in Pakistan compared three treatment strategies for
newly diagnosed sputum smear-positive TB (Walley
et al. 2002). Treatment involved two months of iso-
niazid, rifampin, pyrazinamide, and ethambutol, fol-
lowed by an additional six months of isoniazid and
ethambutol. A total of 497 adults were randomized
to one of 1) treatment directly observed by a health
worker; 2) treatment directly observed by a family
member, who collected medication from the health
center and recorded all doses; or 3) self-administered
treatment, with the patient required to collect his/her
medication from the health center every two weeks.
Outcomes were very similar in the three treatment
arms: cure rates were 64%, 55%, and 62%, and rates
of successful treatment completion were 67%, 62%,
and 65% respectively.
Hence in the setting of a randomized trial, within
a TB control program that incorporated the other ele-
ments of the DOTS strategy, direct treatment obser-
vation itself did not contribute to successful treat-
ment outcomes. However, in other program settings
it may be difficult to separate treatment observation
from other aspects of the DOTS package, particularly
when there are substantial barriers to patient (or
provider) adherence. It is also noteworthy that in all
three arms of the Pakistan study, treatment success
rates were substantially below the 85% benchmark.
Finally, the ultimate costs and impact of patient non-
adherence, when new rifampin monoresistance is the
result, cannot be captured by this type of study.
Volmink and colleagues identified several co-inter-
ventions that frequently accompany directly observed
therapy, but have not been explicitly labeled as part of
the 'official' WHO DOTS strategy (Volmink et al. 2000).
Examples include provision of incentives or enablers,
patient-centered approaches to treatment and care,
tracing of defaulters (and potential legal action), spe-
cific measures to promote motivation and supervision
of health workers, strengthening of other program
components, and additional financial support.
These additional interventions contribute to high
treatment success rates, beyond direct observation
alone. Although direct treatment observation itself is
a relatively cheap intervention on a per-case basis, the
large number of patients involved makes it a major
source of expense. Clearly, additional operational
862
research is urgently required to identify the most
cost-effective strategies in various resource-poor set-
tings - that is, the strategies which provide the high-
est cure rates at the lowest possible cost.
48.5
Other T8 Control Interventions
48.5.1
Active Case-Finding in Resource-Poor Areas
In resource-poor areas, the highest priority remains
efficient and effective passive case-finding, case-hold-
ing, and treatment, for persons with smear-positive
disease. However, simple symptom screens among
particularly high-risk groups can identify large
numbers of patients with previously undiagnosed
TB. Important targets for active case-finding are close
contacts of smear-positive patients, and HIV-infected
persons - particularly those attending counseling
and testing centers. Symptoms of tuberculosis may
in fact prompt such visits for HIV-related care.
For example, a symptom screen among persons
newly diagnosed with HIV infection at volun-
tary HIV counseling and testing (VCT) centers in
Zambia led to the diagnosis of TB among 10% of
those screened; 42% of these TB-HIV patients were
sputum smear-positive (Ginwalla et al. 2002). Hence
VCT centers represent an important opportunity for
active case finding.
48.5.2
Screening Interventions in Resource-Rich Areas
In resource-rich countries, the goals of tuberculosis
screening interventions are often broadened to include
not only active case-finding for current active TB, but
also the identification of persons with latent infection
with an elevated risk of subsequent active TB. This is
only appropriate to the extent that individuals with
latent infection will receive effective treatment. At the
same time, the identification and treatment of latent
infection must not detract from the management of
persons with active TB, which is of a much higher pri-
ority. Hence emphasis on treatment of latent infection
is only appropriate within the framework of a coherent
TB control program once successful program manage-
ment of active TB has been secured and documented.
As in resource-poor areas, the highest priority
groups for screening activities are close contacts of
K. Schwartzman
contagious patients, and HIV-infected persons. This
reflects the relatively high prevalence of active disease
at screening, but also the high risk of subsequent active
disease among persons who are newly infected with M.
tuberculosis, and among those co-infected with HIY.
Among persons with dual HIV-TB infection the annual
risk of active disease is 5-10% or more (Antonucci et al.
1995; Selwyn et al.1989).Among persons newly infected
with M. tuberculosis, the annual risk of active disease is
approximately 1-2% during the first 2-3 years after
infection (Canadian Lung Association 2000).
Following diagnosis of an infectious TB patient,
the risks of contagion must be assessed, based on
clinical and demographic factors, and the patient's
physical and social environment. Contact investiga-
tion follows the 'concentric circle' approach. House-
hold members and/or others with the most extensive
contact with the index patient - and hence the high-
est risk of new infection -are first evaluated. If there
appears to be an excess prevalence of latent infection
(compared to estimated background probability for
the group in question), or if there are secondary
cases of active TB, then the contact investigation is
extended to workplace, school, and more casual con-
tacts (Menzies et al. 1999).
Although TB contact investigation is considered a
high-priority public health activity in high-income
countries, it can be challenging to execute. Patients
may refuse to identify contacts, and contacts may in
turn be difficult to locate. In a U.S. multi-center study,
13% of patients with active TB yielded no identified
contacts, while 39% of identified contacts failed to
complete the screening process (Reichler et al. 2002).
Depending on local epidemiology, other high-risk
groups may be identified for active case-finding or
for screening for latent TB. Examples include prison-
ers, residents of long-term care facilities, homeless
persons, and aboriginal populations.
While clinical evaluation and chest radiography
are the primary screening tools for active TB, the
diagnosis of latent infection hinges on the tuber-
culin skin test. The limitations of the tuberculin
test are well known (see Chapter 62). They include
inconsistent performance and measurement, limited
sensitivity (notably among the HIV-infected) and
poor specificity (notably where there is substantial
confounding by BCG vaccination after infancy, and/
or by environmental exposure to non-tuberculous
mycobacteria).
The tuberculin skin test does not distinguish per-
sons at high risk of subsequent active TB from those
whose risks are decidedly modest. Hence screening
and treatment activities based primarily on mass
Tuberculosis Control in Developing and Developed Countries
tuberculin skin testing provide very little benefit in
terms of prevention of active disease, unless groups
at higher risk are specifically targeted. Recent changes
in u.s. and Canadian recommendations for identifica-
tion and management of latent tuberculous infection
emphasize the targeted selection of persons for tuber-
culin testing based on priorities and medical risk, so
that only persons at high priority for treatment are
tested (American Thoracic Society 2000; Canadian
Lung Association 2000). "A decision to test is a decision
to treat" (American Thoracic Society 2000).
48.5.2.1
Screening of Migrants to Low-Incidence Countries
The foreign-born account for an increasing propor-
tion of tuberculosis incidence in the United States,
Canada, and Western Europe. For example, in Mon-
treal, Canada 80% of reported cases of active TB now
occur among the foreign-born, while for Canada as
a whole the figure was 65% for the year 2000 (Nault
et al. 2002; Rivest et al. 1998). In the United States
50% ofTB cases in 2001 occurred among the foreign-
born (compared to 27% in 1991), while in the United
Kingdom the proportion now slightly exceeds 50%
(Division of Tuberculosis Elimination 2002; Public
Health Laboratory Service 2002).
Fifty percent, or more, of the excess incidence
among the foreign-born occurs during the first
five years after arrival in the destination country,
although TB incidence remains higher than that
of the native-born population for at least 20 years
(Zuber et al. 1997; McCarthy 1984). Figure 48.2 dem-
onstrates the preponderance of cases within the first
years after arrival, in a group of Asian immigrants to
the United Kingdom (McCarthy 1984).
40,--------------------------,
35
30
25
Fig. 48.2. Tuberculosis cases among
Asian immigrants to West Ham
20
(London), 1976-80. Data are from
McCarthy (1984). McCarthy demon-
15
strated that most cases which occurred
over 10 years after immigration
10
occurred shortly after return visits to
Asia - suggesting that patients became
infected during visits to their coun-
tries of origin. The data shown are for
patients without known TB contacts in 3 4
the United Kingdom
863
Indeed, in many immigrant subgroups the inci-
dence initially mirrors that in their countries of origin
(Rivest et al. 1998; Zuber et al. 1997). British authori-
ties estimated the incidence among United Kingdom
residents born in sub-Saharan Africa to be 230 per
100,000 - nearly 20 times the overall national inci-
dence (Public Health Laboratory Service 2002). Hence,
the United States Institute of Medicine and other
national authorities within low-incidence countries
have identified the foreign-born as a priority group
for screening and prevention activities (United States
Institute of Medicine: Committee on the Elimination
ofTuberculosis in the United States 2000).
Prospective migrants applying for immigrant or
refugee status must undergo medical evaluation.
For example, persons applying from abroad for resi-
dence in the United States or Canada must undergo
a screening chest X-ray (Binkin et al. 1996; Dasgupta
et al. 2000). Screened persons are referred for supple-
mentary microbiological testing (sputum analysis)
where the clinical and/or radiographic evaluation
indicates the possibility of active TB.
Immigrant applicants diagnosed with smear-posi-
tive TB must document smear conversion before they
can enter the US, while applicants to Canada must
complete treatment of active TB prior to arrival.
Refugee claimants who arrive with active disease
must complete treatment before they can be accepted
as refugees.
In addition, prospective migrants who are found
to have inactive TB, meaning stable radiographic
abnormalities and latent infection, are required to
undergo further medical evaluation and surveil-
lance in the destination country. The goals are 1)
to detect additional cases of active TB, which were
either missed or not yet evident at the time of the
• Cases in Persons Who Did Not Subsequently Visit Asia
(128 Cases Total)
o Cases in Persons Who Subsequently Visited Asia (59
Cases Total)
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Over
Years Sinee Arrival 20
864
earlier screen, and 2) the provision of treatment of
latent infection to those without active TB, since both
recent migration and radiographic abnormalities
place them at increased risk of reactivation.
Formal evaluations have questioned the effective-
ness and yield of these programs, which are extremely
complex to administer, and require a high degree
of coordination between far-flung providers and
health authorities. In Montreal, 17 (0.13%) of 12,898
screened immigrant and refugee applicants screened
during 1996-7 had active TB, for an estimated total
program cost of $31,418 Canadian ($20,422 US) per
case detected and treated (Dasgupta et al. 2000).
During the same period, of 644 new permanent resi-
dents in Montreal who were referred for surveillance
of inactive TB,401 (64%) reported for evaluation and
four had active TB. The estimated total surveillance
program cost was $55,728 Canadian ($36,223 US) per
case detected and treated. Based on the treatment
actually administered to these cohorts for latent
TB infection, the immigrant screening and surveil-
lance programs were even more costly per future
active case prevented ($73,000-$155,000 Canadian)
(Dasgupta et al. 2000).
Among 5,739 foreign-born persons screened for
permanent residence (1997-98) after arrival in five US
jurisdictions, no cases of active TB were found (Saraiya
et al. 2002). In Seattle, Washington the prevalence of
active TB among new immigrants referred for surveil-
lance (1992-94) of inactive TB identified by overseas
screening was similar to that in Montreal- 1.1 %
(Wells et al. 1997). However, among immigrants whose
screening chest X-rays suggested active TB, but whose
sputum smears were negative at overseas screening,
the prevalence of active disease at re-examination in
Seattle was 10.5%. In this way, programs that success-
fully target and follow the highest-risk migrants are
likely to be relatively cost-effective, with respect to the
detection and treatment of active or latent TB.
Undocumented migrants are obviously missed by
these screening programs. In areas where migrant
workers and other undocumented foreign nationals
are concentrated, these groups are a major concern
with respect to TB incidence, morbidity, and trans-
mission. This reflects not only the risk ofTB itself, but
also barriers to health care after symptoms develop,
which prolong the time for transmission. Key factors
include migrants' fear of arrest and deportation, and
limited access to suitable care and follow up.
Short-term visitors are also missed, as are
residents of low-incidence countries who travel to
high-incidence areas. Lobato documented strong
associations between latent tuberculous infection
K. Schwartzman
in US-born children in California and a history
of travel to and/or household visitor from higher-
incidence countries, notably Mexico - illustrating
the potential influence of short-term visits on TB
transmission and incidence (Lobato and Hopewell
1998). McCarthy demonstrated a rebound in TB
occurrence among Asian immigrants in London who
subsequently visited their home countries, as shown
in Fig. 48.2 (McCarthy 1984). Cobelens estimated an
infection rate of 3.5 per 100-person-months among
Dutch-born visitors to endemic areas, based on serial
tuberculin testing (Cobelens et al. 2000). This repre-
sents an annual risk of infection of 4.2%, which in
fact equals or exceeds estimated risks for inhabitants
of high-incidence countries.
48.5.2.2
Screening and Control in Specialized Settings
In resource-rich areas, TB control authorities and
providers have emphasized control of transmis-
sion in hospitals, prisons, homeless shelters, and
other institutions where persons with TB are likely
to be housed. The approach involves a 'hierarchy of
controls' - the most important of which relates to
provider behavior. Persons with symptoms, signs, or
radiographic features suggestive of active TB must
promptly undergo suitable diagnostic testing, and
treatment where indicated.
Delays in diagnosis of hospitalized patients lead to
increased mortality and promote nosocomial trans-
mission to other patients as well as staff and visitors.
The dramatic US hospital TB outbreaks reported in
the early 1990s- as well as subsequent systematic
surveys - highlighted the importance of prompt
diagnosis, and reinforced the need for immediate
treatment as well as suitable isolation of highly infec-
tious patients.
In Atlanta, aggressive implementation of hospital
control measures reduced the potential exposure of
staff and patients by nearly 90%; the frequency of
skin test conversions among workers decreased to
the same degree (Blumberg et al. 1995). Unfortu-
nately, broad application of these measures is costly,
and many patients are subjected to unnecessary
isolation. However, simple clinical prediction rules
are of limited benefit in reducing unnecessary use of
isolation facilities (Bock et al.1996).
Hospitalized patients who are smear-positive
should remain in respiratory isolation, in hospital,
until there is clear evidence of bacteriological and
clinical response, and appropriate public health
interventions (e.g. contact investigation, provisions
Tuberculosis Control in Developing and Developed Countries
for follow-up) are underway. Repeat sputum smears
offer the best means to assess microbiologic response.
There is rapid mycobacterial kill (and hence reduc-
tion in contagiousness) with standard intensive-phase
drug therapy. However, the increasing prevalence of
drug resistance limits the validity of empiric discharge
recommendations for smear-positive patients (e.g. the
widely held belief that patients are no longer infec-
tious after two weeks of therapy). Two weeks is likely
sufficient for initially smear-negative patients, when
there is a suitable clinical response.
Nonetheless, patients with presumed, or proven,
active pulmonary TB should remain isolated for as
long as they are hospitalized, as there remains some
potential for transmission - which is of particular
concern given the frequent presence of persons
with HIV, or other immunosuppressive conditions,
on general medical and respiratory wards. Provid-
ers must also consider the patients' home and social
environments when planning discharge, e.g. presence
of young children in the home, homelessness, etc. For
a more detailed review of nosocomial tuberculosis
control consult Schwartzman and Menzies 1999.
Within hospitals,national TB control authorities have
suggested engineering interventions to reduce the risk
of nosocomial transmission. Hence, US and Canadian
authorities recommend the use of negative pressure
rooms, with a minimum of six air changes per hour, to
house infectious TB patients or suspects. Higher rates
of negative pressure ventilation (15 air changes per
hour) are recommended for high-risk procedure areas,
including facilities for cough-inducing procedures (e.g.
bronchoscopy, sputum induction rooms), as well as
autopsy rooms and clinical mycobacteriology labora-
tories (Canadian Lung Association 2000).
Another important technical intervention is the
routine use of high-efficiency masks (which filter at
least 95% of particles in the respirable range) by per-
sonnel in contact with potentially infectious patients,
as well as the patients themselves. Personal respira-
tors afford additional protection in the highest risk
situations, e.g. airway procedures in patients with
suspected or known multi-drug-resistant TB (Fen-
nelly and Nardell 1998). Further reductions in air-
borne mycobacteria can be achieved by filtration of
supply air with high-efficiency particulate air filters
(HEPA filters), and ceiling ultraviolet light fixtures
(Fennelly and Nardelll998; Ko et al. 2001; Riley et al.
1962). Indeed, ultraviolet fixtures have been installed
in some bronchoscopy suites and homeless shelters
in North America.
Outside the scope of large-scale public health
programs, tuberculosis screening is also conducted
865
in specialized occupational and clinical settings, in
resource-rich areas. Health care workers are generally
screened for latent infection upon hiring, and may be
regularly reevaluated for skin test conversion if they
are in close or frequent contact with tuberculosis.
Patients who will be heavily immunosuppressed for
long periods, such as those for whom a solid-organ
transplant is anticipated, represent an important
target for screening and treatment of latent infec-
tion - because of the potential for death from dis-
seminated TB. Another group at risk is patients
undergoing long-term dialysis - because of both the
immunosuppression associated with end-stage renal
disease, and the potential for spread within hospital
dialysis units.
48.5.3
Bacille Calmette-Guerin (BCG) Vaccination
The use of BCG vaccination has been reviewed else-
where (see Chapter 49). A rigorous meta-analysis
estimated an overall protective efficacy of 50-51%
with respect to the development of TB, but individual
studies have shown tremendous variation- from no
effectto 80% protection (Colditz et al.1994). Potential
reasons for this variation include changes in the M.
bovis BCG strains, differences in tuberculosis epide-
miology and ascertainment, and host differences.
Rieder concluded that there was strong evidence
to support the use of BCG vaccination in newborns
and infants in high-incidence areas, because of asso-
ciated reductions in disseminated and meningeal TB
in childhood, as well as reductions in childhood TB
mortality (Rieder 2002). The vaccine, therefore, con-
tinues to be part of the WHO's Expanded Programme
on Immunization, with estimated global coverage of
86% among infants in 2000 (Vaccine Assessment and
Monitoring 2002;World Health Organization 2002).
The vaccine continues to be widely used in most
areas of the world, although the estimated infant cov-
erage rate was lower in sub-Saharan Africa (67% in
2000) than in other regions (Vaccine Assessment and
Monitoring 2002). The WHO does not recommend
repeat administration, or administration to older
children or adults.
As infants are not routinely HIV-tested in most
high-incidence areas, the WHO recommends admin-
istration to all except those with clinical evidence
of immune suppression. However, in the United
Kingdom the use of BCG vaccination is discouraged
among HIV-positive children as well as those with
other forms of immunosuppression (e.g. hemato-
866
logic malignancies), because of the possibility of dis-
seminated BCG disease (Department of Health and
the Health Education Authority 1998).
In North America and Western Europe, the use of
BCG vaccination is variable. It is no longer routinely
administered in the United States or Canada. Cana-
dian authorities recommend its use among aborigi-
nal infants, particularly those living on reserves, and
newborns born to infectious mothers (Canadian
Lung Association 2000). In both countries, it may also
be considered for healthcare workers with a high risk
of repeated exposure to multi-drug-resistant TB.
The rationale for discontinuation of BCG vaccina-
tion relates to the low overall incidence of TB in these
countries, particularly among infants and young
children. Moreover, BCG vaccination administered
after the of age two years confounds the subsequent
interpretation of tuberculin skin tests (Menzies and
Vissandjee 1992). Detection and treatment of latent
infection in higher-risk groups represents an impor-
tant TB control strategy for these countries, so opti-
mal interpretation of tuberculin tests is essential.
In the United Kingdom, TB control authorities
have recommended BCG vaccination: 1) for new-
borns and infants born to parents from high-inci-
dence countries; 2) for tuberculin-negative school-
children at ages 10-14. The U.K. authorities continue
to monitor and re-evaluate the need for vaccination
among school children (Department of Health and
the Health Education Authority 1998).
Development of an improved tuberculosis vac-
cine is a high priority for researchers and TB control
authorities. Of particular interest are genotypic dif-
ferences in BCG vaccines which correlate with vary-
ing reports of efficacy (Behr et al. 1999b). Knowledge
of the key bacterial components responsible for
efficacy will permit refined design of future vaccines
(Mostowy and Behr 2002).
48.5.4
Challenges to Global TB Control
48.5.4.7
Overcoming Neglect
In many areas of the world, TB has been primarily a
disease of the poor and the powerless. In resource-
rich areas, it has increasingly been concentrated
among the foreign-born and among poor urban
residents. Hence TB has not been an important
part of the political agenda. Control activities have
frequently not been emphasized, or accorded stable
K. Schwartzman
human and financial resources. The impact of neglect
was felt acutely in the United States during the early
1990s. Many resource-poor countries lack the mate-
rial resources needed to effectively control TB on
their own, and must rely on foreign aid and other
partners. However, until recently the resource-rich
countries have provided little support for TB control
activities elsewhere.
A major limiting factor has been the cost of TB
medications in populous, resource-poor countries;
this has been the impetus for recent initiatives such
as the Global Drug Facility (discussed below). The
training, supervision, and payment of community
health personnel also represents a substantial cost
to overburdened health ministries where there are
many competing agendas - again underlining the
importance of sustained government commitment
to TB control.
48.5.4.2
The HIV Pandemic
HIV co-infection is the strongest known risk factor
for TB reactivation, with up to a 100-fold increase in
risk. HIV-infected persons who subsequently acquire
tuberculous infection are also at high risk for rapid
progression to active disease. Of an estimated 36.1
million persons living with HIV/AIDS worldwide at
the end of 2000, 25.3 million were believed to reside
in sub-Saharan Africa and 5.8 million in Southeast
Asia - both extremely high-burden areas for TB
(UNAIDS 2002). In addition, there is grave concern
about rapid acceleration of the HIV pandemic in
China and India.
The HIV epidemic has been linked to rapid
increases in TB incidence in countries where the TB
control programs were previously highly successful
- notably in sub-Saharan Africa. In many sites within
that region, 30-40% of persons newly diagnosed
with smear-positive pulmonary TB have been HIV-
seropositive; in some locations the proportion has
reached 70% (Maher et al. 2002). It is also noteworthy
that 'mini-epidemics' of TB reported in the United
States in the early 1990s were frequently associated
with HIV infection (Alland et al. 1994; Coronado et
al. 1993; Daley et al. 1992; Fischl et al. 1992).
The increasing prevalence of HIV infection has a
number of implications for TB control. Most obvi-
ously, the high risks of both progressive post-primary
disease and reactivation have greatly increased the
incidence of TB in many areas - often overwhelming
existing treatment and control resources. In turn, this
perpetuates the transmission of new infection and
Tuberculosis Control in Developing and Developed Countries
the development of subsequent active TB, among
both HIV-seropositive and HIV-seronegative indi-
viduals in these communities.
TB can be difficult to diagnose in the context of
advanced HIV disease, with nonspecific symptoms
and a higher frequency of smear-negative disease,
extra-pulmonary disease, and atypical radiographic
presentations. Hence providers may not obtain
suitable material for diagnosis as promptly as in
HIV-seronegative persons with typical respiratory
symptoms - potentially prolonging community
exposure to contagious cases. Moreover, a diagnos-
tic strategy based largely on sputum smears cannot
directly address the burden of morbidity, mortality,
and transmission attributable to smear-negative
tuberculosis among the HIV-infected.
On the other hand, the prevalence of previously
undiagnosed active TB among persons seeking vol-
untary HIV counseling and testing can be surpris-
ingly high, as discussed earlier.
Another implication of the association between
HIV and active TB is the risk of outbreaks resulting
from transmission within healthcare facilities. Fol-
lowing dramatic reports of TB outbreaks in US facili-
ties, primarily involving HIV-infected patients, infec-
tion control activities were reinforced - including
aggressive protocols for hospital isolation of patients
with respiratory symptoms. In resource-poor areas,
active TB case-finding at HIV care centers represents
an important step in reducing ongoing transmission.
Treatment of latent tuberculosis infection was not
previously part of the standard approach to TB con-
trol in high-burden areas, because of the high c6st
and low yield relative to diagnosis and management
of smear-positive cases. However, with progression
of the HIV pandemic, treatment of latent TB among
HIV-seropositive persons is now being advanced as
a potential control strategy in resource-poor areas
where dual HIV-TB infection is highly prevalent
(Creese et al. 2002; Maher et al. 2002).
In a Brazilian cohort study among HIV patients,
treatment for latent TB was associated not only with
the expected reduction in incidence of active TB, but
also with a 76% reduction in the risk of death (Pinho
et al. 2001). In resource-rich areas, HIV-infected
persons with latent TB represent the highest priority
group for treatment with isoniazid or alternatives,
given the estimated annual reactivation risks of
5-10% (American Thoracic Society 2000; Antonucci
et al. 1995; Canadian Lung Association 2000; Selwyn
et al. 1989).
In North America, the standard regimen for latent
tuberculosis is a nine-month course of isoniazid.
867
However, the optimal treatment regimen for latent
TB with concomitant HIV infection remains uncer-
tain in high-incidence areas. Among HIV-infected
persons in Uganda, the protective efficacy of a six-
month isoniazid course began to wane one year after
treatment completion (Johnson et al. 2001). After
three years, the cumulative incidence of active TB
was equal to that among placebo-treated subjects.
This is very different from earlier findings of twenty
years' benefit among Alaskan aboriginals in the pre-
HIV era (Comstock et al. 1979). The discrepancy
could relate to subsequent reactivation, among those
individuals in whom six months of isoniazid mono-
therapy is insufficient to eradicate latent infection. In
the Ugandan study, the protective effect of combina-
tion treatment for latent infection (three months of
isoniazid!rifampin or three months of isoniazid!
rifampin!pyrazinamide) lasted at least three years
(Johnson et al. 2001).
Another potential explanation is TB re-infection
- in communities with extremely high TB incidence,
individuals may simply become exposed and re-
infected after they have completed treatment for the
original latent infection. In a group of South African
gold miners who completed treatment for active TB,
Sonnenberg and colleagues demonstrated that over
one third of episodes of recurrent TB resulted from
reinfection with distinct strains; virtually all those
with re-infection were HIV-coinfected (Sonnenberg
et al. 2001).
The HIV pandemic has also exacted a consider-
able toll on healthcare workers themselves - notably
in sub-Saharan Africa. Not only does HIV increase
the risk of active TB among those caring for TB
patients - it also decimates the ranks of TB control
staff and community workers in many areas where
they are most urgently needed.
48.5.4.3
Multiple Drug-Resistant Tuberculosis
Drug resistant tuberculosis;' develops as a result of
improper treatment - inadequate prescription, and!
or inadequate adherence to therapy. Isoniazid resis-
tance is the most common resistance phenotype of
clinical concern. The median estimated prevalence
of isoniazid monoresistance among new cases of
tuberculosis at 58 sites surveyed by the WHO and
IUATLD during 1996-99 was 3.0%, but ranged from
zero in New Caledonia (Oceania) to 28.1% in Latvia
(Espinal et al. 2001).
Although isoniazid resistance (without rifampin
resistance) was associated with a slightly lower prob-
868 K. Schwartzman

ability of treatment success, 82% of patients with three months, followed by isoniazid, rifampin, and
isoniazid resistance were successfully treated for an ethambutol for an additional five months. Moreover,
initial episode of TB using a standard empiric WHO 9% of new cases and 11 % of re-treatment cases with
regimen (isoniazid, rifampin, pyrazinamide, plus MDR-TB died, as compared with 2% of new cases and
ethambutol or streptomycin for two months; iso- 3% of re-treatment cases with fully susceptible TB.
niazid and rifampin for an additional four months). The increasing prevalence of multi-drug-resistant
These patients were treated in South Korea, Peru, M. tuberculosis strains poses a serious threat to indi-
Hong Kong, Russia (Ivanovo Oblast), the Dominican vidual health, in terms of treatment failure and mor-
Republic, and Italy. In the same study, the probability tality. It also represents a dire threat to tuberculosis
of treatment success was 85% among patients with control, since standard (and relatively cheap) treat-
a first episode of fully drug-susceptible tuberculosis ment regimens are frequently ineffective, and may
(Espinal et al. 2000). In areas where drug susceptibil- even foster additional drug resistance. For example,
ity testing is routinely available, continuation phase patients with MDR-TB who receive the empirical
therapy is lengthened and modified to account for re-treatment regimen effectively receive ethambutol
isoniazid monoresistance. monotherapy during the continuation phase, which
Of much greater concern is the increasing incidence promotes the development of resistance to ethambu-
of multi-drug resistant TB - defined as TB caused by tol. Hence MDR isolates of M. tuberculosis often dem-
M. tuberculosis organisms that are resistant to both onstrate resistance to all first-line drugs (in addition
isoniazid and rifampin, with or without additional to isoniazid and rifampin) (Frieden et al. 1996; Moss
drug resistance. In the 1996-99 WHO-IUATLD survey, et al. 1997).
the median estimated prevalence of multi-drug-resis- The presence of large numbers of patients with
tant TB was 1.0% among new cases and 9.3% among ineffectively treated MDR-TB promotes further
re-treatment cases (Espinal et al. 2001). However, spread -- a fact demonstrated dramatically in prisons
among new cases the estimated prevalence of multi- of the former Soviet Union (Kimerling et al. 1999).
drug-resistant TB exceeded 10% at two sites (Estonia Yet early identification of such individuals through
- 14.1 % and Henan Province, China - 10.8%), while it routine drug susceptibility testing is impossible for
ranged from 5-10% in 5 additional sites (Latvia, two most high-burden countries, as they simply do not
Russian sites, Iran and Israel) (Espinal et al. 2001). The have the material or financial resources for suscep-
high burden of multi-drug-resistant TB among new tibility testing. Moreover, the cost of effective treat-
patients indicates extensive transmission of multi- ment for MDR-TB has until recently been prohibitive
drug-resistant organisms in these areas - notably - $10,000 US or more per patient, in high-income
within the former Soviet Union. countries (Gupta et al. 2001). In most areas, the treat-
Other countries where the estimated overall ment of MDR-TB with expensive second-line drugs
prevalence of multi-drug resistance exceeds 5% shifts limited public health funds away from other
include Dominican Republic, Ivory Coast, Afghani- basic elements of TB management and control.
stan, Armenia, Azerbaijan, Egypt, Georgia, Kyrgyz-
stan, Mongolia, Pakistan, Papua New Guinea, Sudan, 48.5.4.4
Syria, Tajikistan, Turkmenistan, Ukraine, Uzbekistan The Private Health Care Sector
and Yemen (Dye et al. 2002). The WHO regions with
the highest estimated prevalence of MDR-TB are the A key challenge for health authorities is the integration
Eastern Mediterranean region (7.9%) and the East- of private practitioners into a systematic TB control
ern European region (5.5%). The overall estimated framework - meaning providers who work outside the
prevalence worldwide was 3.2%. government-run public health system. For example, in
US reports of MDR-TB outbreaks highlighted the India 50% of TB patients received some, or all, of their
high risks of treatment failure and mortality among treatment in the private sector (Pathania et al. 1997).
affected individuals, who were frequently HIV-sero- Among patients with TB symptoms in the Philippines,
positive (Frieden et al.1996; Moss et al.I997). In Espi- 11.8% consulted a private practitioner, while 7.5% con-
nal's study of six countries, only 52% of new cases sulted a public health center and 4.4% sought attention
with MDR-TB were successfully treated with the six- at a hospital (Tupasi et al. 2000).
month regimen, while only 29% of re-treatment cases Potential problems with private sector care
with MDR-TB were successfully treated with the include missed or delayed diagnoses, non-standard
empiric re-treatment regimen - isoniazid, rifampin, or inadequate treatment regimens, unreliable drug
pyrazinamide, ethambutol, and streptomycin for availability and preparations, limited follow-up, and
Tuberculosis Control in Developing and Developed Countries
incomplete reporting. In a survey of staff members
at 147 private pharmacies in Ho Chi Minh City, Viet-
nam, only 18% of respondents listed TB as a possible
diagnosis for a fictitious case describing fever and
cough for four weeks. Respondents also estimated
that 24% of their customers who had purchased anti-
tuberculosis drugs during the previous four weeks
had done so without a prescription (Lonnroth et al.
2000).
Gaps in private sector care are not limited to lower
income countries. In a 1992 U.S. national physician
survey, 59% of respondents were able to list a recom-
mended tuberculosis treatment regimen, while 28%
described excessive regimens and 12% described
inadequate regimens. A total of 4% described using
a single drug to treat active tuberculosis. The vast
majority of respondents worked in the private sector
(Sumartojo et al. 1997).
Even when private practitioners can assure optimal
treatment with appropriate supervision (which is in
fact often lacking), it is extremely difficult to collect
systematic case reporting and outcome data in settings
where a large number ofsuch practitioners provide TB
care. Hence public health authorities in diverse loca-
tions have collaborated with private sector providers
in joint efforts to promote systematic diagnosis, treat-
ment, and reporting (Uplekar et al. 2001).
48.5.4.5
Population Movement
Movement of populations, both within and between
countries, represents yet another challenge to TB
control (MacPherson and Gushulak 2001). The model
promoted by the WHO and IUATLD emphasizes
diagnosis, treatment, and surveillance at the local
district level. Yet patients may move between districts
or regions temporarily (e.g. as part of agricultural
work), or permanently (in search of economic oppor-
tunities), so that responsibility for their care shifts
from one jurisdiction to another.
Larger population groups may be internally
displaced by famine, conflict, political or economic
instability, leading to congregation of large numbers
in refugee camps or shantytowns. These represent
formidable challenges for sustained health care
delivery, and also foster the spread of tuberculosis
and other communicable diseases.
Finally, movement of persons between countries
(refugees, immigrants, migrant workers) means
that tuberculosis affects every part of the world. In
many low-incidence countries, foreign-born persons
now account for the majority of TB incidence. The
869
increased emphasis on screening of new arrivals has
been reviewed earlier in this chapter. In addition,
the problem of tuberculosis among foreign-born
persons creates many logistical challenges for health
authorities, including:
~ the need to communicate effectively in a variety
of linguistic and cultural settings
lIl! the need to enlist the cooperation and support of
community leaders
~ the need to overcome patients' fear of authorities,
and often of deportation
• the need to reach, evaluate, and care for persons
without legal status in their destination country
(e.g. undocumented migrants)
• the need to address complex psychosocial issues
(e.g. financial concerns, stigmatization, separation
from family members, post-traumatic distress) if
TB management is to prove successful
48.6
Recent Global Initiatives in T8 Control
Since the WHO declared TB a global health emer-
gency, heightened interest in TB control has fuelled
a number of promising new initiatives. Several ini-
tiatives directly address the key challenges outlined
above.
48.6.1
Global Partnership to Stop T8
This organization (the 'Stop TB Partnership') was
established in 1998, and includes over 200 organi-
zations worldwide. Partner groups include inter-
governmental and governmental organizations, as
well as academic institutions, non-governmental
organizations, and others. The WHO has established
the Stop TB secretariat to support and co-ordinate
the partnership.
Publication of the Global Plan to Stop TB (2001)
was a sentinel event in the partnership's history.
The plan has four primary objectives (Castro et al.
2001b):
fit To expand the DOTS strategy, in order to promote
access to effective diagnosis and treatment
• To adapt the strategy, to address the challenges of
HIV and TB drug resistance
II To improve existing tools by development of new
diagnostics, drugs, and vaccines
ill To strengthen the Stop TB Partnership, so as to
870
advance the application of TB control strategies
The partnership incorporates six working groups,
designed to address the objectives of this plan. The
working groups are devoted to DOTS expansion,
TB-HIV, MDR-TB, new diagnostics, new drugs, and
new vaccines. Each working group has identified and
published its own objectives, targets, mechanisms,
and milestones. For example, the Working Group
on DOTS Expansion has identified one of its targets
as the presence of a national tuberculosis reference
laboratory and an effective network of smear micros-
copy in every country with a TB incidence rate above
20 per 100,000 by 2005.
48.6.2
The Global Fund to Fight AIDS, Tuberculosis
and Malaria
The Global Fund to Fight AIDS, Tuberculosis and
Malaria was established in 2001. Its goal is to obtain
and distribute additional resources for control of these
diseases, in those areas where they are most urgently
needed. As of October 2002, the fund had collected
over US$2.1 billion in pledges, from governments,
foundations, non-profit organizations, and individu-
als. The fund provides grants for worthy programs
which further the control of these diseases, based on
a competitive evaluation process administered by sci-
entific experts, community representatives, and other
key stakeholders (Global Fund 2002).
48.6.3
Global Alliance for TB Drug Development
The Global Alliance for TB Drug Development was
established in 2000, as a public-private partnership
dedicated to the development of new anti-tuberculo-
sis drugs. It has become "the lead agency for the TB
Drug Development Working Group" of the Stop-TB
Partnership (Castro et al. 2001a). The Alliance pro-
motes the development of new drugs by financing
research and development initiatives in the public,
academic, and industrial sectors.
In the past, financial constraints have discouraged
'for-profit' companies from the substantial invest-
ments needed to develop new TB drugs, since the
vast majority of the market is in resource-poor areas.
The Global Alliance reduces this financial disincen-
tive, by actively supporting the costs of research and
development. New drugs will be an essential step for-
ward in global TB control: drugs are urgently needed
K. Schwartzman
to shorten and simplify treatment of susceptible TB,
improve the treatment of MDR-TB, and promote
treatment of latent infection. The Alliance's stated
objective is to have at least one new anti-TB drug
licensed by 2010, and available to high-burden coun-
tries by 2012 (TB Alliance 2002).
48.6.4
Global Drug Facility
Even existing anti-TB drugs - effective for the vast
majority of TB patients - often do not reach the areas
where they are most needed. The Global Drug Facil-
ity (GDF) is an initiative of the Stop TB Partnership,
begun in 2001. It was developed to increase access
to high-quality tuberculosis drugs (Global Drug
Facility 2002). The GDF aims to provide drugs for
the treatment of 10 million patients over five years,
thereby accelerating the expansion of the DOTS
strategy. Countries and non-governmental organi-
zations working within the DOTS framework can
apply to the GDF for grants of TB drugs; they can
also benefit from low-cost purchase arrangements
through pooled buying of high-quality drugs. Hence
the GDF has already achieved a 30% reduction in TB
drug prices.
As of October 2002, 23 countries - including sev-
eral of the highest-burden countries in Africa and
Asia -had been granted drugs from the GDF.
48.6.5
TB-HIV
The WHO and other TB control authorities have
increasingly recognized the close relationship between
the dual HIV and TB epidemics. Beyond the develop-
ment of specific TB control measures in the clinical
context of HIV infection, public health authorities are
now integrating TB and HIV control activities in set-
tings where both are highly prevalent.
For example, within the Stop TB Partnership,
the Working Group on TB/HIV published a "Strate-
gic Framework to Decrease the Burden of TB/HIV"
(Maher et al. 2002). This includes recommended
packages of interventions for HIV-infected persons
in low-, middle-, and high-income countries, based
on the cost-effectiveness of these measures. Some of
the interventions target TB directly: effective use of
the DOTS strategy, case finding, enhanced diagnosis
of smear-negative and extrapulmonary TB, possible
treatment oflatent TB with isoniazid. Some interven-
Tuberculosis Control in Developing and Developed Countries
tions target other infectious complications, such as
cotrimoxazole prophylaxis for pneumonia. It is worth
noting that the recommendation that highly active
antiretroviral therapy be provided to HIV-infected
persons in middle-income countries - indirectly
reducing TB incidence.
48.6.6
DOTS-Plus and the Green Light Committee
DOTS-Plus is a WHO initiative designed to address
the growing problem of multi-drug resistant TB. The
Working Group on DOTS-Plus for MDR-TB is one
of the six working groups of the Stop-TB Partner-
ship. DOTS-Plus builds on the DOTS framework by
incorporating systematic approaches to the detec-
tion, management, and surveillance of MDR-TB
- designed for areas with a high prevalence of MDR-
TB. As of October 2002, elements of the DOTS-Plus
strategy were still under development and testing;
pilot projects were underway in Latvia, Estonia, the
Philippines, Peru, and several areas of the Russian
Federation (DOTS-Plus 2002).
In 2000, the WHO published guidelines for the
establishment of DOTS-Plus pilot projects. Consid-
eration of such pilot projects is only appropriate in
sites where the core tuberculosis control program is
already functioning at a high level; otherwise, imple-
mentation of the key elements needed for detection,
treatment, and surveillance of drug-sensitive TB
remain the first priority.
The DOTS-Plus strategy includes two potential
approaches to the problem of multi-drug resistance:
1) standardized therapy with second-line drugs,
based on the empiric diagnosis of multi-drug resis-
tant TB, and knowledge of local multi-drug resistance
patterns; or 2) individualized therapy, based on drug
susceptibility results for each patient's isolate.
Even more than with drug-sensitive TB, a key limi-
tation to control of MDR-TB is consistent access to
effective, high-quality medications - largely second-
line anti-TB drugs. A crucial accomplishment of
WHO's working group on DOTS-Plus for MDR-TB
was the negotiation of major price reductions for
second-line anti-tuberculosis drugs - up to 99%
in resource-poor settings, e.g. $0.06 US per 250 mg
tablet of ciprofioxacin, vs. a reference price of $8.91
in the US (Gupta et al. 2001).
To promote access to reduced-price second-line
drugs within systematic control programs, the WHO
established the Green Light Committee in 2000, as
a subgroup of the working group on DOTS-Plus
871
(Gupta et al. 2002). The Green Light Committee
reviews projects requesting access to the second-line
drugs at reduced prices, to ensure that such projects
adhere to published guidelines-including the pres-
ence of an established core TB control program. As
of September 2002, the Committee had approved 10
proposals: countrywide projects in Estonia, Latvia,
and Malawi, and district-, state-, and prison-level
projects in Mexico, Peru, the Philippines, and the
Russian Federation. The Green Light Committee
has also provided technical assistance, in the form
of training sessions for the clinical management of
patients with MDR-TB.
Using an individualized treatment approach
within the DOTS-Plus framework, Farmer and col-
leagues from Partners in Health demonstrated an
85% cure rate - based on sputum culture conver-
sion - in a pilot project among 74 Peruvian patients
with MDR-TB, many of whom had resistance to all
first-line drugs (Farmer 2001). The ultimate goal of
the DOTS-Plus strategy is to build and implement a
systematic approach to MDR-TB, based on this type
of successful experience.
48.7
Recent Developments in Tuberculosis
Control: Examples from the Field
A total of22 high-burden countries account for an esti-
mated 79% of the world's nearly 9 million tuberculosis
cases.(Table 48.4) However, resources as well as inci-
dence, outcomes, and recent trends vary widely among
these 22 countries as well as in lower-burden areas. The
final section of this chapter outlines recent develop-
ments in TB control in several high-burden countries,
in the United States, and in England and Wales.
48.7.1
India
Although second only to China in population, India
accounts for the largest number of tuberculosis
patients worldwide (21%), because estimated inci-
dence is much higher in India than in China (184
per 100,000 in India in 2000). Notifications were 111
per 100,000 in 2000 (Blanc et al. 2002).
India is the site of remarkable success in expand-
ing coverage of the DOTS program (Khatri and
Frieden 2002). The Revised National Tuberculosis
Control Program was initiated in 1993, after a review
872 K. Schwartzman

Table 48.4. Estimated incidence of TB: high-burden countries, 2000. Data from Blanc et al. (2002)

Number estimated

All cases Smear positive cases

Country Population Thousands Rate per Thousands Rate per Cumulative

(1000 s) 100,000 pop 100,000 pop. incidence (0/0)

1 India 1,008,937 1,856 184 831 82 21

2 China 1,275,133 1,365 107 588 46 37
3 Indonesia 212,092 595 280 267 126 44
4 Nigeria 113,862 347 305 150 132 48
5 Bangladesh 137,439 332 242 149 109 51
6 Ethiopia 62,908 249 397 105 166 54
7 Philippines 75,653 249 330 112 148 57
8 Pakistan 141,256 247 175 111 78 60
9 South Africa 43,309 228 526 93 214 63
10 Russian Federation 145,491 193 132 87 59 65
11 DR Congo 50,948 163 320 70 138 67
12 Kenya 30,669 149 484 62 201 68
13 Viet Nam 78,137 148 189 66 85 70
14 UR Tanzania 35,119 126 359 54 153 72
15 Brazil 170,406 116 68 52 30 73
16 Thailand 62,806 88 140 39 62 74
17 Uganda 23,300 82 351 35 149 75
18 Myanmar 47,749 80 168 36 76 76
19 Mozambique 18,292 79 433 33 180 77
20 Cambodia 13,104 75 572 33 256 77
21 Zimbabwe 12,627 74 584 30 234 78
22 Afghanistan 21,765 70 321 31 144 79
Total, high-burden countries 3,781,004 6,910 183 3,033 80 79
Global total 6,053,531 8,735 144 3,836 63 100

conducted in 1992 indicated that less than half of TB
patients were accurately diagnosed, and less than half
of those diagnosed were effectively treated.
The revised national program is based on the
DOTS strategy, with diagnosis of TB based primar-
ily on sputum microscopy. The standard treatment
regimen for newly diagnosed patients with positive
sputum smears involves directly observed admin-
istration of isoniazid, rifampin, pyrazinamide, and
ethambutol for two months; this is followed by iso-
niazid and rifampin for an additional four months,
with at least one dose per week directly observed.
The program is run at the district level, with support
and administration at the state and central govern-
ment levels. In addition, the WHO hired, trained, and
assigned physician-consultants to central, state, and
local governments.
After eight years, the program had expanded to 211
districts of 19 states, covering 43% of India's popula-
tion. As a result of the program, nearly 200,000 health
workers had been trained and over 3,000 laboratories
were upgraded. A total of 666,000 patients were regis-
tered in the program, with an 83% treatment success
rate overall. Khatri and Frieden estimated that the
program had prevented over 200,000 deaths, while
the cost of program expansion was $50 million US
over 8 years (Khatri and Frieden 2002).
Despite these milestones in rapid expansion, India
continues to face considerable challenges. Much of
the population remains difficult to reach, and over
half resides outside DOTS program areas. Even in
areas now covered by the revised national program,
at least 40% of TB patients obtain diagnosis and
treatment in the private sector. The HIV epidemic
has reached India, with at least four million HIV-
infected persons, and an estimated prevalence of HIV
infection of 0.8% among adults aged 15-49 (UNAIDS
2002). As in Africa, even the most successful TB con-
trol strategies will be compromised by further expan-
sion of the HIV epidemic.
Based on local survey data, the WHO estimated
that 3.3% of TB cases in India involved multi-drug
resistance. Inadequate treatment and supervision
outside the DOTS program, as well as ongoing
transmission of MDR strains, will only worsen this
problem.
Hence continued DOTS expansion and construc-
tive engagement of the private sector are urgently
Tuberculosis Control in Developing and Developed Countries 873

needed. Indeed, with the support of the World Bank (behind Zimbabwe and Cambodia) among the 22
India will further extend the revised national pro- highest-burden countries. The WHO estimates that
gram, with the goal of DOTS coverage of 80% of the 60% of TB patients aged 15-49 are HIV-infected,
country by 2004 (Blanc et al. 2002). while the estimated overall prevalence of HIV infec-
tion in the same age group was 20.1% (Blanc et al.
2002; UNAIDS 2002). South Africa now accounts for
48.7.2 3% of the world's patients with active TB.
China However, South Africa has made considerable
progress with respect to DOTS coverage. A revised
China is second only to India in numbers of TB cases, national TB control program including the DOTS
and accounts for 16% of the world's TB patients. In strategy was first established in 1996. By the end of
2000, estimated incidence (all types) was 107 per 2000, 77% of the country was covered. An estimated
100,000, with a notification rate of 36 per 100,000; the two-thirds of smear-positive cases were detected
latter had not changed since 1998 (Blanc et al. 2002). under the DOTS program - close to the WHO bench-
Coverage under the DOTS program expanded mark of 70%. Unfortunately, even within the DOTS
considerably during the early 1990s, due largely to program overall treatment success rates have fallen
a World Bank loan which covered half the country. (from 73% in 1997 to 60% in 1999), likely because of
By 2000, 68% of the population resided in areas with HIV (Blanc et al. 2002). The estimated prevalence of
access to the DOTS program, but only an estimated MDR-TB is 1.5% among new cases; while MDR-TB
33% of new smear-positive cases were diagnosed has been relatively infrequent to date, the extent of
under the DOTS strategy. Furthermore, only 13% the HIV epidemic makes an increase in the incidence
of patients diagnosed with TB were referred to TB of MDR-TB likely in the coming years.
dispensaries operating within the DOTS program. The severity of the TB-HIV epidemic has made TB
Patients treated within the program have excellent control a priority for the South African government.
outcomes, with over 90% of new smear-positive The high incidence has also spurred international
patients cured (Feng-Zeng et al.1996). research collaborations dedicated to epidemiology,
China has anticipated an additional expansion of clinical trials, and public health operations.
the DOTS program from 2002 onwards, with funds
from the World Bank and other agencies. However,
challenges to TB control are similar to those encoun- 48.7.4
tered elsewhere. Most patients continue to be diag- Peru
nosed and treated outside the DOTS program.
Limited information is available about the extent Peru has witnessed remarkable successes in TB
of HIV disease is China, but the United Nations has control over the last decade. This reflects sustained
estimated that there were 850,000 persons living with effort at all levels of the health system, beginning with
HlV in China at the end of 2001, corresponding to a the introduction of a revised National Tuberculosis
0.1% prevalence among adults aged 18-49 (UNAIDS Control Program in 1990. The revised program incor-
2002). In Henan province, 10.8% of isolates from newly porates all elements of the DOTS strategy, and was
diagnosed patients were multi-drug resistant during rapidly implemented in all areas of the country. From
1996-99; the prevalence of MDR-TB among new cases 1989 to 2000, the number of sputum microscopy labo-
in four other areas ranged from 1.4-4.5% (Espinal et ratories quadrupled. With respect to diagnostics, the
al. 2001). As in India, MDR-TB will be a growing threat Peruvian approach goes beyond the basic DOTS
to Chinese TB control in coming years. strategy, because patients with persistent respiratory
symptoms and negative sputum smears are referred
for more complete medical evaluation which can
48.7.3 include chest radiography and sputum mycobacte-
South Africa rial cultures (Suarez et al. 2001).
From 1990-93, immediately after introduction of
The HlV pandemic has devastated South Africa, as the revised program, case notifications increased,
it has the other countries of sub-Saharan Africa. As most likely because of expanded access to diagnostic
a result, a dual TB-HlV epidemic is in full force; facilities. However, from 1993-2000 reported cases of
in 2000, the estimated TB incidence was 526 cases pulmonary TB declined by an average of 7.4% annu-
per 100,000, placing South Africa third in incidence ally. Hence in 2000, the reported incidence was 136
874
per 100,000 - compared to 170 per 100,000 in 1990.
In 1990, before adoption of the revised program, only
40% of smear-positive patients successfully com-
pleted treatment. By 2000, the proportion exceeded
90%. The revised program averted an estimated
91,000 deaths, or 70% of the deaths from tuberculosis
that would have been expected, in the absence of the
expanded program (Suarez et al. 2001).
The accomplishments of the Peruvian TB control
program highlight the impact of widespread system-
atic diagnosis and treatment on TB incidence and
mortality. Another key factor was the relative rarity
of HIV infection in Peru - estimated to be present
in only 0.56% of adults (Suarez et al. 2001). However,
multi-drug resistance exists in Peru, as elsewhere;
in 1999, 3% of new patients and 12.3% of retreat-
ment patients harbored multi-drug-resistant isolates
(Espinal et al. 2001).
To address this challenge, Peru developed a
national program for standardized 18-month treat-
ment of 'chronic' patients with suspected MDR-TB
- primarily those who remained sputum smear-pos-
itive despite fully supervised administration of the
standard WHO retreatment regimen. The 18-month
regimen included kanamycin, ciprofloxacin, ethion-
amide, pyrazinamide, and ethambutol; all doses were
provided on an outpatient basis in local health clin-
ics. With the standardized empiric regimen, the over-
all cure rate was 48%, while 12% died. The program
authorities estimated that addition of this regimen
reduced the risk of death by 36% among those treated.
The average cost for a complete course of treatment
was US$2381, of which US$824 was the cost of the
drugs themselves. Other major costs included visits
for directly observed therapy (US$507), food parcels
given to patients weekly (US$390), and physician
consultations (US$213) (Suarez et al. 2002).
While expensive compared to first-line TB control
interventions, the Peruvian program nonetheless
demonstrates the feasibility of a systematic approach
to MDR-TB in a middle-income country. Other
reports from Peru have documented the success of
DOTS-Plus pilot projects which incorporated drug
susceptibility testing and individualized treatment
regimens with second-line drugs, in limited areas
(Farmer 2001). However, programs for treatment of
MDR-TB are only relevant in sites where the basic
TB control system is already functioning with a high
level of success, as was the case in Peru.
The sustained drop in TB incidence has in fact
resulted in Peru's removal from the list of the 22
highest-burden countries - the only country to
accomplish this feat thus far. The accomplishments
K. Schwartzman
of the Peruvian program, over a single decade, serve
as a powerful reminder that successful control of TB
is indeed possible. It hinges on the systematic adop-
tion of the basic DOTS strategy for implementation,
diagnostics, treatment, and surveillance.
48.7.5
United States
The United States witnessed a steady decline in tuber-
culosis incidence and deaths from the early 1950s to
1984. From 1985-92, incidence initially reached a
plateau at 9.3-9.4 cases per 100,000 annually, then in
fact increased to a peak of 10.5 cases per 100,000 in
1992 (Division of Tuberculosis Elimination 2002).
This resurgence during the late 1980s and early
1990s rekindled concern about a disease largely forgot-
ten in the US. Mini-epidemics of multi-drug-resistant
TB in inner cities and hospitals exposed major defi-
ciencies in the TB control infrastructure - notably the
lack of resources available to ensure prompt diagnosis,
initiation and completion of appropriate treatment.
The spread of multi-drug resistance indicated a
breakdown of the public health system for TB control,
so that patients received inadequate and incomplete
treatment - due to non-standard treatment regimens
and/or poor patient adherence - and transmitted
drug-resistant strains to others in their communities.
If the Peruvian experience of the 1990s illustrates the
potential benefits of successful implementation of a
systematic TB control strategy, the US experience of
1985-92 highlights the disastrous consequences of its
abandonment.
Nowhere was this phenomenon better demon-
strated than in New York City. As reported by New
York's TB control team,
"By 1992, the situation in New York City looked
bleak. The number of cases of tuberculosis had
nearly tripled in 15 years. In central Harlem, the
case rate of 222 per 100,000 people exceeded that
of many Third World countries. Outbreaks of multi-
drug-resistant tuberculosis had been documented in
more than half a dozen major hospitals, with case
fatality rates greater than 80 percent, and health care
workers were becoming ill and dying of this disease.
Nearly one in five patients with tuberculosis in New
York City had multi-drug-resistant strains, and the
proportion of patient with multi-drug resistance had
more than doubled in seven years. In the first quarter
of 1991, with three percent of the country's popula-
tion, New York City accounted for a remarkable 61
percent of cases of multi-drug-resistant tuberculosis
Tuberculosis Control in Developing and Developed Countries
in the United States:' (Frieden et al.1995)
New York's TB epidemic occurred in the context
of major shifts in the epidemiology of tuberculosis,
with the growing contribution of the HIV epidemic,
immigration from high-prevalence countries, home-
lessness and poverty. However, the TB epidemic
also reflected drastic reductions in public health
activities; the TB control staff had been severely
reduced, the number of TB clinics had decreased by
two-thirds, and by 1989 less than 50% of patients who
began treatment were cured (Frieden et al.I995).
Interventions which ultimately brought TB under
control in New York included 1) massive expan-
sion of directly observed therapy, with a four-fold
increase in TB control staff from 1988-1994, and a
10-fold budget increase (from US$4 million to US$40
million annually); 2) heightened infection control
measures in hospitals, jails, and homeless shelters; 3)
systematic treatment of TB patients with at least four
anti-tuberculous drugs. Treatment completion rates
increased to 90% by 1994 (Frieden et al.I995).
In New York as elsewhere, renewed investment
in TB management has led to major gains in TB
control. From 1993 onward, US TB incidence again
declined steadily, reaching 5.6 cases per 100,000 in
2001 - a 47% reduction in incidence compared to
1992 (Division of Tuberculosis Elimination 2002).
Figure 48.3 illustrates these recent trends in US TB
incidence. However, the cost of the TB resurgence was
enormous. In New York alone, the TB epidemic was
associated with an estimated excess of 20,000 cases
over 15 years, with an estimated cost of over US$1
billion (Frieden et al.I995).
As TB has abated in the US inner cities, other aspects
of TB epidemiology and control have come under
28000 - = = - - - - - - - - - - - - - - - - - - - - - - , . 1 4
til
24000 ' •.
... ..
= 20000
Gl
()
al
Fig. 48.3. Tuberculosis cases and inci- 15 16000
Q.
dence, United States, 1980-2001. Data ~ '.
are from the Division of Tuberculosis '0 12000
...
Elimination (U.S. Centers for Disease .8
Control and Prevention), 2002. After § 8000
reaching a plateau in the early 1980s, z II:
incidence rose for several years, peak- 4000
ing in 1992. From 1993 through 2001,
the total number of cases decreased by o +---+--+--+--+--+-+-+-+-+-+-+-+-+-+-+-+--+--+--+--+-+ 0
5-7% each year. By 2001, the number
of cases had declined 40% from the ~~##~~~##,~##~#~~#~#~~
peak reached in 1992.
875
increased scrutiny. In 2001, the number of reported
foreign-born TB patients exceeded the number of
reported US-born patients, for the first time; each
group accounted for 49% ofTB incidence overall, with
the remaining 2% of unknown origin. The three top
countries of origin were Mexico (23% of all foreign-
born patients), the Philippines (12%), and Vietnam
(8%) (Division of Tuberculosis Elimination 2002).
In 2001, the incidence of TB among foreign-born
persons in the US was 26.6 per 100,000 - compared to
3.1 per 100,000 among US-born persons. During the
decade from 1991-2001, the nationwide incidence of
TB in the US-born declined by 63%, while among the
foreign-born it fell by a more modest 22%. This was
not surprising, in that intensified control measures in
the inner cities were most effective in reducing ongo-
ing transmission, which largely involved the US-born
(Jasmer et al. 1999).
Hence, the US Institute of Medicine, in its land-
mark report entitled "Ending Neglect: the Elimina-
tion of Tuberculosis in the United States:' identified
intensified screening of new immigrants from high-
incidence countries as a key step toward enhanced
TB control (United States Institute of Medicine:
Committee on the Elimination of Tuberculosis in
the United States 2000). The goal of such screening
is the identification (through tuberculin skin testing)
and treatment of new arrivals with latent infection,
among whom the incidence of TB is highest during
the first years after immigration. Most TB in these
groups reflects infection acquired in the countries
of origin (Jasmer et al. 1999; Kulaga et al. 2002), so
that further reductions in incidence will more likely
occur from treatment of latent infection rather than
intensified contact investigation or other measures
... .. -.. .•..•..•.. .. -.- .. .....
12
c
o
-
'
. 10 'i
:;
• Q.
£
• ••
8 Cl
Cl
..•. g
Cl
." 6 ...
...
Gl
IL
4 Gl
';
2
Year
876
targeting transmission within the US.
However, large-scale implementation of screening
programs based on tuberculin testing is extremely
costly to organize and administer; many of the tar-
geted individuals are in fact at relatively low risk for
reactivation (Schwartzman and Menzies 2000). Ulti-
mately, sustained reductions in TB incidence among
migrants to low-incidence countries will result from
improved TB control in their countries of origin.
48.7.6
England and Wales
Recent trends in tuberculosis in England and Wales
parallel those in the US in the late 1980s and early
1990s. Incidence fell steadily from 1950 to 1988;
in 1988, TB incidence reached a nadir of 9.4 cases
per 100,000. However, from 1988 to 1998 incidence
rebounded, reaching 10.9 cases per 100,000 in 1998.
In 1999 incidence once again fell to 10.2 per 100,000
(Public Health Laboratory Service 2002).
As in the US, the TB resurgence has occurred
mainly in the largest cities - notably London, which
has the highest TB incidence in the United Kingdom.
In 1999, the incidence in London (32.2 per 100,000)
was over three times the national average. Available
reporting data implicate HIV in only a small minor-
ity ofTB cases - 3.3% of British adult patients in 1998,
somewhat more in London (5.4% ofTB patients aged
16-54). Multi-drug resistance has never reached the
levels seen in the US; in 1999,0.8% of initial isolates
(those obtained at start of treatment) in England
and Wales demonstrated multi-drug resistance. In
London, the proportion was 1.8% during 1994-99
(Public Health Laboratory Service 2002).
However, as in New York and other U.S. cities in the
early 1990s, TB in inner-city London is characterized
by a large burden of disease due to recent transmis-
sion; a 12-month molecular epidemiological survey in
inner London indicated that 27% of patients had iso-
lates identical to those obtained from at least one other
patient (Hayward et al. 2002). To date, direct observa-
tion of treatment has been rare, and comprehensive
data on treatment completion and other outcomes
have been unavailable (Hayward and Coker 2000).
The foreign-born account for a higher proportion of
TB cases than in the US, and this proportion has been
increasing. British surveillance data have categorized
patients by ethnicity rather than place of birth. In 1999,
patients from the Indian subcontinent accounted for
38% of all cases, while those from sub-Saharan Africa
accounted for 15%; other 'non-Whites' accounted for
K. Schwartzman
14%. Among persons from sub-Saharan Africa, the
estimated incidence reached 230 per 100,000 in 1999,
compared with 3.7 per 100,000 among 'Whites' (the
term used by public health authorities) (Public Health
Laboratory Service 2002).
As described earlier in this chapter, BCG vaccina-
tion has remained a core element of TB control in the
United Kingdom. To better address recent challenges to
TB control, a system of enhanced national TB surveil-
lance was introduced in 1999. The surveillance system
will also incorporate routine reporting of patient out-
come data - in the same format used by national TB
programs in high-burden countries. Other improve-
ments to the TB control infrastructure will target coor-
dination between local health authorities, and more
systematic screening and follow-up of new arrivals.
48.7.7
Russia
Since the collapse of the Soviet Union, TB morbid-
ity and mortality have increased dramatically in
Russia. The notification rate increased from a nadir
of 34 per 100,000 in 1990 to 95 per 100,000 in 2000;
reported TB mortality reached 20.4 per 100,000 in
2000 (Blanc et al. 2002; Shilova 2000). Russia ranks
tenth in total TB cases among the highest burden
countries, and now accounts for 2% of the world's
TB patients (Blanc et al. 2002).
TB care and control are delivered by a network of
facilities, including dispensaries and hospitals, which
is not integrated into the general health system. The
DOTS strategy has not been implemented, except
within localized pilot projects. Drug and laboratory
supplies have been inconsistent, further hampering
TB control and potentially promoting drug resistance.
The estimated prevalence of multi-drug resistance
among Russian TB patients overall is 6.0%, making
drug resistance a major limiting factor to treatment
success. The situation is even more dramatic in Rus-
sian prisons, where TB notifications were 3,118 per
100,000 in 2000. In a survey of 164 patients housed in
a Siberian TB referral prison, the prevalence of initial
multi-drug resistance was 22.6%, while 35% of all
patients in the prison failed directly observed treat-
ment with the empirical WHO re-treatment regimen
(Kimerling et al. 1999).
While HIV infection was initially uncommon in
Russia, local survey data suggest its prevalence is
now increasing rapidly. In Ore! Oblast (a rural area
300 km southwest of Moscow), HIV seroprevalence
increased from 0.001% in 1996 to 0.12% in 2000
Tuberculosis Control in Developing and Developed Countries
(Kazionny et al. 2001).
Regional pilot projects based on the DOTS strategy
have yielded variable results. In Ivanovo Oblast, where
the strategy was implemented in 1995, a survey con-
ducted from April-June 1999 indicated that only 57%
of smear-positive patients were successfully treated.
None of these patients had MDR-TB (Centers for Dis-
ease Control and Prevention 2001a). After these sub-
optimal outcomes were identified, the program was
enhanced: patients were offered food supplements or
free transportation to clinic, public health staff were
provided with transportation so as to locate treatment
interrupters, and providers were offered bonuses for
treatment completion by their patients.
In Orel Oblast, where the DOTS strategy was
implemented in 1999, treatment success rates were
much higher: 81% for new smear-positive patients
from October 1999-March 2000. Three percent of
patients had MDR-TB (Centers for Disease Control
and Prevention 2001b).
Clearly, TB control in Russia represents an enor-
mous challenge. The Russian president has identified
TB as a national priority. A host of international aid
and public health agencies, as well as nongovernmental
organizations, have aided Russian TB control efforts.
Russia will also receive a US$150 million loan from
the World Bank, to support drug procurement (Munro
2002). Although the DOTS strategy differs from previ-
ous approaches to TB control within Russia and the
Soviet Union, there is evidence of its effectiveness in
the Orel pilot project - as well as in other countries
- which should promote its use in Russia. Ongoing col-
laboration with the WHO and other partners will also
promote a more consistent and integrated approach to
TB control, and improve essential infrastructure.
48.8.
Conclusion
The vast majority of tuberculosis patients can be
cheaply and effectively treated, using standard drug
regimens known for decades. Tuberculosis can be
effectively controlled in large populations, using a
simple yet comprehensive strategy which was first
developed and proven 20-30 years ago -in several
of the poorest countries of the world. Nonetheless,
tuberculosis continues to be a leading global cause
of morbidity and mortality.
Recent years have brought renewed concern
about tuberculosis, fueled by the ravages of the dual
TB-HIV epidemics and of multi-drug-resistant TB.
877
Major new public health and funding initiatives for
TB control are concrete evidence of this heightened
awareness. The last decade has seen dramatic chal-
lenges, political mobilization, innovative programs,
and remarkable success in a number of countries.
These successes offer important lessons, as well as
inspiration toward continued progress.
Just as sustained commitment is essential, compla-
cency remains the worst enemy of TB control - and
even the funds now available will not be sufficient.
The WHO estimates that an additional US$3.8 billion
are needed, if the 22 highest-burden countries are
to execute their current plans for expansion of the
DOTS strategy over the next five years. The author
hopes that readers responsible for TB treatment and
control will continue to act as strong advocates for
their patients-and that in so doing, they ensure that
urgently needed resources indeed become available.
Acknowledgements. The author is supported by a
Chercheur-Boursier Clinicien career award from the
Fonds de Recherche en Sante du Quebec. The author
acknowledges the secretarial assistance of Elizabeth
Lustig and Catherine Michaud.
References
Alland D, Kalkut GE, Moss AR, McAdam RA, Hahn JA, Bos-
worth W, Drucker E, Bloom BR (1994) Transmission of
tuberculosis in New York City. An analysis by DNA finger-
printing and conventional epidemiologic methods. N Engl
J Med 330:1710-1716
American Thoracic Society (2000) Targeted tuberculin testing
and treatment of latent tuberculosis infection. Am J Respir
Crit Care Med 161:S221-S247
Antonucci G, Girardi E, Raviglione MC, Ippolito G, and for the
Gruppo Italiano di Studio Tubercolosi e AIDS (1995) Risk
factors for tuberculosis in HIV-infected persons: a prospec-
tive cohort study. JAMA 274:143-148
Badri M, Wilson D, Wood R (2002) Effect of highly active anti-
retroviral therapy on incidence of tuberculosis in South
Africa: a cohort study. Lancet 359:2059-2064
Behr MA, Warren SA, Salamon H, Hopewell PC, Ponce de Leon
A, Daley CL, Small PM (1999a) Transmission of Mycobac-
terium tuberculosis from patients smear-negative for acid-
fast bacilli. Lancet 353:444-449
Behr MA, Wilson MA, Gill WP, Salamon H, Schoolnik GK, Rane
S, Small PM (1999b) Comparative genomics of BCG vaccines
by whole-genome DNA microarray. Science 284: 1520-1523
Binkin NJ, Zuber PLF, Wells CD, Ahn J, Tipple MA, Castro KG
(1996) Overseas screening for tuberculosis in immigrants
and refugees to the United States: current status. Clin Infect
Dis 23:1226-1232
Blanc L, Bleed D, Dye C, Floyd K, Palmer K, Watt C (2002) Global
tuberculosis control: surveillance, planning, financing. WHO
878
report 2002. World Health Organization, Geneva. Available at
www.who.intlgtb/publications/globrep02lPDFI
Blumberg HM, Watkins DL, Berschling JD, Antle A, Moore P,
White N, Hunter M, Green B, Ray S, McGowan JE (1995)
Preventing the nosocomial transmission of tuberculosis.
Ann Intern Med 122:658-663
Bock NN, McGowan JE, Ahn J, Tapia J, Blumberg HM (1996)
Clinical predictors of tuberculosis as a guide for a respi-
ratory isolation policy. Am J Respir Crit Care Med 154:
1468-1472
Borgdorff MW, Floyd K, Broekmans JF (2002) Interventions to
reduce tuberculosis mortality and transmission in low-and
middle-income countries. Bull WHO 80:217-227
Canadian Lung Association (2000) Canadian tuberculosis
standards (Long Red), 5th edn. Health Canada, Ottawa
Castro A, Hinman A, Kibuga D, Smith I, Zintl P (2001 a) Stop TB
plans. In: The global plan to stop TB. Stop TB secretariat,
Geneva. Available at www.stoptb.org/GPSTB/
Castro A, Hinman A, Kibuga D, Smith I, Zintl P (2001b)
The global partnership to stop TB. In: The global plan
to stop TB. Stop TB secretariat, Geneva. Available at
www.stoptb.orglGPSTB/
Centers for Disease Control and Prevention (2001a) Tuber-
culosis treatment interruptions - Ivanovo Oblast Russian
Federation 1999. JAMA 285:1953-1954
Centers for Disease Control and Prevention (2001b) Evalu-
ation of a directly observed short-course strategy for
treating tuberculosis - Orel Oblast, Russian Federation,
1999-2000. JAMA 285:1954-1956
Cobelens FG, Deutekom H, Draayer-Jansen IWE, Schepp-
Beelen ACHM, Gerven PJHJ, Kessel RPM, Mensen MEA
(2000) Risk of infection with Mycobacterium tuberculo-
sis in travellers to areas of high tuberculosis endemicity.
Lancet 356:461-465
Colditz GA, Brewer TF, Berkey CS, Wilson ME, Burdick E, Fine-
berg HV, Mosteller F (1994) Efficacy of BCG vaccine in the
prevention of tuberculosis. JAMA 271:698-702
Comstock GW, Baum C, Snider DEJ (1979) Isoniazid prophy-
laxis among Alaskan Eskimos: a final report of the Bethel
isoniazid studies. Am Rev Respir Dis 119:827-830
Coronado VG, Beck-Sague CM, Hutton MD, Davis BJ, Nicholas
P, Villareal C, Woodley CL, Kilburn JO, Crawford JT, Frieden
TR (1993) Transmission of multidrug-resistant Mycobacte-
rium tuberculosis among persons with human immunode-
ficiency virus infection in an urban hospital: epidemiologic
and restriction fragment length polymorphism analysis. J
Infect Dis 168:1052-1055
Creese A, Floyd K, Alban A, Guinness L (2002) Cost-effective-
ness of HIV I AIDS interventions in Africa: a systematic
review of the evidence. Lancet 359:1635-1642
Daley CL, Small PM, Schecter GF, Schoolnik GK, McAdam RA,
Jacobs WR Jr, Hopewell PC (1992) An outbreak of tubercu-
losis with accelerated progression among persons infected
with the human immunodeficiency virus. An analysis
using restriction-fragment-length polymorphism. N Engl
J Med 326:231-235
Dasgupta K, Schwartzman K, Marchand R, Tannenbaum TN,
Brassard P, Menzies D (2000) Comparison of cost-effec-
tiveness of tuberculosis screening of close contacts and
foreign-born populations. Am J Respir Crit Care Med 162:
2079-2086
Department of Health and the Health Education Authority,
United Kingdom (1998) BCG Vaccine Factsheet. Health
K. Schwartzman
Education Authority, London. www.immunisation.org.uk1
pdflbcg.pdf. Accessed 28 Nov 2002. Ref type: pamphlet
Division of Tuberculosis Elimination, National Center for HIV
STD and TB Prevention (2002) Reported tuberculosis in
the United States, 2001. US Centers for Disease Control and
Prevention, Atlanta, GA. Available at www.cdc.gov/nchstp/
tb/survlsurv2001/default.htm
DOTS-Plus for Multi-Drug-Resistant Tuberculosis (2002) World
Health Organization, Geneva. www.who.int/gtb/policyrd!
DOTSplus.htm. Accessed 28 Nov 2002. Ref type: pamphlet
Dye C, Espinal MA, Watt q, Mbiaga C, Williams BG (2002)
Worldwide incidence of multidrug-resistant tuberculosis.
J Infect Dis 185:1197-1202
Enarson DA, Rieder HL, Arnadottir T, Trebucq A (2000) Man-
agement of tuberculosis: a guide for low income countries.
International Union Against Tuberculosis and Lung Dis-
ease, Paris
Espinal MA, Kim SJ, Suarez PG, Kam KM, Khomenko AG,
Migliori GB, Baez J, Kochi A, Dye C, Raviglione MC (2000)
Standard short-course chemotherapy for drug-resistant
tuberculosis: treatment outcomes in 6 countries. JAMA
283:2537-2545
Espinal MA, Laszlo A, Simonsen L, Boulahbal F, Kim SJ,
Reniero A, Hoffner S, Rieder HL, Binkin N, Dye C, Wil-
liams R, Raviglione MC (2001) Global trends in resistance
to antituberculosis drugs. N Engl J Med 344:1294-1303
Farmer P (2001) DOTS and DOTS-plus: not the only answer.
Ann NY Acad Sci 953:165-184
Feng-Zeng Z, Murray C, Spinaci S, Styblo K, Broekmans JF
(1996) Results of directly observed short-course chemo-
therapy in 112 842 Chinese patients with smear-positive
tuberculosis. Lancet 347:358-362
Fennelly KP, Nardell EA (1998) The relative efficacy of res-
pirators and room ventilation in preventing occupational
tuberculosis. Infect Contr Hosp EpidemioI19:754-759
Fischl MA, Uttamchandani RB, Daikos GL, Poblete RB, Moreno
IN, Reyes RR, Boota AM, Thompson LM, Cleary TJ, Lai S
(1992) An outbreak of tuberculosis caused by multiple-
drug-resistant tubercle bacilli among patients with HIV
infection. Ann Intern Med 117:177-183
Floyd K, Wilkinson D, Gilks C (1997) Comparison of cost
effectiveness of directly observed treatment (DOT) and
conventionally delivered treatment for tuberculosis: expe-
rience from rural South Africa. BMJ 315:1407-1411
Frieden TR, Fujiwara PI, Washko RM, Hamburg MA (1995)
Tuberculosis in New York - turning the tide. N Engl J Med
333:229-233
Frieden TR, Sherman LF, Maw KL, Fujiwara PI, Crawford JT,
Nivin B, Sharp V, Hewlett D Jr, Brudney K, Alland D, Kre-
iswirth BN (1996) A multi-institutional outbreak of highly
drug-resistant tuberculosis: epidemiology and clinical
outcomes. JAMA 276:1229-1235
Ginwalla R, Ayles H, Sibande NM (2002) Intensive TB case
detection among HIV-positive clients through the volun-
tary counselling and testing service. Int J Tuberc Lung Dis
6 [Suppll):S178
Global Drug Facility (2002) What is the GDF? Stop TB partner-
ship secretariat, Geneva. www.stopTB.orglGDF/ Accessed
28 Nov 2002. Ref type: pamphlet
Global Fund to Fight AIDS, Tuberculosis and Malaria (2002)
www.globalfundatm.org/. Accessed 28 Nov 2002. Ref type:
pamphlet
Gupta R, Kim JY, Espinal MA, Caudron JM, Pecoul B, Farmer
Tuberculosis Control in Developing and Developed Countries
PE, Raviglione MC (2001) Responding to market failures in
tuberculosis control. Science 293:1049-1051
Gupta R, Cegielski JP, Espinal MA, Henkens M, Kim JY,
Weezenbeek C, Lee JW, Raviglione MC, Suarez PG, Varaine
F (2002) Increasing transparency in partnerships for health
- introducing the Green Light Committee. Trop Med Int
Health 7:970-976
Hayward AC, Coker RJ (2000) Could a tuberculosis epidemic
occur in London as it did in New York? Emerg Infect Dis
6:12-16
Hayward AC, Goss S, Drobniewski F, Saunders N, Shaw RJ,
Goyal M, Swan A, Uttley A, Pozniak A, Grace-Parker J,
Watson JM (2002) The molecular epidemiology of tuber-
culosis in inner London. Epidemiol Infect 128:175-184
Jasmer RM, Hahn JA, Small PM, Daley CL, Behr MA, Moss AR,
Creasman JM, Schecter GF, Paz EA, Hopewell PC (1999) A
molecular epidemiologic analysis of tuberculosis trends in
San Francisco. Ann Intern Med 130:971-978
Johnson JL, Okwera A, Hom DL, Mayanja H, Kityo CM,
Nsubuga P, Nakibali JG, Loughlin AM, Yun H, Mugyenyi
n,
PN, Vernon A, Mugerwa RD, Ellner and Whalen CC for
the Uganda - Case Western Reserve University Resarch
Collaboration (2001) Duration of efficacy of treatment of
latent tuberculosis infection in HIV-infected adults. AIDS
15:2137-2147
Kazionny B, Wells CD, Kluge H, Gusseynova N, Molotilov V
(2001) Implications of the growing HIV-l epidemic for
tuberculosis control in Russia. Lancet 358:1513-1514
Khatri GR, Frieden TR (2002) Controlling tuberculosis in
India. N Engl J Med 347:1420-1425
Kimerling ME, Kluge H, Vezhnina N, Lacovazzi T, Demeu-
lenaere T, Portaels F, Matthys F (1999) Inadequacy of the
current WHO re-treatment regimen in a central Siberian
prison: treatment failure and MDR-TB. Int J Tuberc Lung
Dis 3:451-453
Ko G, Burge HA, Nardell EA, Thompson KM (2001) Estima-
tion of tuberculosis risk and incidence under upper room
ultraviolet germicidal irradiation in a waiting room in a
hypothetical scenario. Risk Anal 21:657-673
Kulaga S, Behr M, Musana K, Brinkman J, Menzies D, Bras-
sard P, Kunimoto D, Tannenbaum TN, Thibert L, Joseph JF,
Boivin JF, Schwartzman K (2002) Molecular epidemiology
of tuberculosis in Montreal. CMAJ 167:353-354
Lobato MN, Hopewell PC (1998) Mycobacterium tuberculosis
infection after travel to or contact with visitors from coun-
tries with a high prevalence of tuberculosis. Am J Respir
Crit Care Med 158:1871-1875
Lonnroth K, Lambregts K, Nhien DTT, Quy HT, Diwan VK
(2000) Private pharmacies and tuberculosis control: a
survey of case detection skills and reported anti-tuberculo-
sis drug dispensing in private pharmacies in Ho Chi Minh
City, Vietnam. Int J Tuberc Lung Dis 4:1052-1059
MacPherson DW, Gushulak BD (2001) Human mobility and
population health: new approaches in a globalizing world.
Perspect Bioi Med 44:390-341
Maher D, Mikulencak M (1999) What is DOTS? A guide to
understanding the WHO-recommended strategy known
as DOTS. World Health Organization, Geneva
Maher D, Floyd K, Raviglione MC (2002) Strategic framework
to decrease the burden of TB/HIV. World Health Organiza-
tion, Geneva
McCarthy OR (1984) Asian immigrant tuberculosis - the effect
of visiting Asia. Br J Dis Chest 78:248-253
879
Menzies D, Tannenbaum TN, FitzGerald JM (1999) Tuberculo-
sis: 10. Prevention. CMAJ 161:717-724
Menzies D, Vissandjee B (1992) Effect of bacille Calmette-
Guerin vaccination on tuberculin reactivity. Am Rev Respir
Dis 145:621-625
Moss AR, Alland D, Telzak E, Hewlett D Jr, Sharp V, Chiliade
P, LaBombardi V, Kabus D, Hanna B, Palumbo L, Brudney
K, Weltman A, Stoeckle K, Chirgwin K, Simberkoff M,
Mogzadeh S, Eisner W, Lutfey M, Kreiswirth B (1997) A
city-wide outbreak of a multiple-drug-resistant strain of
Mycobacterium tuberculosis in New York. Int J Tuberc
Lung Dis 1:115-121
Mostowy S, Behr MA (2002) Comparative genomics in the
fight against tuberculosis: diagnostics, epidemiology, and
BCG vaccination. Am J Pharm Genom 2: 189-196
Munro R (2002) Path paved for release of $100 M loan for TB.
The Moscow Times. 21 Oct 2002. Ref type: newspaper
Murray CJL, DeJonghe E, Chum HJ, Nyangulu DS, Salomao
A, Styblo K (1991) Cost effectiveness of chemotherapy
for pulmonary tuberculosis in three sub-Saharan African
Countries. Lancet 338:1305-1308
Nault P, Phypers M, Deeks S (2002) Tuberculosis in Canada
2000: pre-release. Health Canada, Ottawa. Ref type: report
Pathania V, Almeida J, Kochi A (1997) TB patients and for-
profit health care providers in India. World Health Organi-
sation, Geneva
Pinho AMF, Santoro-Lopes G, Harrison LH, Schechter M
(2001) Chemoprophylaxis for tuberculosis and survival of
HIV - infected patients in Brazil. AIDS 15:2129-2135
Public Health Laboratory Service, United Kingdom (2002)
Tuberculosis Epidemiological Data for England and Wales.
Public Health Laboratory Service, London. www.phls.co.uk!
topics_az/tb/TB_tables_figures/. Accessed 28 Nov 2002
Reichler MR, Reves R, Bur S, Thompson V, Mangura BT, Ford J,
Valway SE, Onorato 1M (2002) Evaluation of investigations
conducted to detect and prevent transmission of tubercu-
losis. JAMA 287:991-995
Rieder HL (2002) Interventions for tuberculosis control and
elimination. International Union Against Tuberculosis and
Lung Disease, Paris
Riley R, Mills C, O'Grady F (1962) Infectiousness of air from a
tuberculosis ward - ultraviolet irradiation of infected air:
comparative infectiousness of different patients. Am Rev
Respir Dis 84:511-525
Rivest P, Tannenbaum T, Bedard L (1998) Epidemiology of
tuberculosis in Montreal. CMAJ 158:605-609
Saraiya M, Cookson ST, Tribble P, Silk B, Cass R, Poonja S, Wait-
ing M, Howland N, Paz EA, Cochran J, Moser KS, Oxtoby
MJ, Binkin NJ (2002) Tuberculosis screening among for-
eign-born persons applying for permanent US residence.
Am J Publ Health 92:826-829
Sawert H, Sukhontha K, Payanandana V, Akarasewi P, Nunn
PP, Raviglione MC (1997) Costs and benefits of improving
tuberculosis control: the case of Thailand. Soc Sci Med 44:
1805-1816
Schwartzman K, Menzies D (1999) Tuberculosis: 11. Nosoco-
mial disease. CMAJ 161:1271-1277
Schwartzman K, Menzies D (2000) Tuberculosis screening of
immigrants to low-prevalence countries. A cost-effective-
ness analysis: Am J Respir Crit Care Med 161:780-789
Selwyn PA, Hartel D, Lewis VA, Schoenbaum EE, Vermund SH,
Klen RS, Walker AT, Friedland GH (1989) A prospective
study of the risk of tuberculosis among intravenous drug
880
users with human immunodeficiency virus infection. N
Engl J Med 320:545-550
Shilova MV (2001) Specific features of the spread of tubercu-
losis in Russia at the end of the 20th century. Ann NY Acad
Sci 953:124-132
Sonnenberg P, Murray J, Glynn JR, Shearer S, Kambashi B,
Godfrey-Faussett P (2001) HIV-l and recurrence, relapse,
and reinfection of tuberculosis after cure: a cohort study in
South African mineworkers. Lancet 358:1687-1693
Styblo K (1985) The relationship between the risk of tubercu-
losis infection and the risk of developing infectious tuber-
culosis. Bull Int Union Against Tuberc 60:117-119
Suarez PG, Watt CJ, Alarcon E, Portocarrero J, Zavala 0, Cana-
les R, Luelmo F, Espinal MA, Dye C (2001) The dynamics
of tuberculosis in response to 10 years of intensive control
effort in Peru. J Infect Dis 184:473-478
Suarez PG, Floyd K, Portocarrero J, Alarcon E, Rapiti E, Ramos
G, Bonilla C, Sabogal I, Aranda I, Dye C, Raviglione MC,
Espinal MA (2002) Feasibility and cost-effectiveness of
standardised second-line drug treatment for chronic tuber-
culosis patients: a national cohort study in Peru. Lancet 359:
1980-1989
Sumartojo EM, Geiter LJ, Miller B, Hale E (1997) Can physi-
cians treat tuberculosis? Report on a National Survey of
Physician Practices. Am J Publ Health 87:2008-2011
TB Alliance (2002) The global alliance for TB drug develop-
ment. www.tballiance.org. Accessed 28 Nov 2002. Ref type:
pamphlet
Tupasi TE, Radhakrishna S, Co VM, Villa MLA, Quelapio MID,
Mangubat NV, Sarol IN, Rivera AB, Pascula MLG, Reyes
AC, Sarmeinto A, Solon M, Solon FS, Burton L, Mantala
MJ (2000) Bacillary disease and health seeking behavior
among Filipinos with symptoms of tuberculosis: implica-
tions for control. Int J Tuberc Lung Dis 4:1126-1132
K. Schwartzman
UNAIDS (2002) Joint United Nations Programme on HIV
AIDS. Report on the global HIV I AIDS epidemic. UNAIDS,
Geneva
United States Institute of Medicine (2000) Committee on the
Elimination of Tuberculosis in the United States, Division of
Health Promotion and Disease Prevention. Ending Neglect:
The Elimination of Tuberculosis in the United States (Geiter
LJ ed). National Academies Press, Washington DC
Uplekar M (2002) An expanded DOTS framework for effective
tuberculosis control. World Health Organization, Geneva
Uplekar M, Pathania V, Raviglione MC (2001) Involving private
practitioners in tuberculosis control: issues, interventions,
and emerging policy framework. World Health Organiza-
tion, Geneva
Vaccine Assessment and Monitoring (2002) WHO vaccine-
preventable diseases: monitoring system. 2002 global
summary. World Health Organization, Geneva. Available at
www.who.int/vaccines-documentsl
Volmink J, Matchaba P, Garner P (2000) Directly observed
therapy and treatment adherence. Lancet 355:1345-1350
Walley JD, Khan MA, Newell IN, Khan MH (2002) Effective-
ness of the direct observation component of DOTS for
tuberculosis: a randomised controlled trial in Pakistan.
Lancet 357:664-669
Wells CD, Zuber PLF, Nolan CM, Binkin NJ, Goldberg SV
(1997) Tuberculosis prevention among foreign-born per-
sons in Seattle-King County, Washington. Am J Respir Crit
Care Med 156:573-577
World Health Organization (2002) BCG vaccine factsheet. World
Health Organization, Geneva. www.who.inUvaccineslen/
tuberculosis.shtrnl. Accessed 28 Nov 2002. Ref type: pamphlet
Zuber PLF, McKenna MT, Binkin NJ, Onorato 1M, Castro KG
(1997) Long-term risk of tuberculosis among foreign-born
persons in the United States. JAMA 278:304-307
49 BeG and New Tuberculosis Vaccines
ZHOU XING
CONTENTS
49.1 Introduction 881
49.2 BCG Vaccine 881
49.3 Major Obstacles to Successful Development
of Improved TB Vaccines 883
49.4 Immunological Considerations 883
49.5 Selection of Immunogenic M. tuberculosis Antigens
for Recombinant TB Vaccine Development 885
49.6 Selection of Cytokines As Immune Adjuvants
for TB Vaccine Formulations 886
49.7 New TB Vaccines 887
49.7.1 Mycobacterial-based Vaccines 887
49.7.2 Subunit-based Vaccines 888
49.7.3 Plasmid DNA-based Vaccines 889
49.7.4 Viral-based Vaccines 889
49.8 TB Vaccine Target Populations and Potential Need
for Different TB Vaccine Formulations 890
References 891
49.1
Introduction
Tuberculosis (TB) remains a global epidemic. TB
is primarily a pulmonary infectious disease and it
is caused by exposure to airborne M. tuberculosis
bacilli. Approximately one third of the world popu-
lation has been infected by M. tuberculosis and 10%
of these people may develop active TB at some point
of their lives when their host defense is weakened.
Thus, each year eight million people develop TB
worldwide and two million of people die from the
disease (Kaufmann 2001). The risk of developing
active TB in HIV-infected patients is 30-fold greater
than HIV-negative patients. More than 10 million
people are co-infected with both M. tuberculosis
Z.XING,MD
Associate Prof. Pathology & Molecular Medidine, Head, Divi-
sion of Infectious Diseases, Centre of Gene Therapeutics,
McMaster University Health Sciences, 1200 Main Street W,
Hamilton, Ontario L8N 3ZS, Canada
M. Monir Madkour et al. (eds.), Tuberculosis
© Springer-Verlag Berlin Heidelberg 2004
and HIV worldwide and at least a half million of
these people die from TB each year, thus bringing
the annual TB-related death toll to 2.5 million. The
increasing emergence of multi-drug-resistant (MDR)
strains of M. tuberculosis further compounds the TB
epidemic, as a result of poor compliance with anti-TB
therapeutic regimens (50 million people have been
infected with MDR-TB). Normally, a combination of
several anti-TB antibiotics needs to be taken daily
for a minimum of 6-12 months. A similar regimen
is also recommended for preventive chemotherapy
in TB-infected HIV patients who have not yet devel-
oped active TB. It is noteworthy that the difficulty
in completing such lengthy regimens is encountered
widely, not only in the developing countries, but also
in disadvantaged groups in the developed countries
(Kaufmann 2001; Ginsberg 2000).
49.2
BCGVaccine
Currently, the only TB vaccine in use is BCG, an
attenuated strain of M. bovis, which has been used
for 80 years and is now the most widely administered
vaccine in the world with a global coverage of more
than 80% of world population (WHO 2000). There
are a small number of countries including the USA,
Canada, Australia and some developed European
countries where BCG has not been a part of their
national immunization programs for a long time;
some other countries that have recently ceased the
use of BCG on a national scale. Regardless of the
substantial global coverage of BCG vaccination, the
current ongoing TB epidemic indicates the inef-
fectiveness of the BCG vaccine in controlling this
disease. It has been estimated that BCG prevents
only 5% of all vaccine-preventable deaths by TB.
There has been much controversy surrounding the
widely varying results obtained from a number of
BCG immunization trials (0-77% protection rate)
(Roche et al. 1995).
882 Z.Xing

While they remain largely to be verified, many
reasons may have contributed to the ineffectiveness
of the BCG vaccine (Xing 2001a). These include:
1) M. tuberculosisand M. bovis are different, about
100 genes were found to be missing from all the
strains of BCG including the gene encoding an
immunogenic antigen ESAT-6. (Immunogenic
MPT-64 is also missing in a large number of strains
of BCG). 2) Many varying immunization schedules
are used in different countries. These range from
one immunization to four immunization protocols,
spanning from the time of birth to adolescent ages
(Table 49.1) (WHO 2000). There is no scientific
justification for these variations. It is also unclear
whether the intradermal route of immunization is
better than paracutaneous route or vice versa. 3)
Differences in the host genetic background, nutri-
tional levels and M. tuberculosis strains used may
significantly affect, or obscure, the relative efficacy
of BCG. Regardless of the genes missing in the BCG
genome, inclusion of an immune adjuvant formula-
tion to BCG vaccine may help overwrite the differ-
ence in immunogenicity of the BCG vaccine among
genetically diverse human populations. 4) BCG does
not confer a long-lasting immunity which wanes in
about 10-15 years. There is evidence that in some
countries the immunity may wane even faster. This
further favours the use of an immune adjuvant
together with BCG and the implementation of a vac-
cination regimen that involves repeated immuniza-
tions with a mycobacterial, or recombinant, booster.
Recent experimental evidence suggests that pre-
existing anti-mycobacterial immunity decreases
the ability of BCG to replicate upon inoculation.
This therefore favours the use of a non-mycobac-
terial-based booster. 5) The immune history of
BCG vaccinated individuals may also affect the
immunogenicity of BCG vaccine. This represents
a very understudied area. Conceivably, infection
(heterologous infection) that occurs before or after
BCG vaccination will influence either negatively,
or positively, the longevity and quality of type 1
immune responses triggered by BCG vaccine. Such
infections could be parasitic, or viral, in nature, and
are particularly prevalent in developing countries.
The nature of such heterologous infection perhaps
constitutes an important mechanism. A type 2 infec-
tion may weaken the type 1 immunity triggered by
BCG. However, a type 1 infection may also nega-
tively affect BCG-mediated immunity because of
immune competition and suppression. 6) Although
BCG is not recommended for use in symptomatic
HIV-infected patients due to safety concerns, it is
still being given to potentially HIV-infected infants
in many countries. As the HIV epidemic worsens,
it is likely that more people will not receive BCG
immunization. It is believed that if the current BCG
vaccine is ineffective in protecting from adult TB, its
immunogenicity will be even worse in patients with
varying degrees of CD4 T-cell impairment.
Therefore, we need improved TB vaccines. The
development of such vaccines represents a daunt-
ing challenge to the TB vaccine research community
and requires a long-term global commitment , by
scientists, governments, healthcare personnel and
commercial partners. The paradox is that the devel-
oped countries usually have a much lower incidence
of active TB and yet naturally have much greater
resources for TB vaccine development than under-
developed countries. Thus, some of these countries
often demonstrate apathy towards TB vaccine pro-
grams. The perception that TB may be stopped at the
border must be abandoned. It would be ludicrous ifwe
ignore that in the modern world infectious diseases
travel across borders. On the optimistic side, there is
no better time to push forward TB vaccine research
programs: the entire genomic sequence of M. tuber-
culosis (or BCG) is now available; the WHO has been
reinforcing the global TB prevention/treatment pro-
grams; a Blueprint for TB Vaccine Development was
released by the NIH (US) together with the Advisory
Committee for Elimination of TB and the National
Vaccine Program Office; the European Commis-
sion, the NIH and private foundations including the
Sequella Global Tuberculosis Foundation have esca-
lated their support for TB vaccine development; and

Table 49.1. Varying BeG immunization regimens adopted by different countries

One immunization Two immunizations Three immunizations Four immunizations

Birth birth/6-7 years Birth/6-7 year/ll-12 year birthl2 years/6-7 years/14-1syears

15 day birth/1l-12 years Birth/ll month/6-7 years
1 month birthl7 years (if no scar) Birth/8 y/1s years
3 months birth/14 month 1 month/1 year/12 years
<1 year
From WHO Vaccine Preventable Disease Monitoring System (2000) Global summary.
BeG and New Tuberculosis Vaccines
a U.S. Millennium Vaccine Initiative was announced
to boost vaccine development for HIV, malaria and
TB (Ginsberg 2000).
49.4
49.3 Immunological Considerations
Major Obstacles to Successful Development
of Improved T8 Vaccines M. tuberculosis is an intracellular pathogen of mac-
While the global effort in TB vaccine development
has escalated remarkably in the past decade, we are
still facing many practical and scientific hurdles
(Xing 2001a; Orme 1999). 1) Almost all of the cur-
rent vaccines in use provide protection through
antigen-specific antibodies. We have relatively little
experience in the rational design of vaccines capable
of adequately stimulating T-cell-mediated immunity.
2) An improved TB vaccine may require an immune
adjuvant (s) and this may be true for both mycobacte-
rial-based and recombinant forms of a vaccine. The
current adjuvant approved for human use is alu-
minium that is inappropriate for TB immunization.
The potential immune adjuvant may be chemical, or
biological, molecules including immune modulatory
cytokine proteins/transgenes which can be mixed
and delivered together with TB vaccine. The bacte-
rial plasmid or viral backbone of a recombinant TB
vaccine serves as immune adjuvant. 3) The current
BCG vaccine has been used for 80 years and given
to infants in most countries. A new, or improved, TB
vaccine has to be proven to be more effective than
BCG. Ideally this should be both in terms of the level
and duration of anti-TB immunity before it replaces
the current BCG. It may take decades before this can
happen worldwide. It is possible that the current BCG
vaccine will continue to be used but a boosting regi-
men will be introduced. However, the nature of such
a booster vaccine, and the regimen, remain yet to be
established. 4) A total of 1/ 3 of the world population
has been infected with M. tuberculosis. Vaccination
strategies must be developed to prevent active TB not
only in infants but also in populations with latent TB.
While considerable experience has been accumulated
through administering BCG to infants, little is known
about how to effectively immunize infected patients
with, or without, previous BCG immunization. 5)
The HIV epidemic and poor compliance with anti-
TB chemotherapy cause an ever-increasing number
of drug-resistant strains of M.tuberculosis. To effec-
tively control the TB epidemic, there is a need to
develop therapeutic vaccines that can be used as
adjuncts to anti-TB chemotherapy. Such therapeutic
883
vaccines may not only shorten the course of chemo-
therapy but also help combat drug-resistant TB.
rophages. Macrophages take up mycobacteria via
the surface complement and mannose receptors and
'toll-like' receptors 2/6. It is believed that alveolar
macrophages are the primary cells within the airway
space (including alveoli) that phagocytose mycobac-
teria. However, it is known that there are abundant
dendritic cells within the airway epithelium and
the bronchial-associated lymphoid tissue (Schon-
Hegrad et al. 1991); these cells could also phagocytose
mycobacteria as bacteria cross the airway epithelium
(Henderson et al. 1997). The extent to which myco-
bacteria enter the airway tissue before phagocytosis
by macrophages is unclear. Presumably in the initial
stage of infection (by a small number of mycobacte-
rial bacilli) only a few bacilli will be phagocytosed by
the dendritic cells. However, as the infection evolves,
and if the host immunity fails to contain infection, the
balance may shift and free mycobacteria may travel
across the epithelial barrier and subsequently may
be phagocytosed by dendritic cells residing within,
and beneath, the epithelium. Infected dendritic cells
mature and are activated to release, or express, a
number of immune modulatory molecules on their
surface. Unfortunately, relatively little is known
regarding the functional connection and difference
between alveolar macrophages and dendritic cells
upon mycobacterial infection. Phagocytosed myco-
bacteria reside in the phagosomes, or endocytic vac-
uoles, of macrophages and evade the anti-bacterial
activities through preventing phagosome-lysosome
fusion and acidification (Hingley-Wilson et al. 2000).
Cytokine interferon (IFN)-gamma-activated macro-
phages acquire phagolysosomal endocytic vacuoles
of lower pH, mycobacteria and MHC II molecules
(Schaible et al. 1998). How exactly the host eventu-
ally controls the intracellular replication of mycobac-
teria and eradicates these pathogens, is still unclear.
Nitric oxide seems important in this process. While it
is believed that macrophages are important effectors
with direct killing capacity, both T and NK cells are
able to lyse infected macrophages and directly kill
mycobacteria in vitro (Xing 2001a).
The initial T-cell activation during primary TB
infection, or after vaccination, is believed to take place
884
in the local draining lymph nodes. Such activation
requires the interaction between antigen-presenting
cells including both dendritic cells and macrophages,
and antigen-specific T cells. Both CD4 and CD8 T
cells are activated through MHC class II and MHC
class I-restricted pathways, respectively. Exogenous
antigens can also be processed and presented via the
MHC I pathway by a mechanism called 'acrosspre-
sentation' (Albert et al. 1998). It is not yet entirely
clear how certain antigens leak from phagosomal,
or endocytic vacuoles, into the cytosol where they
are degraded into peptides (in the proteosomes) and
are subsequently transported into the endoplasmic
reticulum and loaded onto MHC class I molecules by
the TAP heterodimer (composed of type 1 and type
2 transporter associated antigen processing [TAP]).
Activation of CD8 T cells following intramuscular
vaccination with recombinant DNA-based TB vac-
cines is the best support for this mechanism (Seder
and Hill 2000). In this case, transfected myoblasts
express, and release, an M. tuberculosis antigen which
is taken up by antigen-presenting cells. antigen-spe-
cific T cells activated in this way, or those that picked
up apoptotic myocytes, migrate to the lymph nodes
to activate both CD4 and CD8 T cells.
Both antigen presentation by antigen-specific
T cells and the interaction of B7 molecule on APC
with CD28 on T cells are required for T-cell activa-
tion. In addition, type 1 cytokine interleukin-(IL)-
12 released from activated antigen-specific T cells
also plays a crucial role in the initiation of type 1
T-cell differentiation during primary mycobacterial
infection or vaccination (Wakeham et al. 1998; Xing
2001 b). Thus, adequate activation of antigen-specific
T cells is essential to antigen-specific T-cell expan-
sion and activation. Activated antigen-specific T cells
demonstrate higher levels of MHC class I and II and
co-stimulatory molecules including B7 molecules,
enhanced antigen presentation and IL-12 release. In
addition, antigen-specific T cells also release other
cytokines which participate in orchestrating optimal
immune responses. The quality and longevity of such
immune responses are dictated by a blend of these
immunologically active molecules, both soluble and
membrane-bound, present at the site of infection and
antigen presentation. The make-up of this blend of
molecules is determined by the nature and dose of
the pathogen, the character of the tissue site where
the pathogen enters, the host's genetic make-up and
the history of the host's immune system (which is
dependent on its exposure to various pathogens (het-
erologous infection), or other immunogens before or
after TB infection or vaccination) (Bentwich et al.
Z.Xing
1999). The use of immune adjuvant in TB vaccine
formulations may help enhance antigen-specific T
cells activation and antigen presentation and turn
the unfavourable tissue micro-environment into the
one that favours the activation of anti-TB type 1 T
cells (Xing 2001a; Matzinger 1998).
After activation the majority of T cells die and a
small fraction may become memory T cells and live
much longer (Seder and Hill 2000). Antigen-specific
memory T cells undergo quick expansion and acti-
vation upon re-exposure to M. tuberculosis or their
antigens. Both CD4 and CD8 T cell subsets and B
cells may assume the memory phenotype. Recent
evidence suggests that the gamma/delta T-cell subset
may also possess the memory T cell function after
BCG vaccination (Shen et al. 2002). Much less is
known about the mechanisms of the development
of anti-TB memory T cells after BCG immunization
or following a natural TB infection. It is possible that
a similar set of factors involved in the initial T-cell
activation are also important in the generation of TB
specific memory T cells. Experimental evidence sug-
gests that both CD4 and CD8 memory T cells can per-
sist in the absence of specific antigen presentation, or
in the absence of MHC class II and class I molecules.
This is in sharp contrast to the homeostasis of naive
T cells (Bevan and Goldrath 2000). However, the
immune quality of these memory T cells may have
been altered in the absence of continuous antigenic
stimulation and thus the immune quality does not
necessarily correlate with their number and longev-
ity (Kassiotis et al. 2002). The immune protection, if
any, triggered by BCG vaccination, diminishes within
10-15 years or even sooner in certain countries
(Orme 1999). This coincides with the loss ofPPD skin
reaction which wanes after 5-10 years aftre neonatal
immunization. Successful generation of long-term
memory type 1 T-cell responses is central to the
development of improved TB vaccines.
In summary, there are several important immu-
nological considerations that need to be taken into
account for the design of new TB vaccines (Xing
2001a): 1) adequate levels of exposure of host's
immune system to immunogenic TB antigens which
may ensure effective epitope display and antigen
presentation; 2) effective reciprocal co-stimulation
between antigen-presenting cells and T cells; and 3)
creation of a favourable tissue micro-environment
rich in pro-immune cytokines and other molecules.
Experimental evidence using recombinant BeG vac-
cine supports the notion that increased amounts
of a single immunogenic M.tuberculosis antigen
could make a big difference in immune protection
BeG and New Tuberculosis Vaccines
(Horwitz et al. 2000). Cytokines play an important
role in creating an environment favourable to the
generation of long-term memory T cells. T cells with
a memory phenotype are believed to have a more
rapid turn-over rate and are sensitive to the effect
of T-cell active cytokines. Certain cytokines such
as IL-15 can stimulate the expansion of CD8 T cells
of memory phenotype in the absence of a specific
antigen. Unfortunately, little is known about the
nature of such cytokines involved in the generation
of anti-TB memory T cells. It is likely that some of
these cytokines may overlap with those involved in
the initiation of primary immune responses against
mycobacterial infection. In this regard, IL-6 was
found to be required for the optimal level of primary
immune responses to mycobacterial infection but
not for the maintenance of memory immunity. On
the other hand, IL-12 was found to be required for
sustaining the response of Th1 cells to secondary
leishmania infection. The timing of the presence of a
given cytokine is also important. For instance, IL-2 is
a T-cell growth factor and when present at the time of
initiation of immune response, it enhances the level
of memory T-cell response whereas when present in
the maintenance stage, it downsizes the population
of memory T cells.
49.5
Selection of Immunogenic M. tuberculosis
Antigens for Recombinant 18 Vaccine
Development
Mycobacterium tuberculosis has approximately 4000
gene products and it is anticipated that more TB
antigens with previously unknown functions will
turn out to be immunogenic. A challenge to TB vac-
cine researchers is to find out whether these new
antigens are superior to other previously known
immunogenic antigens and to select a few of the best
candidates to be incorporated into recombinant TB
vaccines. At present, the majority of recombinant TB
vaccines only express a single TB antigen. This will
result in the activation of limited antigen-specific T
cell clones. Thus, concurrent expression of multiple
immunogenic antigens in a multivalent TB vaccine
will be highly desirable (Xing 2001a).
Criteria for Selection. There has not been a set of
standards developed to aid the selection of antigens.
However, several immunological aspects are worth-
while for consideration (Xing 200Ia):
885
1) Secretory antigens are better candidates for the
activation of both CD4 and CD8 T cells. This is
perhaps why up to now, the majority of immu-
nogenic TB antigens are those present in the
culture filtrate of M. .tuberculosis, or BCG, and live
organisms are superior to dead ones for immu-
nogenicity.
2) Dominant T-cell antigens ought to be those that
account for a major fraction (s) of TB-specific
T cells or memory-T cell clones found both in
infected humans and mice, and these include the
Ag85 complex antigens and ESAT-6. In humans,
these antigen-specific T cells may be detected in
hosts who were infected with M. tuberculosis but
have not developed TB and those who have active
TB.
3) Universal immunogenic antigens. Multiple epit-
opes may span the large portion of a given anti-
genic protein. Thus, antigen-presenting cells from
hosts of different HLA haplotypes will be able
to present one or more immunogenic epitopes
from the same antigen. M. tuberculosis Ag85A
(Smith et al. 2000), ESAT-6 (Mustafa et al. 2000a),
Ag85B(Mustafa 2000), MPT64 (Mustafa 2000) and
hsp70 all fall into this group of antigens.
4) Promiscuous immunogenic epitopes. An epitope
of a given antigen may be presented by APC of
mismatching haplotypes. Ag85A, Ag85B, ESAT-6
and MPT64 possess multiple epitopes recogniz-
able by T cells in a HLA-non-restricted manner.
Such promiscuity of M. tuberculosis epitopes was
also observed in rodents (Vordermeier et a1.1994).
In contrast, each epitope of hsp70 is presented
only by one particular HLA molecule (Oftung et
al. 1994).
5) Dominant T cell antigens that are released from
M. tuberculosis but are missing from BCG vac-
cines, which include ESAT-6 and MPT-64 (Behr et
aI. 1999). Among M. tuberculosis antigens each of
which has been expressed in a DNA vaccine, are
Ag85A, Ag85B, Ag85C, ESAT-6, hsp70 (or hsp65
from M. leprae), Pst-I, Pst-2, PstS-3, KatG, HBHA,
MPT-63 or MPT-64, MPT83. While Ag85C, Pst-
1, Pst-2 are weak immunogenic antigens, other
antigens expressed in DNA-based vaccines have
been shown to be immune-protective to varying
degrees in mouse models of T8. Among all of the
tested immunogenic antigens, Ag85A or Ag85B,
ESAT-6 or hsp65 appears superior to the others.
The M. TB Ag85 Complex (30-32 kDa). This group of
outer wall antigens include Ag85A, Band C. They are
the major secreted proteins in M. tuberculosis or BCG
886
culture filtrate. These antigens are highly conserved
among all mycobacterial species. Their function is
to bind to fibronectin and thus may be involved in
the macrophage phagocytic process. In experimental
models,Ag85A and Bare among a handful of memory
T cell-stimulating antigens. These antigens stimulate
a strong CD4 and CD8 T cell response both in humans
and experimental animals (Smith et al. 2000). Ag85B
(MPT59) contains multiple human T-cell epitopes
that scatter throughout the entire protein and sev-
eral epitopes are promiscuous T-cell epitopes and
thus could stimulate T cells by HLA-DR-mismatched
antigen-presenting cells (Mustafa et al. 2000b). Thus,
at least one of Ag85A and B antigens should be incor-
porated into a multivalent TB vaccine.
ESAT-6 (6 kDa). This is a small molecular weight,
secreted antigen. The gene coding for ESAT-6 is not
present in any BCG strain although it is in virulent
M. bovis and M. tuberculosis. Of importance, a large
proportion of memory T cells in M. tuberculosis-
infected mice are ESAT-6-reactive (Andersen et al.
1995). Furthermore, this antigen also stimulates a
strong human T-cell response and similar to Ag85,
contains multiple T-cell epitopes that span the entire
ESAT-6 protein (Mustafa et al. 2000a). Thus, ESAT-6
is a dominant human T-cell antigen which could be
presented by antigen-presenting cells of a wide spec-
trum of HLA-DR haplotypes. This antigen warrants
serious consideration for a multivalent formulation
of TB vaccine.
MPT64 (26 kDa). This antigen is also among the
major secreted proteins. The gene coding for this
protein is present only in some BCG strains (Tokyo,
Russia, Sweden and Morau) but not in other strains
(Danish 1331, Glaxo, Pasteur and Tice). MPT64 is
also one of human dominant T-cell antigens (Mus-
tafa 2000) and can be recognized by T cells from the
majority of TB patients. DNA vaccine expressing
MPT64 conferred protection in mouse models of
TB. Thus, this antigen is worthy of consideration for
recombinant TB vaccines.
hsp proteins (l0, 18, 60, 65, 70 kDa). This represents
a group of secreted heat shock proteins of diverse
molecular weights. They are widely conserved in
prokaryotes and eukaryotes, involved in the protein
folding, unfolding and assembly processes. These
proteins are induced both in infected antigen-pre-
senting cells and M. tuberculosis under any stressful
conditions including elevated temperature. The ini-
tial success in using DNA-based TB vaccines express-
Z. Xing
ing mycobacterial hsp to protect mice with TB infec-
tion has generated a considerable level of excitement
(Tascon et al. 1996). With regard to hsp70, multiple
human T-cell epitopes have been identified but each
epitope is presented only by antigen-presenting cells
of a particular HLA molecule, hence the lack of pro-
miscuity (Oftung et al. 1994). Due to the high homol-
ogy of these proteins with mammalian counterparts,
expression of hsp proteins in TB vaccines raises a
serious concern regarding their potential to trigger
an autoimmune response. Indeed, a recent study
demonstrates an autoimmune-like tissue response in
guinea pigs immunized with DNA vaccines express-
ing M. tuberculosis hsp antigens and subsequently
challenged by M. tuberculosis (Turner et al. 2000).
49.6
Selection of Cytokines As Immune
Adjuvants for TB Vaccine Formulations
Cytokines that can act at a point along the cascade
of anti-TB immunity may be used as an adjuvant
for both live mycobacterial-based and recombinant
forms of TB vaccines (Xing and Wang 2000; Xing
2001a). The rationale is to create a cytokine-rich
tissue microenvironment at the site of initial vac-
cine inoculation or antigen expression/presentation.
As such, antigen presentation may be enhanced via
increased recruitment and activation of antigen pre-
senting cells (APC) and increased epitope display on
APC, and more T cells may be recruited to the site
of immune activation. Improved antigen presenta-
tion will lead to activation of a larger repertoire of
T cells while activation of APC facilitates the type
1 polarization of these antigen specific T cells and
perhaps the generation of memory T cells.
Cytokines Active on Antigen-presenting Cells. These
include GM-CSF, IL-3, Flt-3 and MIP-3 alpha. GM-
CSF is a hematopoietic cytokine capable of multiple
functional activities with a prominent effect on the
activation of myeloid dendritic cells and macro-
phages (Xing et al.1996; Wang et al. 2000). Thus, GM-
CSF may enhance the immunogenicity of TB vaccine
by acting on the early events of immune cascade.
Indeed, by using a GM-CSF-expressing adenoviral
vector as an adjuvant, we were able to enhance the
immunogenicity of BCG vaccine in genetically weak
responders Balb/c mice (Wang et al. 2002). Flt-3 is
also a hemotopoietic cytokine capable of mobilizing
and activating dendritic cells. MIP-3 alpha is a che-
BeG and New Tuberculosis Vaccines
mokine chemo-attractive to both dendritic cells and
memory T cells (Cook et al. 2000).
Type 1 Differentiating Cytokines. IL-I2 is an indis-
pensable type I-differentiating cytokine, particularly
in host defense against non-viral intracellular infec-
tions (Xing 200Ib). The inclusion of this cytokine
or its transgene formulation has been shown to
enhance the immunogenicity of vaccines against
a number of intracellular infectious diseases in
experimental models. Recombinant IL-I2 increased
immune protection by BCG vaccination against M.
tuberculosis infection. We have recently found that
subcutaneous, or intranasal, co-administration of
an adenoviral vector expressing IL-I2 with BCG
enhances the immunogenicity of BCG (Xing 200Ia).
IL-I8 and IL-23 are also type I-like cytokines capa-
ble of potentiating a number of biological activities
mediated by IL-I2. IL-IS, however, differs from IL-I2
as it can be released by a variety of cell types and by
itself it cannot directly activate type I differentiation
since expression of the receptor for IL-IS on T cells
is dependent on the action of IL-I2.
Lymphocyte- or Memory T Cell-stimulating Cytokines.
These include IL-2, IFN-gamma and such cytokines as
IL-IS capable of stimulating memory CDS T cells (Ku
et al. 2000). IL-2 is a T-cell-proliferation growth factor
and has a dual role in the generation and maintenance
of CDS memory T cells (Dai et al. 2000). This cytokine
may not be an ideal adjuvant for TB vaccine due to its
notorious cytotoxicity. At the moment, relatively little
is known about the nature of the cytokines that are
important in the generation of anti-TB memory T
cells post-vaccination since the majority of cytokine
biological studies in the TB arena, or in any other
intracellular infection, have been carried out in models
of primary infection. However, evidence suggests that
the size of the initial antigen-specific effector pool is
proportionate to the size of memory T-cell population
(Seder and Hill 2000). Thus, conceptually, cytokines
that potently activate the development of primary T
cell activation will subsequently have a positive impact
on the memory T-cell pool.
Chemokines Active on Lymphocytes and Monocytesl
Macrophages. Lymphocyte chemo-attractive cyto-
kines include RANTES and lymphotactin. Many of
these chemokines are also active on monocytes/
macrophages (Hogaboam et al. 2000). These cyto-
kines may enhance the immunogenicity of TB vac-
cine by recruiting more antigen-specific T precursors
to the site of antigen presentation and creating a less
887
dangerous environment for strong type I differentia-
tion. MIP-3 alpha is a chemokine active on both den-
dritic cells and memory T cells (Cook et al. 2000).
Choice of Cytokine Formulations. Recombinant cyto-
kines are not only expensive but also short-lived in
vivo (Xing and Wang 2000).And the storage conditions
are tricky as well which will hinder their use in many
developing countries. The best strategy is to express
cytokine in vivo in the form of a transgene which is
not only economical but renders a sustained level of
cytokine in the vicinity of antigen delivery/expression.
Currently there are three strategies for consideration:
1) Engineered live organisms stably expressing a
cytokine. 2) Plasmid-DNA based Cytokines may be
expressed by a separate cassette or co-expressed in
the same TB-DNA vaccine. The latter is more logical
and practical. While this represents an attractive strat-
egy for DNA-based TB vaccination, it will be less ideal
for other formulations of TB vaccines since the DNA
vector is expressed best only in the muscle and this
will detract from the co-delivery when TB vaccines are
given mucosally or para-cutaneously. 3) Viral-based.
This strategy allows not only the efficient expression
of cytokine but flexible routes of administration. For
instance, we have shown that adenoviral gene trans-
fer vector efficiently transduce the transgene at the
muscle, intradermal or subcutaneous compartment,
or mucosa. Furthermore, viral cytokine vector may be
mixed with either live organism TB vaccine or viral TB
vaccine. The dose of such viral-based vectors can be
kept to a minimum since expression is highly efficient
and only a local effect is desired.
49.7
New T8 Vaccines
49.7.1
Mycobacterial-based Vaccines
Current BCG Vaccine and Vaccination Regimen. It
will be difficult to abandon the use of current BCG
on a global scale in the near future. It may be much
easier to change the regimen or to add a straightfor-
ward immune adjuvant to the current BCG vaccine.
Currently, BCG is being given paracutaneously in
humans. The potential advantage of the mucosal
route of immunization via the airway has not been
extensively studied. It is interesting that oral BCG
immunization does not seem to convert a PPD skin
test (Hoft et al. 2000). Furthermore, BCG immuniza-
888
tion by aerosolization in rhesus monkeys generated
minimal side effects and only a small skin reaction to
PPD while it protected against airway M. tuberculosis
infection (Barclay et al. 1973). Therefore, it is possible
that the benefit of respiratory mucosal immunization
may outweigh potential side effects.
The current BCG immunization schedules vary
widely across the world and there is a lack of logic
as to why BCG should be administered only once or
several times, or how long the intervals should be if
given several times. Studies are urgently needed to
investigate whether repeated BCG vaccinations help
maintain the memory response. If not, other booster
formulations need to be defined. Not only is BCG
vaccine ineffective in protecting humans, but also a
natural TB infection may not necessarily protect the
host from a second TB infection (van Rie et al. 1999).
This strongly suggests that the current BCG vaccine,
if it is to be continued to be used for the next several
decades, needs an adjuvant to heighten the level and
longevity of immunity.
Recombinant BCG. Plasmid DNAs containing trans-
genes encoding cytokines were introduced by stable
transformation into the BCG genome. Cytokines
expressed by such recombinant BCG organisms
include IL-2, IFN-gamma and GM-CSF (Murray et al.
1996). These cytokine-secreting BCG vaccines were
shown to enhance the immunogenicity of BCG in
mouse models. One of these vaccines secreting IL-2
was also recently evaluated in deer with no increased
immunogenicity. These vaccines have not yet been
extensively tested in TB challenge models. Whether
the random insertion of plasmid DNA into the BCG
genome will have any effect on the expression of domi-
nant mycobacterial antigens is unknown. It is also of
potential concern that cytokine release will continue
as long as the BCG lives and this may cause some
undesired effects. A M. tuberculosis Ag85B gene was
stably transformed into BCG and such recombinant
BCG vaccines were found to provide a better protec-
tion than BCG in a susceptible guinea pig model of
TB. Since M. tuberculosis Ag85B and BCG Ag85B differ
only in two contiguous amino acids, this finding sug-
gests that it is the level of this antigen that is crucial to
the difference in the level of immune protection.
Auxotrophic Mutants. Taking advantage of the fact
that mycobacteria depend on their own synthesis of
certain amino acids for replication, BeG or M. tuber-
culosis auxotrophs have been developed by deleting
the mycobacterial genes required for the biosynthesis
of these amino acids. As a result, the mutated organ-
Z.Xing
isms cannot replicate, and survive, like their wild
type counterparts and will persist only long enough
to activate immune components. Some of these auxo-
trophs have been shown to confer a level of immune
protection similar to the wild-type BCG in mice and
guinea pigs (Guleria et al.1996).At the moment, these
live, replication-deficient mutants represent the best
candidate vaccines to be used in immuno-compro-
mised hosts. However, little is known about whether
any live mycobacterial organism-based vaccine will
be effective in immune-compromised hosts such as
HIV-infected patients who suffer impaired CD4 T cell
functions. Basic research and assessment are urgently
needed in this regard. The childhood immunization
with BCG seems to reduce the risk of disseminated TB
in HIV-infected adults (Marsh et al. 1997). A vaccine
may be needed for the adults who suffer HIV infec-
tion with varying degrees of CD4 T-cell deficiency
but who were never immunized against TB. In this
case, even with an auxotrophic form of TB vaccine, an
adjuvant formulation will be highly desired to target
CD8 T cell-based immune mechanisms.
Rapid-growing Mycobacteria. Fast-growing myco-
bacterial species including M. smegmatis and M.
vaccae are among potential TB vaccine candidates.
The concept of using a strain of mycobacterium
cross-reactive to the pathogenic one in order to
overcome the unresponsive state of immune system,
is supported by recent successful immunotherapy for
lepromatous leprosy (Talwar 1999). Heat-inactivated
M. vaccae appeared to provide some immune protec-
tion against M. tuberculosis challenge in experimental
models. It also seems immunogenic in HIV-infected
humans. However, a recent clinical trial indicates that
when used as a therapeutic vaccine in patients with
TB, it provides no clinical benefit (Durban Immuno-
therapy Trial Group 1999). While the fast-growing
mycobacterial species share some immunogenic
antigens with BCG and M. tuberculosis, it is hard to
conceive that they will be any better than BCG if no
adjuvant is used. Moreover, use of dead M. vaccae is
unlikely to be advantageous over auxotrophic BCG
or M. tuberculosis. It is worth noting that the NIH
has supported Phase lIII trials involving the use of
inactivated M. vaccae and recombinant IL-2.
49.7.2
Subunit-based Vaccines
These vaccines consist of defined secreted M. tuber-
culosis proteins/peptides or simply the entire reper-
BeG and New Tuberculosis Vaccines
toire of unfractionated M. tuberculosis culture filtrate
proteins (Orme 1999). The safety is a prominent fea-
ture of these vaccines. Furthermore, subunit-based
vaccines do not seem to convert skin PPD reaction.
However, the major limitations to the application of
such vaccines include: 1) an adjuvant formulation
such as incomplete Freund's adjuvant (IFA), dimethyl
dioctadecyl ammoniumbromide (DDA) and Quil-A
saponin, and multiple doses are required for immu-
nogenicity; 2) in general these vaccines are poor
stimulators of CD8 T cells; and 3) they are relatively
expensive to produce. Furthermore, there has been
a lack of comparative studies addressing the poten-
tial difference in antigen presentation, T-cell subset
activation and the longevity of immune activation
between these vaccines and live organism-based or
recombinant vector-based vaccines. Experimental
studies demonstrate that protein-based TB vaccines
can protect mice, or guinea pigs, from M. tuberculo-
sis challenge but the protection level generally is not
greater than BCG. Notwithstanding, the EAST-6 sub-
unit vaccine, with DDA and MPL (monophosphoryl
lipid A) as an adjuvant, elicited a level of protection
against M. tuberculosis comparable to that of BCG
(Brandt et al. 2000).
49.7.3
Plasmid DNA-based Vaccines
Bacterial plasmid DNA is rich in CpG motifs which
directly or indirectly activate the innate and adap-
tive immune components and thus have a potent
immune adjuvant effect in mammalian organisms.
The CpG motif consists of an unmethylated cytidine-
phosphate-guanosine dinucleotide with appropriate
flanking regions (Gurunathan et al. 2000). Although
non-living, plasmid DNA vectors allow the expres-
sion of foreign transgene(s) under appropriate
control and subsequent production of the transgene
protein by transfected cells in vivo - a process simi-
lar to that by live intracellular infection. As a result,
plasmid DNA-based vaccines trigger both humoral
and cell-mediated immune responses by activating B,
CD4 and CD8 T cells, without the need for additional
adjuvant. DNA-based TB vaccines, similar to subunit-
based vaccines, do not convert the skin PPD reaction.
Compared to such viral vectors as adenoviral vector,
plasmid DNA vectors elicit lower levels of expression
due to Iowa transfection efficiency. Human clinical
trials are underway to test the safety and efficacy of
DNA-based vaccines. Although the effect of DNA vac-
cination has not been reported, there have been suc-
889
cessful uses of DNA-based vector for topical cytokine
delivery in treating cardiovascular diseases.
The transgene coding for a number of immu-
nogenic M. tuberculosis antigens including Ag85A,
Ag85B, ESAT-6, MPT64, PstS-1I2/3, KatG, hsp65 and
hsp70 have been expressed in plasmid DNA vector.
These DNA vaccines, when used individually, or in
combination, have demonstrated impressive immu-
nogenicity in mice but the level of conferred immune
protection against subsequent M. tuberculosis infec-
tion hardly exceeds that by BCG (Tascon et al. 1996).
A DNA vaccine expressing Ag85A was shown to elicit
a stronger and broader CD8 T cell response specific to
Ag85A than a natural M. tuberculosis infection (Denis
et al.1998). Secreted forms of these TB antigens from a
plasmid DNA vector appear to elicit a better immune
response than non-secreted forms. The best route of
DNA-TB vaccination is intramuscular injection. It is
worth noting that these DNA vaccines tend to work
effectively in mice while they are much less effective
in guinea pigs. This finding casts a shadow over their
effectiveness in humans. In addition to a prophylactic
effect, some of DNA-TB vaccines (but not a single
dose of BCG) have been proven therapeutic in mouse
models of primary i.v. M. tuberculosis infection and
reactivation (Lowrie et al. 1999). This represents an
important step forward in rationalizing the use of
DNA-based TB vaccines in humans.
49.7.4
Viral-based Vaccines
Vaccinia Virus. Vaccinia virus is a commonly used pox-
viral vector. Vaccinia virus replicates in the cytoplasm
and is normally highly toxic to the infected cells. Since
only poxviral, but not strong viral, promoters, can be
used for the expression of heterologous genes in poxvi-
ral vector (Hitt and Gauldie 2000), transgene expression
is usually very low. Like adenoviral-based vectors, the
vaccinia viral genome does not integrate into the host
genome. Recombinant vaccinia viral vector can harbor
up to 25 kb of heterologous transgene sequences.
Poxvirus-based vaccines have been widely used to
immunize cattIe, chickens, raccoons and foxes against
viral infections. Recent human clinical trials using vac-
cinia vector expressing immune-modulatory cytokine,
or HIV envelope proteins, have provided promising
results. Vaccinia virus vector has also been exploited to
express mycobacterial antigens and demonstrated cer-
tain protective effects in mouse models (Zhu et al.1997).
However, this vector is unlikely be a major platform
for a TB vaccine due to low expression of TB antigens.
890
Notwithstanding, this vector may be potentially useful
as a boosting vaccine following BCG, or DNA-based
priming, vaccination (McShane et al. 2001).
Adenovirus. The potential for adenovirus for gene
therapy, or gene transfer, both in experimental animals
and human trials has widely been explored (Hitt and
Gauldie 2000). Experimental evidence also supports
the potential use of adenovirus-based vaccines for pre-
venting infectious diseases (Imler 1995). However, this
viral vector has not been explored for TB vaccination.
We have initiated a program to develop adenoviral-
based recombinant TB vaccines and this program is
built upon our expertise in study of adeno-vectors and
gene transfer (Hitt and Gauldie 2000; Xing et al. 1996).
Our initial results indicate that adenoviral TB vaccine
hold great promise to be a mainstay, or booster, TB
vaccine platform (Xing 2001a).
Recombinant adenoviral-based vaccines are effec-
tive in protecting against several infectious diseases in
animals including chimpanzees. Although originally
designed for treating genetic disease, the adenoviral
vector, when used as a vaccine, is associated with sev-
eral advantages:
1) Ample safety data are available from well-docu-
mented human immunization programs (Gurwith
et al. 1989) and the adenoviral genome does not
integrate into the host genome and thus does not
cause mutagenesis;
2) Adenoviral vector has been genetically rendered
unable to replicate, hence it is safe;
3) Compared to other vectors including plasmid
DNA, the adenoviral vector is the most efficient
in driving transgene expression for 2-3 weeks;
4) Adenovirus has a broad range of target cells in
vivo. This allows the use of adenoviral vaccine at
a preferred tissue site of choice;
5) Adenoviral vector may harbor up to 8.3 kb of
foreign DNA sequences. This allows expression
of multiple TB antigens in one vaccine (multi-
valent);
6) Adenoviral vector per se is a type 1 immune adju-
vant; and
7) Evidence obtained from clinical trials by us, and
others, suggests that adenoviral gene transfer is
feasible in spite of pre-existing anti-adenoviral
antibodies in some subjects. Adenoviral-based TB
vaccines have the potential to be used not only
to immunize newborns but to boost the immune
response in BCG vaccinated patients who have
been infected with M. tuberculosis. Compared to
BCG, they will be a safer vaccine for potential use
in HIV-infected hosts.
Z.Xing
49.8
T8 Vaccine Target Populations and Potential
Need for Different T8 Vaccine Formulations
In order to effectively control the current global TB
epidemic, there are several human groups who may
need TB vaccine of different forms (Xing 2001a):
1) Neonates who need to be protected from TB
immediately after birth;
2) BCG-immunized adults who may have been
infected with M. tuberculosis and are no longer
protected by the initial BCG vaccination in adult-
hood;
3) Non-immunized populations who were infected
with M. tuberculosis in a developed nation;
4) High risk populations who were never immunized
with BCG including health workers, the elderly,
the homeless and prisoners in the developed
countries;
5) BCG immunized or non-immunized immune-
compromised individuasl; and
6) TB patients. The vaccine designed for the immu-
nization of the newborn has to meet the conven-
tional expectations for a vaccine used on a global
scale (simple formulation and regimen, easy to
produce, transport, store and administer, and
cheap). The vaccine to be used for people who
have previously been immunized with BCG and
infected by M. tuberculosis, may not be the same
vaccine used for newborns. Such a vaccine should
prove effective in previously immunized hosts
and it may be a BCG vaccine with an adjuvant or
a recombinant viral TB vaccine. The vaccine used
for non-immunized but TB-infected patients, and
non-immunized high-risk patients, are mostly in
the developed countries where BCG is not part of
immunization program. This vaccine may be the
same vaccine used for newborns if it proves effec-
tive in TB-infected hosts, or it may be a strength-
ened formulation that is different from the one
used for newborns. The vaccine designed for
use in HIV patients has to be the one tailored to
target CD8 and other subset T cells. Such a vaccine
ought to be safe and is likely to require a potent
adjuvant formulation. The vaccine designed as an
immune therapeutic adjunct to chemotherapeu-
tics in patients with active TB, particularly drug-
resistant TB, is likely different from the one used
in newborns and could also be different from the
one designed for HIV patients. This vaccine ought
to be able to overwrite the anergy that allows the
development of TB after infection both in BCG-
immunized and non-immunized individuals.
BCG and New Tuberculosis Vaccines
References
Albert ML, Sauter B, Bhardwaj N (l998) Dendritic cells acquire
antigen from apoptotic cells and induce classl-restricted
CTLs. Nature 392:86-89
Andersen P, Andersen AB, Sorensen AL, Nagai S (1995) Recall
of long-lived immunity to Mycobacterium tuberculosis
infection in mice. J ImmunoI154:3359-3372
Barclay WR, Busey WM, Dalgard DW, Good RC, Janicki BW,
Kasik JE, Ribi E, Ulrich CE, Wolinsky E (l973) Protection
of monkeys against airborne tuberculosis by aerosol vac-
cination with bacillus Calmette-Guerin. Am Rev Respir Dis
107:351-358
Behr MA, Wilson MA, Gill WP, Salaman H, Schoolnik GK,
Rane S, Small PM (l999) Comparative genomics of BCG
vaccines by whole-genome DNA microarray. Science 284:
1520-1523
Bentwich Z, Kalinkovich A, Weisman Z, Borkow G, Beyers N,
Beyers AD (1999) Can eradication of helminthic infec-
tions change the face of AIDS and tuberculosis? Immunol
Today.20:485-487
Bevan MJ, Goldrath AW (2000) T-cell memory: you must
remember this... Curr BiollO:R338-R340
Brandt L, Elhay M, Rosenkrands I, Lindblad EB, Andersen P
(2000) ESAT-6 subunit vaccination against Mycobacterium
tuberculosis. Infect Immun 68:791-795
Cook DN et al (2000) CCR6 mediates dendritic celliocaliza-
tion, lymphocyte homeostasis, and immune responses in
mucosal tissue. Immunity 12:495-503
Dai Z, Konieczny BT, Lakkis FG (2000) The dual role of IL-2 in
the generation and maintenance of CD8+memory Tcells. J
ImmunoI165:3031-3036
Denis 0, Tanghe A, Palfliet K, Jurion F, van den Berg TP,
Vanonckelen A, Ooms J, Saman E, Ulmer JB, Content J,
Huygen K (l998) Vaccination with plasmid DNA encoding
mycobacterial antigen 85A stimulates a CD4+ and CD8+
T-cell epitopic repertoire broader than that stimulated
by Mycobacterium tuberculosis H37Rv infection. Infect
Immun 66:1527-1533
Durban Immunotherapy Trial Group (1999) Lancet 354:116
Ginsberg AM (2000) A proposed national strategy for tuber-
culosis vaccine development. Clin Infect Dis 30 [Suppl):
S233-S242
Guleria I, Teitelbaum R, McAdam RA, Kalpana G, Jacobs WR
Jr, Bloom BR (l996) Auxotrophic vaccines for tuberculosis.
Nat Med 2:334-337
Gurunathan S, Klinman DM, Seder RA (2000) DNA vaccines:
immunology, application, and optimization. Annu Rev
ImmunoI18:927-974
Gurwith MJ, Horwith GS, Impellizzeri CA, Davis AR, Lubeck
MD, Hung PP (l989) Current use and future directions of
adenovirus vaccine. Semin Respir Infect 4:299-303
Henderson RA, Watkins SC, Flynn JL (l997) Activation of
human dendritic cells following infection with Mycobac-
terium tuberculosis. J ImmunoI159:635-643
Hitt M, Gauldie J (2000) Gene vectors for cytokine expression
in vivo. Curr Pharm Des 6:613-632
Hingley-Wilson SM, Sly LM, Reiner NE, McMaster WR (2000)
The immunobiology of the mycobacterial infected macro-
phagesModAsp. Immunobiology 1:96-101
Hogaboam CM, Bone-Larson C, Matsukawa A, Steinhauser
ML, Blease K, Lukacs NW, Kunkel SL (2000) Therapeutic
use of chemokines. Curr Pharm Des 6:651-663
891
Hoft DF,Brown RM,Belshe RB (2000) Mucosal bacille Calrnette-
Guerin vaccination of humans inhibits delayed-type hyper-
sensitivity to purified protein derivative but induces myco-
bacteria-specific interferon-gamma responses. Clin Infect
Dis 30 [SuppI3]:S217-S222
Horwitz MA, Harth G, Dillon BJ, Maslesa-Galic S (2000) Recom-
binant bacillus Calmette-Guerin (BCG) vaccines expressing
the Mycobacterium tuberculosis 30-kDa major secretory
protein induce greater protective immunity against tubercu-
losis than conventional BCG vaccines in a higWy susceptible
animal model. Proc Nat! Acad Sci USA 97:13853-13858
Imler JL (l995) Adenovirus vectors as recombinant viral vac-
cines. Vaccine 13:1143-1151
Kaufmann SH (2001) How can immunology contribute to the
control of tuberculosis? Nature Rev Immunol 1:20-30
Kossiotis G, Garcia S, Simpson E, Stockinger B (2002) Impair-
ment of immunological memory in the absence of MHC
despite survival of memory T cells. Nat ImmunoI3:244-250
Ku CC, Murakami M, Sakamoto A, Kappler J, Marrack P (2000)
Control of homeostasis of CD8+ memory T cells by oppos-
ing cytokines. Science 288:675-678
Lowrie DB, Tascon RE, Bonato VLD, Lima VMF, Faccioli LH,
Stalvropoulos E, Colston MJ, Hewinson RG, Moelling K,
Silva CL (l999) Therapy of tuberculosis in mice by DNA
vaccination. Nature 400:269-271
Marsh BJ, von Reyn CF, Edwards J, Ristola MA, Bartholomew
C, Brindle RJ, Gilks CF, Waddell R, Tosteson AN, Pelz R, Sox
CH, Frothingham R, Arbeit RD (l997) The risks and ben-
efits of childhood bacille Calmette-Guerin immunization
among adults with AIDS International MAC study groups.
AIDS 11 :669-672
Matzinger P (l998) An innate sense of danger. Semin Immunol
10:399-415
McShane H, Brookes R, Gilbert SC, Hill AV (2001) Enhanced
immunogenicity of CD4(+) t-cell responses and protective
efficacy of a DNA-modified vaccinia virus Ankara prime-
boost vaccination regimen for murine tuberculosis. Infect
Immun 69:681-686
Murray PJ, Aldovini A, Young RA (l996) Manipulation and
potentiation of antimycobacterial immunity using recom-
binant bacille Calmette-Guerin strains that secrete cyto-
kines. Proc Nat! Acad Sci USA 93:934-939
Mustafa AS (2000) HLA-restricted immune response to
mycobacterial antigens: relevance to vaccine design. Hum
ImmunoI61:166-171
Mustafa AS, Amoudy HA, Wiker HG, Abal AT, Ravn P, Oftung
F, Andersen P (1998) Comparison of antigen-specific T-cell
responses of tuberculosis patients using complex or single
antigens of Mycobacterium tuberculosis. Scand J Immunol
48:535-543
Mustafa AS, Oftung F, Amoudy HAl, Madi NM, Abal AT, Shaban
F, Krands IR, Andersen P (2000a) Multiple epitopes from
the Mycobacterium tuberculosis ESAT-6 antigen are rec-
ognized by antigen-specific human T cell lines. Clin Infect
Dis 30 [SuppI3):S201-S205
Mustafa AS, Shaban FA, Abal AT, Al-Attiyah R, Wiker HG,
Lundin KA, Oftung F, Huygen K (2000) Identification and
HLA restriction of naturally derived Thl-cell epitopes
from the secreted Mycobacterium tuberculosis antigen
85B recognized by antigen-specific human CD4(+) T-cell
lines. Infect Immun 68:3933-3940
892
Oftung F, Geluk A, Lundin KA, Meloen RH, Thole JER, Mustafa
AS, Ottenhoff THM (1994) Mapping of multiple HLA class
II-restricted T-cell epitopes of the mycobacterial70-kilo-
daltonheatshockprotein. Infect Immun 62:5411-5418
Orme 1M (1999) New vaccines against tuberculosis. The status
of current research. Infect Dis Clin North Am 13:169-185
Roche PW, Triccas JA, Winter N (1995) BCG vaccination
against tuberculosis: past disappointments and future
hopes. Trends Microbiol 3:397-401
Schaible UE, Sturgill-Koszycki S, Schlesinger PH, Russell
DG (1998) Cytokine activation leads to acidification and
increases maturation of Mycobacterium avium-contain-
ing phagosomes in murine macrophages. J Immunol160:
1290-1296
Schon-Hegrad MA, Oliver J, McMenamin PG, Holt PG (1991)
Studies on the density, distribution, and surface phenotype
of intraepithelial class II major histocompatibility complex
antigen (Ia)-bearing dendritic cells (DC) in the conducting
airways. J Exp Med 173:1345-1356
Seder RA, Hill AVS (2000) Vaccines against intracellular infec-
tions requiring cellular immunity. Nature 406:793-798
Shen Y, Zhou D, Qiu L, Lai X et al (2002) Adaptive immune
responses of Vgamma2Vdelta2 T cells during mycobacte-
rial infections. Science 295:2255-2258
Smith SM, Klein MR, Malin AS, Sillah J, Huygen K, Andersen
P, McAdam KPWJ, Dockrell HM (2000) Human CD8+ CTL
specific for the mycobacterial major secreted antigen 85A.
J ImmunoI165:7088-7095
Talwar GP (1999) An immunotherapeutic vaccine for multi-
bacillary leprosy. Int Rev ImmunoI18:229-249
Tascon RE, Jo Colston M, Ragno S, Stavropoulos E, Gregory D,
Lowrie DB (1996) Vaccination against tuberculosis by DNA
injection. Nat Med 2:888-892
Turner OC, Roberts AD, Frank AA, Phalen SW, McMurray
DM, Content J, Denis 0, D'Souza S, Tanghe A, Huygen K,
Orme 1M (2000) Lack of protection in mice and necrotizing
bronchointerstitial pneumonia with bronchiolitis in guinea
pigs immunized with vaccines directed against the hsp60
molecule of Mycobacterium tuberculosis. Infect Immun
68:3674-3679
Z.Xing
Van Rie A, Warren R, Richardson M, Victor TC, Gie RP, Enar-
son DA, Beyers N, van Heiden PD (1999) Exogenous rein-
fection as a cause of recurrent tuberculosis after curative
treatment. N Engl J Med 341:1174-1179
Vordermeier HM, Harris DP, Moreno C, Ivanyi J (1994) Pro-
miscuous T cell recognition of an H-2 lA-presented myco-
bacterial epitope. Eur J ImmunoI24:2061-2067
Wakeham J, Wang J, Magram J, Croitoru K, Harkness R, Dunn
P, Zaganiacz A, Xing Z (1998) Lack of both types 1 and
2 cytokines, tissue inflammatory responses, and immune
protection during pulmonary infection by Mycobacterium
bovis bacille Calmette-Guerin in IL-12-deficient mice. J
ImmunoI160:6101-6111
Wang J, Snider D, Hewlett B, Lukacs NW, Gauldie J, Liang H,
Xing Z (2000) Transgenic expression of GM-CSF induces
the differentiation and activation of a novel dendritic cell
population in the lung. Blood 95:2337-2345
Wang J, Zganiacz A, Xing Z (2002) Enhanced immunogenecity
of BCG vaccine by using a viral-based GM-CSF transgene
adjuvant formulation. Vaccine 20:2887-2898
WHO (2000) WHO vaccine preventable diseases - global
summary
Xing Z (2001a) The hunt for new tuberculosis vaccines: anti-
TB immunity and rational design of vaccines. Curr Pharm
Des 7:1015-1037
Xing Z (2001b) Breach of IL-12 monopoly in the initiation
of type 1 immunity to intracellular infections:IL-12 is not
required for host defense against viral infection. Cell Mol
Bioi 47:689-694
Xing Z, Wang J (2000) Consideration of cytokines as therapeu-
tics agents or targets. Curr Pharm Des 6:599-611
Xing Z, Braciak T, Ohkawara Y, Sallenave J-M, Foley R, Sime PJ,
Jordana M, Graham FL, Gaudie J (1996) Gene transfer for
cytokine functional studies in the lung: the multifunctional
role of GM-CSF in pulmonary inflammation. J Leukoc Bioi
59:481-488
Zhu X, Venkataprasad N, Ivanyi J, Vordermeier HM (1997) Vac-
cination with recombinant vaccinia viruses protects mice
against Mycobacterium tuberculosis infection. Immunol-
ogy 92:6-9
50 Tuberculosis in Animals
P. L. NICOLETTI

CONTENTS cattle were the same. Later studies by many work-

ers concluded they were different. The relationship
50.1 Introduction 893 between bovine tuberculosis, milk and human dis-
50.2 Agents 893
50.2.1 Classification 893
ease has been known for over a century. Hundreds of
50.2.2 Morphology 894 children have died of tuberculosis meningitis, or the
50.2.3 Properties 894 miliary form, as a result of ingestion of contaminated
50.3 Hosts, Pathogenesis and Diagnosis 894 milk. These and other observations led to campaigns
50.3.1 Hosts 894 in many countries to eliminate bovine tuberculosis.
50.3.2 Pathogenesis 895
50.3.3 Diagnosis 898
In industrialized countries, animal tuberculosis
50.3.3.1 Cattle 898 control and eradication programs have drastically
50.3.3.2 Deer 899 reduced M. bovis infection in both animals and
50.3.3.3 Non-human Primates 899 humans. In Europe and North America, only 0.5-
50.3.3.4 Swine 900 1.0% of human cases are believed to be from M. bovis.
50.3.3.5 Elephants 900
50.4 Public Health: Epidemiology and Control 900
Estimates of previous decades were between 5-20%.
50.4.1 Public Health 900 In developing countries, however, animal tuberculo-
50.4.2 Epidemiology 900 sis is widely distributed and control measures poorly
50.4.3 Control 901 applied. The incidence of HIV-attributable tubercu-
References 902 losis cases have increased in many African and Asian
countries.

50.1
Introduction
Tuberculosis is one of the oldest recognized diseases
of man and animals. DNA containing fragments
suggestive of tuberculosis has been extracted from
the bones of the extinct long-horned bison which
lived over 17,000 years ago (Rothschild et al. 2001).
Ancient Judaic teaching from the Talmud warned that
any animal carcass showing adhesions between the
lungs and the pleura was unsatisfactory for human
consumption. This admonition is believed to be due
to tuberculosis and suggests that animal to human
transmission has long been recognized.
In 1882, Koch identified Mycobacterium tuberculo-
sis and demonstrated that it could be transmitted to
animals. He believed that organisms in humans and
P. 1. NICOLETTI, DVM, MS
College ofVeterinary Medicine, Department ofVeterinary Pathol-
ogy, P.O. Box 110880, Gainseville, Florida 32611 0880, USA
50.2
Agents
50.2.1
Classification
Mycobacteriosis is a chronic infectious disease of
animals, birds, reptiles, fish and humans caused by
Mycobacterium spp and characterized by inflamma-
tory and necrotic processes. The mycobacteria are
divisible into a few obligate pathogens and a much
larger number of environmental pathogens and non-
pathogens.
The Genus Mycobacterium has been classified in
several ways. The Mycobacterium tuberculosis com-
plex (MTC) (M. bovis, M. microti, and M. tuberculo-
sis) and the Mycobacterium avium complex (MAC)
(M. intracellulare and three subspecies of M. avium;
M. avium subsp. avium. M. avium subsp. paratuber-
culosis and M. avium subsp. silvaticum). In addition,
other mycobacteria that may infect animals include

M. Monir Madkour et al. (eds.), Tuberculosis

© Springer-Verlag Berlin Heidelberg 2004
894
M. marinum (fish), M. ulcerans (animal models, e.g.
mice), and M. leprae (armadillo). It has been sug-
gested that the MTC produces disease should properly
be called 'tuberculosis' while other infections caused
by the MAC should be called 'mycobacteriosis'.
50.2.2
Morphology
Mycobacteria are small, aerobic, gram-positive, acid-
fast, non-motile and non-spore-forming bacilli. They
may survive in phagocytic cells (Fig. 50.1).All species
have a high cell-wall lipid content. The growth rate
is important in classification and speciation. Special
media are required and it may take up to several
months to develop visible colonies (Fig. 50.2). The
optimum incubation temperature is 28-33°C and a
pH of 5-5.5. It may be difficult to determine species
and often requires the submission of isolates to refer-
ence laboratories.
Fig. 50.1. Intracellular mycobacteria
Fig. 50.2. Mycobacteria colonies of solid medium
P. 1. Nicoletti
50.2.3
Properties
The degree of susceptibility to infection by certain
mycobacteria among animal species varies and has
been a criterion for differentiation of mycobacterial
species. Other criteria include the photochromogenic
property, biochemical analyses, and DNA probes.
Mycobacteria can survive in soil and fomites for days
to months depending upon local environmental condi-
tions such as sunlight. A review of studies on survival
of M. bovis in the environment has been published by
Morris et al. (1994). Examples cited are 150-332 days at
12-24°C in shade, 18-31 days at 24-34°C in sunlight, in
deep soil at depths of 5 cm for up to 2 years and in tap
water for 236 days at 18-24°C. Cresylic, or substituted-
phenol, disinfectants should be used in concentrations
of 2-5% to kill mycobacteria (Thoen et al.1984).
The literature provides no evidence that there are
any measurable differences in virulence between
strains of M. bovis with the notable exception of the
artificially attenuated strain BCG (Morris et al. 1994).
50.3
Hosts, Pathogenesis and Diagnosis
50.3.1
Hosts
In general, members of the genus Mycobacterium have
one of the highest host ranges of any pathogen. M. bovis
causes tuberculosis in a large number of mammalian
hosts including cattle and other ruminants, felids,
canids, lagomorphs, porcids, camelids, cervids and pri-
mates including humans. It has been reported to cause
disease in over 40 species of wild and exotic animals.
The degree of susceptibility to infection varies
widely from one animal species to another. Genetic
resistance has not been conclusively demonstrated.
Age, sex, reproductive status and other character-
istics have no proven influence on transmission or
course (Morris et al. 1994).
Tuberculosis (MTC) in non-human primates is most
commonly found in the Old World species. Tuberculo-
sis in cattle and other large ruminants is almost exclu-
sively caused by M. bovis. Natural cases of MTC infec-
tion in sheep and goats are rare. Dogs are susceptible
to both M. bovis and M. tuberculosis. Cats appear to be
more resistant to M. tuberculosis than M. bovis since
infections in the former are rare (Hines et al. 1995). M.
bovis infections have now been established in badgers
Tuberculosis in Animals 895

(Meles meles) in the United Kingdom and Ireland, in

brush-tailed possums (Trichosurus vulpecula) and
farmed deer in New Zealand, and in captive and feral
deer (Odocoileus virginianus) in the United States.
MAC infections occur in a large number of animal
hosts. The most common in cattle and other large
ruminants is caused by M. avium subsp . paratuber-
culosis. Other MAC infections in cattle do not result
in clinical disease. MAC infections are common in
swine causing a localized lymphadenitis.
All domestic fowl and many wild species of birds
are commonly infected with MAC but are generally
resistant to MTC. The incidence of MAC infections
is highest in zoological parks and aviaries (Hines
et al. 1995). Some non-human primates have been Fig. 50.3. Granulomas in soft tissue
reported to be infected with M. leprae and in the late
1960s it was reported that the nine-banded armadillo
(Dasypus novemcinctus) could be experimentally
infected. Natural infections were later discovered in
Texas and Louisiana and the armadillo became a host
for laboratory studies.

50.3.2
Pathogenesis

Mycobacterial infections are transmitted by inhala-

tion, direct contact with mucous membranes (inges-
tion) or through broken skin. Eighty to ninety per
cent of cattle are infected by inhalation (Morris et al.
1994). Only about 1% of calves born to tuberculosis
cows have congenital infection. Transmission of M.
avium subsp paratuberculosis is via direct ingestion
of feces, environmental contaminants, colostrum or
the placenta. Ingestion of other MACs most com-
monly leads to extra-pulmonary disease. Some sites
include the lymph nodes, skeletal, genitourinary and
central nervous system.
The pathogenesis of mycobacterial infections is
influenced by species of the genus and animal host,
host factors such as immune status, and the route and
numbers of exposure. The disease depends upon the
ability of the mycobacteria to infect and grow within
host cells. M. bovis is phagocytized by macrophages
and carried to lymph nodes and other sites. They sur-
vive within macrophages by fusion of phagosomal
compartments and are thus able to multiply, destroy
phagocytes and escape into intracellular spaces. This
stimulates accumulation of other phagocytes creat-
ing typical histopathological lesions called granulo-
mas (Figs. 50.3, 50.4).
High numbers of acid-fast bacilli may be seen
with Ziehl-Neelsen staining. Nodules of 1-5 cm may
Fig. 50.4. Granulomas in thorax of a cow
be seen in many tissues such as enlarged mesenteric,
hepatic and mediastinal lymph nodes or in the pleura
where the presence of hard white nodules has been
called 'pearl disease'.
There may be caseous necrotic coalescence. Calcifi-
cation is common (Neill et al.1994). Data from a large
number of cattle with lesions of tuberculosis showed
that 57% had lesions confined to the bronchial and/or
mediastinal nodes. Only 3.2% had mesenteric lymph
node lesions (Neill et al. 1994). These data may have
been influenced by the early removal of tuberculosis
infected cattle prior to the spread of the infection to
other body areas. Though primary genital infections
are not common, the uterus or epididymis may be
infected. The mammary gland may also be infected
resulting in shedding of organisms in milk.
Both antibody and cell-mediated immune responses
can develop following mycobacterial infections. It is
generally accepted that the cell-mediated immune
system has the most significant role in protective
896
immunity. There appears to be an inverse relationship
between cell-mediated and humoral responses from
natural M. bovis infection in cattle (Neill et al.1994).
The pathogenesis of M. bovis in other animal
species ranges from a single lymph node to rapidly
spreading fulminating disease (Figs. 50.6, 50.7). The
lymph nodes of the nasopharynx, lung or mediasti-
num are affected.
Tuberculosis in non-human primates is usually
due to M. tuberculosis and is often transmitted to
them by infected humans. The route of infection may
be cutaneous, or via the alimentary or respiratory
tracts. When the alimentary route is involved, there
may be pharyngeal involvement (scrofula), intestinal
ulceration and mesenteric lymphadenopathy with
caseous necrosis.
In the respiratory form, tubercles are pres-
ent within the lung (Fig.50.8), intercostal spaces
Fig. 50.5. Granulomas in the pleura of a cow
Fig. 50.6. Multiple granulomas in the lungs of an elk
P. L. Nicoletti
(Fig. 50.9), and bronchial lymph nodes. Primary
tubercles may be seen in the liver, or spleen, with
or without pulmonary involvement. Other foci may
include the kidney, brain, meninges, bone, skin,
endocrine glands and eye (Hines et al. 1995). Lesions
caused by M. bovis have been described in numer-
ous animal species besides ruminants, e.g. badgers
and brush-tailed possums. Lesions in the latter are
mostly in the respiratory tract (Morris et al. 1994).
The superficial lymph nodes may be infected and
result in discharges.
There is progressive disease and nodules have cen-
tral necrotic areas. Tuberculosis in badgers is primar-
ilya chronic respiratory disease (de Lisle et al. 2001).
The severity varies from no lesions to generalized
disease with massive excretion of M. bovis.
There have been several recent reports of M. tuber-
culosis infections in elephants, which were infected
Fig.50.7. Acute tuberculosis with necrosis and suppuration
in an elk
Fig. 50.8. Tubercles in the lung of a monkey
Tuberculosis in Animals 897

Fig. 50.9. Tubercles in the intercostal areas of a monkey Fig. 50.10. Granulomas in bronchial lymph nodes of an Asian
elephant

by humans. Between August 1996 and June 2000, 17

cases were confirmed in North America in Asian
elephants (Mikota et al. 2000). Non-specific signs
include weight loss and lethargy. Granulomatous
nodules may be found in bronchial lymph nodes
(Fig. 50.10). Elephants may die from caseocalcareous
and cavitating lesions with pulmonary abscesses.
MAC infections have received the most attention
as pathogens of avian species. These have an initial
involvement in the intestines and liver (Fig. 50.11).
Other organs may be affected such as spleen, lungs
and air sacs and bone marrow. The disease may result
in weight loss, depression, diarrhea and polyuria. The
tubercles are non-caseated, non-mineralized and con-
tain a large number of acid-fast intracellular bacilli.
MAC infections in swine are usually in the lymph
nodes of the head and/or mesentery (Fig. 50.12).
Generalized disease is rare. In cattle, lesions due to Fig. 50.11. M. avium infection of an avian liver
the MAC are usually transient and indistinguishable
from those caused by M. bovis. The primary veteri-
nary importance is sensitization which leads to posi-
tive intradermal tests using M. bovis tuberculin.
Some non-human primates are susceptible to
infections of M. leprae and are used as laboratory
models. In the lepromatous form there are small
lesions on the cooler areas of the body such as hands
and face. In the tuberculoid form, dermal granulo-
mas appear. Naturally occurring leprosy in armadil-
los begins with hepatomegaly and splenomegaly with
subsequent lesions in numerous tissues.
M. avium subsp paratuberculosis causes a chronic
disease of cattle and other ruminants (Johne's disease)
which results in a malabsorption syndrome with diar-
rhea,weight loss and reduced milk production. The ini-
tial infection occurs primarily in young calves but cIini-
cal signs appear most frequently in cows aged 3-6 years. Fig. 50.12. M. avium complex infections of swine lymph nodes
898
Infected cattle develop chronic granulomatous enteri-
tis, primarily in the ileum. There is no caseation, cal-
cification, or fibrosis associated with lesions of para-
tuberculosis in cattle. Organisms may be excreted in
feces, milk, semen, and urine.
Occasional cases of MAC infections in horses have
been reported. There is a proliferative enteritis with
nodular lesions in the lungs, spleen, and liver.
50.3.3
Diagnosis
50.3.3.1
Cattle
In control programs, diagnosis of possible tuberculo-
sis usually begins with a postmortem examination of
selected tissues (Fig. 50.13). Lymph nodes of the head
and thorax are sliced and observed for the presence
of tubercles.
Possible further gross pathology lesions may be
observed in lungs, spleen, kidney, udder or male
reproductive tissues. Specimens are collected for his-
topathological and bacteriological studies. A definite
diagnosis depends upon isolation of Mycobacterium
spp. A major problem is the long period required for
culture confirmation. Further, routine postmortem
examinations may fail to detect approximately half of
infected carcasses while more thorough procedures
may detect up to 90% (Corner 1994). Proper care
and storage of tissues is very important along with
culture techniques. The preferential sites for M. bovis
infection specimen collection, and the microbiologi-
cal procedures have been published (Corner 1994).
Fig.50.13. Bronchial lymph node of a cow with suspected
tuberculosis
P. 1. Nicoletti
Diagnosis of mycobacterial infections in live
animals may be difficult. The disease is difficult to
distinguish at the initial clinical, radiological or
pathologic examinations. It is interesting that over
100 years since Koch described the delayed type
hypersensitivity test (DTH), it remains the primary
method of diagnosis of individual animals. The pro-
cedure has many disadvantages.
Questions have often been raised about the
sensitivity, specificity, standards and quality of the
tuberculin used. False-positive reactions may be due
to sensitization caused by non-tuberculous mycobac-
teria. Tuberculins are a complex mixture of soluble
antigens produced by M. bovis or M. avium. Purified
protein derivative (PPD) is the antigen of choice and
is prepared by chemical fractionation. It is used as a
100 III volume either in the caudal fold (Fig. 50.14),
single intradermal cervical, or intradermal compara-
tive cervical (Fig. 50.15) locations.
Increases in skin thickness (Fig. 50.16) are detected
by manual palpation and/or measurements using
calipers (Fig. 50.17) at 72 h post-injection. These pro-
cedures require handling of animals at least twice. Due
to possible desensitization, testing of the same animal
within 60 days is not recommended.
The diagnostic efficacy of the skin test in cattle
ranges from 70-90% in sensitivity while estimated
specificity varies from 75-90% (Adams 2001). Many
factors affect the accuracy of skin tests such as the
interpretation of responses, quality and dose of
tuberculins, timing of prior tuberculin tests, cross-
reactions due to environmental mycobacteria and
other organisms, skill of application, and others.
A presumptive diagnosis of tuberculosis based
upon responses to tuberculins is further made
Fig.50.14. Injection of tuberculin into the caudal fold of a
cow
Tuberculosis in Animals
Fig. 50.15. Comparison of reactions of separate tuberculins
Fig. 50.16. Reaction to tuberculin in the caudal fold of a cow
through histopathology, Cultures and other tech-
niques such as immunohistochemical staining using
M. bovis-specific probes (Adams 2001). In some labo-
ratories, PCR techniques have provided a useful tool
for the rapid detection of slow growing pathogens
(Wards et al. 1995). These have been shown to be
faster and more sensitive than traditional methods
including histopathology.
Procedures for application of DTH tests in cattle
vary from country to country. Clearly, improved
tests are needed. Many false-positive reactions are
observed. Recently it has been found that infected
cattle may fail to respond to intradermal tuberculins
for many days (anergy). Poor nutrition and preg-
nancy have been suggested as factors.
A sensitive, and specific, serological test to detect
antibodies to M. bovis antigen in livestock and syl-
vatic populations would be very useful to replace or
899
Fig. 50.17. Measurement of response to tuberculin
supplement the intradermal tests. Antibody produc-
tion is variable and largely undetectable during the
early subclinical stages of infection. It usually occurs
when CMI tests may be negative (Adams 2001). Anti-
body detection procedures are based mostly on the
enzyme-linked immunosorbent assay.
In Australia, Wood and Rothel (1994) developed a
gamma interferon (IFNd) assay as a simple and rapid
in vitro cellular assay for the diagnosis of bovine
tuberculosis. The specificity in over 6000 cattle varied
between 96.2-98%. The sensitivity among cattle in
herds, which were de-populated because of tubercu-
losis was between 76.8-93.6%.
50.3.3.2
Deer
Conventional intradermal tuberculin tests may be
unsatisfactory (Griffin and Macintosh 2000). False-
positive reactions are common and infected deer
may be anergic. The above authors have developed
a cell-based lymphocyte transformation (LT) test
with a reported sensitivity of greater than 95% and
a specificity of greater than 98%.
50.3.3.3
Non-human Primates
Intradermal tuberculosis of M. tuberculosis or M.
bovis are used of apparent equal efficacy. There is
a much lower response than in humans. The eyelid
(palpebra) and peri-umbilical skin of the abdomen
are sites of injection. Animals with advanced tuber-
culosis may be anergic. Cultures may be obtained
from throat swabs or gastric lavage. DNA probes
900
and nucleic acid amplification procedures along with
ELISA have been used.
50.3.3.4
Swine
The diagnosis of MAC is with 0.1 ml of PPD-A in the
dorsal surface of the ear with an observation at 48 h.
A swelling of 3 em or greater in diameter is usually
considered positive. The diagnosis of MAC infec-
tion in birds may be with intradermal PPD-A. Also
slide agglutination and ELISA have been used. In
the future, DNA probes and PCR used in specimens
from humans may prove to be useful.
50.3.3.5
Elephants
Intradermal tests correlate poorly with cultures.
There is a high degree of culture positive/test nega-
tive results. Currently, cultures of respiratory secre-
tions obtained by trunk lavage are recommended.
Details have been published by Montali et al. (2001).
For the diagnosis of paratuberculosis in cattle and
other large ruminants, an ELISA screening test may
be used. There are many false-positives and cultures
of ELISA positive animals range from 33-84% (Wells
et al. 2002). The sensitivity of cultures of fecal samples
in sub-clinically infected cattle has been estimated to
be only 40-50%.
50.4
Public Health: Epidemiology and Control
50.4.1
Public Health
Tuberculosis in humans is usually caused by M.
tuberculosis and has no direct zoonotic relation-
ship. However, an unknown proportion of cases are
due to M. bovis. This is particularly true in countries
without organized control programs. It is estimated
that 5-10% of cases in developing countries may be
due to M. bovis (O'Reilly and Daborn 1995). In Latin
American, it is estimated that 2% of pulmonary and
8% of extrapulmonary cases are caused by M. bovis
(Cosivi et al. 1998).
M. bovis infections in humans are clinically indis-
tinguishable from those caused by M. tuberculosis.
Most cases caused by M. bovis occur in young persons
and result from drinking contaminated milk. Cervi-
P. 1. Nicoletti
cal lymphadenopathy, intestinal lesions, chronic skin
(lupus vulgaris) and other non-pulmonary forms are
common. Cases are now rare in developed countries
due to the compulsory pasteurization of milk and the
slaughter of infected cattle.
There is published evidence of aerosol transmis-
sion of M. bovis and direct contact with tuberculous
animals. Fanning and Edwards (1991) described a
case of M. bovis in a human who was in contact with
an infected elk. There were 446 persons with identi-
fied elk contacts and 81 of 394 were skin-test positive.
Of these, 50 had been in contact with culture-positive
animals. The results of tests may have been influenced
by previous BCG vaccination in several persons.
Humans who routinely ingest pork have no differ-
ences in hypersensitivity rates using different PPDs
compared with those who do not consume pork. There
appears to be no proof of a direct relationship between
swine and human mycobacteriosis. Intravenous injec-
tion of serotypes of M. avium-intracellular complex
have not resulted in isolations from muscle tissue.
M. avium infections in humans have increased in the
past 20 years or so, principally in human immunodefi-
ciency virus-infected patients. Disseminated infections
occur frequently in the late stages of AIDS. The poten-
tial zoonotic risk of MAC infections in humans which
are derived from avian species is somewhat unclear.
The risk of transmission appears to be extremely small.
In general, M. avium infections in birds and mammals
(including humans) are environmentally derived and
not passed from bird to mammal.
Clearly, some forms of tuberculosis in humans are
of animal origin (e.g. from non-human primates)
and these zoonotic infections can best be controlled
by reducing possible exposure through direct or
indirect contacts with infected animals and through
pasteurization of milk.
50.4.2
Epidemiology
Morris et al. (1994) have published an excellent and
extensive review of many aspects of the epidemiology
of M. bovis infections. They especially place emphasis
on the role of wildlife as maintenance hosts and the
difficulties that this causes in controlling and eradi-
cating the disease in domestic animals. In addition,
the survival of the organisms in the environment for
substantial periods further complications elimina-
tion of the disease.
In many countries, the increasing demand for
animal protein and milk has resulted in larger herds
Tuberculosis in Animals
with greater animal density. This and the frequent
and close contacts between humans and animals have
resulted in a higher incidence of tuberculosis in both
animals and humans.
50.4.3
Control
Cosivi et al. (1998) have published an excellent review
of the status of tuberculosis due to M. bovis in devel-
oping countries. Of 55 African countries, only seven
apply disease control measures. Approximately 85%
of cattle are in areas where bovine tuberculosis is
either partly controlled or not al all. Among 36 Asian
nations, only seven apply control measures and only
6% of cattle are in countries which attempt to con-
trol tuberculosis. Of 34 Latin American or Caribbean
countries, only 12 apply a test and slaughter policy.
However, 76% of the cattle are in countries where test
and slaughter policy is used.
During the 1800s and early 1900s, bovine tuberculo-
sis was the most prevalent infectious disease of cattle in
the USA. Over 80% of the carcass condemnations were
from tuberculosis. It is estimated that 5% of cattle were
condemned. The National Cooperative State-Federal
Tuberculosis Eradication Program was initiated in
1917 and continues. The program is similar to those in
many other developed countries and is based upon sur-
veillance at postmortem, tests and slaughter of tuber-
culin positive animals, and subsequent investigation of
herds epidemiologically connected to infected animals
and herds. Depopulation of infected herds is optimal.
Nearly all areas are now accredited and prevalence is
less than 0.02%. Similar data exist from many other
countries. However, total eradication of cattle tubercu-
losis has been difficult in several countries due to wild-
life reservoirs such as badgers, wild and domesticated
deer, and possums. It is important to understand the
role of a potential host in maintenance and transmis-
sion of M. bovis within animal populations. Additional
problems are anergy and false-positive tuberculin test,
which create livestock owner distrust and permits
infected animals to go undetected.
The difficulties of eliminating animal tuberculosis
through test and slaughter have led to suggestions
of possible alternate methods, or adjuncts such as
treatment and/or vaccination. Many anti-tuberculosis
drugs have been studied in selected animal species.
These have included isoniazid, pyrazinamide, rifampin
and ethambutol. Combination of streptomycin and
isoniazid have been used successfully for M. tuber-
culosis in non-human primates (Ward et al.1985).
901
Unlike test and slaughter procedures for livestock,
current guidelines allow for treatment of tuberculous
and suspect elephants under conditions of quaran-
tine and travel restrictions (Mikota et al. 2000).
In very valuable avian species, there have been
reported successes in treating MAC infections using
drugs such as ethambutol, rifampin, amikacin, strep-
tomycin, ciprofloxacin and enrofloxacin.
To date, the possible treatment of animals with
tuberculosis has been limited to very valuable, mostly
exotic species and with varying success. Clearly, treat-
ment is not an alternative to slaughter of tuberculous
livestock due to costs, lengthy regimens and limited
success.
The implementation of national bovine tuber-
culosis programs in many industrialized countries
based upon regular tuberculin testing and removal of
infected animals has led to successful eradication of
major reductions in the incidence of cattle and farmed
deer tuberculosis. In some countries, alternative strat-
egies are needed since test and slaughter is economi-
cally and socially unacceptable. One possible strategy
is vaccination.
The requirements for a satisfactory vaccine are
many such as immunogenicity, no causation of hyper-
sensitivity to dermal tuberculins, minimal adverse
host response, practical application, and acceptance
by other countries. Skinner et al. (2001) have reviewed
the subject of vaccination of animals against M. bovis.
Experiments using the bacillus of Calmette-Guerin -
(BeG) in cattle have been summarized and the authors
concluded the results of experiments have been gener-
ally disappointing. The conferred resistance to natural
or artificial infections has been short lived. An added
disadvantage was the sensitization to tuberculins. The
authors concluded that a reduced dose of BCG may
preferentially stimulate cellular immunity with mini-
mal effect on skin test.
Buddie et al. (1999) summarized experiments in
deer, possums, badgers and ferrets. In deer, two doses
ofBCG produced significant protection against infec-
tion and disease with live but not heat-killed organ-
isms. Lower doses were superior. An orally adminis-
tered BCG produced a 100-fold lower lung bacterial
counts in possums. BCG vaccinated (10 6 CFU) bad-
gers had higher lymphocyte blastogenic responses
than controls. They also shed fewer bacilli. The
authors concluded that future studies may develop
improved vaccine and methods of administering
them in selected hosts. The increase in knowledge
of mycobacterial genetics and immune responses in
recent years suggests that development of an effective
vaccine for cattle, deer and wildlife is a realistic goal
902
(BuddIe et al. 1999). Although protection may not be
complete, vaccination may reduce economic losses in
developing countries. Clearly, the elimination of res-
ervoir hosts such as badgers in the UK is impractical
and unacceptable to many concerned organizations.
When there is an interplay between infection in wild-
life and domestic animals, eradication of the disease
becomes impractical (Morris et al. 1994).
In the absence of a wildlife reservoir of M. hovis,
tuberculosis is a readily controllable disease (Morris
et al. 1994). Test and removal of infected animals at
intervals of less than a year can eliminate the infection
from herds provided that an epidemiologically sound
control policy is followed. The first step is to motivate
livestock owners to co-operate. However, in many coun-
tries there are severe restraints. These include limited
financial resources, lack of trained professionals, lack
of laboratory support and since, the disease is not usu-
ally epidemic, more dramatic diseases receive the most
attention. Conversely, there must be a commitment by
farmers and governmental officials to control measures.
An organized meat inspection program to detect cases
must exist. Adequate funding for a long-term effort is
necessary. Control measures have significant economic
consequences: These are expensive and labor inten-
sive which include testing of animals, slaughter of test
positives, disinfection and public education. It seems
obvious that the many technical and logistical factors
involved in the control of tuberculosis will assure the
persistence of the disease for many decades.
References
Adams LG (2001) In vivo and in vitro diagnosis of Mycobacte-
rium bovis infection. Rev Sci Tech Off Int Epiz 20:304-324
BuddIe BM et al (1999) Differentiation between Mycobacte-
rium bovis BCG-vaccinated and M. bovis-infected cattle by
P. 1. Nicoletti
using recombinant mycobacterial antigens. Clin Diagn Lab
Immun 6:1-5
Corner LA (1994) Post mortem diagnosis of Mycobacterium
bovis infection in cattle. Vet Microbiol 40:53-63
Cosivi 0 et al (1998) Zoonotic tuberculosis due to Mycobac-
terium bovis in developing countries. Emerg Infect Dis 4:
59-70
De Lisle GW et al (2001) Mycobacterium bovis in free-living
and captive wildlife, including farmed deer. Rev Sci Tech
Off Int Epiz 20:86-111
Fanning A, Edwards S (1991) Mycobacterium bovis infection
in human beings in contact with elk (Cervus elaphus) in
Alberta, Canada. Lancet 338:1253-1255
Griffin JFT,Macintosh CG (2000) Tuberculosis in deer: percep-
tions, problemsand progress. Vet J 160:202-219
Hines ME et al (1995) Mycobacterial infections of animals:
pathology and Pathogenesis. Lab Anim Sci 45:334-351
Mikota SK et al (2000) Tuberculosis in elephants in North
America. Zoobiology 19:393-403
Morris RS et al (1994) The epidemiology of Mycobacterium
bovis infections. Vet MicrobioI40:153-177
Neill SD et al (1994) Pathogenesis of Mycobacterium bovis
infections in cattle. Vet MicrobioI40:41-52
O'Reilly LM, Daborn CJ (1995) The epidemiology of Myco-
bacterium bovis infections in animals and man; a review.
Tubercle Lung Dis 16 [SuppI1]:1-46
Rothschild BM et al (2001) Mycobacterium tuberculosis com-
plex DNA from an extinct bison 17,000 years before the
present. Clin Infect Dis 33:35-311
Skinner MA et al (2001) Vaccination of animals against Myco-
bacterium bovis. Rev Sci Tech Off Int Epiz 20:112-132
Thoen CO et al (1984) Mycobacterium avium complex. In:
Kubica GP, Wayne LG (eds) The Mycobacteria, a source-
book, part B. Dekker, New York, pp 1251-1257
Ward GS et al (1985) Use of streptomycin and isoniazid during
a tuberculosis epizootic in a rhesus cynomolgus breeding
colony. Lab Anim Sci 35:395-399
Wards BJ et al (1995) Detection of Mycobacterium bovis in
tissues by polymerase chain reaction. Vet Microbiol 43:
227-240
Wells SJ et al (2002) Sensitivity of test strategies used in the
Voluntary Johne's Disease Herd Status Program for detec-
tion of Mycobacterium paratuberculosis infection in dairy
cattle herds. J Am Vet Med Assoc 220:1053-1057
Wood PR, Rothel JS (1994) In vitro immunodiagnostic assays
for bovine tuberculosis. Vet MicrobioI40:125-135
Subject Index

A Adherence to treatment 809

A60 Antigen 189 Adhesive arachnoiditis 541
A-a 02 gradient 389 Adhesive ependymitis 539
Abdomen, shielding of the 307 Adjunctive treatment 474
Abdominal imaging 670 Administrative support 858
Abdominal mass 717 Adolescence, tuberculosis during 255
Abdominal pain 703,717 Adrenal enlargement, bilateral 753
Abdominal tuberculosis 659,661,682 Adrenal gland 752
- direct spread from 702 - MRI 753
Abscess Adrenal
- cold (empyema necessitatis) 375,484,522 - insufficiency 752
- formation 482,692, 711, 716 - irregularly enlarged 753
- metastatic tuberculous (gumma) 639 - steroids 145
- multiloculated paravertebral 514 - tuberculosis 751
- orbital 736 Adrenalitis 752
- paraspinal 485,486,538 - tuberculous 752
- paravertebral 486, 494 Adult respiratory distress syndrome (see also ARDS) 276
- periappendiceal 662 - and disseminated intravascular coagulation 292
- prostatic tuberculous 712 Adult-type apical lesions 255
- retropharyngeal 743 Adult-type progressive tuberculosis 330
- subcutaneous 630 Advocacy 807
- tuberculous of the brain 537 Aerosol transmission 900
- tuberculous of the breast 639 AFB
- urethral abscess 703 - and culture 448
Abscess drainage 710 - smear 199
- CT guided 486 - smear-negative 199
Accuprobe system 124 - smear-positive 450
Achilles tendon 599 After treatment, recurrence 472
Acid Ag85b (MPT59) 236
- pyrazinoic 789 Agar-based media 121
- tuberculostearic 188 Agent, single 784
Acne agminata (lupus miliaris disseminatus facie!) 647 AhpC gene 790
Actinomycosis 342 AIDS
Actinomyces 115 - co-infected with 349
Active case-finding 862 - epidemic 536
- in resource-poor areas 862 - impact 37
Active disease - pandemic 455
- absence 413 - patients 660
- extent of 415 - - therapy in patients with 675
- progression to 35 - related deaths 456
ADA, pleural 353 Air-bone 266
Addison 751 - infection 315
Addison's disease 297 - transmitted 104
Adenoids 447 Air filters 865
Adenopathy Air-fluid levels 373
- mediastinal 380 Airway infiltration, direct 331
- mesenteric 684 Airway obstruction, mechanisms of 387
- paratracheal 340,451 Albucasis 18
Adenosine deaminase activity 434 Alcoholic 789
Adenovirus 890 Alcoholism, malnutrition, diabeter 275
- based vaccine 890 Alexandria, Egypt 481
904 Subject Index

Alimentary route 896 -low-molecular-weight ESAT-6 (6 KDa) 237

Al Kindi 18 - for recombinant TB vaccine 885
Alkyhydroperoxide reductase 790 - released mycobacterial 847
Alternative therapies 825 Antigen 5 (38 kDa Antigen) 188
Alveolar arterial oxygen gradient 389 Antigen p90 189
Alveolar macrophages 883 Antigen presenting cells (APe) 884,886
Ambulatory chemotherapy 805 Antigen specific T cells 884
Amebiasis 662 Anti-glycolipids IgG antibody 846
Amenorrhea 719 Anti-LAM 846
American Academy of Pediatrics 786 Anti-leukemic therapy 218
American Thoracic Society 822 Antimicrobial susceptibility testing 126
Amikacin 792, 825 Antinuclear antibodies 790
Aminoglycosides 790 Antiretrovirals 791
Amplicor M. tuberculosis ( PCR) Test 125 - with anti-tuberculosis drugs 470
Anal TB 688 Anti-retroviral therapy for poor developing countries 469
Ancesort, common 79 Anti-ST-CF IgG antibody 848
Ancient Egypt, tuberculosis 3 Anti-TB immunity 886
Ancient Egyptian art 5 Anti-tuberculosis chemotherapy 342
Anemia 215,216,794 - prophylactic 473
Anergic response 458 - for spinal tuberculosis 490
Aneurysm 371 Anti-tuberculous drugs, endocrine and metabolic effects 766
- mycotic 771 APC (antigen presenting cells) 884,886
- subaortic 440 Apoptotic debris 459
- submitral 440 Apoptotic mycocytes 884
- tuberculous 771 Appendicitis, subacute 670
- - mycotic 771 Appendicular fistulas 702
- ventricular 440 Appendicular tuberculosis 672
Angiographic and postmortem studies 551 Appendix, tuberculosis 686
Angiography Apple jelly nodules 746
- cerebral 542 Aqueous smear 737
- and dynamic CT 557 Arachnoiditis
Angular deformity 523 - adhesive 541
Animal - spinal 574
- pathogen 78 ARDS (see also adult respiratory distress syndrome) 274,
- tuberculosis 893 292,389
- - eliminating 901 - diagnosis of 293
Ankh-my, tomb of 5 - and tuberculosis 319
Ankle, tuberculosis of the 592,621 Arterial hypoxemia 389
Ankylosis 611 Arteriography 321,360
Annual incidence 856 Arteritis 540,551
Annual risk 856 - tuberculous 772
Antenatal 254 Artery, tuberculous 320
Anterior debridement 522 Arthritis, tuberculous 588
Anterior decompression and fusion 523 Arthrodesis 595
Anterior radical debridement and fusion 519 - surgical 596
Anterior radical surgery 523 ASC
Anterior spinal cord decompression 524 - on day 8 848
Anterior spinal surgical technique 493 - during the treatment 848
Anthracofibrosis, bronchial 330 Ascitic fluid 667
Antibiotics - adenosine deaminase 669
- prophylactic 465 Asherman's syndrome (synechia uteri) 721
- with surgery 518 Aspergilloma (mycetoma) 319,386
- without surgery (MRC Trials) 518 Aspergillus fumigatus 319,378
Antibody 899 Aspergillus hyphae 378
- antinuclear 790 Aspirate, nasopharyngeal 253
- complexed 847 Aspiration biopsies, imaging-guided 694
Antibody card test, rapid 192 Aspiration, gastric 251,253,339
Antibody levels, specific 843 Asthma 329
Antibody secreting B cells (ASC) 847 - bronchial 342
Antibody secreting Cells 843 Atelectasis 370
Anti-DAT antibodies 844 Atrophy
Anti-diuretic hormone secretion, inappropriate 540 - cortical 552
Antigen - optic 735
Subject Index 905

ATS recommendation 784, 785 Betiatide MAG3 419

ATS/CDC/IDSA Committee 451 Bifacetal dislocation 530
ATS-CDC recommendation 785 Birnaristan 18
Automated systems 817 Bioinformatics 82
Autonephrectomy 701 Biopsy 622
Autopsy studies 465 - endometrial 720, 723
Auxotrophic mutants 888 - endoscopic 681,686
Avenbrugger, Leopold Joseph 19 - - mucosal 664
Avicenna 17 - laparoscopic 669
Axillary 447 - pleural 353
-lymph 447 - surgical 448
Azoospermia, obstructive 717,712 - testicular 717
- trans-sphenoidal 759
B Biosafety 105
Bacilli Calmette Guerin (see also BCG) 628 Biphasic media 121
- intravesicular 735 Birds 893
- vaccination 260,865 Bladder augmentation 710
Bacilli-laden macrophages 153 Bladder capacity, reduced 704
Bacilliuria 700 Bladder symptoms, irritative 712
- intermittent 703 Bladder ulceration 701
BACTEC 352,817 Blastomycosis 565
- system 339,817 Bleeding, gastrointestinal 673
BACTEC 460 TB instrument 121 Blindness 792
BACTEC MGIT 960 122 Blood cultures 225
Bactericidal activity 847 BM culture 225
Bacteriological relapse 840 BM histology 225
Bacteriology 339 Body morphotype 171
Badgers 896 Bone
BAE (bronchial artery embolization) 317 - ethmoid 447
Ball of wind 662 - small round 597
Balloon dilatation, endoscopic 343 Bone and joint tuberculosis 422
Barium Bone marrow infiltration 222
- enema 664 Bone-strut grafting, autologus (Medical Research Council
- studies 681, 682 Working Party) 518
- swallow 672, 683 Bony ankylosis 611
Basal cisterns 549 Booster 235,882
Basal contrast enhancement 549 - vaccine 883
Basal inflammatory exudate 554 Boosting regimen 883
Basalleptomeninges 549 Bovine tubercle bacillus 116
Bayle 20 Bovine tuberculosis 75,315
B-cell response, immunological 842 - control of 659
BCG (see also Bacilli Calmette-Guerin, disseminated) 160, Brachytherapy 346
447 Brain 573
- administration, intravesical 772 - abscess, tuberculous 560
- in cattle 901 - infarction 550
- discontinuation of 866 Brain stem infarction 553
- genome 888 Breast feeding 787
- history 27 - and its contraindication 307
-lymphadentis 261 Breast milk, heating 308
- protective efficary 145 British Medical Research Council (BMRC) 801
- recombinant 884, 888 Bronchial anthracofibrosis 330
- therapy, intravesical 274,277,773 Bronchial artery embolization (BAE) 317
- and new tuberculosis vaccines 881 Bronchial aspirate, microbiology of 321
- vacccination 187, 259 Bronchial asthma 342
- - and cutaneous tuberculosis 641 Bronchial carcinoma 316
- - and HIV serostatus 475 Bronchial dilatation 369
-vaccine 253,881,884 Bronchial involvement, location of 336
- - cytokine-secreting 888 Bronchial lumen, obstruction of the 333
- - irnmunogenicity of 882 Bronchial obstruction 331
- - ineffectiveness of 881, 882 Bronchial stenosis 337,395
BDProbe Tec ET 124 Bronchial wall
Beny Hassan 5 - infection 318
Beta-endorphin 764 - thickening 341,369,372
906 Subject Index

Bronchiectasis 246, 318, 330, 346, 363, 372, 386, 388 Cauda equina 572
Bronchitics, nonspecific 332,336 Cavernous sinus 758
Bronchogenic dissemination 363 -lesion of the 561
Bronchogenic spread 315,371 Cavitation 246,370
Bronchography 321,360,373 - miliary 379
Broncholglandular fistula 335 Cavity, tuberculous 315
Broncholith 346 CC, cerebral 565
Bronchoplastic surgery 346 Cd-l- restricted T cells 137
Bronchopleural fistula 315,375,386,395,401,406 CD4+
Bronchopneumonia 369 - cell count 222
Bronchoscopic biopsy 330 - defects in the 458
Bronchoscopic diagnosis 321 - lymphocytes, depletion of 458
Bronchoscopy 269,330,339,341,399,450 - T cells 459
- broncho-alveolar lavage 461 CD8+
Bronchostenosis 329,332,343,363 - cytotoxic T cells 136
Bronchovascular markings, accentuated 369 - cytotoxin T lymphocytes 351
Bronchus - IFN-y-seceting T cells 136
- contours of large 341 CD8 T cells, activation of 884
- erosion 447 Cecum, contraction 685
- thickening of 372 Cecum, retracted 664
Bullate formation 370 Cell-mediated immune response (CMI) 266,276
Bursitis Cell wall 116
- tuberculous 621 Cell-mediated immunity (CMI) 351
- - of the greater trochanter 599 Cellular immune response 459
Buttock sinuses, discharging 594 Cell-wall biosynthesis 819
Central nervous system 159
Centrifugation 542
C Century of tuberculosis 25
Cairo Museum 5 Cerebritis 553
Calcification 246,370,379,447,630,701,722 - subcortical 550
- parenchymal 362 Cerebrospinal fluids (see also CSF) 277,537
- pleural 375 Cervicitis, tuberculous 721
Calcified primary focus 267 Cervico-thoracic spine 530
Caliectasis 705 Cervix 718
Calvarium 563 Chancre, tuberculous 629
Calyces 704 Chemokines 351
Calyx 701 - active on lymphocytes and monocyteslmacrophages 887
Canon of Medicine 17 Chemoprophylaxis 97,259,857
Capreomycin 794 - regime 825
Captain of all of these men of death 455 Chemotherapy
Carbamazepine 790 - and debridement 518
Carbohydrate antigen, purified 846 - history 27
Carcinoma - preventive 839,881
- bronchial 316 - regimen of 807
- colon 662 - supervised 807
Carcinomatosis, meningeal 576 - with immobilization 518
Cardiac catheterization 438 - with mobilization 518
Cardiac mass 440 Chest computed tomography 437
Cardiac tamponade 431,433 Chest pain 432
Care, postoperative 400 Chest radiography 253,293,307,341,859
Carotid artery 744 Chest wall tuberculosis 618
Carpal tunnel syndrome 599 Chest X-ray 436
Carpectomy 595 Chiasm, optic 758
Case finding 804 Child bearing age 301
- active 839 Childhood respiratory diseases, the dilemma of multiple
- passive 839,856,859,862 aetiology of 465
Caseation Childhood tuberculosis 193,243,251
- actively 332,333 - clinical manifestations of 2245
- central 320 - epidemiology 243
Caseous focus 273 - and HIY/AIDS 251
Caseous serofibrinous phase 610 - incidence of 243
Catheterization 434 Children 786
Cattle 897 - facial palsy in children 564
Subject Index 907

China 873 Confirmation

Chorionic villi, amniotic fluid 302 - histologic 330
Chronic cases 803 - microbiogical 269
Chronic tuberculous postprimary disease 330 - rapid 297
CIC (see also circulating immune complex), total 847 Conjunctival tuberculosis 732
Cicatrization 370 Conjunctivits, phlyctenular 630
Ciprofloxacin 204,793 Consequences for others 829
Circle of Willis 552 Consolidation 246, 340, 369
Circulating antibodies 842 - segmental 369
Circulating ASC 847 Constrictive pericarditis 435
Circulating hormones 764 Constrictive pericitis 431
Circulating immune complexes (CIe) 842,846 Contact 255,338,825
Circumcision 629 - and outbreak investigations 62
Clarissmus Galenus 481 - study 100
Classification 332,894 Contagion, risk of 862
Clavicle 597 Contagious spread 660
- lateral end of the 598 Contraceptives, oral 791
Clinical and bacteriological monitoring 839 Control 901
Clinical feature 351,447,483,588,630-632,635,636,639, - inadequate 825
642-645,647-653,712,713 - programmes 824
- atypical 268 Convention cultures 353
- of adrenal TB 752 Conventional tomography 486,514
- and complications 315 Cor pulmonale 389
- and diagnosis 268, 599 Cord
- of female genital tuberculosis 719 - displacement of 486
- of miliary/disseminated tuberculosis 277 - Factor 190
- of penile tuberculosis 713 Corneal disease 733
- of tuberculosis in these groups 724 Correctional facilities 98
- of tuberculosis prostatitis and seminal vesiculitis 712 Cortical atrophy 552
- of tuberculous epididymo-orchitis 712 Corticosteroid 297,543,786
- other 703 Corticosteroid treatment 343
Clinical history 338 Corticotropin releasing hormone (CRH) 764
Clinical manifestation 736,662,666,670 Cortico-white matter junction 555
Clinical pathophysiology 439 Cortisol 764
- and epidemiology 433 Cost-effective approach 805
Clinical presentation 608 682 Cost-effective method 235
Clinics in the USA 824 Cost-effectiveness 468
Clustered regularly interspaced short palindromic repeats - of tuberculosis treatment 857
(CRISPRs) 81 Costotransversectomy 524
Clustering 100 Cough, persistent 252
CMI (cell-mediated immunity) 351 Course 630,632,633,635,636,639,642-644,646,647
CNS, tuberculosis 420,535 Course of action when the diagnosis is unclear 674
Coagulation abnormalities 226 Cow's milk, infected 746
Co-infection rate 824 Cr51EDTA 707
Cold abscess (empyema necessitatis) 375,484,522 Cranial nerve 537
- formation 694 - palsies 539,758
Colitis Cranial neuropathies 540
- tuberculous 687 Craniocervical junction 484
- ulcerative 662,680 - tuberculosis 483
Collapse therapy 395 Creatine kinase 600
Colon, tuberculosis 661 CRH (corticotropin releasing hormone) 764
Colonoscopy 664 CRISPRs (clustered regularly interspaced short palindromic
Color vision 792 repeats) 81
Comb sign 686 Crohn's disease 659,662,685
Community epidemiology 65 Cross-reactivity 187
Community prevalence 784 Cryotherapy 346
Compliance 257,543,823 Cryptic angiomatous malformation 569
- poor 310 Cryptococcus neoformans infection 565
Complications, postoperative 401 CSF (see also cerebrospinal fluids)
Computed tomography (see also CT) 340,360,614,663, - adenosine deaminase concentrations 249
682 - polmerase chain reaction 249
- cranial 249 - in TBM 248
- scan 705 - white cellcount 249
908 Subject Index

CT (see also computer tomography) 541,752,716,758 Destruction, avoidance of 142

- angiography 557 Developing countries 831
- combined with water-soluble myelography 572 Diabetes insipidus 758
- dynamic 557 Diabetes mellitus, tuberculosis 765
- enhanced 451 Diabetics 52,93
- features 514 Diagnosis 306,352,436,440,441,487,541,632,653,662,666,
- - of adrenals tuberculosis 752 670,736,897
- findings 667 - criteria 252,266
- guided procedures 622 - cytological 448,522
- myelography 575 - and differential diagnosis 631, 634, 637,639, 641, 643, 644,
- of prostatic tuberculosis 716 646-653
- of the thyroid 757 - histological 757
- scanning 486, 549 - rapid 296
Culture 119,466,899 - serological 129
- conversion 840 Diagnostic investigation 252
- media 120 Diagnostic methods 841
- positive 257 Diagnostic test 436
Cure rate Diastolic filling, rapid 435
- greatly increased 802 DIC (disseminated intravascular coagulation) 215,226,293
- of surgical treatment 402 Diffusing capacity 389
Cycloserine 790,791,794 Digital subtraction angiography (DSA) 552,561
Cystic erosion 622 1,25 dihydroxycholecalciferol (1,25 OH) 761
Cysticercosis 565 Dilatation
- in the spine 581 - bronchial 369
Cystoscopy 707 - cystoscopic 710
- of urinary tuberculosis 707 - post-stenotic 372
Cytokines 266, 885, 888 Direct spread 718
- active on antigen-presenting cells 886 Directly observed treatment 801
- expression 138 Directly observed treatment strategy (see also DOTS) 801
- IFN-y 276 Disc destruction 486
- IL-12 276 Disc tubercle, optic 735
- as immune adjuvants for TB vaccine formulations 886 Disease
- recombinant 887 - activity of 413
- TNF-a 276 - inactive 411
Cytologic findings 448, 707 - risk of 34
Cytopenias 215 Disseminated anergic tuberculosis 458
Disseminated Bacille Calmette-Guerin (BCG) 160
Disseminated cerebral coccidioidomycosis (DCC) 565,578
D Disseminated intravascular coagulation (01C) 215, 226, 293
Dactylitis 620 Disseminated NTM infection 165,175
Dairy herds 315 - with AIDS 175
DAT 843 Disseminated tuberculosis 389,463,661,691
Daughter lesions 561 Distant septic foci 425
DCC (disseminated cerebral coddidioidomycosis) 565,578 Distant skip lesion 494
De Quervain's-like disease 599 Division of tuberculosis elimination 874
Deafness 540 DNA
Death, sudden 441 - based TB vaccination 887
Death, vaccine-preventable 881 - fingerprinting 35,467,810
Debris, apoptotic 459 - probe analyses 199
Decalcification 494 - vaccine 886
Decidua 302 Documentation and surveillance 859
Decompression 523 Dominant T-cell antigens 885
Decortication 406 Doppler
- pleural 319 - discharging 432
Deep vein thrombosis 226 - echocardiography 438
Deer 899 - hemodynamics shown by 433
Defination 330 - tracings 435
Deformed cone-shaped 664 Dose modification 725
Deir el-Bahri temple 12 DOTS (see also directly observed treatment strategy) 451
Delayed type hypersensitivity test (see also DTH) 351,898 - assessing results 807
Demographic changes 40 - impact advantages of 802
Demonstration area 806 - and its impact in India 811
Demyelination 541 - implementation, approaches to 806
Subject Index 909

- plus and the green light committee 871 Echocardiogram 432,440

- quality control of 807 Echocardiography 434
DOTS strategy 39,810,811 - and doppler studies 438
-limitations of 808 - two-dimensional 438
- status of the 860 Echo-doppler techniques 434
Doughy sensation 719 Echo-free space 433
Drainage 623 Economic risk 820
- of abscesses 522 Edematous-hyperemic type 332,333
- of hydronephrosis 710 Effective and appropriate therapy, barriers to 471
- lymphnodes 662 Egyptian art, ancient 5
- sinuses 715 Egyptian mummies 8
Drawings 5 Ehrlich 24
Dripping candle 574 Eighth nerve toxicity 792
Droplet nucleus 267 Ejaculatory duct obstruction 717
DR-RFLP 88 Elbow joint tuberculosis 595
Drug absorption, poor 827 Elbow tuberculosis 621
Drug Elderly 52
- abusers 97 Electrocautery 343
- cost of 828 Electrolyte abnormalities 754
- induced liver disease 788 Electrosurgery 346
- interaction 789 - endobronchial 346
- most important 791 Elephants 896,900
- second line drugs 828, 831 ELISA 129, 188,842,843
- sensitivity patterns 825 - sensitivities 846
- side effects, increased 471 - tests 844
Drug regimen 308 ELISPOT 843
- initial 829 - technique 849
- standard 674 EmbAB gene 792
- for treatment of tuberculosis in HIV-negative individuals Emergency decompression 524
468 Emergency operative management 710
Drug resistance 472,782,785 EMG 600
- acquired 811 Emphysema, paraciatricial 373
- epidemiology of 51 Empyema 315
- in M. tuberculosis 202 - persistencet 386
- patterns 816 - tuberculous 403
- in prisons 816 Encephalitis 539
- risk of 803 Encephalomalacia 552
Drug-resistant tuberculosis in childhood 257 Encephalopathy, tuberculous 541
Drug-susceptibility 787 Endemic areas 306
- testing 810 Endobronchial 334
- - conventional 202 Endobronchial disease 371
DSA (digital subtraction angiography) 552,561 Endobronchial involvement 315
DTH (tissue-damaging delayed hypersensitivity) 266,276, Endobronchial tuberculosis (see also tuberculosis) 329,386,
899 387,681
- reaction 232 - incidence of 330
DTPA 419,421 Endocarditis 431
Duodenal obstruction 683 - tuberculous 439
Duodenal tuberculosis 673 Endocrine perturbations related to systemic tuberculosis
Duodenocolic fistula 683 761
Duodenum, tuberculosis 683 Endocrine tuberculosis see tuberculosis
Duration of treatment see treatment Endogenous reactivation 313,314
Dyspareunia 722 Endogenous reinfection 330
Dysregulation of the cortisol-cortisone 145 Endometrial aspirates 720
Endometritis, tuberculous 720
Endometrium 718
E Endopthalmitis 735
Eales' disease and tuberculin hypersensitivity 735 Endoscopic examination 662
Ear, tubercular infection 747 Endoscopy, upper gastrointestinal 672
Early bactericidal activity test (EBA) 840 End-plate destruction 514
Early changes of the disc 518 Enema, small bowel 664
EBA (early bactericidal activity test) 840 Engineered live organisms 887
Ebers papyrus 10,445 England 876
EeG 432 Enhancement 557
910 Subject Index

Enhancement of nerve roots 575 F

- of nerve roots 541 Facial palsy in children 564
Enzyme-inducing effect 793 Fallopian tube 718
Eosinophilia 792 - tuberculosis 719
Ependymitis 550 Familial syndrome 171
- adhesive 539 Farm worker 106
Epidemiological typing 78 Fastrack diagnostic service 819
Epidemiology 301,330,349,445,482,536,588,599,627,659, FDA 787
700,710,724,771 Feasibility 473
- in developed countries 76,77 Features and investigations 719,721
- in Saudi Arabia 45 Female genital tract 158
Epididymal enlargement 713 Female genital tuberculosis 717
Epididymitis 711 - epidemiology 717
- tuberculous 712 Fertility, primary 722
Epididymo-orchitis 711,713 Fetal ingestion 303
Epidural granulomatous mass, isolated 573 FGTB, treatment of 723
Epidural tuberculous infection 539 Fibrosis 144,332
Episodic hematogenous spread 276 - parenchymal 363, 704
Epithelial barrier 883 - pulmonary 379
Epitope - and strictures 705
- multiple 885 Fibrostenotic 332-334
- promiscuous immunogenic 885 Fibrous stricture 662
Epituberculosis 331 Financial problems 828
Epsilometer ( E-) test 128 Fine needle aspiration (FNA) 160,322,448,716,756
Erythema induratum 644 - and biopsies 713
Erythema nodosum 597,645 - transrectal 716
ESAT-6 (6 kDa) 882,886 - transthoracic image guided 450
Esophageal tuberculosis 672 - under ultrasound guidance 716
Esophagus Fine needle aspiration cytology (FNAC) 448,742
- phrenic nerve 269 First dorsal retinacular compartment 599
- tuberculosis 683 First trimester 307
Ethambutol 204,789,792 Fish 893
- emission of 821 Fish tank granuloma 175
- inclusion of 821 Fistula 683
- resistance 792 - appendicular 702
Ethionamide 794 - bronchoglandular 335
Ethnic and racial difference 274,700,701 - nephrocutaneous 703
Ethnic immigrant 301 - tubointestinal 720
Etiology 645 - urethral 703
Etionamide 787 - vesicovaginal 719
Etiopathogenesis 628 Fistulogram 318,375,694
Euthyroid syndrome, sick 757,764 Fistulous (or gandular) 332
Ex vivo screen 820 FLAIR (fluid attenuation inversion recovery) 550
Excision arthroplasty 596 Flan pain 703
Excretion 419 Flank tenderness 703
Exotic dairy breeds 77 Flat bones 597
Extensive patterns of resistance 399 Fleischner's sign 664, 684
Extent of involvement 422 Flow rates 389
External disease 731 Fluid attenuation inversion recovery (FLAIR) 550
Extrabronchial 332 Fluid bronchogram 369
Extradural granulomatous 486 Flu-like syndrome 791
Extramedullary disease 571 18 fluorodeoxyglucose position emission tomography 418
Extrapleural pneumonectomy 406 Fluorodeoxyglucose position emission tomography 415,421
Extra-pulmonary disease 445 Fluoroquinolones 204, 793
Extrapulmonary sites 304 FNA (fine needle aspiration) 160,322,448,716,756
Extrapulmonary tuberculosis, incidence of (see also FNAC (fine needle aspiration cytology) 448,742
tuberculosis) 48 Focal epilepsy 556
Extraspinal bone, diagnosis 590 Foci, metastatic 363
Extrinsic compression 341 Foot, tuberculosis 592, 618
Eyelid Foreign-born 863,876
- disease 732 Fortress mycobacterium, breaching the wall of 819
- tuberculosis 732 Fructose measurement 717
Function 419
Subject Index 911

Functional imaging 411 Government commitment 805,858

Funds 856 Gradenigo's syndrome 747
Fungal disease 578 Granular type 332,335
Fungal infection 342,461 Granulocyte macrophage colony-stimulating factor
Fungoid cheesy form 610 (GM-CSE) 178,886
Fungus ball 378 Granuloma 154
- caseating 275
G - miliary 701
Gadolinium enhanced images 550 - subependymal 537
Galen of Pergamon 15,481 - subpial 537
Gallium 67 see gallium citrate Granulomatous diseases, other 579
Gallium citrate (67 Ga) 412,416,420,425,432,683 Granulomatous response 459
Gas exchange 389 Grape-duster like 537
Gas trapping 389 Great mimicker 681
Gatifioxacin 793 Great white plague 455
Gelationous material 330 Grepafioxacin 793
Gene expression 821 GRF (glomerular filtration rate) 420
Genes and mutations 817 Groundglass opacities (GGs) 296
Genes encoding enzymes 819 Growth hormone (GH) 758,765
Genetic loci 820 Guidlines 310
Genital tract - and cateria 268
- female 158 Gumma, spinal 579
- tuberculosis 702, 717 GUTB (genitourinary tuberculosis) 699
Genitourinary tract (see also MAC, M. chelonae) 165,175, - a large series of one thousand patients with suspected
893 GUTB 707
Genitourinary tuberculosis (GUTB) 699
Genome microarray technologies 62
Genomes, bacterial 58 H
Genomic sequence 882 HAART (highly active anti-retroviral therapy) 470,471
Genomics 82 Haemophilus infiuezae 565
- comparative 78 Haly Abbas 17
Genotypic detection 818 Hand, tuberculosis 618
Genotyping methods 57 Harun AI-Rashid 18
Geographic distribution 57 Health care
Geographic distribution and dissemination 67 - prenatal 306
GFR (glomerular filtration rate) 420 - quality of 474
GG (groundglass opacity) 296 - workers 101
GH (growth hormone) 758,765 Health sector reform 808
Ghon complex 156,267 Healthy individuals 136
Ghon focus 246,266,330 Heamoptysis 378
Gilbert's syndrome 789 Hearing loss 747
Giovanni Cosimo Bonomo 18 Heart valves, prosthetic 175
Gland, small fibrotic 753 Heart, tuberculosis of the 431
Gliosis 537 Heat shock proteins 821, 886
Global drug facility 866,870 Hematogenous 350
Global epidemiology 33,881 Hematogenous dissemination 273,758
- of drug-resistant TB 810 Hematogenous route 701
Global fund 870 Hematogenous seeding 331
Global TB Hematologic abnormalities, prognostic significance of
- epidemic 890 219
- prevention/treatment 882 Hematologic disorders 226
Glomerular filtration rate (GFR) 420 Hematologic findings 213
Glorious Galen 481 - mechanisms of 217
Glove finger appearance 721 Hematopoiesis 217
Glucocorticoid therapy, adjunctive 675 Hematuria 703,712
Glycolipid - painless 703
- antigen 190,843 Hemisphere, cerebral 538
- purified extracted 188 Hemodialysis 89
GM-CSE (granulocyte macrophage colony-stimulating - patients 53
factor) 178, 886 Hemodynamic alterations 438
- expressing adenoviral vector 886 Hemodynamic changes 432
Gonadotropin 758,764 Hemodynamic findings 438
Gout 791 Hemophagocytic syndrome 217, 223
912 Subject Index

Hemoptysis 319 - serostatus, BCG vaccination 475

- massive 315,395,773 - sputum smear in 467
- - in tuberculosis 316 - TB co-epidemic 808
- in pulmonary tuberculosis 316 - transmission to infants 308
- significant 386 - and tuberculosis 456, 824
Hemorrhage - - governmental responses to the threat posed by 475
- intraocular 735 - tuberculosis and co-infection 455
- intraparenchymal 540 - and tuberculous lymphadenitis 446
Henle, Jacob 18 - via breast milk 307
Hepaticportal 447 HLA haplotypes 885
Hepatitis 785 Hollow viscus 681
- fatal 788,790 Homeless shelters 824
- granulomatous 671 Homosexuals with HIV 711
- of rifampin 791 Hospital control measures 864
- viral 789 Hospital ward, open 825
Hepatosplenomegaly 252 Hospitalization, brief 861
Hepatotoxicity 788,789,791 Hospitals in Europe 824
Herxheimer-type reaction 474 Host 894
Heteroduplex analysis 818 - defense 168,785
High prevalence 38 - immune response 314
High resolution CT 296, 360 - pathogenesis and diagnosis 894
High-burden countries 38 Households contact 252
High-efficiency masks 865 HPA (hybridisation protection assay) 128
Highly active anti-retroviral therapy (HAART) 470,471 Hsp70, multiple human T-cell epitopes 886
High-risk groups 723 Human host 78
Hilar lymph node 331 Human immune defects 169
Hilar lymphadenopathy 269 Human immunodeficiency virus (see also HIV) 52,445
- pleural effusions 462 Humoral and cell-mediated immune responses 889
Hilaradenopathy 268 Hybridisation protection assay (HPA) 128
Hip, tuberculosis 617 Hydrocephalus 277,539,543,550,551
Hippocrates 15 Hydronephrosis, bilateral 710
Hippocratic doctrine 18 Hydropneumothorax 375
Hippuran 419 Hypercalcemia 761,762
Histology 647 - mechanism of 761
- findings 448 - treatment of 762
Histopathological findings 354,487 Hypercapnia 389
Histopathology 353,631-635,637,639,641,643-646,649-653, Hyperinflation 246
899 Hypernatremia 763
- of intracranial tuberculosis 548 Hyperprolactinemia 758
Histoplasmosis 565 Hypersensitivity, cutaneous 471
- gastrointestinal 662 Hypertension, pulmonary 389,399
Historical aspects 15 Hypertonic saline 763
History 648 Hypertrophiclesion 661
- BCG 27 Hyponatremia 540,762
- chemotherapy 27 - treatment of 763
- and physical findings 436 Hypophysitis, granulomatous 758,759
HIVIAIDS (see also human immunodeficiency virus) 445 Hypopituitarism 758
- childhood tuberculosis 464, 251 Hyporeactive response 459
- children 251 Hypothalamo-pituitary-adrenal axis 145,764
- co-infection 448,451,866 Hypothyroidism, primary 756
- epidemic 275,302,304,329,350,468,772 Hypoventilation, regional patterns of 415
- homosexuals 711 Hypoxemia, arterial 389
- impact 37 Hysterosalpingography 720,721
- individuals 193
- infection, hematologic complications 220
- pandemic 855,866 I
- patients 700 123 I (sodium 123 iodide) 415
- - co-infected with M.bovis 315 Ibn barmak 18
- patients with tuberculosis, immune interactions in 459 ICSI (intracycoplasmic sperm injection) 712
- pregnant women 306 ICT tuberculosis test 130,192
- related tuberculosis Identification of distant sites 417
- - in children 465 Idiosyncratic reactions 219
- - clinical features of 460 IFN, systemic 178
Subject Index 913

I1dl kinase defects 170 implantation, direct 331

IFN-y 168,351,842 In vitro antibody production (see also IVAP) 849
- control ofNTM 168 In vitro fertilization 720
- production 236 In vivo phenotype 820
- response 232 - targeting persistent organisms 820
- therapy, aerosolized 178 Incidence 265,274,301,629,632,635,636,642-645
IFN-y receptor 1 deficiency 169 - annual 856
IFN-y receptor 2 deficiency 170 Incision and drainage 743
IgE antibody 460 Incompetent ileocecal valve 664
IgG antibodies 192,843,846 Incubation procedure 122
IL-2 887 111 Indium (111 In) radiopharmaceuticals 415
IL-4 135,139,460 India 811,871
IL-I0 138 Indian subcontinent, China, South East Asia 457
IL-12 135 Infant feeding, affordable 308
- recombinant 887 Infarction
- therapeutic use 178 - cerebral 551
IL-12 p40 deficiency 170 - extensive cerebral cortical 552
IL-12 receptor ~1 deficiency 170 - thalamic infarction 551
IL-15 887 Infected sputum, swallowing of 660
Ileocecal region 661 Infection control 823
Ileum, tuberculosis 684 - policies 824
Ill-defined air space shadowing 379 Infection
Imaging 564 - annual risk of 856
- of brain and spinal cord tuberculosis 547 - exposure to 33,34
- of complications of gastrointestinal TB 694 - from milk 330
- of gastrointestinal tuberculosis 679 - granulomatous 487
- findings 450 - incidence of 34
- of musculoskeletal tuberculosis 605 - intracellular 887
- of tuberculous meningitis 541 -latent 862
Imaging characteristics - -lifelong latent 857
- ofTBM 549 - non-tuberculous 380,627
- of tuberculomas 556 - - other non-tuberculous 381
Imaging features 485,494, 713 - prevalence of 34, 49
- of foot and ankle tuberculosis 593 - primary 349
- of knee tuberculosis 592 - rate, tuberculosis 803
- of pleural tuberculosis 354 - transmission 456
- of post-primary pulmonary TB 320 - tuberculous 808
- of tuberculosis of the hip 591 - worldwide 34
- unusual 269 Infective dose 267
Imaging methods in spinal neurotuberculosis 571 Infertility 712,717,719
Immigrants 100,445 - secondary 702
- illegal 98 infiltrations 369
Immobilization, longed 518 Infiltrative 332
Immune adjuvant 883 Infiltrative-proliferative 332
Immune competition and suppression 882 Inflammation 330
Immune modulatory cytokine proteins/transgenes 883 Infriltrates, pulmonary 380
Immune reconstitution syndrome 471 Inhalation 246, 302
Immunity Injections, painful 792
- long-lasting 882 Inoculation procedure 122
- protective 266 Insidious begining 432
Immunization Instrumentation 523
- mucosal route of 887 Integral membrane antigens of M. habana TMC 5135 192
- protocols 882 intensive phase, initial 468
- repeated 882 Intercostal nerve 483
Immunoassays 235,339 Interferon
Immunocompromised 220 -level 434
Immunogenic antigen ESAT-6 882 - nebulised 825
Immunological considerations 883 Interleukin 351
Immunomagnetic separation 119 Interleukin (IL-12)-12 136,168,884
Immunomodulation 825 Intermetaphyseal communicating vessels 485
Immunosorbent assay, enzyme-linked 130,542 Intermetaphyseal communication 494
Immunosuppression 94,175 Intermittent treatment 257
Immunotherapy 146,177 Internal jugular vein 744
914 Subject Index

International Union Against Tuberculosis and Lung Disease Joint tuberculosis

(IUATLD) 84,801,810,855 - extraspinal bone 590
Interventional management 343 - sacroiliac 594
Intervertebral herniation of disc 494 - sternoclavicular 596
intestinal obstructions, subacute 682
Intestine 157,330 K
Intrabronchial 332 Kanamycin 792
Intrabronchial perforations 334 Kansasii, mycobacterial 381
Intracanalicular spread 711 Kaposi's sarcoma 461
Intracellular killing, poor 459 KatG gene 790
Intracellular pathogen 883 kDa (MPT64) 886
Intracranial pressure 249 30 kDa antigen 189
- increased 542 38KDa antigen b 236
Intracranial sapce-occupying lesions 555 45/47 kDa antigen complex 190
Intracytoplasmic sperm injection (ICSI) 712 916 kDa antigen 189
Intradermal route 882 919 kDa antigen 189
Intramedullary 538,581 Keratitis, interstitial 733
Intraparenchymal 538 Kerley B lines 436
Intrapleural immune response, compartmentalized 351 Kidney, hydronephrotic solitary 710
Intrauterine adhesions 721 Kinetics
Intrauterine transmission 702 - of circulating immune complexes during treatment 46
Intravenous excretory urography 704 - on early treatment outcomes 846
Intravenous urography (IVU) 703,704,713 King's evil 25,445
Invasion, direct 350 King's touch 25,445
Invasive diagnostic intervention (CNS biopsy) 543 Kissing sequestra 611
Investigation 724,776 Knee, tuberculosis of the 591, 616
- of female genital tuberculosis 722 Koch phenomenon 22,134,142,143
Ipsilateral hilar region 369 Koch, Robert 21
Ipwy, tomb of 5 Koch's oostulates 21
Irregular period 719 Kussmaul's sign 436
IS6100 80 Kyphosis 484,518
IS6110-PCR 707 - mobile 524
- assay 201 - progressive 523
IS6110-RFLP 59,85 - stabilization of 523
- typing 60 - surgical treatment of 524
Isemann, Michael 824 - and wedging 494
Islamic medicine 16
Isocitrate lyase 820 L
Isocitrate lyase activity, inhibition of 820 Laboratory
Isocitrate lyase gene 820 - cross-contamination 68
Isolation measures 103 - investigations 646
Isoniazid 28,785,788,790 - testing 339,609
- associated lupus 790 - workers 104
- diagnostic trials 738 Lactic dehydrogenase (LDH) 669
- neuropathy 790 Laennec's
- resistance to 790, 817 - stethoscope 19
Isotope scanning 614 - tubercles 19
IUATLD (International Union Against Tuberculosis and Lung LAM 846
Disease) 84,801,810,855 Langhans giant cells 459,707
IVAP (in vitro antibody production) 849 Laparoscopic examination 723
- and the ELISPOT 843 Laparoscopy 696,720
- mean levels 849 - with directed biopsies 669
IVU (intravenous urography) 703,704,713 Laparotomy 662,720
Laryngeal involvement 660
Laryngeal nerve
J - current 269
Jails 98 - recurrent 447
Jaundice, obstructive 671 Laryngitis, tuberculous 745
Jejunum, tuberculosis 684 Laryngoscopy 746
Joint Laser photoresection 346
- destruction 588 Late generalized tuberculosis (LGT) 273,275,276
- effusion 588 Late stage appearances of tuberculomas 560
- space narrowing 611 LCR (ligase chain reaction) 125,156
Subject Index 915

LDH (lactic dehydrogenase) 669 - biopsy, transbronchial 321

- serum-ascitic fluid ratio of 669 - cancer 320,342
Left heart failure 436 - collapse therapy 26
Leptomeningitis - fibrosis 316
- infective 575 - infiltrates and consolidation 268
- suprasellar 554 - lower lobes of the 460
Lesion - opacities, nodular 296
- calcifying 560 - perfusion scintigraphy 415
- miliary 462 - resection 319,390
- multiple 422 - shadowing, miliary 276
- myelitic 570 - volume, loss of 363
- ulcerative 661 Lung disease
- ulcerohypertrophic 662 - parenchymal 385
Leukemoid reactions 215 - restrictive 385
Leukocytes 418 Lupus, isoniazid-associated 790
Leukopenia 215,793,794 Lupus miliaris disseminatus faciel (acne agminata) 647
Leukorrhea 719 Lupus vulgaris 632,746
Levofloxacin 204,793 - mucosal 633
LGT (late generalized tuberculosis) 273,275,276 Lympadenitis
Lichen scrofulosorum 643 - non-tuberculous mycobacterial 448,451
Lifetime risk 785 - outbreak of 261
Ligase chain reaction (LCR) 125,156 Lymph node 157,246,895
Likehood of activity 370 - aspirates 448
Line probe assay (LIPA) 128 - calcification 450
Linezolid 794 - classification of 362
LIPA (line probe assay) 128,819 - enlargement, paratracheal 269
Lipoarabinomannan 191 - inguinal 447
Lipo-oligosaccharide (LOS) 843 - mediastinal 683
Lipopolysaccharide antigen 192 - regional 331
Lipoproteins 140 - submandibular 598
Liquefaction and decontamination 119 - superficial 896
Liquid automated culture system 121 - supratrochlear 595
Liver Lymph, mediastinal 369
- abscess, tuberculous 671 Lymphadenitis 266
- failure 789 - NTM 173
- involvement 671 - treatment of mycobacterial 451
- normal 788 - tubercular cervical 742
- toxicity 788 Lymphadenopathy 250,252,269,362,464
- transaminases 297 - asymmetrical 463
- transplantation 789 - mesenteric 896
- - and tuberculosis 789 Lymphadentitis 350
- tuberculosis 691 - non-tuberculous 447
Liver disease 787, 789 Lymphatic channels 331
- therapy in patients with 675 Lymphatic spread 318
- and tuberculous peritonitis 660 Lymphocyte or memory T-cell-stimulating cytokines 887
Lobar collapse 246, 363 Lymphocytic hypophysitis 758
Lobar hyper-inflation 372 Lymphopenia and immunosuppression 219
Lobectomy 319,395,400 Lymphotactin 887
Lobular emphysematous 373
Lobulated pleural collections 375
Localized wheezes 316 M
Long flight 267 M. africanum 79,116,587
Longitudinal ligaments 482,485 M.avium 793
Long-term memory T-cells 885 M. avium complex (MAC) 165
Long-term neurological deficits 483 M.bovis 315,446,482,587,701,746,772,893,894
Long-term sick adults 828 - BCG 81
LOS (lipo-oligosaccharide) 843 - developing countries 901
Lowenstein-Jensen (BACTEC, Ln 120,707 - epidemiology 75,900
Low-income high-burden areas 857 - in humans 315,900
Luciferase-based repoter phage 128 - molecular typing 87
Lumbar puncture 542 - multi-drug resistant 826
Luminal narrowing 333 - transmission 88
Lung 156 - wildlife reservoir of 901
916 Subject Index

M. chelonae (genitourinary tract, MAC) 165,175,893 - areas of 811

M. marinum infection 649 - cost of 828
M. microti 79,116,893 - definition of 824
M. tuberculosis 893 - epidemiology 809
- antigens 841 - financial challenges of 827
- drug resistance, molecular basis of 202 - in HIV, sporadic 827
- genome sequence 819 - individual patient with 828
- skin infections, treatment of 654 - isolation of infectious 828
MAC (M. avium complex, M. chelonae, genitourinary tract) - outbreak of 472,824,826
165,175,893 - possible 826
- extrapulmonary 175 - spread of 810
- pulmonary 176 - treatment of 821
Macrolides 175, 176 Mechanical ventilation 297
Macrophage 459 Mediastinal 447
- apoptosis 141 Mediastinal adenopathy 380
- CD4+ T cell 276 Mediastinallymph 369
- depletion of 458 Mediastinal lymph nodes 683
- epithelioid 458 Mediastinoscopy 450
- function 106 Mediastinum 483
- functional 141 Medicalmanagement 747
- mycobactericidal mechanisms within 140 Medical papyrus 10
- phagocytic 351 Medical papyrus of Ebers (Ebers papyrus) 10,445
- phagocytic process 885 Medical Research Council (MRC) 482
Macroscopic pathology 156 Medical Research Council of the United Kingdom 549
Magnetic resonance arteriography (MRA) 552 Medical treatment, preoperative 396
Magnetic resonance imaging (MRI) 249,437,440,486,541, Medulla oblongata 573
550,557,572,616,663,682,707,716,758 Memory immunity 885
MAl (mycobacterial a-intercellular complex) 380 Memory T cells 884
Maintaining airway patency 344 Meningeal-parenchymal type 538
Major blood vessels, compression of 477 Meningioma 561
Major obstacles to successful development of improved TB Meningitis 277,538,561
vaccines 883 - spinal 569
Malabsorption 474 - and tuberculomas in miliary/disseminated tuberculosis 277
Malawi 457 Menopause 721
Male genital tract 158 Menorrhagia 719
Male genital tuberculosis 710 Menstrual disturbances 717
- diagnosis of 713 Mesentery 447,663
Malignant diseases 96 Mesothelial cell count 352
Malignant tumors 565 Mesue 17
Malnutrition 97 Metastatic lesions 363
Management 432,434,439-441,543 MGIT (mycobacterial growth indicator tube) 127
- considerations in HIV-positive individuals 470 MIC (minimum inhibitory concentration) 817
- proactive 856 Microbiology 536
- surgical 346 - findings 448
Mandible 598 Microlaryngoscopy 746
Mantoux test 233,841 Micronodules, pulmonary 273
Marginal erosions 611 Microscopic examination 118
Marker, phenothypic 57 Microscopic pathology 154
Marrow infiltrate 224 Microscopy 466
Masks, negative-pressure ventilation 856 Middlebrook 7HI0 and 7H11 media 120
Mass-like lesion 340 Midexpiratory flow 389
Mastoiditis 564 Migrants, undocumented 864
Matting 447 Miliary cerebral tuberculomas 562
- of bowelloops 666 Miliary form 671,747
Maximum voluntary ventilation (MVV) 390 Miliary granulomas 701
MB/BacT system 121 Miliary lesions 462
MBBacT microbial detection system 127 Miliary tuberculosis (see also tuberculosis) 159,269,273,276,
100-mBq99mDMSA 707 363,364,389,440,441
200-mBqTc99m DTPA 707 - epidemiology 274
MDR-T, transmission of 832 - laboratory diagnosis 296
MDR-TB (multi-drug resistant tuberculosis) 202,472,808, - notification of 275
817,877 - of the skin, acute 631
- in Africa 812 - with multiple small tuberculi 439
Subject Index 917

Military spread 380 Mutant strain 821

Milk Mutations 790
- borne infection 77 MVV (maximum voluntary ventilation) 390
- consumption, raw milk 587 Mycetoma 316
- contamination 33,315,660,893 - formation 378
- and human disease 893 Mycetoma Rasmussen 371
- infection from 330 Mycobacteria 115
- pasteurization of 659 - characteristics of 116
Mineralocorticoid deficiency 763 - classification of 628
Mines 106 - environmental 145
Minimum inhibitory concentration (MIC) 817 - nontuberculous 163,187,379
MIRU-typing 85 - rapid growing 165, 177, 888
M-Mode echocardiography 438 - slowly growing 165
Mnasoura University Hospital 707 Mycobacteria other than tuberculosis (MOTT) 163
Molecular beacons 206 Mycobacterial a-intercellular complex (MAl) 380
Molecular diagnostic techniques 817 Mycobacterial culture and sensitivity testing 859
Molecular epidemiological studies 314 Mycobacterial cutaneous infections, atypical 648
Molecular epidemiology 57,78 Mycobacterial growth indicator tube (MGIT) system 127
- future 69 Mycobacterial infection and human immunocompromised
Molecular genotyping 57 virus (HIV) 379
Molecular methods 466 Mycobacteriological burden 789
Molecular techniques 738 Mycobacteriosis, classification of nontuberculous 163
- for identification of mycobacteria 123 Mycobacterium avium complex (MAC) 165,175,893
Monarthritis, tuberculous 588 Mycobacterium avium-intracellular complex infection 51
Monitoring treatment efficacy 839 Mycobacterium bovis 75,107,274,277,747
Morbidity 35,254 - genotyping 86
- increased rate 470 Mycobacterium fortuitum complex infection 652
Morphology 894 Mycobacterium haemophilum infection 653
Mortality 36, 254 Mycobacterium kansasii infection 650
- during infancy 254 Mycobacterium scrifulaceum infection 651
- increased rate 470 Mycobacterium tuberculosis 700
- obstetric 254 - genetic approach to the detection of drug resistance 127
Moth eaten appearance 746 - genetics 78
Mother, nursing 308 Mycobacterium tuberculosis complex (MTC) 893
MOTT (Mycobacteria other than tuberculosis) 163 Mycobacterium ulcerans infections 649
Movement disorder 539 Mycobacterium, non-tuberculous 711,720
Moxifloxacin 793 Mycocytes, apoptotic 884
MPB64 antigen, dermal response to 192 Mycotic aneurysm, tuberculous 771
MPT59(Ag85b) 236 Myelitis 539
MPT64 (26 kDa) 235,236,886 Myelofibrosis 215
MRA (magnetic resonance arteriography) 552 Myelography 486,514,571
MRC (Medical Research Council) 482 Myelopathy 541
MRI (magnetic resonance imaging) 249, 437,440, 486, 541, Myeloradiculopathy, tuberculous 571
550,557,572,616,663,687,707,716,758 Myocarditis 431
- of the adrenal glands 753 - tuberculous 440
- of epididymal and testicular tuberculosis 716 Myositis, tuberculous 599,600,610
- features 514
MTC (mycobacterium tuberculosis complex) 893
MTD test 124 N
Mucosal route of immunization 887 Napkin ring sign 687
Mucosal ulceration 664 National Jewish Center group 396
Dr. Muhny ad-Din at-Tatawi 16 National tuberculosis control 782
Multi-drug resistance 826 National Tuberculosis Control Programmes 476
- impact of 38 National Tuberculosis Programme (NTP) 805,860
Multi-drug resistant tuberculosis see Tuberculosis, multi- Nationwide coverage 805
drug resistant and MDR- TB Native-born 863
Multidrug-resistant strains 329,396 Natural Course of Endobronchial Tuberculosis 337
Multiple intrahepatic biliary strictures 671 Natural human knockouts 136
Multiple lesions 422 Nd-YAG laser resection 343
Multiple sites 447 Neck abscess, tuberculous 742
Multiple sclerosis 582 Necrosis
Multiple small nodular opacities 371 - caseating 157
Mummy of Nespapheran 6 - gummatous 560
918 Subject Index

Needle biopsy (under fluoroscopic of CT guidance) 514,522 Nucleic acid probes 124
Neelsen 24 Nursing Mother 308
Negative pressure room 823
Negative pressure unit 829 o
Neoplasia 320 Obstruction
Nephrectomy, partial 710 - bronchial 331
Nephrostomy, percutaneous 710 - endobronchial 341
Nephrotoxicity 792 - pure 389
Nephroureterectomy 710 - subacute 670
Nerve root Obstructive 331,385
- enhancement 575 Obstructive-restrictive pattern 387,389
- entrapment 541 Occipitocervial junction 486, 514
- systems 571 Occupational hazards 93,101
Neural arch tuberculous infection 494 Occupational therapy 595
Neuritis, optic 739,792 Ocular tuberculosis 731
Neurobrucellosis, intracranial 567 Oculomotor palsies 540
Neurocysticerosis 581 Ofloxacin 204, 825
Neurological deficits 483 1,25 (OH)2D3 766
Neurological deterioration, acute 519 25 (OH)D3 766
Neuromyelitis optica syndrome 541 Old tuberculin (OT) 231
Neuropathy, optic 735 Omental cakes (omental thickening) 666, 688
Neuroretinitis, tuberculous 735 Omental thickening (omental cake) 666, 668
Neurosarcoidosis 566,579 Omentum 663
Neurosyphilis 579 Operative procedures, types of 400
Neurotoxicity 792 Operative treatment 519
New cell-wall inhibitors 820 Optic 540
New smear-negative pulmonary tuberculosis 822 Oral cavitiy, tubercular infections of 746
New sputum smear-positive tuberculosis 821 Orange discoloration 791
New York City 457 Oropharyngeal mucosa 447
New York's TB control team 874 Oropharynx 746
New York's TB epidemic 875 Osteolytic lesion 610
NFlCB essential modulator (NEMO) 170 Osteomyelitis
Nocardia 115 - extraspinal tuberculous 597
Node - tuberculosis 494
- mesenteric nodes 667 Osteomyelitis of the chest wall, tuberculous 597
- retroperitoneal 667 Osteomyelitis of the ribs, tuberculous 597
Nodular disease 671 Osteomyelitis of the sternum, tuberculous 597
Nodulation, miliary 379 Osteoporosis 611
Nodules Osteosclerosis 611
- diffuse 380 OT (old tuberculin) 231
- erythematous subcutaneous 631 Otitis media, tuberculous 563,747
- miliary 296 Otorhinolaryngeal manifestations 742
Non-eadiometic methods of drug susceptibility testing 127 Otorrhoea 747
Non-human primates 899 Ototoxicity 309
Non-MDR resistance, management of 822 Outcome, postoperative 390
Non-rigid kyphosis 526 Outpatient facilities 259
Nonsteroidal anti-inflammatory drug 433 Ovary, tuberculosis 719
Nontuberculids 646 Overcoming neglect 866
Non-tuberculous infections 380 Overcrowding 101
Non-tuberculous lymphadentitis 447 Over-inflation 363
Non-tuberculous mycobacteria 163,187,379
Non-tuberculous mycobacteriallymphadentitis 448,451 P
Nose, tubercular infection of the 746 P32 antigen 190
Nosocomial transmission 865 Pachymeningitis 554
Novel drug discovery programme 820 Pachymenix 549
Novel inhibitory compounds 820 PAl (primary adrenal insufficiency) 751
NTM disease, radiographic appearance of 172 Pain
NTM infection - abdominal 703
- cutaneous 175 - flank 703
- pulmonary 172 - readicular 538
NTM lymphadenitis 173 Pancreatic head 694
NTM therapy 176 Pancytopenia 216
NTP (National Tuberculosis Programmes) 805,860 Papillae, sloughing of the 704
Subject Index 919

Papilledema 735 - analysis assays 204

Para-aminosalicyclic acid (PAS) 28,781,794 PCR-SSCP see PCR-single strand conformation polymor-
Paracutaneous route 882 phism
Paradoxical phenomenon 560 Pelvic examination, bimnual 719
Paradoxical reactions 471,840 Pelvis, tuberculosis 618
Paradoxical response 451, 543 Pelviureteric junction 701,704
Paradoxical transient worsening 270,361 Pelviureteric obstructions 704
Paradoxical transient worsening phenomenon 365 Pentavelent dimercaptosuccinic acid [(V) DMSA) 415
Paraparesis 485 Penumonia, obstructive 372
Parapharyngeal space abscess 743 Percutaneous debridement 524
Paraplegia 485 Perforation and fistulae 662
- decompressive techniques and treatment of 524 Perfusion 419
Paratracheal 331,362 - abnormality 415
Parenchymal cerebral tuberculosis 554 Pericardial effusion 431
Parenchymal disease 361,369 - and cardiac tamponade 433
Parenchymal infiltration 268 Pericardial friction rub 432
Parenchymal tuberculosis, differential diagnosis 565 Pericardial thickening 437
Paronychia, painless 630 Pericarditis 363,431,463
Parotid gland 747 - acute 431,432
Parotid, tuberculosis 419 - calcific constrictive 431
Partition test 421 - tuberculous 431
PAS (para-aminosalicyclic acid) 28,781,794 Pericardium 269
Pasteur strain 261 - calcification 432,436
Pasteurization 315 - tuberculosis 431
Patella 597 Perilymphadenitis 742
Pathogen, virulent intracellular 459 Peripancreatic 447
Pathogenesis 245,302,330,446,482,536,588,599,629,631, Peripheral nervous system 569
632,635,636,639,642-645,647,649,652,661,666,670,680, Peritoneal biopsy, blind 669
711,772,895 Peritoneum 663
Pathogenesis and pathology of renal tuberculosis 701 Peritonitis
- of female genital tuberculosis 718 - plastic 666
- and pathology 313 - purulent 666
- of primary pulmonary tuberculosis 266 - tuberculous 417, 660, 666, 696, 702
- of tuberculous pleural effusion 350 Persistent unilateral wheezing 339
Pathogenetic mechanisms 762 Persistently culture positive 399
Pathology 331,537,606 Peru 873
- and pathogenesis of miliary/disseminated tuberculosis 275 PET imaging 419
- of tuberculosis 153 Peyer's patches 670
- - in HIV-infected individuals 458 PFGE (pulsed field gel electrophoresis) 79
Pathozyme TB complex test 130 PGLTB1 843
PATHOZYME-MYCO PGRS-RFLP 81
- IgA test 130 PGRS typing 59
- IgG test 130 Phage genome 818
- IgM test 130 Phagocytosis, impaired 93
Patient Phagosome 883
- compliance 785,840 Pharyngeal involvement 896
- immunocompromised 433 Pharynx, tubercular infections of 746
- - with HIV 722 Phenotypic methods 818
- immunosuppressed 663 Phenytoin 790
PCP (pneumocystis carinii pneumonia) 461 - concentrations 790
PCR (see also polymerase chain reaction) 542,899 Phlyctenulosis 733
- in ancient Egyptian population 8 Photodynamic therapy 346
- based methods 59 Photoresection 343
- detection of drug resistance 204 Phthisis 15
- diagnosis 199 Physical examination 339
- - impact of 206 Physical findings 436
- DNA sequencing 204 Physiotherapy 595
- heteroduplex assay 205 Pia-arachnoid 571
- LiPA 206 Pituitary tuberculoma, diagnosis 759
- of M. tuberculosis 200 Pituitary tumor, non-functioning 758
- of pleural tissue 354 Pituitary-adrenal axis 765
PCR-single strand conformation polymorphism (PCR-SSCP) Placenta 302
450,818,840 - tuberculous 702
920 Subject Index

Plain pelvic radiography 722 Posterior pharyngeal wall 743

Plain radiographs 610,682 Posterior spinal fixation 523
Plain Radiography 485,494,703 Posterior spinal fusion 493,523
Plasmid DNA 888 Posterior subligamentous abscesses 514
- based cytokines 887 Posterior synechiae 733
- based vaccines 889 Posterior-element infection 494
Pleura, calcified 403 Post-genomic era 819
Pleural disease Post-genomic tools 820
- and fibrothorax 386 Post-mortem 441
- in post-primary tuberculosis 375 - examination 898
- in primary tuberculosis 364 Postnatal 254
Pleural effusion 268,269,380 Post-primary disease 275,330
- bilateral 436 Post-primary pulmonary tuberculosis 313,349
Pleural fluid Postprimary TB 370
- culture 353 Post-Rifampin era and emergence of drug-resistant tubercu-
- cytokines 353 losis 396
- examination 352 Post-treatment prophylaxis 472
- loculated 86 Post-tuberculous bronchiectasis 315,316
- for microbiological examination 352 Potential drug recovery 820
Pleural mesothelial cells 351 Pott, Sir Percival 481,493
Pleural rub 352 Pott's disease 519
Pleural space Pott's puffy tumor 563
- drainage 405 Poverty 40, 302,456
- surgery for tuberculous infection of the 403 Power doppler sonography 450
Pleural thickness 386 PPD (purified protein derivative) 232,322,354,467,738,862
Pleural tuberculosis, primary 349 - induration 233
Pleurisy 463 - strenght 232
Plombage 27 PPD skin test 306
pncA genes, mutated 791 - conversion of 266
Pneumoconiousis 107 PPD testing
Pneumocystis carinii pneumonia (PCP) 461 - adverse reaction 235
Pneumonectomy 319,395,400 - methodology 233
- extrapleural 406 PPD-S 232
Pneumonia 341 Practical problems DOTS faces in India 812
- tuberculous 369 Pre-antibiotic era 274
Pneumonitis, obstructive 331 Predynastic period 3
Pneumothorax Pre-existing chronic obstructive pulmonary 379
- bilaterial 293 Pregnancy 301,787
- as complication of post-primary pulmonary TB 318 - outcome of 301
- in miliary/disseminated tuberculousis 293 - termination of 309
- spontaneous 371 - treatment
Political change 808 - - of active TB during 308
Political commitment 856,858 - - of multidrug-resistant TB 309
Political will 856 - tuberculosis 301
Polycythemia 215 - - interrelation 303
Polymerase chain reaction (see also PCR) 297,434, 450, 707, Preoperative evaluation 390,399
713 Presentation
- and genetic probes 339 - unusual 269
Polymorphism 171 - usual 269
Polyneuropathy 571 Presenting symptoms 719
Polyphosphonate 422 Pretreatment era 660
Polyps and masses 687 Preventive measures 103
Poncet's disease 596 Pre-vertebral soft tissue 743
Poor developing countries, anti-retroviral therapy 469 Primary adrenal insufficiency (PAl) 751
Population Primary complex 330,701
- based molecular epidemiological study 63 Primary hematogenous dissemenation 537
- certain 51 Primary pulmonary tuberculosis 331
- movement 869 Prisoner of war 98
Positive smears 542 Private health care sector 868
Post-bronchoscopy sputum 339 Private sector 40
Posterior decompression and internal fixation 524 Productivity, lost 828
Posterior fixation and fusion 519 Prognosis 543
Posterior instrumentation 523 - and predictors 825
Subject Index 921

Progressive primary 275 - tuberculosis 130

Prolapse, vaginal 717 Rash 790, 792
Properties 894 Rasmussen aneurysm 317,773
Prophylaxis regimen 260 Raw milk consumption 587
Proportion methods 817 REA (restriction enzyme analysis) 79,86
Proptosis 736 Reactivation 701
Prostate, tuberculosis 712 - of primary complex lesions 458
Prostatic mass, cystic 712 - tuberculosis 330
Protection versus immunopathology 134 Reactive nitrogen intermediates 140
Protective clothing 105 Reactive oxygen intermediates 140
Protective measures 103 Reanastomosis 710
Protein concentration 249 Recent conversion 785
Protein levels 542 Reconstruction, urethral 710
Pseudo-aneurysm 773 Reconstructive surgery 709,710
- mycotic 317 Rectal examination 712
Psoas abscess 494,522 Rectal tenesmus 713
Psoas muscles, absence 485 Reduced transmission 803
Psychosis 794 Reflux, ureterovesical 704
Pubic symphysis 594 Regimen, shorter combination 473
Public health 900 Re-infection 57,458,701
- and the wider economy, implications for 830 - exogenous 66, 313, 314
Public health issues 830 Rejection 789
Pulmonary artery, peripheral 317 Relapse 57
Pulmonary circulation and Ibn-an-Nafis 16 - rate 822
Pulmonary disease, atypical 460 Renal allografts 702
Pulmonary function test 322,340,399 Renal cortex 701
Pulmonary gangrene, tuberculous 320 Renal failure 789,792
Pulmonary tuberculosis, incidence of 46 Renal function 705
Pulsed field gel electrophoresis (PFGE) 79 - restoration of 710
Punched-out defect 563 Renal parenchyma 157
Puosalpinx, tuberculous 702 - healing 701
Purified protein derivative (PPD) 232 Renal pelvis 701
Purpura, thrombotic thrombocytopenic 227 - calyceal 701
Pyrazinamide (PZA) 203,789,791,825 Renal salt wasting replacement 763
Pyridoxine 790 Renal scintigraphy 707
Pyuria 703 Renal transplant 95
PZA (Pyrazinamide) 203,,789,791,825 Renal tuberculosis 419,702
- bilateral 704
- congenital 702
Q Replication, intracellular 883
Q-IFN 236
Reptiles 893
- assay 236
Resection
- kit 236
- pulmonary 27,402
- test 237
- - for multidrug-resistant tuberculosis 401
Quadriceps muscle, wasting of the 594
Resistance 792
Quality of health care 474
- acquired 810
Quinolone 787,789
- among previously treated cases 810
- isolated 822
R - ratio 817
Racail and ethnic difference 274,700,701 Resource 855
Radiculomyelopathy, tuberculous 570,571 Resource-rich areas 857
Radiculopathy 485,541 Respiratory failure 319,389
Radiographic finding, unusual 268 Respiratoryobstruction 744
Radiography, convertional 359 Restriction enzyme analysis (REA) 79,86
Radioisotope studies 610 Restriction fragment length polymorphism (RFLP) 58
Radiologic presentation 616 Restriction 385
Radiological features, unusual 269 - pure 389
Radiometic method of drug susceptibility testing 126 Retina, tuberculosis 735
Radionuclide thyroid scans 757 Retreatment regimen, empirical 859
Radioscopic scans 683 Retroperitoneal approach 524
RANTES 887 Retroperitoneal tissue planes 494
Rapid test Retropharyngeal mass 483,485
- in routine use 819 Reverse transcriptase PCR (RT-PCR) 819
922 Subject Index

Revised national program 872 - tuberculosis 595

RFLP (restriction fragment length polymorphism) 58 Shunt, ventriculoperitoneal 540
RFLP stain typing 253, 258 Shwartzman reaction 143
Rhazes 17 SIADH (syndrome of inappropriate secretion of anti-diuretic
Rib destruction 375 hormone) 762
Rice bodies 588, 599 Sialogram 419
Rifabutin 792 Side effects 785, 788
Rifampicin 28 Silicosis 106, 785
- resistance 817,818 Sinogram 612
Rifampin 790 Sinonasal 447
- alone 786 Sinus
- based treatments 785 - charging 598
- resistance 791 - cutaneous 448
- RMP 203 - discharging 447,487,630
Rifapentine 793 - sphenoid 758
Right heart catherization 399 - tract 713
Right-sided heart failure 435 - urethral 703
Rigid deformity 526 Sinus formation 447
Ring enhancing lesions 557 - and fistula 683
Ritual circumcision 713 Skeletal tuberculosis, active 422
Roentgenogram, simple 340 Skeleton 425
Rosacea-like tuberculid 647 Skin
rpoB genes 791 - lesions 463
16S rRNA-PCR 707 - and soft-tissue infection 173
RT-PCR (reverse transcriptase PCR) 819 - spinal involvement 486
Russia 876 - test conversion 864, 865
- test interpretation, tuberculosis 234
- testing 231
S - thickness 898
SAAG (serum-ascites albumin gradient) 669 - tuberculosis 627,628
Saccular true aneurismal dilatation 773 - - acute miliary 631
Salivary glands, tuberculosis 747 Skull vault 563
Sanatorial treatment 11 Slim disease 463
Sanatorium-Bomaristan or Deir el-Bahri 26 Small bowel follow-through 664
Satellite granulomas 665 Smear-negative disease 256
Scarring 447 Smear-positive disease 858
Schistosoma 580 Smears and cultures 77
Sciatica 484 Social-economic reform 808
Scintigraphy 486 Socio-economically disadvantaged communities 457
Screening Sodium 123 iodide (123 1) 415
- and control in specialized settings 864 Soft tissue 425
- interventions in resource-rich areas 862 - calcification 486
- of migrants to low-incidence countries 863 Solid media 120
Scrofuloderma 636 Solid organ transplantation 54, 175
Scrotal mass 713 Somatostatin 765
Secretory antigens 885 Sonographically echogenic debris 688
Segmentectomy 395 South Africa 873
Seizures 541 - goldmines 827
Selection, criteria 885 South-East Asia 37,866
Seminal vesicles 713,716 Spacer interspersed direct repeats (SPIDRS) 81
Septicaemia, tuberculous 245 Sparfioxacin 793
Sequestrum formation 483 Special groups and occupational hazards 93
Serodiagnosis 434 Speciation 894
Serological 466 Specimen collection 117
Serum ADA 353 Sperm retrieval 712
Serum-ascites albumin gradient (SAAG) 669 SPIDRS (spacer interspersed direct repeats) 81
Seshen Nufer, tomb of 5 Spinal canal 483,486,514
Sexual intercourse 711,718 - encroachment 486, 514
Sexual transmission 702 Spinal fixation, external 524
Shadowing, parenchymal 361 Spinal fusion 523
Shelters 100 Spinal instabilit) 519,530
Shoulder joint Spinal meningitis, isolated 575
-lesion 595 'ipinal neurotuberculosis 571
Subject Index 923

Spinal tuberculosis 3,350,484,487,489 Suprasellar areas 549

- surgical management 493 Surgery
Spine - indications for 399,519,675
- cysticerosis 581 - role of surgery in the pre-Rifampin era 395
- humped 5 Surgical and collapse procedures 386
Spirochetal disease 565, 579 Surgical findings 758
Spirometric evidence 387 Surgical intervention 451
Spleen, tuberculosis 691 Surgical technique 399
Spoligotyping 59,81 - approach 522
Spondylitis, tuberculous 538 Surgical treatment 402,405
Spread Surveillance, accurate 856
- hematogenous 275,302,660,701,711,718,736 Susceptibility
- miliary 691 - genetics of 136
- subligamentous 518 - testing 822, 840
- submucosal lymphatic 335 SVC (superior vena cava) obstruction 269,419
Sputum Swimming pool granuloma 175
- induction 253 Swine 900
- microbiology of 321 Symphysis pubis 596
- microscopy 859 Symptoms
- monitoring 840 - constitutional 315
- sample of 117 - frequency of 316
- smear in HIV-positive persons 467 Synacthen stimulation 755
Sputum culture 255 Syndrome of inappropriate secretion of anti-diuretic hor-
- conversion 840 mone (SIADH) 762
Stabilization surgery 523 Synechia uteri (Asherman's syndrome) 721
Stable funding 858 Synergistic 792
Staining properties 116,819 Synovectomy, surgical 595
Standard diagnostic approach 858 Systemic tuberculosis, endocrine perturbations related to 761
Standard drug regimen 674
Stenosis 330 T
- bronchial 337,395 T cells
- cicatrical 332 - gammaldelta (r/8) 137
- with fibrosis 332 - regulatory 138
- spinal 485 Tachypnea 277
Stent Tacrolimus 791
- tracheobronchial 344 Tamponade 432
- types 344 Target sign 559
Sternum 598 Targeted populations, specific 192
Steroids 433 Taste, unpleasant 794
Stierlin's sign 664,684 TB control
Stomach, tuberculosis 683 - cervical 530
Straight leg raising test 594 - challenges to global 866
Strain typing methodology 79 - drug development, global alliance for 820,870
Streptomycin 27,787,789 - global partnership to stop TB 869
- and other aminoglycosides 792 - patient, compliance of the treated 840
- STR 204 - vaccine formulations, cytokines as immune adjuvants for
Strictures 664 886
Stridor 743 TBM (tuberculous meningitis) 248,537,539,548
Strut graft 523 - differential diagnosis 565
Studies - imaging characteristics of 549
- angiographic 551 - tuberculoma 554
- postmortem 551 TcDTPA 420
Subarachnoid 551 T-cell
Subluxation 494,530 - activation 883
Subpleural tuberculous lung parenchymal 350 - epitopes, multiple 886
Sub-Saharan Africa 37,302,457,824,855,863,866 - response, immunological 842
Substance abusers 824 Technetium (99 TcM) radiopharmaceuticals 414,420
Subtile radiographic changes 784 TEE (transesophageal echo) 438
Sudden death 441 Tel El-Amarna 5
Superior vena cava (SVC) 269 Ten commandments 824
- obstruction 419 Tendon sheaths
Supervised provision of standard treatment regimen 859 - bones 175
Suppuration 459 - bursae 175
924 Subject Index

- joints 175 - of infection 456

Tenosynovitis, tuberculous 599,610 - mode of 350
Tentorial edge 561 - sexual 702
Teratogenic effect 308, 309 Transpedicular instrumentation 524
Test Transplacental 254
- antimicrobial susceptibility 126 Transplant patients 54
- diagnostic 436 Transplantation 94
- immunochromatagraphic 130 Transthoracic approach 524
- susceptibility 822,840 Transvalvular flow velocities 434
Testis, hypoechoic 715 Traveling 267
TGF-p 138 Treatment 270,354,591,646,647,649-653,674,738,776,825
Thallium chlorid (201 TI) 124,414 - conservative 484,785
Therapeutic outcome 332 - daily 257
Therapeutic response 414 - duration of 787,808,825
- evalution 414 - evolution 25
Therapy - failure 822
- alternative 825 - monitoring 846
- before 846 - - of the response to 738
- climatological and sanatorium therapy 25 - operative 519
- directly observed, short-course 858 - response 324,425
- during 846 - short-course 807
-longer 786 - six-month 786
- preventive 99,310,857 - surgical 405
- - for tuberculosis 473 Tree-on bud appearance 372,373
- prophylactic 473 True worsening 368
Thoracoplasty 26, 406 TV (tuberculin unit) 233,322
Thoracoscopy in tuberculous patients 354 Tubal obstruction 720
Thoracotomy 405,450 Tubercle 20, 115
Thrombocytopenia 215,216,794 Tubercle bacillus, the discovery of the 21
Thrombosis 540 Tubercle, choroidal 733
Thrombotic occlusion 552 Tuberculamate, intracranial 250
Thyroid gland, tuberculosis 747 Tuberculid 642
Thyroid, FNA 757 - lichenoid 648
Thyroiditis - papulonecrotic 642
- subacute 756 Tuberculin 464
- tuberculous 747 - conversion 266, 268
201 TI (thallium chlorid) 124,414 - response and protection 142
Tissue biopsy, CT guided 487 - testing 253
Tissue section 155 Tuberculin skin test (PPD) 322,354,467,738,862
Tissue-damaging delayed hypersensitivity (DTH) 266,276 Tuberculin test 340
TLR (toll-like receptors) 140,170 - negative 253
T-lymphocyte cell 97 Tuberculin unit (TV) 233,322
TNF 351 Tuberculin-positive 245
TNF-a (tumour necrosis factor) 136 Tuberculoma 315,320,361,374,494,537,538,550
- role of 138, 144 - of the cord 572
Toll-like receptors (TLR) 140,170 - imaging characteristics of 556
Tomb of Ankh-my 5 - intracranial 758
Tomb of Seshen Nufer 5 - intrasellar 538
Tonsils 330,447,746 - medullary 538
Toxicity 791, 792 - miliary cerebral 562
- monitoring: plasma levels; audiograms, costs of 828 - myelitic 569
Toxoplasmosis 566 - parenchymal 554
Traction bronchiectasis 370,372 - 'en plaque' lesions 561
Training course 806 - suprasellar 565
Tranction bronchitasis 318 - to treatment, paradoxical response of 560
Transbronchial spread 318 Tuberculosis (see also) 600, 841
Transcriptional signals 819 - activity of pulmonary 379
Transesophageal echo (TEE) 438 - during adolescence 255
Transgene coding 889 - in adult, primary 265
Transgene encoding cytokines 888 - adult-type progressive 330
Transient radiographic progression 379 - in AIDS, miliary/disseminated 223
Transmissibility 69 - in ancient Egypt 3
Transmission 85,259 - in animals 893
Subject Index 925

- antigen, single 885 - - in pregnancy 787

- appendicular 672 - - new trends in diagnosing latent TB 235
- arthritis 609 - - therapy 785
- assessment of pulmonary 379 - - treatment 867
- bilateral choroidal 733 -leptomeningeal 570
- bronchopleural fistula 319 - lung cancer and other neoplasia 320
- century 25 - meningitis 803
- CNS 420 - mesenteric lymph nodes 670
- colonic 687 - microbiology of drug-resistant 816
- congenital 641 - miliary (see also miliary tuberculosis) 159,269,273,276,
- contact investigation 862 363,364,389,440,441
- contagious 856 - - epidemiology 274
- control 801,855 - - in HIV 224
- - global 805 - - treatment 297
- - recent developments in 871 - mortality 802
- cryptic disseminated 273,276 - multi-drug resistant 396,794,809,822,859,867
- cryptic miliary 274 - - epidemiology 809
- development of new drugs in the treatment 819 - - treatment during pregnancy 309
- diagnosis - multifocal 598
- - of miliary /disseminated 293 - - osteoarticular 600
- - of post-primary pulmonary 320 - muscles 610
- disseminated (see also disseminated tuberculosis) 273 - mycobacterial 380
- distribution of the burden of HIV-related 457 - neonatal 302
- effect on circulating hormones 764 - nonspecific tuberculids 644
- endobronchial (see also endobronchial tuberculosis) 329, - obstructive 717
386,387,681 - ocular manifestations of 731
- - in HIV-infected patients 337 - orbital 735
- endocrine 751 - osteomyelitis 609
- - and metabolic manifestations 751 - otorhinolaryngeal aspects 741
- epidemic 802 - outbreak of 64
- esophageal 672 - pancreatic 693
- extrapulmonary (see also extrapulmonary tuberculosis) - parathyroid 760
194,315,446 786 - paucibacillary 784
- - in HIV-positive adults 463 - penile 713
- extraspinal musculoskeletal 587 - peritoneal 688
- future trends 38 - pituitary 758
- gastric 672,673 - portal vein of disseminated 691
- gebitourinary 699 - post primary 368,370
- genital 702 - post-primary pulmonary, treatment 322
- hepatobiliary 671,691 - pregnancy 301
- hepatosplenic 692 - prenatally acquired 702
- histology of adrenal TB 752 - prevalence 803
- HIV 870 - preventive therapy 473
- - co-infection with 455,824 - primary 360
- - endemic areas, control of 474 - - in adult, epidemiology 265
- - epidemiological considerations 456 - - group at risk 267
- - global emergencies 456 - - progressive 266
- - governmental responses to the threat posed by 475 - - pulmonary 361
- - infected children 464 - - testicular 712
- - positive individuals, diagnosis 466 - prospective antibodies study of patients with 843
-HRCT 364 - prostatic and seminal vesicle tuberculosis 711
- hypertrophic ileocecal 662 - pulmonary 304
- iatrogenic miliary/disseminated 274 - - in HIV-positive adults 461
- IgA EIA 130 - - resection for multidrug-resistant 401
- ileocecal 684 - reinfection 330
- imaging, mesenteric 681 - respiratory failure/hemodynamics 389
- infection 841 - retinal 735
- - rate 803 - retrudescent 330
- intestinal 702 - risk factors for 456
- intracranial 554 - scleral 733
-laryngeal 745,856 - serologic testing for 187
- latent 134, 821 - skeletal 482
- - and paucibacillary 783 - smear-negative vs smear-positive 194
926
Tuberculosis (Continued)
- smear-positive 801
- - cavitating pulmonary 243
- - pulmonary tuberculosis 805
- soft tissue 416
- spinal 3) 350,487) 493
- - versus spinal brucellosis 489
- - with neurological deficits 484
- splenic 671
- surgery for multidrug-resistant pulmonary 396
- suspects 859
- test, rapid 130
- thoracic surgery for 395
- thyroid 756
- tonsillar 746
- toxicity of anti 739
- treatment of 468
- - of active TB during pregnancy 308
- - of adrenal 753
- - of genitourinary 725
- - cost-effectiveness of 857
- - of in HIV-positive individuals 468
- - prophylactic 785
- - of renal 709
- - of spine 518
- - of upper limb joint 595
- and tuberculomas in miliary! disseminated 277
- vaccine
- - multivalent 885
- - new 887
- vaginal 718
- verrucosa cutis 635
- vitamin D3 metabolism 144
- vulval 722
- water and sodium balance in 762
-meningitis in miliary! disseminated 277
- microbiology of 115
Tuberculosis in childhood
- diagnosis 251
- prevention and control of 258
- treatment of 256
Tuberculosis of the ankle 592,621
Tuberculosis of the appendix 686
Tuberculosis of the calvarium with eNS lesions 563
Tuberculosis of the central nervous system 535
Tuberculosis cutis orificialis 637
Tuberculosis of the duodenum 683
Tuberculosis of the elbow 621
Tuberculosis in the elderly 193
Tuberculosis of esophagus 683
Tuberculosis of the eyelid 732
Tuberculosis of the fallopian tube and ovary 719
Tuberculosis of the foot 592,618
Tuberculosis, of other gastrointestinal sites 672
Tuberculosis of the hand 618
Tuberculosis of the heart 431
Tuberculosis of the hip 617
Tuberculosis of the hip joint 590
Tuberculosis in the immunocompromised, cutaneous 648
Tuberculosis of the jejunum and ileum 684
Tuberculosis of the joints of the upper extremity 594
Tuberculosis of the knee 591,616
Tuberculosis of the liver 691
Subject Index
Tuberculosis lymphadenitis 690
Tuberculosis and mixed pattern of obstructive and restrictive
abnormality 389
Tuberculosis and obstructive lung disease 387
Tuberculosis osteomyelitis 494
Tuberculosis in parotid 419
Tuberculosis in patients
- with chronic renal failure 723
- on hemodialysis 723
- with renal transplant 723
Tuberculosis of the pelvis and sacroiliac joints 618
Tuberculosis of the penis) primary 712
Tuberculosis of the pericardium 431
Tuberculosis of the prostate 712
Tuberculosis of the retina 735
Tuberculosis of the salivary glands 747
Tuberculosis of the shoulder joint 595
Tuberculosis of the skin 627, 628
- due to M. bovis 629
- due to M. tuberculosis 629
- test interpretation 234
Tuberculosis of small bowel and colon 661
Tuberculosis of solid abdominal organs 671
Tuberculosis-specific antigens 841
Tuberculosis of the spleen 691
Tuberculosis of the stomach 683
Tuberculosis of the temporal bone 563
Tuberculosis of the thyroid gland 747
Tuberculosis of the uterine cervix 721
Tuberculosis of the uterus 720
Tuberculosis of the vagina 721
Tuberculosis of the vulva 722
Tuberculosis of the wrist 594
Tuberculous 537,482
Tuberculous infection) localizing sites of 413
- preventing 856
Tuberculous lesion, debridement of 522
Tuberculous lymphadenitis, and HIV 446
Tuberculous mass, atypical 561
Tuberculous meningeal mass 561
Tuberculous meningitis (TBM) 248,537,539,548
Tubular bones
- long bones 597
- short 597
Tubular disease 671
Tumor 330
- malignant 565
Tumorous 332
Tumorous type 334
Tumour necrosis factor (TNF-) 136
Tunica vaginalis testis 711
Tunnel syndrome 595
Thrban shaped 746
Two-dimensional echocardiography 438
Type 1 differentiating cytokines IL-12 887
Type 1 response 135
Type 2 cytokine profile 139
Type 2 responses 135, 138
Typing Methods 59
Subject Index 927

U - design 146
mcer 330 - immunogenecity of 887
- segmental 687 - mycobacterial-based 887
-vulval 719 - preventable death 881
Ulceration 330 - recombinant adenoviral-based 890
- intestinal 896 - recombinant forms of 883
- mucosal 664 - subunit-based 888
Ulcerative 332 - therapeutic 883
Ulcerative type 335 - viral-based 889
Ulcerative-granulative 332 Vagina 718
mtrasound 614,663, 682, 704 - tuberculosis 721
- and color doppler 713 Vaginal discharge 719,722
- transrectal 716 Vaginal prolapse 717
Ultraviolet light 856, 865 Vaginitis, atuberculous 722
UNAIDS 465,471,873 Vas deferens 712
UNESCO 7 Vascular complications 540
UNICEF 308 Vascular occlusion 540
Unilateral mass 753 Vasculitis 250,539
United States 874 - granulomatous 317
Unstable kyphotic deformities, bony bridge compression 485 - pulmonary 320
Upperlobe infiltration 314 - tuberculous 771
Upper urinary tract, reconstruction of 710 Vasography 717
Ureter 701 Vein
- ulceration and fibrosis 704 - pulmonary 331
Ureteral replacement 710 - umbilical 303
Ureteric stricture 710 Ventilation
Ureterocolic implant 710 - altered 415
Ureteroureterostomy 710 - mechanical 319,389
Ureterovesical junction stricture 710 - perfusion scan 399
Uric acid metabolism 791 Ventriculitis 561
Urinary bladder Vertebral body wedging 482
- tuberculosis 702 Vertebral damage 486
- thickness of the 704 Villemin 21
Urinary dribbling 722 Visitor, short-term 864
Urinary retention, recurrent 713 Visual acuity 792
Urinary tract 157 Vital capacity 389
Urinary tuberculosis Vitamin D 761
- cystoscopy 707 - metabolism 766
- diagnosis 703 Vitamin D3 metabolism in tuberculosis 144
- general clinical features 702 Vitreous aspirate 737
- signs of 703 Vitritis 735
- surgery of 709 VNTR
Urine culture 703 - MIRU loci 83
Urine samples 118 - technique 84
Urine sediments, staining of 703 Voiding, irritative 713
US guidance 622 Volume reduction 387
US guided procedures 622 Vulva
US mini epidemics 874 - tuberculosis 722
USA, clinics 824 - ulcers 717
Using molecular epidemiology 62
Uterine cervix, tuberculosis 721 W
Uterus 718 W strain 68, 824
- tuberculosis 720 Wales 876
Utonephrectomy 704 Warfarin 791
Uveitis 733,793 Waterdrop-shape 334
Watermelon skin sign 716
Water-soluble contrast myelography (WSCM) 571
v Water-soluble myelography, combined with CT 572
V(V) DMSA (pentavelent dimercaptosuccinic acid) 415 Watery phase 610
Vaccina virus 889 WBC, total 352
Vaccination 145 Weekly dosing 793
- strategies 883 Wheezing 254,339
Vaccine White blood cell abnormalities 216
928 Subject Index

White plague 20 Y
White-stripe 494 Y desents 436
WHO 810 Yersinia enterocolitis 662
WHO DOTS strategy 475,858 Yield of bone marrow aspiration biopsy 224
WHO/IUATLD global surveillance 811
WHOm notification to 35
Wind filled sail sign 620 Z
World Bank 820,873 Zambia 457
World Health Organization (WHO) 302,855 Zebu cattle 77
Wrist, tuberculosis 594 Ziehl 23
WSCM (water-soluble contrast myelography) 571 Ziehl-Neelsen stain 8,23,118, 155,297,321,450,895
- of urine 703
x Zimbabwe 465
X desents 436 Zoonotic disease 15
About the Editor

M.MoNIR MADKOUR, MD,DM,FRCP (London)

is a consultant physician at Riyadh Military Hospital, Saudi Arabia. He was born in Kafr
Mishlah (Gharbiah), Egypt. He qualified from Alexandria University in 1968 and served
in the Rural Medical Unit in his own village. He was then appointed Senior Resident at
AI-Hussein University Hospital (AI-Azhar University), Cairo.
In 1971, he traveled to Great Britain and was appointed Registrar and later Research
Fellow at the University Department of Medicine, Glasgow. In 1977 he traveled to work in
Saudi Arabia and was appointed Assistant Professor of Medicine at King Saud University,
Riyadh, later joining the Military Hospital.
In 1979 Dr. Madkour and three colleagues founded the first scientific medical journal
in Saudi Arabia, the Saudi Medical Journal, and Dr. Madkour served as Associate
Editor for 20 years (1979-1998). He collated some 100 years (1887-1984) of medical
and paramedical research publications about Saudi Arabia into the Saudi Medical
Bibliography, published in three volumes by Churchill Livingstone in 1983 and 1986.
In 1986, Dr. Madkour established the first medical clinic devoted to the investigation
of brucellosis. Experiences and scientific data gained from this clinic were published in
the first edition of his book, "Brucellosis" (1989) by Butterworths. In 2001, the second
edition, "Madkour's Brucellosis" was published by Springer-Verlag. Dr. Madkour has
many other medical publications to his credit. He is the author of chapters on brucellosis
in "Harrison's Principles ofInternal Medicine", 14th edn. (1998) and 15th en. (2001), and
in the "Oxford Textbook of Medicine", 3rd edn. (1996) and 4th edn. (2003).