1

Adverse Drug Reactions

Adverse drug reactions are defined as unwanted effects occurring at therapeutic doses of the
drug. Drugs are great mimics of disease and adverse drug reactions present with diverse
clinical signs and symptoms. The classification proposed by Rawlins and Thompson (1977)
divides reactions into type A and type B (Table 1).
TYPE A are adverse reactions which are a consequence of the drugs normal
pharmacological effect and are therefore predictable; they are dose-related with a low
mortality. Such reactions are usually due to incorrect dosage (too much or for too long) or to
disordered pharmacokinetics, usually a failure of drug elimination. The term side effects is
often applied to minor type A reactions.
TYPE B reactions are not predictable from the drugs known action, are not dose-
related and have a considerable mortality. The underlying pathophysiology of type B
reactions is by definition poorly if at all understood, and may have a genetic or immunological
basis. Indeed as more is known about the pharmacology of a drug, an adverse effect may have
to be reclassified type B to type A. A good example of this is hemolytic anemia due to
primaquine which used to be regarded as a rare type B effect of possible immune etiology until
its association with glucose-6-phosphate dehydrogenase (G-6PD) deficiency was established. It
is now a predictable (type A) reaction in such genetically predisposed individuals. Type B
reactions occur infrequently (1:1000-1:10 000 treated subjects being typical).

Table 1 Some examples of type A and type B reactions

Drug Type A reaction Type B reaction
Ampicillin Pseudomembranous colitis Interstitial nephritis
Chlorpromazine Sedation Hepatotoxicity
Naproxen Gastrointestinal hemorrhage Agranulocytosis
Practolol Bradycardia Oculomucocutaneous syndrome
Thiazides Gout Thrombocytopenia
Warfarin Bleeding Breast necrosis

Three further minor categories of adverse drug reaction have been proposed:

1. Type C: continuous reactions due to long-term use, such as tardive dyskinesia or analgesic
nephropathy.
2. Type D: delayed reactions such as carcinogenesis or teratogenesis.
3. Type E: end of use reactions such as adrenocortical insufficiency following withdrawal of
corticosteroids, or withdrawal syndromes following discontinuation of treatment with
clonidine, benzodiazepines, tricyclic antidepressants or -adrenoreceptor antagonists.

A random survey found that approximately 80% of adults took some kind of medication during
a 2-week period in UK. Exposure to drugs in the population is thus substantial and the
incidence of adverse reactions must be viewed in this context. Adverse drug reactions are most
commonly of type A (80%) causing up to 3% of acute hospital admissions and 2-3% of
consultations in general practice. In hospital their incidence is estimated to be 10-20%, causing
increased use of hospital resources and a mortality of 0,3-1%. They are most frequent and
most severe in neonates or the elderly (>60 years), in women, patients with hepatic or renal
disease, and those with a history of previous adverse drug reactions. Adverse drug reactions
occur most commonly early in therapy (1-10 days), and the drugs most commonly implicated
are digoxin, antimicrobials, diuretics, potassium, analgesics, sedatives and tranquillizers,
insulin, aspirin, glucocorticosteroids, antihypertensives and warfarin.
Factors involved in the etiology of adverse drug reactions can be classified as shown in
Tab. 2.
 2

Table 2. Factors involved in adverse drug reactions.

Patient factors
Intrinsic Extrinsic
Age neonatal, infant and elderly Environment - sun
Sex hormonal environment Xenobiotics (e.g. drugs, herbicides)
Genetic abnormalities, e.g. enzyme defects Malnutrition
Previous adverse drug reactions, allergy, atopy
Presence of organ dysfunction disease
Personality and habits alcoholic, drug addict, nicotine, compliance

Prescriber factors
Incorrect drug or drug combination
Incorrect route of administration
Incorrect dose
Incorrect duration of therapy

Drug factors
Drug-drug interactions
Pharmaceutical - batch problems, shelf-life, incorrect dispensing

MONITORING/SURVEILLANCE (PHARMACOVIGILANCE)

The ideal method would identify adverse drug reactions with high sensitivity and specificity
and respond rapidly. It would detect rare adverse common ones, the incidence of which it
would quantify together with predisposing factors. Continued surveillance is mandatory
after a new drug has been marketed since it is inevitable that the preliminary testing of
medicines during drug development, although excluding many ill effects, cannot identify
uncommon adverse effects. A variety of early detection methods have been introduced to
discover these actions as swiftly as possible.

Phase I/II/III trials

Early (phase I/II) trials are important in assessing the tolerability and dose-response of new
drugs, but they are very insensitive at detecting adverse reactions because they are performed
on relatively few subjects (perhaps 200-300). Phase III clinical trials can, however, establish
the incidence of common adverse reactions and relate this to therapeutic benefit. Analysis of
the reasons given for dropping out of phase III trials is particularly valuable in establishing
whether common events such as headache, constipation, lethargy or male sexual dysfunction
are truly drug related The problem of detection is several orders of magnitude worse if the
adverse drug reaction resembles spontaneous disease in the population, such that physicians
are unlikely to attribute the reaction to drug exposure.

Yellow card scheme and postmarketing surveillance

Untoward effects that have been missed in early clinical trials become apparent when the drug
is used on a wider scale. Case reports in the literature, which may often stimulate further
reports, remain the most sensitive means of detecting rare but serious and unusual adverse
effects. In the UK, a Register of Adverse Reactions was started in 1964. Doctors and dentists
were asked to report any adverse effects which they considered due to drugs. The Committee
of Safety of Medicines (CSM) and the Medicines Control Agency (MCA) operate a system of
spontaneous reporting by doctors on prepaid yellow postcards. The yellow card scheme
consits of three stages: (1) collection of data; (2) analysis of data; and (3) feedback of
information to doctors.
 3

Such methods of surveillance have proved useful as an early warning system, but the
major difficulty is underreporting. On average probably not more than 10% of adverse
reactions are reported. This may be due partly to confusion about what events to report, or in
deciding the precise relationship of a drug to a specific adverse event or sometimes which of
several drugs being taken by the patient is responsible. A further problem is that if the drug is
responsible for an increase in the incidence of some disease which is already common, for
example gallbladder disease or carcinoma of the bronchus, the change in incidence must be
very large before this can be detected, since cases due to drug therapy are obscured by cases
due to other (often unknown) causes. Doctors are inefficient at detecting adverse reactions to
drugs and those reactions reported are in general the obvious or previously described and well-
known reactions. Several initiatives are currently underway to try to improve this situation by
both education and involvement of specially trained pharmacists in and out of hospitals.
The CSM introduced a system of high vigilance for newly marketed drugs. Any newly
marketed drug has on its data sheet and its entry in the British National Formulary a black
triangle for its first 2 years in the general market. This conveys to the prescriber that any
unexpected event occurring to a patient prescribed this drug should be reported by the yellow
card system. The pharmaceutical company marketing a new drug is also responsible for
obtaining accurate reports on all patients treated up to an agreed number. The scheme was
successful in the case of benoxaprofen, an antiinflammatory analgesic. Following its release,
there were spontaneous reports to the CSM of photosensitivity and onycholysis. Further
reports appeared in the elderly, in whom its half-life is prolonged, of cholestatic jaundice and
hepatorenal failure, which was fatal in eight cases. Benoxaprofen was subsequently taken off
the market when 3500 adverse drug reaction reports were received with 61 fatalities. The
yellow card black triangle scheme was instrumental in the early identification of urticaria and
cough as adverse effects of angiotensin-converting enzyme inhibitors. Although, potentially, the
population under study by this system comprises all the patients using a drug, in fact
underreporting yields a population that is not uniformly sampled and thus data that are
unrepresentative and difficult to work with statistically. This accounts for the paucity of
accurate incidence data for adverse drug reactions.
Systems such as the yellow card one are relatively inexpensive and easy to manage and
facilitate ongoing monitoring of all drugs, all consumers and all types of adverse reaction.
Reports from drug regulatory bodies of 22 countries are collated by the World Health
Organization (WHO) Unit of Drug Evaluation and Monitoring in Geneva. Rapid access to
reports from other countries should be of great value in detecting rarer types of adverse
reaction, although the same reservations as apply to national systems apply to this register.
Additionally, this database might be able to yield information concerning different prescribing
habits and even differences in the pattern of untoward drug effects in different countries.

Prescription event monitoring

Successful studies have been carried out by the Drug Surveillance Research Unit at the
University of Southampton using prescription event monitoring. Prescriptions for certain drugs
are identified by the prescriptions pricing office in Edinburgh. These are followed up and the
prescribing doctor asked to fill in a simple questionnaire recording any medical event from the
patient notes. This has the advantage that the prescriber does not have to judge causality
between the event and the drug. This scheme identified a new event associated with enalapril,
i.e. deafness in 19/12 500. This is in excess of an expected five patients developing deafness in
the general population. This needs further investigation, but highlights the potential usefulness
of the approach as well as one of its main limitations, namely the absence of a control goup.

Case-control studies
One of the great difficulties in monitoring drugs for adverse effects is that if they occur
relatively rarely (perhaps 1 in 1000 patients), very many patients have to be monitored before a
 4

particular side effect becomes apparent. An alternative approach is to study a group of patients
who have developed a particular symptom or syndrome and to compare the frequency of
exposure to possible etiologic agents with a control group. A prior suspicion (hypothesis) must
exist to prompt the setting up of such a study: examples are the possible connection between
irradiation or environmental pollution and certain malignancies, especially where they are
observed in clusters. Although this method is comparatively easy to carry out, artifacs occur as
a result of faulty selection of patients and controls, and the approach remains highly
controversial among epidemiologists, public health physicians and statisticians. It has had some
apparent successes, such as the linking of stilbestrol with vaginal adenocarcinoma and of
licomycin with pseudomembranous colitis, and, recently, the association between fenoterol use
and an increased number of deaths in asthmatics. In its most simple form it is really an
extension of the idea of the alert doctor, but with the considerable amount of data now
available in such studies as the Boston Drug Surveillance Program, it should be possible to use
this method on a wider scale and in a more sophisticated manner.

Patient questionnaires
Self-administered questionnaires have been used for outpatients attending hypertension and diabetic clinics
and have detected previously unsuspected adverse effects, for example headache and weakness in the legs as
effects of metformin. They have also been used to show an absence of effects, for example that propranolol is
not associated with any of the eye symptoms caused by practolol.

Intensive monitoring
In the last few years a number of intensive monitoring programs, usually hospital-based, have
been started. The Aberdeen-Dundee monitoring system abstracts data from some 70 000
hospital admissions each year, storing these on a computer file before analysis. The Boston
Collaborative Drug Surveillance Program (BCDSP) which involves selected hospitals in
several countries, is even more comprehensive. In the BCDSP all patients admitted to specially
designated general wards are included in the analysis. Specially trained personnel obtain the
following information from hospital patients and records:

1. Background information, i.e. age, weight, height, etc.

2. Patients medical history.
3. Patients drug exposure.
4. Side effects of drugs.
5. Outcome of treatment and changes in laboratory tests during hospital admission.

A unique feature of comprehensive drug-monitoring systems lies in their potential to

follow-up and investigate adverse reactions suggested by less sophisticated detection systems
or by isolated case reports in medical journals. Furthermore, the frequency of side effects can
be determined more cheaply than by a specially mounted trial to investigate a simple adverse
effect. Thus, for example, the risk of developing a rash with ampicillin was found to be around
7% both by clinical trial and by the BCDSP which can determine such facts almost
automatically from data on its files. New adverse reactions or drug interactions are sough by
multiple correlation analysis of the data. Thus, when an unexpected relationship arises, such as
the 20% incidence of gastrointestinal bleeding in severely ill patients treated with ethacrynic
acid compared to 4,3% of similar patients treated with other diuretics, there is little chance of
bias arising from awareness of the hypothesis during data collection since the data are collected
before the hypothesis is proposed. It is also possible to isolate factors predisposing patients to
a particular complication; in the case of ethacrynic acid these were female sex, a high blood
area, previous heparin administration and intravenous administration of the drug. An important
aspect of this type of approach is that lack of clinically important associations can also be
detected. Thus no significant association between aspirin and renal disease was found, whereas
 5

long-term aspirin consumption is associated with a decreased incidence of myocardial

infarction, an association which has been shown to be of therapeutic importance in randomized
clinical trials. There are plans to extend intensive drug monitoring to cover other areas of
medical practice.
In terms of new but uncommon adverse reactions, however, the numbers of patients
undergoing intensive monitoring while taking a particular drug will inevitably be too small for
the effect to be detectable. Such monitoring therefore can provide information only about
relatively common, early reactions to drugs used under hospital conditions. Patients are not in
hospital long enough for detection of delayed effects, which are just the reactions least likely to
be recognized as such even by an astute clinician.

Monitoring from national statistics

A great deal of information is available from death certificates, hospital discharge diagnoses
and similar records. It may be possible from these data to pick out a change in disease trends
and relate this to drug therapy. Perhaps the best-known example of this is the increased death
rate in young asthmatics noted in the mid-1960s, which was connected with overuse of
bronchodilator inhalers containing non-specific -adrenoreceptor agonists (e.g. adrenaline
and/or isoprenaline). Although relatively inexpensive the difficulty of this method is obvious.
Large numbers of patients must suffer before the change is detectable, particularly in diseases
with an appreciable mortality. Data interpretation is particularly difficult when hospital
discharges are used as a source of information, since sometimes diagnosis is provisional and
changes as the patients disease progresses.

Record linkage
Record linkage is presently under investigation as a method of monitoring adverse drug
reactions and may well produce some important results. The basic idea is that the medical
records from a defined population (both general practice and hospital sources) are kept from
cradle to grave and analyzed. This method could be particularly useful when looking for really
long-term ill effects from drugs or from other factors, for instance the increased incidence of
leukemia in those receiving radiation in utero or the occurrence of diseases such as cancer or
mental retardation in individuals whose mothers had received drugs during pregnancy. Analysis
of data is becoming more sophisticated with the use of computers to store information, but one
must know what information is worth storing, and ask the correct questions to avoid the well-
known garbage in - garbage out situation.
The main types of inquiry can be made:

1. Follow-up of individuals who have received selected drugs to determine adverse reactions
attributable to the drug. This approach is of most value when a drug is already suspected of
producing some particular effect and is not particularly suited to the discovery of unexpected
adverse effects. A defect is the absence of a control group. It has been convincingly
demonstrated that untreated subjects or those receiving placebo often experience symptoms
commonly listed as side effects of drugs. Therefore the results of such studies can be biased by
the preconceived ideas of the investigator or the effects on the patient of his or her enhanced
interest.
2. Follow-up of patients with specific diseases rather than those treated with specific drugs is
sometimes easier to carry out since most clinicians categorize their patients under disease
rather than treatment. Thus it is possible to investigate the frequency of adverse reactions to,
say, phenytoin, by selecting the records of epileptics. One of the largest disease-oriented
systems is that of the USA Perinatal Study which received data on drug exposure and fetal
outcome from over 50 000 consecutive pregnancies. From this has come evidence linking
malformations and maternal exposure to phenytoin.
 6

Feedback
There is no use in collecting vast amounts of data on adverse reactions unless they are analyzed
and important information reported back to prescribing doctors. In addition to articles in the
medical press the CSM and MCA circulate to all doctors a Current Problems information
sheet which deals with important and recently discovered adverse reactions. If an acute and
serious problem is recognized doctors will usually receive notification from the CSM and often
from the pharmaceutical company involved.

ALLERGIC ADVERSE DRUG REACTIONS

Immune mechanisms are involved in a number of adverse effects caused by drugs. The
development of allergy implies previous exposure to the drug or to some very closely related
substance. Most drugs are of low molecular weight (less than 1000) and thus are not antigenic.
They can, however, combine with substances of high molecular weight, usually proteins, acting
as haptens so that the conjugate thus formed is antigenic.
The factors that determine the development of allergy to a drug are no fully
understood. Some drugs (e.g. penicillin) are more likely to cause allergic reactions than others,
and type I (immediate anaphylactic) reactions are more common in patients with a history of
atopy. A correlation between allergic reactions involving immunoglobulin E (IgE) and human
leukocyte antigen (HLA) serotypes have been reported and so genetic factors may also be
important. There is some evidence that drug allergies are more common in older people, in
women, and in those with a previous history of drug reaction. This may, however, merely
represent increased frequencies of drug exposure in these groups, and prevalence figures
(expressed relative to the appropriate denominator) are currently not available.

Fig. 11.1 The immune response to drugs.

Type IV
Central immune response
apparatus

Sensiti-
Drugs or its Cell T lymp- zed lym-
metabolites membrane hocytes phocytes

Marcrophages

Type I, II &
III
Drugs or its Protein Antigen responses
metabolites
B lymp-
hocytes Humoral
antibodies

Drug
(large molecule)
 7

Types of allergy

Drugs cause a variety of allergic responses (Fig. 11.1.) and sometimes a single drug can be
responsible for more than one type of allergic response.

TYPE I

Type I reactions (e.g. to penicillin) are due to the production of reagenic antibodies known to
consist predominantly of class IgE. The antigen-antibody reaction on the surface of mast cells
causes degranulation and release of pharmacologically active substances. It occurs commonly
with foreign serum or penicillin but may also occur with streptomycin and some local
anesthetics. With penicillin it is believed that the penicilloyl moiety of the penicillin molecule is
responsible for the production of antibodies. Occasionally the development of symptoms is
delayed for some hours after the drug has been taken. It seems probable that antigen-antibody
aggregates are involved. The antibody in this mechanism consists of IgG. The therapy of
anaphylactic shock is described in chapter 47.

TYPE II

These are due to antibodies of class IgG and IgM which on contact with antibodies on the
surface of cells are able to fix complement, causing cell lysis, for example penicillin and
methyldopa causing Coombs positive hemolytic anemia.

TYPE III, IMMUNE COMPLEX, ARTHUS REACTIONS

Circulating immune complexes can produce several clinical allergic states including serum
sickness and immune complex glomerulonephritis, and a syndrome resembling systemic lupus
erythematosus. The onset of serum sickness is delayed for some days until the symptoms -
fever, urticaria, arthropathy, lymphadenopathy and eosinophilia - develop. Proteinuria occurs
frequently. Recovery takes a few days. Causative agents include foreign serum, penicillin,
sulfonamides, streptomycin and propylthiouracil. Amiodarone lung and hydralazine induced
systemic lupus syndrome are also possibly mediated by immune complex-related mechanisms,
although these reactions are less well understood.

TYPE IV

Type IV are delayed hypersensitivity reactions, the classical example of which is contact
dermatitis (e.g. to topical antibiotics such as penicillin). The mechanism here is that the drug
applied to the skin forms an antigenic conjugate with dermal proteins, stimulating formation of
sensitized T lymphocytes in the regional lymph nodes with a resultant rash if the drug is applied
again. Drug photosensitivity is due to a photochemical combination between the drug (e.g.
tetracycline, amiodarone) and dermal protein. Delayed sensitivity can also result from the
systemic administration of drugs.
 8

EXAMPLES OF ALLERGIC AND OTHER KINDS OF ADVERSE DRUG

REACTIONS

Adverse drug reactions can be manifest in any organ system, or involve multiple organ systems,
and in extraordinarily diverse forms. Specific instances are dealt with throughout this book.
Some examples to illustrate the diversity of adverse drug reactions are also given here.
Rashes
These are common manifestations of drug reactions. A number of immune mechanisms may be
involved which produce many types of rash. In some cases (e.g. ampicillin) the rash is more
commonly due to non-allergic factors; in others the cause is not known (Table 11.4).

Serum sickness

This is caused by circulating immune complexes and the onset is delayed for some days until
the symptoms - fever, urticaria and arthropathy develop. Proteinuria, cosinophilia and
lymphadenopathy occur frequently. Recovery takes a few days. Causative agents include
foreign serum, penicillin, sulfonamides, streptomycin and propylthiouracil, possibly amiodarone
lung and hydralazine-induced systemic lupus syndrome.

Lymphadenopathy

Lymph-node enlargement can result from taking drugs (e.g. phenytoin). The mechanism is not
known, but allergic factors may be involved. The reaction may be confused with a lymphoma
and inquiry about chronic drug taking is important in patients with lymphadenopathy of
unknown cause.

Blood dyscrasias

Thrombocytopenia, anemia (aplastic, iron deficiency, macrocytic, hemolytic) and

agranulocytosis can all be produced by drugs.
Thrombocytopenia can occur with many drugs and in many instances it is direct
suppression of the megakarocytes rathter than immune processes which is important. Quinine
thrombocytopenia may be an allergic phenomenon. The following are some of the drugs which
most commonly cause thrombocytopenia:
Thiazides, chloramphenicol.
Quinidine.
Heparin.
Sulfonamides, quinine, gold.
Most cytotoxic agents.

Hemolytic anemia can be caused by a number of drugs and sometimes immune

mechanisms are responsible. Glucose-6-phosphate dehydrogenase deficiency predisposes to
non-immune hemolysis (e.g. to primaquine). Immune mechanisms include:

1. Combination of the drug with the red cell membrane with the conjugate acting as an antigen.
This has been shown to occur with penicillin-induced hemolysis, and may also occur with
chlorpromazine, chlorpropamide and sulfonamides.

2. Alteration of the red-cell membrane by the drug so that it becomes autoimmunogenic. This
may happen with methyldopa, and a direct positive Coombs test develops in about 20% of
patients who have been treated with this drug for more than a year. Is only a small proportion
 9

does hemolysis actually occur. Similar changes can take place with levodopa and mefenamic
acid.

3. Causing non-specific binding of plasma protein to red cells and thus causing hemolysis. This
is believed to occur with cephalosporins.
Aplastic anemia as an isolated entity is not common but may occur either in isolation
or as part of a general depression of bone-marrow activity (pancytopenia); examples include
chloramphenicol and, commonly and predictably, cytotoxic drugs.

Agranulocytosis can be caused by many drugs. Several different mechanisms are

implicated and it is not known whether allergy plays a part. Among the drugs most frequently
implicated are:
Antithyroid drugs.
Phenothiazines, tolbutamide, most cytotoxic agents.
Sulfonamides, antidepressants (especially mianserin)
Clozapine.

Systemic lupus erythematosus

Several drugs (including procainamide, isoniazid, hydralazine, chlorpromazine,

anticonvulsants) produce a syndrome that resembles systemic lupus together with a positive
antinuclear factor test. The development of this is closely related to dose, and in the case of
hydralazine also depends upon the rate of acetylation of the drug, which is genetically
controlled. There is some evidence that the drugs act as haptens combining with DNA and
forming antigens. Symptoms usually disappear when the drug is stopped, but recovery may be
slow.

Vasculitis

Both acute and chronic vasculitis can result from taking drugs, and may have an allergic basis.
Acute vasculitis with purpura and renal involvement occurs with penicillins and the
sulfonamides. A more chronic form can occur with phenytoin.

Renal

All clinical expressions of renal disease can be caused by drugs: common culprits are non-
steroidal anti-inflammatory drugs and angiotensin-converting enzyme inhibitors (which cause
functional and usually reversible renal failure in susceptible patients). Nephrotic syndrome
results from several drugs (e.g. penicillamine, high-dose captopril, gold) which cause a variety
of immune-mediated glomerular injuries. Interstitial nephritis can be caused by several drugs
including non-steroidal anti-inflammatory drugs and penicillins. Aminoglycoside antibiotics and
vancomycin cause direct tubular toxicity. Many drugs cause electrolyte or acid-base
disturbances via their predictable direct or indirect effects on renal electrolyte excretion (e.g.
hypokalemia and hypomagnesemia from loop diuretics, hyperkalemia from potassium-sparing
diuretics and converting enzyme inhibitors, proximal renal tubular acidosis from carbonic
anhydrase inhibitors) and some cause unpredictable toxic effects on acid-base balance (e.g.
distal renal tubular acidosis from amphotericin). Obstructive uropathy can be caused by uric
acid crystals consequent on initiation of chemotherapy in patients with hematological
malignancy, and rarely poorly soluble drugs such as sulfonamides can themselves cause clinical
problems consequent on crystaluria.
 10

Other reactions

Fever is a common manifestation of drug allergy and should be remembered in patients with
fever of unknown cause.
Liver damage (hepatitis with or without obstructive features) as a side effect from
drugs is important. It may be insidious, leading slowly to end-stage cirrhosis (e.g. during
chronic treatment with methotrexate) or acute and fulminant (as in some cases of isoniazid,
halothane or phenytoin hepatitis). Chlorpromazine commonly causes liver involvement
characterized by raised alkaline phosphatase (obstructive pattern). Gallstones (and
mechanical obstruction) can be caused by fibrates and other lipid-lowering drugs, and by
octreotide, a somatostatin analog used to treat a variety of enteropancreatic tumors including
carcinoid syndrome and VIPomas*. Immune mechanisms are implicated in some forms of
hepatic injury by drugs, but are seldom solely responsible.
---------------------------------------------------------------------------------------------------------------
* VIP = vasoactive intestinal polypeptide
---------------------------------------------------------------------------------------------------------------

IDENTIFICATION OF THE DRUG AT FAULT

It is often difficult to decide whether a clinical event is drug related; even when this is probable
it may be difficult to determine which drug is responsible, as patients are often taking more
than one. One or more of several possible approaches may be appropriate:

1. A careful drug history is essential, but may be inconclusive, because although allergy to a
drug implies previous exposure the antigen may have occurred in foods (e.g. antibiotics are
often fed to livestock and drug residues remain in the flesh), in drug mixtures or in some casual
fashion.
2. Provocation tests. This means giving a very small amount of the suspected drug and seeing
if a reaction ensues. The commonest method is skin testing, where a drug is applied as a patch,
or is pricked or scratched into the skin, or injected intradermally. Unfortunately prick and
scratch testing is less useful for assessing the systemic reaction to drugs than it is for the more
usual atopic antigens (e.g. pollens) and both false positives and false negatives can occur.
Intradermal injection can provoke serious systemic anaphylaxis and fatalities have been
recorded. Patch testing is safe and is useful for the diagnosis of contact sensitivity but does not
reflect systemic reactions. It may also itself cause allergy. Provocation tests can also involve
giving small doses of the drug by inhalation, by mouth or parenterally and should only be
undertaken under expert guidance, after obtaining informed consent, and with full facilities for
resuscitation available. The initial dose should not exceed 1,0 mg by mouth or 1,0 ng by
injection. The dose is increased at intervals until a therapeutic dose is reached or a reaction
occurs.
3. Serological testing is rarely helpful as the demonstration of circulating antibodies does not
mean that they are necessarily the cause of the symptoms.
4. Sensitized lymphocytes. The demonstration of transformation occurring when lymphocytes
are exposed to a drug suggests that they are T lymphocytes sensitized to the drug, but
interpretation of results can be difficult in a clinical context. In this type of reaction the hapten
itself will often provoke lymphocyte transformation as well as the conjugate.
5. Often it is necessary to stop all the drugs that a patient is taking and reintroduce them one
by one until the drug at fault is discovered. This should only be done if the reaction is not
serious or if the drug is essential and no chemically unrelated alternative is available. Drug
allergies should be recorded in the case notes and the patient informed of the risks involved in
taking the drug again.
 11

PREVENTION OF ALLERGIC DRUG REACTIONS

Although it is probably not possible to avoid allergic drug reactions completely, the following
can decrease the incidence:
1. Taking a drug history is essential whenever drug treatment is anticipated, particularly with
antibiotics and other drugs with a high allergy potential. A history of atopy, although not
excluding the use of drugs, should make one wary.
2. Drugs given orally are less likely to cause severe allergic reactions than when given by
injection.
3. Prophylactic skin testing is not usually practicable and a negative test does not exclude the
possibility of an allergic reaction. Such testing may, however, be appropriate where a suspect
drug is potentially life saving and there is no equally effective alternative. It probably reduces
the risk of anaphylaxis if not of other less severe reactions.
4. Desensitization (hyposensitization). This should only be used when continued use of the
drug is essential. It consists of giving a very small dose of the drug and increasing the dose at
regular intervals, sometimes under cover of a glucorticosteroid and 2-adrenoreceptor agonist.
An antihistamine may be added if a drug reaction occurs, and equipment for resuscitation and
therapy of anaphylactic shock must be close at hand. It is often successful, though little is
known of the mechanism by which it is achieved.

Reference
Rawlins MD, Thompson JW: Pathogenesis of adverse drug reactions. In: Davis DM, ed.
Textbook of Adverse Drug Reactions. Oxford: Oxford University Press, 1977: 44.
 12

Adverse Drug Reactions

 13

Adverse drug reactions are defined as unwanted effects occurring at therapeutic doses of the
drug. Drugs are great mimics of disease and adverse drug reactions present with diverse
clinical signs and symptoms. The classification proposed by Rawlins and Thompson (1977)
divides reactions into type A and type B (Table 1).
TYPE A are adverse reactions which are a consequence of the drugs normal
pharmacological effect and are therefore predictable; they are dose-related with a low
mortality. Such reactions are usually due to incorrect dosage (too much or for too long) or to
disordered pharmacokinetics, usually a failure of drug elimination. The term side effects is
often applied to minor type A reactions.
TYPE B reactions are not predictable from the drugs known action, are not dose-
related and have a considerable mortality. The underlying pathophysiology of type B
reactions is by definition poorly if at all understood, and may have a genetic or immunological
basis. Indeed as more is known about the pharmacology of a drug, an adverse effect may have
to be reclassified type B to type A. A good example of this is hemolytic anemia due to
primaquine which used to be regarded as a rare type B effect of possible immune etiology until
its association with glucose-6-phosphate dehydrogenase (G-6PD) deficiency was established. It
is now a predictable (type A) reaction in such genetically predisposed individuals. Type B
reactions occur infrequently (1:1000-1:10 000 treated subjects being typical).

Table 1 Some examples of type A and type B reactions

Drug Type A reaction Type B reaction
Ampicillin Pseudomembranous colitis Interstitial nephritis
Chlorpromazine Sedation Hepatotoxicity
Naproxen Gastrointestinal hemorrhage Agranulocytosis
Practolol Bradycardia Oculomucocutaneous syndrome
Thiazides Gout Thrombocytopenia
Warfarin Bleeding Breast necrosis

Three further minor categories of adverse drug reaction have been proposed:

1. Type C: continuous reactions due to long-term use, such as tardive dyskinesia or analgesic
nephropathy.
2. Type D: delayed reactions such as carcinogenesis or teratogenesis.
3. Type E: end of use reactions such as adrenocortical insufficiency following withdrawal of
corticosteroids, or withdrawal syndromes following discontinuation of treatment with
clonidine, benzodiazepines, tricyclic antidepressants or -adrenoreceptor antagonists.

A random survey found that approximately 80% of adults took some kind of medication during
a 2-week period in UK. Exposure to drugs in the population is thus substantial and the
incidence of adverse reactions must be viewed in this context. Adverse drug reactions are most
commonly of type A (80%) causing up to 3% of acute hospital admissions and 2-3% of
consultations in general practice. In hospital their incidence is estimated to be 10-20%, causing
increased use of hospital resources and a mortality of 0,3-1%. They are most frequent and
most severe in neonates or the elderly (>60 years), in women, patients with hepatic or renal
disease, and those with a history of previous adverse drug reactions. Adverse drug reactions
occur most commonly early in therapy (1-10 days), and the drugs most commonly implicated
are digoxin, antimicrobials, diuretics, potassium, analgesics, sedatives and tranquillizers,
insulin, aspirin, glucocorticosteroids, antihypertensives and warfarin.
Factors involved in the etiology of adverse drug reactions can be classified as shown in Tab. 2.

Table 2 Factors involved in adverse drug reactions.

Patient factors
 14

Intrinsic Extrinsic
Age neonatal, infant and elderly Environment - sun
Sex hormonal environment Xenobiotics (e.g. drugs, herbicides)
Genetic abnormalities, e.g. enzyme defects Malnutrition
Previous adverse drug reactions, allergy, atopy
Presence of organ dysfunction disease
Personality and habits alcoholic, drug addict, nicotine, compliance

Prescriber factors
Incorrect drug or drug combination
Incorrect route of administration
Incorrect dose
Incorrect duration of therapy

Drug factors
Drug-drug interactions
Pharmaceutical - batch problems, shelf-life, incorrect dispensing

MONITORING/SURVEILLANCE (PHARMACOVIGILANCE)

The ideal method would identify adverse drug reactions with high sensitivity and specificity
and respond rapidly. It would detect rare adverse common ones, the incidence of which it
would quantify together with predisposing factors. Continued surveillance is mandatory
after a new drug has been marketed since it is inevitable that the preliminary testing of
medicines during drug development, although excluding many ill effects, cannot identify
uncommon adverse effects. A variety of early detection methods have been introduced to
discover these actions as swiftly as possible.

Phase I/II/III trials

Early (phase I/II) trials are important in assessing the tolerability and dose-response of new
drugs, but they are very insensitive at detecting adverse reactions because they are performed
on relatively few subjects (perhaps 200-300). Phase III clinical trials can, however, establish
the incidence of common adverse reactions and relate this to therapeutic benefit. Analysis of
the reasons given for dropping out of phase III trials is particularly valuable in establishing
whether common events such as headache, constipation, lethargy or male sexual dysfunction
are truly drug related The problem of detection is several orders of magnitude worse if the
adverse drug reaction resembles spontaneous disease in the population, such that physicians
are unlikely to attribute the reaction to drug exposure.

Yellow card scheme and postmarketing surveillance

Untoward effects that have been missed in early clinical trials become apparent when the drug
is used on a wider scale. Case reports in the literature, which may often stimulate further
reports, remain the most sensitive means of detecting rare but serious and unusual adverse
effects. In the UK, a Register of Adverse Reactions was started in 1964. Doctors and dentists
were asked to report any adverse effects which they considered due to drugs. The Committee
of Safety of Medicines (CSM) and the Medicines Control Agency (MCA) operate a system of
spontaneous reporting by doctors on prepaid yellow postcards. The yellow card scheme
consits of three stages: (1) collection of data; (2) analysis of data; and (3) feedback of
information to doctors.
Such methods of surveillance have proved useful as an early warning system, but the
major difficulty is underreporting. On average probably not more than 10% of adverse
reactions are reported. This may be due partly to confusion about what events to report, or in
deciding the precise relationship of a drug to a specific adverse event or sometimes which of
several drugs being taken by the patient is responsible. A further problem is that if the drug is
 15

responsible for an increase in the incidence of some disease which is already common, for
example gallbladder disease or carcinoma of the bronchus, the change in incidence must be
very large before this can be detected, since cases due to drug therapy are obscured by cases
due to other (often unknown) causes. Doctors are inefficient at detecting adverse reactions to
drugs and those reactions reported are in general the obvious or previously described and well-
known reactions. Several initiatives are currently underway to try to improve this situation by
both education and involvement of specially trained pharmacists in and out of hospitals.
The CSM introduced a system of high vigilance for newly marketed drugs. Any newly
marketed drug has on its data sheet and its entry in the British National Formulary a black
triangle for its first 2 years in the general market. This conveys to the prescriber that any
unexpected event occurring to a patient prescribed this drug should be reported by the yellow
card system. The pharmaceutical company marketing a new drug is also responsible for
obtaining accurate reports on all patients treated up to an agreed number. The scheme was
successful in the case of benoxaprofen, an antiinflammatory analgesic. Following its release,
there were spontaneous reports to the CSM of photosensitivity and onycholysis. Further
reports appeared in the elderly, in whom its half-life is prolonged, of cholestatic jaundice and
hepatorenal failure, which was fatal in eight cases. Benoxaprofen was subsequently taken off
the market when 3500 adverse drug reaction reports were received with 61 fatalities. The
yellow card black triangle scheme was instrumental in the early identification of urticaria and
cough as adverse effects of angiotensin-converting enzyme inhibitors. Although, potentially, the
population under study by this system comprises all the patients using a drug, in fact
underreporting yields a population that is not uniformly sampled and thus data that are
unrepresentative and difficult to work with statistically. This accounts for the paucity of
accurate incidence data for adverse drug reactions.
Systems such as the yellow card one are relatively inexpensive and easy to manage and
facilitate ongoing monitoring of all drugs, all consumers and all types of adverse reaction.
Reports from drug regulatory bodies of 22 countries are collated by the World Health
Organization (WHO) Unit of Drug Evaluation and Monitoring in Geneva. Rapid access to
reports from other countries should be of great value in detecting rarer types of adverse
reaction, although the same reservations as apply to national systems apply to this register.
Additionally, this database might be able to yield information concerning different prescribing
habits and even differences in the pattern of untoward drug effects in different countries.

Prescription event monitoring

Successful studies have been carried out by the Drug Surveillance Research Unit at the
University of Southampton using prescription event monitoring. Prescriptions for certain drugs
are identified by the prescriptions pricing office in Edinburgh. These are followed up and the
prescribing doctor asked to fill in a simple questionnaire recording any medical event from the
patient notes. This has the advantage that the prescriber does not have to judge causality
between the event and the drug. This scheme identified a new event associated with enalapril,
i.e. deafness in 19/12 500. This is in excess of an expected five patients developing deafness in
the general population. This needs further investigation, but highlights the potential usefulness
of the approach as well as one of its main limitations, namely the absence of a control goup.

Case-control studies
One of the great difficulties in monitoring drugs for adverse effects is that if they occur
relatively rarely (perhaps 1 in 1000 patients), very many patients have to be monitored before a
particular side effect becomes apparent. An alternative approach is to study a group of patients
who have developed a particular symptom or syndrome and to compare the frequency of
exposure to possible etiologic agents with a control group. A prior suspicion (hypothesis) must
exist to prompt the setting up of such a study: examples are the possible connection between
irradiation or environmental pollution and certain malignancies, especially where they are
 16

observed in clusters. Although this method is comparatively easy to carry out, artifacs occur as
a result of faulty selection of patients and controls, and the approach remains highly
controversial among epidemiologists, public health physicians and statisticians. It has had some
apparent successes, such as the linking of stilbestrol with vaginal adenocarcinoma and of
licomycin with pseudomembranous colitis, and, recently, the association between fenoterol use
and an increased number of deaths in asthmatics. In its most simple form it is really an
extension of the idea of the alert doctor, but with the considerable amount of data now
available in such studies as the Boston Drug Surveillance Program, it should be possible to use
this method on a wider scale and in a more sophisticated manner.

Patient questionnaires
Self-administered questionnaires have been used for outpatients attending hypertension and diabetic clinics
and have detected previously unsuspected adverse effects, for example headache and weakness in the legs as
effects of metformin. They have also been used to show an absence of effects, for example that propranolol is
not associated with any of the eye symptoms caused by practolol.

Intensive monitoring
In the last few years a number of intensive monitoring programs, usually hospital-based, have
been started. The Aberdeen-Dundee monitoring system abstracts data from some 70 000
hospital admissions each year, storing these on a computer file before analysis. The Boston
Collaborative Drug Surveillance Program (BCDSP) which involves selected hospitals in
several countries, is even more comprehensive. In the BCDSP all patients admitted to specially
designated general wards are included in the analysis. Specially trained personnel obtain the
following information from hospital patients and records:

1. Background information, i.e. age, weight, height, etc.

2. Patients medical history.
3. Patients drug exposure.
4. Side effects of drugs.
5. Outcome of treatment and changes in laboratory tests during hospital admission.

A unique feature of comprehensive drug-monitoring systems lies in their potential to

follow-up and investigate adverse reactions suggested by less sophisticated detection systems
or by isolated case reports in medical journals. Furthermore, the frequency of side effects can
be determined more cheaply than by a specially mounted trial to investigate a simple adverse
effect. Thus, for example, the risk of developing a rash with ampicillin was found to be around
7% both by clinical trial and by the BCDSP which can determine such facts almost
automatically from data on its files. New adverse reactions or drug interactions are sough by
multiple correlation analysis of the data. Thus, when an unexpected relationship arises, such as
the 20% incidence of gastrointestinal bleeding in severely ill patients treated with ethacrynic
acid compared to 4,3% of similar patients treated with other diuretics, there is little chance of
bias arising from awareness of the hypothesis during data collection since the data are collected
before the hypothesis is proposed. It is also possible to isolate factors predisposing patients to
a particular complication; in the case of ethacrynic acid these were female sex, a high blood
area, previous heparin administration and intravenous administration of the drug. An important
aspect of this type of approach is that lack of clinically important associations can also be
detected. Thus no significant association between aspirin and renal disease was found, whereas
long-term aspirin consumption is associated with a decreased incidence of myocardial
infarction, an association which has been shown to be of therapeutic importance in randomized
clinical trials. There are plans to extend intensive drug monitoring to cover other areas of
medical practice.
 17

In terms of new but uncommon adverse reactions, however, the numbers of patients
undergoing intensive monitoring while taking a particular drug will inevitably be too small for
the effect to be detectable. Such monitoring therefore can provide information only about
relatively common, early reactions to drugs used under hospital conditions. Patients are not in
hospital long enough for detection of delayed effects, which are just the reactions least likely to
be recognized as such even by an astute clinician.

Monitoring from national statistics

A great deal of information is available from death certificates, hospital discharge diagnoses
and similar records. It may be possible from these data to pick out a change in disease trends
and relate this to drug therapy. Perhaps the best-known example of this is the increased death
rate in young asthmatics noted in the mid-1960s, which was connected with overuse of
bronchodilator inhalers containing non-specific -adrenoreceptor agonists (e.g. adrenaline
and/or isoprenaline). Although relatively inexpensive the difficulty of this method is obvious.
Large numbers of patients must suffer before the change is detectable, particularly in diseases
with an appreciable mortality. Data interpretation is particularly difficult when hospital
discharges are used as a source of information, since sometimes diagnosis is provisional and
changes as the patients disease progresses.

Record linkage
Record linkage is presently under investigation as a method of monitoring adverse drug
reactions and may well produce some important results. The basic idea is that the medical
records from a defined population (both general practice and hospital sources) are kept from
cradle to grave and analyzed. This method could be particularly useful when looking for really
long-term ill effects from drugs or from other factors, for instance the increased incidence of
leukemia in those receiving radiation in utero or the occurrence of diseases such as cancer or
mental retardation in individuals whose mothers had received drugs during pregnancy. Analysis
of data is becoming more sophisticated with the use of computers to store information, but one
must know what information is worth storing, and ask the correct questions to avoid the well-
known garbage in - garbage out situation.
The main types of inquiry can be made:

1. Follow-up of individuals who have received selected drugs to determine adverse reactions
attributable to the drug. This approach is of most value when a drug is already suspected of
producing some particular effect and is not particularly suited to the discovery of unexpected
adverse effects. A defect is the absence of a control group. It has been convincingly
demonstrated that untreated subjects or those receiving placebo often experience symptoms
commonly listed as side effects of drugs. Therefore the results of such studies can be biased by
the preconceived ideas of the investigator or the effects on the patient of his or her enhanced
interest.
2. Follow-up of patients with specific diseases rather than those treated with specific drugs is
sometimes easier to carry out since most clinicians categorize their patients under disease
rather than treatment. Thus it is possible to investigate the frequency of adverse reactions to,
say, phenytoin, by selecting the records of epileptics. One of the largest disease-oriented
systems is that of the USA Perinatal Study which received data on drug exposure and fetal
outcome from over 50 000 consecutive pregnancies. From this has come evidence linking
malformations and maternal exposure to phenytoin.

Feedback
There is no use in collecting vast amounts of data on adverse reactions unless they are analyzed
and important information reported back to prescribing doctors. In addition to articles in the
 18

medical press the CSM and MCA circulate to all doctors a Current Problems information
sheet which deals with important and recently discovered adverse reactions. If an acute and
serious problem is recognized doctors will usually receive notification from the CSM and often
from the pharmaceutical company involved.

ALLERGIC ADVERSE DRUG REACTIONS

Immune mechanisms are involved in a number of adverse effects caused by drugs. The
development of allergy implies previous exposure to the drug or to some very closely related
substance. Most drugs are of low molecular weight (less than 1000) and thus are not antigenic.
They can, however, combine with substances of high molecular weight, usually proteins, acting
as haptens so that the conjugate thus formed is antigenic.
The factors that determine the development of allergy to a drug are no fully
understood. Some drugs (e.g. penicillin) are more likely to cause allergic reactions than others,
and type I (immediate anaphylactic) reactions are more common in patients with a history of
atopy. A correlation between allergic reactions involving immunoglobulin E (IgE) and human
leukocyte antigen (HLA) serotypes have been reported and so genetic factors may also be
important. There is some evidence that drug allergies are more common in older people, in
women, and in those with a previous history of drug reaction. This may, however, merely
represent increased frequencies of drug exposure in these groups, and prevalence figures
(expressed relative to the appropriate denominator) are currently not available.

Fig. 11.1 The immune response to drugs.

Type IV
Central immune response
apparatus

Sensiti-
Drugs or its Cell T lymp- zed lym-
metabolites membrane hocytes phocytes

Marcrophages

Type I, II &
III
Drugs or its Protein Antigen responses
metabolites
B lymp-
hocytes Humoral
antibodies

Drug
(large molecule)
 19

Types of allergy

Drugs cause a variety of allergic responses (Fig. 11.1.) and sometimes a single drug can be
responsible for more than one type of allergic response.

TYPE I

Type I reactions (e.g. to penicillin) are due to the production of reagenic antibodies known to
consist predominantly of class IgE. The antigen-antibody reaction on the surface of mast cells
causes degranulation and release of pharmacologically active substances. It occurs commonly
with foreign serum or penicillin but may also occur with streptomycin and some local
anesthetics. With penicillin it is believed that the penicilloyl moiety of the penicillin molecule is
responsible for the production of antibodies. Occasionally the development of symptoms is
delayed for some hours after the drug has been taken. It seems probable that antigen-antibody
aggregates are involved. The antibody in this mechanism consists of IgG. The therapy of
anaphylactic shock is described in chapter 47.

TYPE II

These are due to antibodies of class IgG and IgM which on contact with antibodies on the
surface of cells are able to fix complement, causing cell lysis, for example penicillin and
methyldopa causing Coombs positive hemolytic anemia.

TYPE III, IMMUNE COMPLEX, ARTHUS REACTIONS

Circulating immune complexes can produce several clinical allergic states including serum
sickness and immune complex glomerulonephritis, and a syndrome resembling systemic lupus
erythematosus. The onset of serum sickness is delayed for some days until the symptoms -
fever, urticaria, arthropathy, lymphadenopathy and eosinophilia - develop. Proteinuria occurs
frequently. Recovery takes a few days. Causative agents include foreign serum, penicillin,
sulfonamides, streptomycin and propylthiouracil. Amiodarone lung and hydralazine induced
systemic lupus syndrome are also possibly mediated by immune complex-related mechanisms,
although these reactions are less well understood.

TYPE IV

Type IV are delayed hypersensitivity reactions, the classical example of which is contact
dermatitis (e.g. to topical antibiotics such as penicillin). The mechanism here is that the drug
applied to the skin forms an antigenic conjugate with dermal proteins, stimulating formation of
sensitized T lymphocytes in the regional lymph nodes with a resultant rash if the drug is applied
again. Drug photosensitivity is due to a photochemical combination between the drug (e.g.
tetracycline, amiodarone) and dermal protein. Delayed sensitivity can also result from the
systemic administration of drugs.

EXAMPLES OF ALLERGIC AND OTHER KINDS OF ADVERSE DRUG

REACTIONS
 20

Adverse drug reactions can be manifest in any organ system, or involve multiple organ systems,
and in extraordinarily diverse forms. Specific instances are dealt with throughout this book.
Some examples to illustrate the diversity of adverse drug reactions are also given here.
Rashes
These are common manifestations of drug reactions. A number of immune mechanisms may be
involved which produce many types of rash. In some cases (e.g. ampicillin) the rash is more
commonly due to non-allergic factors; in others the cause is not known (Table 11.4).

Serum sickness

This is caused by circulating immune complexes and the onset is delayed for some days until
the symptoms - fever, urticaria and arthropathy develop. Proteinuria, cosinophilia and
lymphadenopathy occur frequently. Recovery takes a few days. Causative agents include
foreign serum, penicillin, sulfonamides, streptomycin and propylthiouracil, possibly amiodarone
lung and hydralazine-induced systemic lupus syndrome.

Lymphadenopathy

Lymph-node enlargement can result from taking drugs (e.g. phenytoin). The mechanism is not
known, but allergic factors may be involved. The reaction may be confused with a lymphoma
and inquiry about chronic drug taking is important in patients with lymphadenopathy of
unknown cause.

Blood dyscrasias

Thrombocytopenia, anemia (aplastic, iron deficiency, macrocytic, hemolytic) and

agranulocytosis can all be produced by drugs.
Thrombocytopenia can occur with many drugs and in many instances it is direct
suppression of the megakarocytes rathter than immune processes which is important. Quinine
thrombocytopenia may be an allergic phenomenon. The following are some of the drugs which
most commonly cause thrombocytopenia:
Thiazides, chloramphenicol.
Quinidine.
Heparin.
Sulfonamides, quinine, gold.
Most cytotoxic agents.

Hemolytic anemia can be caused by a number of drugs and sometimes immune

mechanisms are responsible. Glucose-6-phosphate dehydrogenase deficiency predisposes to
non-immune hemolysis (e.g. to primaquine). Immune mechanisms include:

1. Combination of the drug with the red cell membrane with the conjugate acting as an antigen.
This has been shown to occur with penicillin-induced hemolysis, and may also occur with
chlorpromazine, chlorpropamide and sulfonamides.

2. Alteration of the red-cell membrane by the drug so that it becomes autoimmunogenic. This
may happen with methyldopa, and a direct positive Coombs test develops in about 20% of
patients who have been treated with this drug for more than a year. Is only a small proportion
does hemolysis actually occur. Similar changes can take place with levodopa and mefenamic
acid.
 21

3. Causing non-specific binding of plasma protein to red cells and thus causing hemolysis. This
is believed to occur with cephalosporins.
Aplastic anemia as an isolated entity is not common but may occur either in isolation
or as part of a general depression of bone-marrow activity (pancytopenia); examples include
chloramphenicol and, commonly and predictably, cytotoxic drugs.

Agranulocytosis can be caused by many drugs. Several different mechanisms are

implicated and it is not known whether allergy plays a part. Among the drugs most frequently
implicated are:
Antithyroid drugs.
Phenothiazines, tolbutamide, most cytotoxic agents.
Sulfonamides, antidepressants (especially mianserin)
Clozapine.

Systemic lupus erythematosus

Several drugs (including procainamide, isoniazid, hydralazine, chlorpromazine,

anticonvulsants) produce a syndrome that resembles systemic lupus together with a positive
antinuclear factor test. The development of this is closely related to dose, and in the case of
hydralazine also depends upon the rate of acetylation of the drug, which is genetically
controlled. There is some evidence that the drugs act as haptens combining with DNA and
forming antigens. Symptoms usually disappear when the drug is stopped, but recovery may be
slow.

Vasculitis

Both acute and chronic vasculitis can result from taking drugs, and may have an allergic basis.
Acute vasculitis with purpura and renal involvement occurs with penicillins and the
sulfonamides. A more chronic form can occur with phenytoin.

Renal

All clinical expressions of renal disease can be caused by drugs: common culprits are non-
steroidal anti-inflammatory drugs and angiotensin-converting enzyme inhibitors (which cause
functional and usually reversible renal failure in susceptible patients). Nephrotic syndrome
results from several drugs (e.g. penicillamine, high-dose captopril, gold) which cause a variety
of immune-mediated glomerular injuries. Interstitial nephritis can be caused by several drugs
including non-steroidal anti-inflammatory drugs and penicillins. Aminoglycoside antibiotics and
vancomycin cause direct tubular toxicity. Many drugs cause electrolyte or acid-base
disturbances via their predictable direct or indirect effects on renal electrolyte excretion (e.g.
hypokalemia and hypomagnesemia from loop diuretics, hyperkalemia from potassium-sparing
diuretics and converting enzyme inhibitors, proximal renal tubular acidosis from carbonic
anhydrase inhibitors) and some cause unpredictable toxic effects on acid-base balance (e.g.
distal renal tubular acidosis from amphotericin). Obstructive uropathy can be caused by uric
acid crystals consequent on initiation of chemotherapy in patients with hematological
malignancy, and rarely poorly soluble drugs such as sulfonamides can themselves cause clinical
problems consequent on crystaluria.

Other reactions
 22

Fever is a common manifestation of drug allergy and should be remembered in patients with
fever of unknown cause.
Liver damage (hepatitis with or without obstructive features) as a side effect from
drugs is important. It may be insidious, leading slowly to end-stage cirrhosis (e.g. during
chronic treatment with methotrexate) or acute and fulminant (as in some cases of isoniazid,
halothane or phenytoin hepatitis). Chlorpromazine commonly causes liver involvement
characterized by raised alkaline phosphatase (obstructive pattern). Gallstones (and
mechanical obstruction) can be caused by fibrates and other lipid-lowering drugs, and by
octreotide, a somatostatin analog used to treat a variety of enteropancreatic tumors including
carcinoid syndrome and VIPomas*. Immune mechanisms are implicated in some forms of
hepatic injury by drugs, but are seldom solely responsible.
---------------------------------------------------------------------------------------------------------------
* VIP = vasoactive intestinal polypeptide
---------------------------------------------------------------------------------------------------------------

IDENTIFICATION OF THE DRUG AT FAULT

It is often difficult to decide whether a clinical event is drug related; even when this is probable
it may be difficult to determine which drug is responsible, as patients are often taking more
than one. One or more of several possible approaches may be appropriate:

1. A careful drug history is essential, but may be inconclusive, because although allergy to a
drug implies previous exposure the antigen may have occurred in foods (e.g. antibiotics are
often fed to livestock and drug residues remain in the flesh), in drug mixtures or in some casual
fashion.
2. Provocation tests. This means giving a very small amount of the suspected drug and seeing
if a reaction ensues. The commonest method is skin testing, where a drug is applied as a patch,
or is pricked or scratched into the skin, or injected intradermally. Unfortunately prick and
scratch testing is less useful for assessing the systemic reaction to drugs than it is for the more
usual atopic antigens (e.g. pollens) and both false positives and false negatives can occur.
Intradermal injection can provoke serious systemic anaphylaxis and fatalities have been
recorded. Patch testing is safe and is useful for the diagnosis of contact sensitivity but does not
reflect systemic reactions. It may also itself cause allergy. Provocation tests can also involve
giving small doses of the drug by inhalation, by mouth or parenterally and should only be
undertaken under expert guidance, after obtaining informed consent, and with full facilities for
resuscitation available. The initial dose should not exceed 1,0 mg by mouth or 1,0 ng by
injection. The dose is increased at intervals until a therapeutic dose is reached or a reaction
occurs.
3. Serological testing is rarely helpful as the demonstration of circulating antibodies does not
mean that they are necessarily the cause of the symptoms.
4. Sensitized lymphocytes. The demonstration of transformation occurring when lymphocytes
are exposed to a drug suggests that they are T lymphocytes sensitized to the drug, but
interpretation of results can be difficult in a clinical context. In this type of reaction the hapten
itself will often provoke lymphocyte transformation as well as the conjugate.
5. Often it is necessary to stop all the drugs that a patient is taking and reintroduce them one
by one until the drug at fault is discovered. This should only be done if the reaction is not
serious or if the drug is essential and no chemically unrelated alternative is available. Drug
allergies should be recorded in the case notes and the patient informed of the risks involved in
taking the drug again.

PREVENTION OF ALLERGIC DRUG REACTIONS

 23

Although it is probably not possible to avoid allergic drug reactions completely, the following
can decrease the incidence:
1. Taking a drug history is essential whenever drug treatment is anticipated, particularly with
antibiotics and other drugs with a high allergy potential. A history of atopy, although not
excluding the use of drugs, should make one wary.
2. Drugs given orally are less likely to cause severe allergic reactions than when given by
injection.
3. Prophylactic skin testing is not usually practicable and a negative test does not exclude the
possibility of an allergic reaction. Such testing may, however, be appropriate where a suspect
drug is potentially life saving and there is no equally effective alternative. It probably reduces
the risk of anaphylaxis if not of other less severe reactions.
4. Desensitization (hyposensitization). This should only be used when continued use of the
drug is essential. It consists of giving a very small dose of the drug and increasing the dose at
regular intervals, sometimes under cover of a glucorticosteroid and 2-adrenoreceptor agonist.
An antihistamine may be added if a drug reaction occurs, and equipment for resuscitation and
therapy of anaphylactic shock must be close at hand. It is often successful, though little is
known of the mechanism by which it is achieved.

Reference
Rawlins MD, Thompson JW: Pathogenesis of adverse drug reactions. In: Davis DM, ed.
Textbook of Adverse Drug Reactions. Oxford: Oxford University Press, 1977: 44.