Sie sind auf Seite 1von 4

Case Report

Pleural Effusion and Fever in an


Immunocompromised Patient
Alexander W. Kay, Megumi Itoh, Jessica Valdez, Sharon F. Chen, Roshni Mathew, and Hayley A. Gans
Department of Pediatrics, Division of Infectious Diseases, Lucile Packard Childrens Hospital at Stanford University,
Palo Alto, California

Corresponding Author: Alexander W. Kay, MD, Department of Pediatrics, Division of Infectious Disease, 300 Pasteur Drive,
Rm G312, Stanford, CA 94305. E-mail: awkay@stanford.edu.
Received November 5, 2013; revisions received January 03, 2014; accepted February 25, 2014; electronically published
March 19, 2014.

Key words. Mycobacterium bovis; pleural effusion; pleurisy.

CASE REPORT
tests, and complete blood count. An infectious laboratory
A 15-year-old female with a past medical history signi- work-up was performed including 2 negative antibody
cant for autoimmune hepatitis on chronic immunosuppres- screens for Coccidioides immitis by immunodiffusion, pos-
sive medications presented to a community hospital with a itive tuberculin skin test with 25-mm induration, positive
24-hour history of shortness of breath, right-sided pleuritic QuantiFERON-Gold IT assay, negative human immuno-
pain, cough, fever, and headache. Review of systems was deciency virus (HIV) antibody, elevated Mycoplasma
otherwise negative. She was diagnosed with pneumonia pneumoniae serologies immunoglobulin (Ig)M 1:128 and
and admitted for empiric treatment with ceftriaxone, clin- IgG 1:128, and a negative respiratory virus panel. A previ-
damycin, azithromycin, and uconazole. She remained fe- ous tuberculin skin test in 2006 had 0 mm of induration.
brile and was transferred after 72 hours to a quaternary On day 5 of illness, a chest tube was placed draining
childrens hospital. Her past medical history was signi- 750 mL of exudative serous uid with 1240/L nucleated
cant for autoimmune hepatitis diagnosed in 2005, and cells and 92% lymphocytes. The adenosine deaminase
home medications included 6-mercaptopurine and allopu- was 5.1 U/L (09.4). Acid-fast bacilli (AFB), fungal, and
rinol. The patient was born and lives in coastal central bacterial stains were negative as were a mycobacterial and
California. She traveled 1 month before presentation to mycoplasma polymerase chain reaction (PCR). Cytology
Arizona and 1 year prior to visit family in Mexico, both was negative for malignant cells. Her fevers resolved the
trips were 2 weeks in duration. She had no known sick day after chest tube placement. A pleural biopsy was per-
contacts. formed on day 8 of her illness, and brinous material
Her physical exam upon transfer was notable for a gen- was noted in the pleural space intraoperatively. Acid-fast
erally well appearance. Her vital signs were as follows: bacilli, fungal, and bacterial stains were negative as was
blood pressure 102/67; heart rate 80; respiratory rate 21; bacterial sequencing, which can identify Mycobacterium
SpO2 97% on room air; and temperature 38.1C. She tuberculosis complex. Pathology revealed necrotizing gran-
had decreased breath sounds and dullness to percussion ulomatous inammation but no well formed granulomas.
on the right. There was no cervical, axillary, or inguinal Three induced sputa were AFB smear negative. She re-
lymphadenopathy. Her abdominal exam was benign. The mained afebrile and was discharged with presumed latent
patient was continued empirically on ceftriaxone, clinda- tuberculosis infection but without a clear diagnosis for her
mycin, azithromycin, and uconazole. A chest radiograph disease after completing a 10-day course of ceftriaxone and
(Figure 1A) demonstrated a large, free-owing, right-sided 5 days of azithromycin. Fluconazole was also discontinued.
pleural effusion, and a chest computed tomography On follow-up 1 month after discharge she remained afe-
(Figure 1B) demonstrated the same without parenchymal brile; however, she had shortness of breath with exertion.
disease or intrathoracic lymphadenopathy. Her initial lab- Her pleural effusion was unchanged on chest radiograph
oratory studies demonstrated an elevated c-reactive protein from the time of discharge. Her M pneumoniae serologies
8.6 (00.9 mg/dL) and normal electrolytes, liver function were IgM 1:256 and IgG 1:256 and Histoplasma serologies

Journal of the Pediatric Infectious Diseases Society, Vol. 4, No. 1, pp. e6e9, 2015. DOI:10.1093/jpids/piu018
The Author 2014. Published by Oxford University Press on behalf of the Pediatric Infectious Diseases Society.
All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
Pleurisy Due to Mycobacterium bovis e7

Figure 1. (A) Posterior-anterior chest radiograph of a large, free-owing, right-sided pleural effusion. (B) Large, right-sided pleural effusion with compression atelectasis
and no intrathoracic lymphadenopathy or parenchymal disease.

were negative. All AFB cultures from sputum, pleural uid, ingestion typically presents with intra-abdominal disease
and pleural biopsy remained negative. or cervical lymphadenopathy. In the absence of a source
case, a microbiologic diagnosis should be aggressively pur-
DISCUSSION sued if there is concern for mycobacterial disease.
Diagnosis: Pleural Tuberculosis due to Mycobacterium bovis Treatment for latent tuberculosis infection should not be
The patients pleural biopsy culture became positive for initiated until disease has been denitively excluded or an
M bovis after 35 days. All other cultures were nalized neg- alternate diagnosis has been established. In this case, both
ative. She was then treated with isoniazid, rifampicin, and the patients tuberculin skin test and QuantiFERON-Gold
ethambutol for 2 months followed by 7 months of isonia- IT assay were positive. The QuantiFERON-Gold IT detects
zid and rifampicin, with resolution of her shortness of infection by M tuberculosis complex organisms, M tuber-
breath and pleural effusion. Her exposure history revealed culosis, M bovis, and Mycobacterium africanum, but not
ingestion of cheese brought from Mexico by her grandfa- M bovis-derived Bacillus CalmetteGurin (BCG) strains.
ther 6 months before presentation, which placed her at Pleurisy secondary to M tuberculosis, or in this case
risk for M bovis infection. M bovis, can be a challenging diagnosis. Cultures and
We present a case of a moderately immunosuppressed PCR-based testing of pleural uid is rarely helpful in mak-
adolescent who ultimately was diagnosed with pleural tu- ing the diagnosis, and sputum cultures are positive in only
berculosis secondary to M bovis. The patient presented 50% of cases. The pleural effusion represents an immune
acutely with only a 1-day history of fever and chest pain response to mycobacterium in the pleural tissue, resulting
possibly consistent with community-acquired pneumonia in the low diagnostic yield. Pleural uid adenosine deami-
and parapneumonic effusion. However, she had minimal nase has a reported sensitivity of >90% [1], but it was neg-
airspace disease and was well appearing given the size of ative in this case. Pleural biopsy is the critical test in the
the effusion. Her well appearance, exposures, lymphocytic diagnosis of pleural tuberculosis. In contrast to pleural
pleural uid, and necrotizing granulomatous inammation uid culture, the sensitivity of a pleural biopsy culture is
pointed toward a more indolent infection such as coccidioi- greater than 90%. Granulomata on pathology are consis-
domycosis, histoplasmosis, or mycobacterial disease. tent with tuberculosis; however, granulomatous inamma-
Noninfectious etiologies such as lymphoma were also con- tion is a more nonspecic nding indicative of chronic
sidered and excluded. inammation. The sensitivity of PCR testing of pleural tis-
The patient was not treated empirically for culture- sue remains undened, although from sputa the sensitivity
negative tuberculosis disease because of her high risk for in our laboratory is greater than 90% in smear-positive
medication-related complications, her resolution of fever, cases, which this patients was not.
and negative preliminary testing. This decision must be In this patients case, the identication of M bovis
made on a case-by-case basis. Her exposure to unpasteur- occurred within 24 hours of growth in liquid media and
ized dairy products was concerning, but M bovis after conrmatory M tuberculosis complex PCR. This rapid
e8 Kay et al

identication was possible using a PCR assay designed to that patients who carry an increased risk for disease due
identify regions of difference- and species-based genomic to M tuberculosisbecause of immunosuppression for
deletions within the M tuberculosis complex [2]. However, solid organ transplantation, stem cell transplantation, or
this assay is not available in most clinical laboratories, and autoimmunityalso have an increased risk for disease
the identication of M bovis is often made presumptively due to M bovis. However, perhaps due to the relative rarity
based on the isolates drug sensitivities before laboratory of disease due to M bovis in resource-rich countries, there
send-out results return. Mycobacterium bovis is intrinsically are only case reports describing M bovis disease in solid
resistant to pyrazinamide; however, monoresistance to organ transplant and stem cell transplant recipients [11, 12].
pyrazinamide does not exclude M tuberculosis or M bovis- In populations where M bovis is common, there are clin-
derived BCG, when clinically relevant, and therefore identi- ical implications when treating culture-negative tuberculo-
cation to the species level should be obtained [3]. sis. Mycobacterium bovis is intrinsically resistant to
Mycobacterium bovis is part of the M tuberculosis com- pyrazinamide, and therefore a 9-month treatment course
plex and currently causes 1%2% of reported tuberculosis is recommended. This difference in treatment underlines
cases in the United States. The low rate in the United States the importance of obtaining a diagnosis by culture or
and other industrialized nations is attributed to screening PCR. In patients with possible tuberculosis disease and po-
and culling of cattle infected with M bovis in addition to tential exposure, M bovis should be considered in the dif-
pasteurization; however, this practice is rare in developing ferential diagnosis because of differences in treatment.
countries, and bovine tuberculosis is considered enzoo- In summary, it is important to assess for zoonotic risk
notic throughout much of the world [4]. Polymerase factors in patients presenting with clinical disease concern-
chain reaction analysis of culture-positive tuberculosis iso- ing for tuberculosis but without an identied tuberculosis
lates from Mexico identied M bovis in 28% of the human contact. Patients, especially those who are immunocom-
samples in a location where as many as 16% of cattle are promised, should be counseled to avoid unpasteurized
thought to be infected with M bovis [5]. A report from San dairy products, particularly from countries where bovine
Diego, where there is a large bi-National Hispanic popula- tuberculosis is endemic. Lastly, identication of M tubercu-
tion, identied M bovis in 6% of culture-positive adult losis complex isolates to the species level should be at-
cases and 35% of culture-positive pediatric cases [6]. tempted if possible because the presence of M bovis can
Mycobacterium bovis is also the etiology for an increased have treatment and epidemiologic signicance.
proportion of tuberculosis cases in urban Hispanic com-
munities far removed from the Mexican border [7]. Acknowledgments
The primary mechanism for infection in the United Potential conicts of interest. All authors: No reported conicts.
States is ingestion of imported unpasteurized dairy prod- All authors have submitted the ICMJE Form for Disclosure of
Potential Conicts of Interest. Conicts that the editors consider rele-
ucts. Although M bovis can cause pulmonary tuberculosis, vant to the content of the manuscript have been disclosed.
extrapulmonary disease, scrofula, and intra-abdominal
disease in particular are most common in US cases (51% References
and 19%, respectively, of cases in the San Diego series)
1. Porcel JM. Tuberculous pleural effusion. Lung 2009; 187:
[6]. Intra-abdominal disease may manifest as gastrointesti-
26370.
nal enteritis or colitis, peritonitis with peritoneal nodules, 2. Pinsky BA, Banaei N. Multiplex real-time PCR assay for rapid
or intra-abdominal lymphadenopathy. Pulmonary disease identication of Mycobacterium tuberculosis complex members
to the species level. J Clin Microbiol 2008; 46:22416.
is rare in children in the United States who are typically ex-
3. Kurbatova EV, Cavanaugh JS, Dalton T, et al. Epidemiology of
posed via ingestion. However, among M bovis cases origi- pyrazinamide-resistant tuberculosis in the United States, 1999
nating in people born in endemic countries, 44% present 2009. Clin Infect Dis 2013; 57:108193.
with pulmonary disease, which suggests that aerosolized 4. Cosivi O, Grange JM, Daborn CJ, et al. Zoonotic tuberculosis
due to Mycobacterium bovis in developing countries. Emerg
exposure is more common either through exposure to in- Infect Dis 1998; 4:5970.
fected cattle or person-to-person spread [8]. 5. Portillo-Gmez L, Sosa-Iglesias EG. Molecular identication of
Person-to-person spread is well documented in individu- Mycobacterium bovis and the importance of zoonotic tuberculo-
sis in Mexican patients. Int J Tuberc Lung Dis 2011; 15:140914.
als infected with HIV; however, it has been described in im-
6. Dankner WM, Davis CE. Mycobacterium bovis as a signicant
munocompetent patients as well and manifests more cause of tuberculosis in children residing along the United
typically as pulmonary disease [9]. As with M tuberculosis, States-Mexico border in the Baja California region. Pediatrics
2000; 105:e79.
nosocomial transmission and rapid progression from infec-
7. Centers for Disease Control and Prevention. Human tuberculosis
tion to disease has been documented in HIV-infected pa- caused by Mycobacterium bovisNew York City, 20012004.
tients infected with M bovis [10]. It is generally thought MMWR Morb Mortal Wkly Rep 2005; 54:6058.
Pleurisy Due to Mycobacterium bovis e9

8. Hlavsa MC, Moonan PK, Cowan LS, et al. Human tuberculosis HIV-infected patients. A case-control study. AIDS 1993; 7:
due to Mycobacterium bovis in the United States, 19952005. 145360.
Clin Infect Dis 2008; 47:16875. 11. Buyck HC, Holliman R, Else J, et al. An unusual cause of acute
9. Evans JT, Smith EG, Banerjee A, et al. Cluster of human tubercu- abdomen following allogeneic transplantation: a zoonotic disease
losis caused by Mycobacterium bovis: evidence for person-to- revisited. Bone Marrow Transplant 2008; 41:106970.
person transmission in the UK. Lancet 2007; 369:12706. 12. Chen D, Gutierrez K. Mycobacterium bovis disease in a ped-
10. Bouvet E, Casalino E, Mendoza-Sassi G, et al. A nosocomial out- iatric renal transplant patient. Pediatr Infect Dis J 2006; 25:
break of multidrug-resistant Mycobacterium bovis among 5579.

Das könnte Ihnen auch gefallen