10.

Drug allergy
Rebecca S. Gruchalla, MD, PhD Dallas, Tex

Adverse drug reactions are common, but only 6% to 10% are

immunologically mediated. Unlike most adverse drug reac- Abbreviations used
tions, allergic drug reactions are unpredictable. Whereas some ADR: Adverse drug reaction
drug-induced allergic reactions may be easily classified into DRESS: Drug reaction with eosinophilia and systemic
one of the four Gell and Coombs hypersensitivity categories, symptoms
many others that appear to have an immunologic component HSS: Hypersensitivity syndrome
cannot be classified because of our lack of mechanistic infor- NSAID: Nonsteroidal anti-inflammatory drug
mation. Theoretically, any drug can induce an immune PPL: Penicilloyl polylysine
response. However, some drugs are more likely to elicit clini- SJS: Stevens-Johnson syndrome
cally relevant immune responses than are others. Drugs in this TEN: Toxic epidermal necrolysis
category include antimicrobial drugs, anticonvulsants,
chemotherapeutic agents, heparin, insulin, protamine, and bio-
logic response modifiers. After a drug-disease connection is Drug-induced adverse events were well characterized for
established, it must be determined whether the reaction was each patient, and cutaneous reactions that were acceptable
immunologically mediated. Subsequently, confirmatory tests, if for further analysis consisted of only those reactions that
available, should be used to determine the allergic status of the
were known or presumed to be allergic in nature.
patient. If these tests are not available, a graded challenge or
Three hundred fifty-eight cutaneous reactions occurred
desensitization may be considered, depending on the type of
clinical reaction previously demonstrated and the need for in 347 patients, yielding a mean reaction rate of 2.2%.
drug readministration. Education of the patient and primary Most of the reactions (94%) were morbilliform eruptions.
care physician is an important component of patient manage- Five percent were urticarial. For each of the 51 drugs
ment. (J Allergy Clin Immunol 2003;111:S548-59.) studied, the number of reactions per 1000 drug adminis-
trations was determined, and reaction rates were calculat-
Key words: Adverse drug reaction, allergic drug reaction, hyper-
ed. Reaction rates were as follows: amoxicillin, 5.1%;
sensitivity syndrome, hapten hypothesis, antimicrobial drugs, peni-
cillins, skin testing, graded challenge, desensitization trimethoprim-sulfamethoxazole, 3.4%; ampicillin, 3.3%;
blood products, 2.2%; cephalosporins, 2.1%; semisyn-
EPIDEMIOLOGY OF DRUG-INDUCED thetic penicillins, 2.1%; erythromycin, 2.0%; and peni-
ALLERGIC REACTIONS cillin G, 1.8%. Thus, in this study, antibiotics accounted
for most drug-induced cutaneous allergic reactions.
Allergic drug reactions are one type of adverse drug More recently, the epidemiology of drug-induced ana-
reaction (ADR). An ADR has been defined by the World phylaxis has been evaluated. In 1999, Laxenaire3 pub-
Health Organization as any noxious, unintended, and unde- lished the fourth French survey of anaphylaxis during
sired effect of a drug that occurs at doses used for preven- general anesthesia. Members of the Perioperative Ana-
tion, diagnosis, or treatment.1 Although ADRs are com- phylactoid Reactions Study Group identified and evalu-
mon, it has been difficult to determine their true frequency ated patients in whom anaphylaxis had occurred during
because adequate reporting mechanisms do not exist. For anesthesia from July 1994 to December 1996. One thou-
similar reasons, determining the incidence and prevalence sand six hundred forty-eight patients were identified and
of allergic drug reactions has been challenging as well. skin tested. An IgE-mediated mechanism accounted for
Although it has been difficult to determine the frequen- 692 of the reactions (characteristic clinical symptoms
cy of drug-induced allergic reactions specifically, it is and positive skin test results), and another 611 were
known that they account for only a small proportion of judged to be anaphylactoid in nature (characteristic clin-
ADRs. To determine the frequency of cutaneous allergic ical symptoms and negative skin test results). The
reactions to drugs introduced after 1975, Bigby et al2 ana- remaining 345 cases could not be categorized. The most
lyzed data on 15,438 consecutive medical inpatients who common agents that provoked IgE-mediated anaphylac-
had been monitored by the Boston Collaborative Drug tic reactions were muscle relaxants and latex.
Surveillance Program from June 1975 to June 1982. Fatalities as a result of drug-induced anaphylaxis have
been recorded in several studies. On the basis of informa-
tion gathered by the Danish Committee on Adverse Drug
Reactions and the Central Death Register, 30 cases of
From the University of Texas Southwestern Medical Center, Dallas. fatal drug-induced anaphylaxis were identified in Den-
Reprint requests: Rebecca S. Gruchalla, MD, PhD, Associate Professor of mark from 1968 through 1990. The most common causes
Internal Medicine, Chief, Division of Allergy and Immunology, UT South- were contrast media, antibiotics, and allergenic extracts.4
western Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-8859.
© 2003 Mosby, Inc. All rights reserved.
More recently, Pumphrey5 searched death certificates in
0091-6749/2003 $30.00 + 0 the United Kingdom for anaphylactic causes of death dur-
doi:10.1067/mai.2003.93 ing the interval from 1992 through 1998. One hundred
S548
J ALLERGY CLIN IMMUNOL Gruchalla S549
VOLUME 111, NUMBER 2

sixty-four anaphylactically induced fatalities were identi- initiated. Although bioactivation is typically mediated by
fied, and 39% of these were drug induced. The most com- cytochrome P450 enzymes in liver hepatocytes, it may also
mon drug causes were anesthetics (27 cases), antibiotics occur at other sites, such as skin keratinocytes.
(16 cases), and contrast media (8 cases). Bioactivation is usually followed by a bioinactivating
process. In some cases, however, genetic or environmen-
CLASSIFICATION OF DRUG-INDUCED tal factors may perturb the balance between these two
ALLERGIC REACTIONS processes, leading to increased formation or decreased
elimination of reactive drug metabolites. Once formed,
ADRs are grouped into two broad categories, those that these reactive species may do one of several things. (1)
are predictable, common, and related to the pharmacolog- They may bind to macromolecules and cause direct cel-
ic actions of the drug (type A reactions) and those that are lular damage. (2) They may bind to nucleic acids to pro-
unpredictable, uncommon, and usually not related to the duce an altered gene product. (3) They may bind cova-
pharmacologic actions of the drug (type B reactions).6 lently to larger macromolecular targets, form an
Approximately 80% of ADRs fall into the first category, immunogenic complex, and induce an immune response.
and typical examples are drug-induced toxicity, side
effects, secondary effects, and drug interactions. Immune- IMMUNE REACTIONS TO ANTIMICROBIAL
mediated or allergic drug reactions fall into the second DRUGS
category. Like other type B reactions, these reactions are Penicillin and other β-lactam drugs
uncommon, comprising only 6% to 10% of all ADRs.7
Unlike type A reactions, type B reactions often are not Allergy to β-lactam drugs is commonly reported,
manifested until after a drug is marketed. Moreover, their especially penicillin allergy. The most common β-lac-
development appears to be dependent on both genetic tam–induced drug reactions are maculopapular or mor-
and environmental factors. Reactions in this category billiform and urticarial eruptions. However, severe ana-
include drug intolerance (an undesired drug effect pro- phylactic reactions can and do occur on rare occasions. A
duced by the drug at therapeutic or subtherapeutic review of penicillin-induced anaphylaxis that was done
dosages), idiosyncratic reactions (uncharacteristic reac- in the late 1960s evaluated data from both published and
tions that are not explicable in terms of the known phar- unpublished reports and found an occurrence rate of 1.5
macologic actions of the drug), and allergic or hypersen- to 4 cases per 10,000 treated patients.10 Subsequently, a
sitivity reactions (reactions that are dependent on one or prospective international study was performed to deter-
more immunologic mechanisms). mine the incidence of allergic reactions to monthly intra-
Allergic reactions can be further classified according muscular benzathine penicillin injections that were
to the Coombs and Gell8 classification system into administered to prevent rheumatic fever recurrences. One
immediate-type hypersensitivity reactions (mediated by thousand ninety patients from 11 countries were
drug-specific IgE antibodies), cytotoxic and immune enrolled. After 32,430 injections during 2736 patient
complex reactions (mediated by drug-specific IgG or years of observation, 57 of 1790 patients (3.2%) had an
IgM antibodies), and delayed-type hypersensitivity reac- allergic reaction, and 4 of these reactions were anaphy-
tions (mediated by drug-specific T lymphocytes). lactic (incidence of 0.2%; 1.2 cases/10,000 injections).11
Although these categories seem relatively straightfor- Despite the fact that penicillin-induced anaphylaxis is
ward, classifying most drug reactions into one or more of rare, this drug continues to be the most common cause of
them can be quite a challenge because of our lack of anaphylaxis in human beings, accounting for approxi-
mechanistic information regarding these reactions. mately 75% of fatal anaphylactic cases in the United
States each year.12,13
PATHOGENESIS OF DRUG-INDUCED The penicillins have been the most extensively studied
ALLERGIC REACTIONS antibiotic family, and for this reason much is known
about their immunochemistry. All penicillins contain
Because of their macromolecular form, some drugs, both a β-lactam ring and a thiazolidine ring. In addition,
such as peptide hormones, are intrinsically immunogenic. each can be distinguished by the nature of the R side-
Many drugs, however, have a molecular mass of less than chain group (Fig 1). Whereas most other haptenic drugs,
1000 daltons and are incapable of inducing an immune such as the sulfonamides, must be metabolized before
response in their native state. For these agents to become they react with proteins to form immunogenic complex-
effective immunogens, they not only must bind covalent- es, penicillin is intrinsically reactive because of its β-lac-
ly to high–molecular weight proteins but also must under- tam ring. Because of its instability, this ring structure
go successful antigen processing and presentation. readily opens, allowing the carbonyl group to form amide
Our understanding of the immune response to drug anti- linkages with amino groups of lysine residues on nearby
gens is based primarily on the hapten hypothesis.9 Some proteins.14 Because approximately 95% of penicillin
drugs, such as penicillin, can be directly chemically reac- molecules bind to proteins in this manner, the antigenic
tive as a result of the instability of their molecular struc- determinant formed, benzyl penicilloyl, has been termed
ture. Others, however, must be metabolized, or bioactivat- the major penicillin determinant. After its identification,
ed, to a reactive form before an immune response can be penicilloyl determinants were coupled to a weakly
S550 Gruchalla J ALLERGY CLIN IMMUNOL
FEBRUARY 2003

FIG 1. General structure of penicillins and structures of the major and minor penicillin antigenic-allergenic
determinants, showing point of attachment of the drug to the carrier. From Baldo B. Penicillins and
cephalosporins as allergens—structural aspects of recognition and cross-reactions. Clin Exp Allergy
1999;29:745. Used with permission.

immunogenic polylysine carrier to form penicilloyl In addition to the antigenic determinants that are
polylysine (PPL), which is now commercially available. formed from the β-lactam ring structure, the side-chain
In addition to the penicilloyl determinant, several group that distinguishes the different penicillins also may
other minor penicillin determinants are formed, and these elicit the production of IgE antibodies that are clinically
too have been shown to elicit IgE-mediated responses in significant. Thus, specific tests for individual penicillins
human beings. Because of their importance, not only may be needed, as opposed to simply using major and
should PPL be used as a testing reagent when evaluating minor determinant preparations made from benzylpeni-
patients for the presence of penicillin-specific IgE anti- cillin. The importance of side-chain–specific antibodies
bodies, but a mixture of minor determinants should also was recently demonstrated in a study by Baldo23 in
be used. The original minor determinant mixture that was which the IgE-binding specificity was evaluated in
developed and analyzed consisted of benzylpenicillin, its patients who had reacted to flucloxacillin. Quantitative
alkaline hydrolysis product (benzylpenicilloate), and its hapten inhibition studies demonstrated that only
acid hydrolysis product (benzylpenilloate).15 dicloxacillin, cloxacillin, and oxacillin (penicillins that
It has been well documented that patients with a posi- have an R group similar to that in flucloxacillin) were
tive history but negative skin test results with PPL and able to strongly inhibit IgE binding. Penicillins that did
minor determinant mixture rarely have IgE-mediated not possess a methyl-phenyl-isoxazolyl side-chain deter-
reactions on penicillin readministration.16-19 If such reac- minant were poor inhibitors. These results indicate that,
tions do occur, they are mild and self-limited, and ana- at least for some β-lactam–allergic persons, the IgE anti-
phylaxis has never been reported in a person with a neg- bodies that are formed may be directed toward the R
ative penicillin skin test.20 group of the β-lactam drug and not to the determinants
PPL (Pre-Pen) is the only commercially available formed by the β-lactam or the thiazolidine rings. This
penicillin skin test reagent. Unfortunately, the use of this finding suggests that different penicillins may be cross-
reagent alone could cause as many as 25% of all poten- reactive, not only by virtue of their shared β-lactam and
tial positive skin test reactions to be missed.18 If fresh thiazolidine rings but also by virtue of shared or similar
(not aged) benzylpenicillin G is used (at a concentration side-chain determinants. Because we do not have any
of 10,000 U/mL) as the only minor determinant (along skin test reagents for the semisynthetic penicillins in this
with PPL), 5% to 10% of potential positive skin test reac- country and thus do not have side-chain–specific
tions will be missed.18,21 Some of the missed persons reagents, it is helpful to have knowledge of the side-
may be at risk for development of anaphylaxis if peni- chain. Fig 2 lists the various semisynthetic penicillins
cillin is readministered.22 and their structural similarities.
J ALLERGY CLIN IMMUNOL Gruchalla S551
VOLUME 111, NUMBER 2

FIG 2. Structural similarities and differences of penicillin. From Baldo B. Penicillins and cephalosporins as
allergens—structural aspects of recognition and cross-reactions. Clin Exp Allergy 1999;29:745. Used with
permission.

In contrast to the penicillins, our understanding of the 135 patients with positive skin test results who under-
immunochemistry of the cephalosporins is even more went challenge, 6 had reactions (reaction rate of 4.4%),
limited. Thus, our knowledge of the relevant whereas only 2 of 351 (reaction rate of 1.3%) patients
cephalosporin antigenic determinants is sparse, and for with negative skin test results reacted. Although these
this reason their degree of cross-reactivity is unknown. data indicate that patients who have known penicillin-
Also, we are still unable to answer an important age-old specific IgE antibodies may be at increased risk for a
question: Can penicillin-allergic patients safely receive reaction to cephalosporins, other studies have shown that
cephalosporins? Although these two drug classes share a this risk is actually minimal.26,27
β-lactam ring (cephalosporins also have a unique dihy- Like the penicillins, cephalosporins too can induce
drothiazine ring), clinically relevant cross-reactivity is immune responses. Side-chain–specific antibodies can be
not common. Lin24 found in a review of the literature that formed, as well as antibodies directed toward the ring
of 15,987 patients who were treated with cephaloridine, structures. Thus, the principles of allergenic cross-reac-
cephalexin, cephalothin, cefaxolin, or cefamandole, tions between cephalosporins are similar to those that per-
8.1% of those with a history of penicillin allergy had tain to the penicillins. If IgE antibodies are directed
reactions, versus 1.9% of persons who did not have such toward the core ring structures, cross-reactivity may exist
a history. More recently, Kelkar and Li25 summarized all among all the cephalosporins. If antibodies exist to the R1
published studies that evaluated the risk of administering or R2 side-chain group, however, the situation becomes
a cephalosporin to a penicillin-allergic patient. In 8 of the much more complex. Cross-reactions may occur through
studies evaluated, penicillin skin testing was performed. R1 recognition of identical (cefaclor, cephalexin,
In 3 of these studies, both persons who had positive skin cephaloglycin) or similar (cefaclor and cefadroxil) side-
test results and those who had negative skin test results chains, or they may occur through R2 recognition
underwent challenge; in 4, only those with positive skin (cephalothin and cefotaxime).23 Current recommenda-
test results underwent challenge, and in 1, only those tions for patients with a demonstrated cephalosporin sen-
with negative skin test results underwent challenge. Of sitivity are as follows. If a patient who has a history of a
S552 Gruchalla J ALLERGY CLIN IMMUNOL
FEBRUARY 2003

FIG 3. Structural similarities and differences of cephalosporins. From Baldo B. Penicillins and
cephalosporins as allergens—structural aspects of recognition and cross-reactions. Clin Exp Allergy
1999;29:745. Used with permission.

cephalosporin allergy requires another cephalosporin, one over a period of hours to days and is a process whereby a
of two approaches may be considered. (1) Perform a grad- drug-allergic person is converted from a drug-sensitive
ed challenge with a cephalosporin that does not share state to a state in which the drug is tolerated. Not only is
side-chain determinants with the original cephalosporin. the desensitized state antigen specific, it also is antigen
(2) Perform cephalosporin skin testing, although such dependent, requiring the continuous presence of antigen.
skin testing is not standardized and the negative predictive Penicillin desensitization is commonly performed, and
value is unknown.28 Fig 3 lists the various cephalosporins either the oral or the intravenous route may be used.
and their side-chain structural similarities. Once the starting dose has been determined,30 drug doses
In addition to IgE-mediated reactions, one cephalosporin, are doubled every 15 minutes. Vital signs, the physical
cefaclor, has been shown to cause a serum sickness–like examination, and peak flow values are monitored
syndrome. Because circulating immune complexes have not throughout the procedure. Although most of our experi-
been found, these reactions are not considered to represent ence with drug desensitization has been derived from
true serum sickness or immune complex reactions. penicillin, this principle has been successfully applied to
Although the mechanism of these reactions is not clearly numerous other drugs as well.31,32
known, Kearns et al29 have demonstrated that they may
result from hepatic biotransformation of the parent drug. Sulfonamides
Patients who have known or presumed IgE antibodies A sulfonamide is any compound that contains a sulfon-
to a β-lactam drug may undergo desensitization if that amide (SO2NH2) moiety. Sulfonamide antimicrobial
drug is required for treatment. Acute drug desensitization agents are different from other sulfonamide-containing
involves the administration of incremental doses of a drug medications, such as furosemide, thiazide diuretics, and
J ALLERGY CLIN IMMUNOL Gruchalla S553
VOLUME 111, NUMBER 2

celecoxib, by virtue of the presence of an aromatic amine

at the N4 position, which is necessary for antibacterial
activity. The presence of the aromatic amine allows sul-
fonamide agents to be placed into one of two groups, those
that contain that contain aromatic amines (ie, sulfonamide
antimicrobials) and those that do not. In addition to the
aromatic amine, sulfonamide antimicrobials also contain a
substituted ring at the N1 position. This group is not found
in the nonaromatic amine-containing sulfonamides.
Reactions to sulfonamide antimicrobial agents are
usually cutaneous in nature, and they occur in approxi-
mately 2% to 4% of healthy persons but in as many as
50% to 60% of patients with AIDS. The clinical reactions
exhibited are diverse and include anaphylaxis, urticaria,
erythroderma, fixed drug eruption, erythema multiforme,
macular exanthems, and even more severe cutaneous
reactions, such as Stevens-Johnson syndrome (SJS) and FIG 4. Sulfonamide metabolism. From Gruchalla R, Pesenko, RD,
toxic epidermal necrolysis (TEN). Do, TT, Skiest, DJ. Sulfonamide-induced reactions in patients with
Sulfonamides are metabolized in the liver by N-acety- AIDS—the role of covalent protein haptenation. J Allergy Clin
lation, yielding nontoxic metabolites, and by cytochrome Immunol 1998;101:372. Used with permission.
P450–catalyzed N-oxidation, yielding reactive hydroxy-
lamines33 that then oxidize to nitroso species.34 These dorzolamide, sumatriptan) that do not have this group. It
reactive nitroso metabolites are reduced by glutathione appears that cross-reactivity between sulfonamide antibi-
and then excreted. When the capacity for glutathione otics and these other sulfonamide drugs is theoretical
conjugation is exceeded, however, these metabolites may only. Despite this, however, the Physicians Desk Refer-
be directly cytotoxic, or through the formation of the N4- ence (2002) cautions against the use of some of these
sulfonamidoyl hapten they may haptenate protein carri- agents in patients with a history of a sulfonamide allergy.
ers, forming immunogenic complexes (Fig 4).33,35-37 Allen49 summarized the product labeling advice for var-
The fact that sulfonamide-induced reactions usually ious sulfonamide drugs with regard to their use in
occur after several days of treatment suggests that an patients with a history of sulfonamide allergy. This infor-
immune mechanism may be involved in at least some of mation is presented in Table II.
the reactions demonstrated. More direct evidence of this
involvement has come from studies that have demon- Other antimicrobial agents
strated the presence of both specific T lymphocytes38 and Unfortunately, little is known about the mechanisms
pro-inflammatory cytokines39 in the involved skin of responsible for most antimicrobial reactions. Although
patients with sulfonamide-induced cutaneous reactions. most of these reactions are not immunologically mediat-
In addition, IgE antibodies have been detected40,41 in ed, some of them most certainly are. Reactions consistent
patients reporting immediate-type hypersensitivity reac- with an IgE-mediated mechanism, such as urticaria,
tions to sulfamethoxazole. angioedema, and anaphylaxis, are typically easy to rec-
At this time we have few diagnostic tools for evaluat- ognize. Because there are no valid skin testing reagents
ing patients with sulfonamide-induced reactions. Despite available to confirm the presence of drug-specific IgE
the fact that little is known about the mechanisms respon- antibodies, however, we are currently unable to prove
sible for sulfonamide reactions in patients with AIDS, their existence and thus confirm the diagnosis. The more
several “desensitization” protocols have been empirical- nonspecific drug eruptions are even more problematic.
ly developed and used, quite successfully, for these Not only are the mechanisms responsible for many of
patients.42-48 The term “desensitization” is placed in quo- these reactions not known, but also the drug determinants
tation marks because in some instances the protocols that responsible for their elicitation have in most cases not
have been developed are more typical of a cautious grad- been identified. For these reasons, we have few in vivo or
ed challenge than of the rapid desensitization methods in vitro diagnostic tests for drug allergy.50
that are typically used in antibiotic desensitization. Sev- Despite our limited tools and knowledge, for those
eral of the published protocols are listed in Table I. antimicrobial reactions that are presumed to be IgE
Desensitization protocols should not be attempted in mediated, skin testing with the native drug may provide
patients who have had severe drug-induced cutaneous some useful information.51,52 Because the native drug
reactions, such as SJS or TEN.28 may not contain all the relevant antigenic determinants
Another important clinical problem with respect to that may elicit IgE-mediated reactions, a negative skin
sulfonamides is the degree of cross-reactivity between test result must be interpreted with caution. In contrast, a
the sulfonamide antibiotics that have an arylamine group positive skin test result may be extremely helpful. If the
at the N4 position of the benzene ring and other sulfon- dose used to elicit the positive wheal and flare response
amide derivatives (diuretics, sulfonylureas, celecoxib, does not elicit an irritant response in a healthy control
S554 Gruchalla J ALLERGY CLIN IMMUNOL
FEBRUARY 2003

TABLE I. Sulfonamide desensitization protocols for patients with AIDS

Desensitization
Study procedure Indication Outcome and comments

Absar et al,42 1994 10 d PCP prophylaxis, isosporiasis 23/27 successfully desensitized

Moreno et al,43 1995 TMP-SMX, 8 h; Central nervous system toxoplasmosis, 1/2 TMP-SMX and 7/11 sulfadiazine
sulfadiazine, nocardiosis, isosporiasis desensitizations successful; of 8 successfully
2.5 h desensitized, 6 were premedicated
Palusci et al,44 1996 4h PCP or PCP prophylaxis 5 children and infants with IgE-mediated reaction
to TMP-SMX evaluated; 3/5 successfully
desensitized
Caumes et al,45 1997 3d PCP prophylaxis 37/48 successfully desensitized; factors predictive
of failure were higher CD4+ cell percentage and
higher CD4+/CD8+ ratio
Rich et al,46 1997 8d PCP prophylaxis 18/22 successfully desensitized
Demoly et al,47 1998 6h PCP prophylaxis 44/44 successfully desensitized on day of
procedure; overall success rate at 1 mo 91%
(40/44)
Yoshizawa et al,48 2000 5d PCP prophylaxis 15/17 patients successfully desensitized

PCP, Pneumocystis carinii pneumonia; TMP-SMX, trimethoprim-sulfamethoxazole.

TABLE II. Use of sulfonamide-containing drugs in patients with a history of sulfonamide allergy
Product labeling
advice for use in
Drug sulfonamide allergy Comments

Acetazolamide (Diamox) None One case report of cross-reactivity in 1955

Amprenavir (Agenerase) Caution advised —
Brinzolamide (Azopt) None Labeling warns that because brinzolamide is a sulfonamide with systemic absorption,
serious allergic reactions are possible
Bumetanide (Bumex) None Labeling notes sulfonamide-allergic patients may be allergic to bumetanide
Captopril (Capoten) None Not a sulfonamide; no concerns about use
Celecoxib (Celebrex) Contraindicated —
Chlorpropamide None —
Chlorthalidone Contraindicated —
Chlorothiazide (Diuril) Contraindicated —
Dorzolamide (Trusopt) None Labeling warns that because dorzolamide is a sulfonamide with systemic absorption,
serious allergic reactions are possible
Furosemide (Lasix) Caution advised —
Glipizide (Glucotrol) None —
Glyburide (DiaBeta) None A few postmarketing reports of cross-reactivity
Hydrochlorothiazide Contraindicated —
(HydroDiuril)
Indapamide (Lozol) None —
Metolazone (Zaroxolyn) None Cross-sensitivity has not been reported but is a theoretical concern
Rofecoxib (Vioxx) None Not a sulfonamide; no concerns about use

From Allen J. Which medications to avoid in patients with sulfa allergy. Stockton (CA): Pharmacist’s Letter and Prescriber’s Letter; 2000. Used with permis-
sion.

subject, then it can be presumed that the positive skin test that consists of fever, rash, lymphadenopathy, and differ-
response is due to the presence of drug-specific IgE anti- ing degrees of internal organ involvement. Typically the
bodies. It is important to realize that when skin testing is symptoms begin several weeks into therapy. The rash,
performed with intravenous antibiotic formulations, which initially is often a benign morbilliform eruption,
some antibiotics produce marked irritant responses if may develop into frank exfoliative dermatitis. Because
they are not extensively diluted.52 this syndrome is oftentimes associated with eosinophilia
and systemic symptoms, it is also termed drug reaction
IMMUNE REACTIONS TO with eosinophilia and systemic symptoms (DRESS).
NONANTIMICROBIAL DRUGS Other drugs, in addition to the anticonvulsants, that have
Anticonvulsants been associated with this syndrome, include dapsone and
the sulfonamides, allopurinol, and minocycline.
Phenytoin, phenobarbital, and carbamazepine are Recently, data supporting the involvement of an
known to cause a severe hypersensitivity syndrome (HSS) immune mechanism in this syndrome have been accu-
J ALLERGY CLIN IMMUNOL Gruchalla S555
VOLUME 111, NUMBER 2

TABLE III. Criteria used in the classification of severe drug eruptions

SJS TEN HSS/DRESS

Mucous membrane >90% >90% <30%

Erosions Several sites Several sites Mouth and lips
Detachment of epidermis Yes, <10% BSA* Yes, ≥30% BSA* No
Hyperkeratosis, desquamation No No Usual
Neutropenia No 30% No
Eosinophilia No No 90%
Atypical lymphocytes No No 30%-40%
Respiratory tract Bronchial erosions, ARDS Bronchial erosions, ARDS Interstitial pneumonitis
Liver Hepatitis 10% Hepatitis 10% Hepatitis 60%
Heart No No Myocarditis
Lymph node enlargement No No Usual

From Bachot N, Roujeau J-C. Pathophysiology and treatment of severe drug eruptions. Curr Opin Allergy Clin Immunol 2001;1:293-8. Used with permission.
BSA, Body surface area; ARDS, adult respiratory distress syndrome.
*When detachment involves 10% to 29% of BSA, case is classified as SJS-TEN overlap.

mulating. The fact that an induction period appears to be Chemotherapy agents

required during the first course of treatment only but not
after subsequent exposures supports the involvement of Hypersensitivity reactions have been reported for all
an immune mechanism. chemotherapeutic agents in use. However, the most com-
The aromatic anticonvulsants are metabolized in part mon agents that elicit such reactions are the taxanes, plat-
by cytochrome P450 enzymes to reactive arene oxide inum compounds, asparaginases, and epipodophyllotox-
metabolites, and it is thought that these metabolites may ins. Reactions range from mild cutaneous eruptions to
be responsible for the pathologic process demonstrated respiratory arrest, cardiac collapse, and even death.
in drug-induced HSS.53 Defective detoxification may In the chemotherapy literature, the term hypersensitiv-
lead to excessive formation of arene oxides, which may ity reaction is often used. However, convincing support-
then contribute to the disease process either through ing mechanistic data often is not provided. Because
direct cellular necrosis or through the generation of an many of the associated reactions consist of flushing,
immune response after haptenation of protein carriers. heart rate and blood pressure changes, bronchospasm,
Treatment of HSS is similar to the treatment of other and pruritus, it is thought that these reactions may be
severe drug eruptions and is basically supportive. First and mediated by an immediate-type hypersensitivity mecha-
foremost, the causative agent should be rapidly withdrawn. nism. Because some of these agents or their excipients
For patients who have a marked exfoliative dermatitis, may elicit mast cell mediator release directly, however,
treatment often requires specialized care in intensive care nonimmunologic mechanisms probably account for at
units or in burn units. The main goals of care are similar to least some of these reactions.
those for patients with burns and other similar severe drug- The taxane paclitaxel is used to treat lung, breast, and
induced reactions (eg, SJS and TEN) and include volume gynecologic malignancies. Docetaxel, its sister agent, is
replacement, antibiotics, nutritional support, warming of a newer semisynthetic taxane. In the early phase I and
the environment, and extensive skin care. Corticosteroids phase II studies of the taxanes, as many as 42% of
may be administered if symptoms are severe. Although patients who received paclitaxel had a so-called “hyper-
there have been case reports of dramatic improvement of sensitivity” reaction, and 2% of these were severe.56
visceral manifestations of HSS or DRESS after administra- Most reactions occur during first exposure, and they con-
tion of moderate to high systemic doses of steroids, relaps- sist of dyspnea or bronchospasm, urticaria, flushing,
es have been seen to occur when steroids are tapered.54 hypotension, and angioedema. Although these symptoms
Table III compares and contrasts the characteristics of the are consistent with an IgE-mediated mechanism, there is
most common severe drug-induced cutaneous eruptions. no period of sensitization, so the more likely mechanism
Cross-reactivity rates among the aromatic anticonvul- is direct mast cell degranulation. In the case of paclitax-
sants may be as high as 75%. Patients who have had HSS el reactions, the induced reactions may be due to the
or DRESS should avoid all aromatic anticonvulsants. For drug’s excipient, Cremophor EL, a nonionic surfactant
these patients potential alternative drugs are valproic acid derived from castor oil that has been shown to lead to his-
(not in the acute phase because of the risk of hepatitis), tamine release and hypotension in dogs.57 Another tax-
gabapentin, vigabatrin, and benzodiazepines.55 There is no ane, docetaxel, is also associated with a high incidence of
evidence that cross-reactivity exists between lamotrigine reactions yet does not contain Cremophor EL, however,
and the aromatic anticonvulsants. However, this agent too so it is likely that the taxane moiety may be the etiologic
has been associated with HSS, as well as with TEN and agent responsible for these reactions.
SJS. Also, because there is a familial tendency toward Prophylaxis is routinely done before taxane adminis-
hypersensitivity to anticonvulsants, family counseling is tration because of its effectiveness in reducing the inci-
an important component of patient management.55 dence and severity of paclitaxel- and docetaxel-induced
S556 Gruchalla J ALLERGY CLIN IMMUNOL
FEBRUARY 2003

“hypersensitivity” reactions.56 Essentially all patients thematous plaques and skin necrosis), and heparin-asso-
who have “hypersensitivity” reactions to paclitaxel or ciated thrombocytopenia II. The mild thrombocytopenia
docetaxel are able to undergo successful retreatment that is noted with heparin therapy is probably not
again, as long as they are pretreated with antihistamines immunologically mediated and is reversible after the
and corticosteroids.58 For patients who have recurrent drug is discontinued.68 However, the more severe sudden
reactions despite premedication, a desensitization proto- and massive thrombocytopenia, thrombosis, and necrosis
col has been developed.58 seen after about 5 days of treatment are due to immune
The platinum compounds, both cisplatin and carbo- complexes composed of heparin-dependent IgG antibod-
platin, commonly induce hypersensitivity reactions. ies specific for platelet factor 4.
Unlike with the taxanes, however, a period of exposure Additional antithrombotic agents have been devel-
(several courses of drug administration) is required before oped, including the low–molecular weight heparins
a reaction, typically anaphylactic, is elicited. Thus, these (enoxaparin [Lovenox] and dalteparin [Fragmin]), dana-
reactions may be truly immunologic in nature. Skin test- paroid sodium (Orgaran), and the direct thrombin
ing has been used to identify patients at risk for develop- inhibitors argatroban and lepirudin (Refludan, a hirudin).
ment of reactions, and a negative test result has been Skin testing does not appear to be useful in evaluating
shown to have an extremely high (96%) negative predic- immediate-type hypersensitivity responses to heparin.
tive value.59 Desensitization protocols have been devel- However, positive delayed cutaneous skin test responses
oped for patients with positive skin test results, but these have been demonstrated in patients who have erythema-
have not been uniformly successful.58,60,61 tous plaques develop at heparin injection sites.69,70 In
Asparaginase, a bacterial (from Escherichia coli) addition, it has been shown that patients who have posi-
polypeptide protease that depletes tumor cells of tive delayed skin test responses also frequently have
asparagine, is used for treatment of acute lymphoblastic “heparin-induced” IgG antibodies. Because cross-reac-
leukemia. Approximately 25% to 35% of patients who tivity has been demonstrated between heparin, the
receive this agent have a hypersensitivity reaction that low–molecular weight heparins, and danaparoid sodium,
consists of anaphylactoid symptoms.62 Most patients it is recommended that the new alternative direct throm-
require repeated exposure before a reaction develops, and bin inhibitors be used in patients allergic to heparin.69,71
antiasparaginase antibodies have been found in some
persons.63 Intradermal testing is generally performed Insulin and protamine
before initial administration of asparaginase and before Reactions to insulin therapy have occurred since the
subsequent doses if a week or more has elapsed since the introduction of animal insulin in 1922. With the intro-
previous dose.64 duction of recombinant human insulin (Humalog), how-
Several additional asparaginase preparations (Erwinia ever, the incidence of insulin-induced allergic reactions
carotovora asparaginase and polyethylene glycol–modi- has decreased. Although the most common insulin-
fied E coli asparaginase) have been developed, and these induced reactions are local reactions at the injection site,
may be used to treat persons who have had asparaginase- systemic reactions may occur as well, although they are
induced reactions. Although a rapid desensitization pro- rare. IgE antibodies have been demonstrated in both
tocol for asparaginase hypersensitivity has been pub- types of reaction.72,73 Some local reactions may be con-
lished,65 this procedure is not used routinely. trolled by premedication with antihistamines or cortico-
The epipodophyllotoxins etoposide and teniposide are steroids.72 For patients who do not have a response with
antimitotic agents that are used in the treatment of testic- this approach, subcutaneous insulin infusion may be ben-
ular and ovarian germ cell tumors, small cell lung carci- eficial.72 For patients who have had systemic reactions,
nomas, non-Hodgkin lymphomas, and many other can- desensitization has been successful.73
cers. The incidence of etoposide- and teniposide-induced Reactions to protamine-containing insulins may be
reactions ranges from 6% to 41%, and there is a 0.7% to caused by the protamine component in the insulin prepa-
14% incidence of anaphylaxis.66,67 Reactions typically ration and not by the insulin itself. Protamine sulfate is a
consist of fever, chills, hypotension, dyspnea, and bron- low–molecular weight polycationic protein that is used to
chospasm. Because some of these reactions occur with reverse the anticoagulant properties of heparin, and it is
the first dose, it is not clear that they are immunological- also complexed to insulin (neutral protamine Hagedorn
ly mediated. Rather these agents, like paclitaxel, may insulin) to delay absorption. Dykewicz et al74 reported on
cause direct mast-cell mediator release. There are no two patients with diabetes who had anaphylaxis in
standard prophylaxis regimens, and fewer than half of response to neutral protamine Hagedorn human insulin
those who have a drug-induced reaction are able to toler- but who tolerated regular insulin. Both patients had neg-
ate the drug on readministration.64,66 ative skin test reactions to regular insulin but positive
reactions to neutral protamine Hagedorn insulin and to
Heparin protamine, indicating that protamine-specific IgE anti-
Heparin, a mucopolysaccharide with a molecular bodies were responsible for the anaphylactic reactions.
weight of 6000 d to 20,000 d, can elicit several types of Because it has been shown that patients with diabetes
immunologically mediated reactions, such as urticaria, who receive protamine-containing insulins are at a much
asthma, anaphylaxis,68 delayed cutaneous eruptions (ery- greater risk for development of anaphylaxis in response
J ALLERGY CLIN IMMUNOL Gruchalla S557
VOLUME 111, NUMBER 2

to intravenous protamine, these patients should be care- MANAGEMENT OF THE PATIENT WITH DRUG
fully evaluated before the performance of cardiopul- ALLERGY
monary bypass procedures that involve that agent.
Currently, there are few tools available to help us in
Biologic response modifiers the evaluation and management of patients who are seen
Biologic and other novel therapies that target specific with drug-induced reactions. These tools do not yet exist,
pathogenic processes are being developed at a rapid rate. because we have limited knowledge of the pathophysiol-
These agents provide new therapeutic options for patients ogy and factors that predispose toward the development
with chronic debilitating diseases. New immunomodula- of most of these reactions. Despite these inadequacies,
tors include interferons, TNF-α inhibitors, growth factors, patients must be managed.
monoclonal antibodies that inhibit T- and B-cell activa- The approach to the drug-allergic patient must be
tion, complement protein inhibitors, and many other methodic. First, a drug-disease connection must be estab-
agents. Although many of these agents have been shown lished. Once this has been done, the reaction type must
to have therapeutic benefits, adverse reactions, some of be determined if at all possible. For reactions in the type
which may be immunologic, are not uncommon. A category, dosage modification may be all that is neces-
General symptoms that have been seen include fever, sary before drug readministration. Toxicity, as well as
chills, and malaise, but more severe reactions may be drug-induced side effects and secondary effects, may
seen with specific agents. Among the more mild reac- resolve at lower drug dosages.
tions, injection site reactions are the most common. For type B drug intolerance-type reactions, the impli-
These have been demonstrated with both the interferons cated drug may be readministered if the previous reaction
and TNF-α inhibitors. In a recent retrospective study that was mild (eg, tinnitus with aspirin). For idiosyncratic
evaluated the incidence of injection site reactions in reactions, however, more caution is advised. For severe
patients receiving the TNF-α inhibitor etanercept, 20% or life-threatening reactions, the drug should not be read-
of treated patients reported reactions within the first 2 ministered. For less severe reactions, however, a provo-
months of therapy. These reactions were characterized by cational challenge may be considered.
an inflammatory infiltrate that consisted of predominant- For type B immunologically mediated reactions, the
ly activated mature cytotoxic T lymphocytes, but inter- management option depends on the mechanism responsible
estingly they waned overtime.75 for the reaction. If confirmatory tests are available and they
As more and more biologic response modifiers are have been validated, these tests should be used to determine
developed, further investigation will be necessary to bet- the allergic status of the patient (eg, testing for penicillin-
ter understand the nature of the adverse reactions that specific IgE antibodies with Pre-Pen and minor determinant
occur with their administration. Some of these most cer- mixture). When such tests are not available, however, and in
tainly will be immunologic in nature. Not only will it be most cases they are not, several approaches can be taken.
important to identify the antigenic stimulus responsible The simplest approach is to avoid the drug if an alternative
for these immunologically mediated reactions, but also agent is available. If an alternative drug does not exist, a
tests that predict their occurrence should be developed. graded challenge with the implicated agent can be done if
the previous reaction was not consistent with an IgE-medi-
ASPIRIN AND OTHER NONSTEROIDAL ANTI- ated reaction, and it was not life-threatening. If the previous
INFLAMMATORY DRUGS reaction was consistent with an IgE-mediated reaction,
Most reactions to aspirin and other nonsteroidal anti- however, then desensitization should be considered.
inflammatory drugs (NSAIDs) are nonimmunologically
mediated. However, some anaphylactoid reactions to CONCLUSIONS
aspirin and NSAIDs have features that are consistent
with IgE-mediated reactions (they occur after two or In addition to educating patients about their drug reac-
more exposures to aspirin, and they are specific to one tions, it is imperative that allergists educate the primary
NSAID only). To date, however, aspirin-specific IgE care physicians caring for these patients, as well. Most
antibodies have only rarely been demonstrated. For drug-induced reactions are nonallergic in nature. Both the
patients with aspirin-sensitive asthma, a recent study patient and the referring physician must be made aware of
demonstrated that one of the new cyclooxygenase 2 this fact. All too often the referring physician terrifies the
inhibitors, rofecoxib, is well tolerated.76 For an overview patient by stating that, in light of the patient’s multiple drug
of aspirin- and NSAID-induced reactions, see the 1998 “allergies,” few treatment options exist. It is no wonder
review by Stevenson and Simon.77 that, in the case of antibiotic reactions, the patient believes
that he or she is doomed should an infection arise. As stat-
NONIMMUNOLOGIC DRUG REACTIONS ed in this review, options, though limited, do exist. It must
be remembered, however, that both time and patience are
Reactions to angiotensin-converting enzyme inhibitors, needed to develop the optimal treatment approach.
local anesthetics, NSAIDs, opioid antagonists, and radio- REFERENCES
contrast media are not immunologically mediated in most 1. World Health Organization. International drug monitoring: the role of the
instances. They therefore are not discussed in this article. hospital. Geneva: The Organization; 1966.
S558 Gruchalla J ALLERGY CLIN IMMUNOL
FEBRUARY 2003

2. Bigby M, Jick S, Jick H, Arndt K. Drug-induced cutaneous reactions: a 28. Bernstein I, Gruchalla R, Lee R, Nicklas R, Dykewicz M. Disease man-
report from the Boston Collaborative Drug Surveillance Program on agement of drug hypersensitivity: a practice parameter. Joint Task Force
15,238 consecutive inpatients, 1975 to 1982. JAMA 1986;256:3358-63. on Practice Parameters, the American Academy of Allergy, Asthma and
3. Laxenaire M. Epidemiology of anesthetic anaphylactoid reactions: fourth Immunology, the American Academy of Allergy, Asthma and Immunol-
multicenter survey (July 1994-December 1996). Ann Fr Anesth Reanim ogy, and the Joint Council of Allergy, Asthma and Immunology. Ann
1999;18:796-809. Allergy Asthma Immunol 1999;83:665-700.
4. Lenler-Petersen P, Hansen D, Andersen M, Sorensen H, Bille H. Drug- 29. Kearns GI, Wheeler JG, Childress SH, Letzig LG. Serum sickness-like
induced fatal anaphylactic shock in Denmark 1968-1990. Ugeskr Laeger reactions to cefaclor: role of hepatic metabolism and individual sensitiv-
1996;158:3316-8. ity. J Pediatr 1994;125:805-11.
5. Pumphrey R. Lessons for management of anaphylaxis from a study of 30. Gruchalla R. Acute drug desensitization. Clin Exp Allergy 1998;28:63-4.
fatal reactions. Clin Exp Allergy 2000;30:1144-50. 31. Sullivan T. Drug allergy. In: Middleton E Jr, Reed CE, Ellis EF, Adkin-
6. Rawlins M, Thompson W. Mechanisms of adverse drug reactions. In: son NF Jr, Yunginger JW, Busse WW, editors. Allergy: principles and
Davies D, editor. Textbook of adverse drug reactions. New York: Oxford practice. St Louis: Mosby; 1993. p. 1726-46.
University Press; 1991. p. 18-45. 32. Patterson R, DeSwarte R, Greenberger P, Grammar L, Brown J, Choy A.
7. Borda I, Slone D, Jick H. Assessment of adverse reactions within a drug Drug allergy and protocols for management of drug allergies. Providence
surveillance program. JAMA 1968;205:645-7. (RI): Oceanside Publications; 1995. p. 1-27.
8. Coombs R, Gell PG. Classification of allergic reactions responsible for 33. Cribb A, Spielberg S. Hepatic microsomal metabolism of sulfamethoxa-
clinical hypersensitivity and disease. In: Gell P, Coombs RR, Lachman zole to the hydroxylamine. Drug Metab Dispos 1990;18:784-7.
PJ, editors. Clinical aspects of immunology. Oxford, UK: Blackwell Sci- 34. Naisbitt D, Hough S, Gill H, Pirmohamed M, Kitteringham N, Park B.
entific Publications; 1975. p. 761. Cellular disposition of sulphamethoxazole and its metabolites: implica-
9. Naisbitt D, Williams D, Pirmohamed M, Kitteringham N, Park K. Reac- tions for hypersensitivity. Br J Pharmacol 1999;126:1393-407.
tive metabolites and their role in drug reactions. Curr Opin Allergy 35. Gruchalla R, Pesenko, RD, Do, TT, Skiest, DJ. Sulfonamide-induced
Immunol 2001;1:317-25. reactions in patients with AIDS—the role of covalent protein haptena-
10. Idsoe O, Guthe T, Wilcox R, deWeck A. Nature and extent of penicillin tion. J Allergy Clin Immunol 1998;101:371-8.
side-reactions, with particular reference to fatalities from anaphylactic 36. Reilly T, Lash L, Doll M, Hein D, Woster P, Svensson C. A role for bioac-
shock. Bull WHO 1968;38:159-88. tivation and covalent binding within epidermal keratinocytes in sulfon-
11. International Rheumatic Fever Group. Allergic reactions to long-term amide-induced cutaneous drug reactions. J Invest Dermatol
benzathine penicillin prophylaxis for rheumatic fever. Lancet 2000;114:1164-73.
1991;337:1308-10. 37. Naisbitt D, Gordon S, Pirmohamed M, Burkhart C, Cribb A, Pichler W,
12. Joint Task Force on Practice Parameters, American Academy of Allergy, et al. Antigenicity and immunogenicity of sulphamethoxazole: demon-
Asthma and Immunology, American College of Allergy, Asthma and stration of metabolism-dependent haptenation and T cell proliferation in
Immunology, and the Joint Council of Allergy, Asthma and Immunology. vivo. Br J Pharmacol 2001;133:295-305.
The diagnosis and management of anaphylaxis. J Allergy Clin Immunol 38. Pichler W, Schnyder B, Zanni M, Hari Y, von Greyerz S. Role of T cells
1998;101(6 Pt 2):S465-528. in drug allergies. Allergy 1998;53:225-32.
13. Neugut A, Ghatak A, Miller R. Anaphylaxis in the United States: an 39. Yawalkar N, Shrikhande M, Hari Y, Nievergelt H, Braathen L, Pichler W.
investigation into its epidemiology. Arch Intern Med 2001;161:15-21. Evidence for a role for IL-5 and eotaxin in activating and recruiting
14. Levine B. Immunologic mechanisms of penicillin allergy. A haptenic eosinophils in drug-induced cutaneous eruptions. J Allergy Clin Immunol
model system for the study of allergic disease in man. N Engl J Med 2000;106:1171-6.
1966;275:1115-25. 40. Carrington D, Earl H, Sullivan T. Studies of human IgE to a sulfonamide
15. Levine B, Redmond A. Minor haptenic determinant-specific reagins of determinant. J Allergy Clin Immunol 1987;79:442-7.
penicillin hypersensitivity in man. Int Arch Allergy Appl Immunol 41. Gruchalla R, Sullivan T. Detection of human IgE to sulfamethoxazole by
1969;35:445-55. skin testing with sulfamethoxazole poly-L-tyrosine. J Allergy Clin
16. Adkinson N, Thompson W, Maddrey W, Lichtenstein L. Routine use of peni- 1991;88:784-92.
cillin skin testing on an inpatient service. N Engl J Med 1971;285:22-4. 42. Absar N, Daneshvar H, Beall G. Desensitization to trimethoprim/sul-
17. Green G, Rosenblum R, Sweet L. Evaluation of penicillin hypersensitiv- famethoxazole in HIV-infected patients. J Allergy Clin Immunol
ity: value of clinical history and skin testing with penicilloyl-polylysine 1994;93:1001-5.
and penicillin G. A cooperative prospective study of the Penicillin Study 43. Moreno J, Poblete R, Maggio C, Gagnon S, Fischl M. Rapid oral desen-
Group of the American Academy of Allergy. J Allergy Clin Immunol sitization for sulfonamides in patients with the acquired immunodefi-
1977;60:339-45. ciency syndrome. Ann Allergy Asthma Immunol 1995;74:140-6.
18. Sullivan T, Wedner H, Shatz G, Yecies L, Parker C. Skin testing to detect 44. Palusci V, Kaul A, Lawrence R, Haines K, Kwittken P. Rapid oral desen-
penicillin allergy. J Allergy Clin Immunol 1981;68:171-80. sitization to trimethoprim-sulfamethoxazole in infants and children.
19. Sogn D, Evans RE, Shepherd G, Casale T, Condemi J, Greenberger P, et Pediatr Infect Dis J 1996;15:456-60.
al. Results of the National Institute of Allergy and Infectious Diseases 45. Caumes E, Guermonprez G, Lecomte C, Katlama C, Bricaire F. Efficacy
Collaborative Clinical Trial to test the predictive value of skin testing and safety of desensitization with sulfamethoxazole and trimethoprim in
with major and minor penicillin derivatives in hospitalized adults. Arch 48 previously hypersensitive patients infected with human immunodefi-
Intern Med 1992;152:1025-32. ciency virus. Arch Dermatol 1997;133:465-9.
20. Weiss M, Adkinson, NF, Jr. Immediate hypersensitivity reactions to peni- 46. Rich J, Sullivan T, Greineder D, Kazanjian P. Trimethoprim/sul-
cillin and related antibiotics. Clin Allergy 1988;18:515-40. famethoxazole incremental dose regimen in human immunodeficiency
21. Parker C. Drug therapy. N Engl J Med 1975;292:511-4. virus-infected persons. Ann Allergy Asthma Immunol 1997;79:409-14.
22. Gorevic P, Levine B. Desensitization of anaphylactic hypersensitivity 47. Demoly P, Messaad D, Sahla H, Fabre J, Faucherre V, Andre P, et al. Six-
specific for the penicilloate minor determinant of penicillin and carben- hour trimethoprim-sulfamethoxazole-graded challenge in HIV-infected
cillin. J Allergy Clin Immunol 1981;68:267-72. patients. J Allergy Clin Immunol 1998;102:1033-6.
23. Baldo B. Penicillins and cephalosporins as allergens—structural aspects 48. Yoshizawa S, Yasuoka A, Kikuchi Y, Honda M, Gatanaga H, Tachikawa
of recognition and cross-reactions. Clin Exp Allergy 1999;29:744-9. N, et al. A 5-day course of oral desensitization to trimethoprim/sul-
24. Lin R. A perspective on penicillin allergy. Arch Intern Med 1992;152:930-7. famethoxazole (T/S) in patients with human immunodeficiency virus
25. Kelkar P, Li J. Cephalosporin allergy. N Engl J Med 2001;345:804-9. type-1 infection who were previously intolerant to T/S. Ann Allergy
26. Anne S, Reisman R. Risk of administering cephalosporin antibiotics to Asthma Immunol 2000;85:241-4.
patients with histories of penicillin allergy. Ann Allergy Asthma Immunol 49. Allen J. Which medications to avoid in patients with sulfa allergy. Stock-
1995;74:167-70. ton (CA): Pharmacist’s Letter and Prescriber’s Letter; 2000.
27. Novalbos A, Sastre J, Cuesta J, De Las Heras M, Lluch-Bernal M, 50. Weiss ME, Adkinson NF Jr. Diagnostic testing for drug hypersensitivity.
Bombin C, et al. Lack of allergic cross-reactivity to cephalosporins Immunol Allergy Clin North Am 1998;18:731-44.
among patients allergic to penicillins. Clin Exp Allergy 2001;31:438-43. 51. Brandt M, Gruchalla R, Sullivan T. Skin testing and in vitro testing to
J ALLERGY CLIN IMMUNOL Gruchalla S559
VOLUME 111, NUMBER 2

detect IgE to antimicrobial drugs [abstract]. J Allergy Clin Immunol 64. Zanotti K, Markman M. Prevention and management of antineoplastic-
1993;91:263. induced hypersensitivity reactions. Drug Saf 2001;24:767-79.
52. Empedrad R, Earl H, Gruchalla R. Determination of nonirritating skin 65. Bonno M, Kawasaki H, Hori H, Umemoto M, Komada Y, Sakurai M.
test concentrations of commonly used antimicrobial drugs [abstract]. J Rapid desensitization for L-asparaginase hypersensitivity. J Allergy Clin
Allergy Clin Immunol 2000;105:803. Immunol 1998;101:571-2.
53. Shear N, Spielberg SP. Anticonvulsant hypersensitivity syndrome: in 66. O’Dwyer P, King S, Fortner C, Leyland-Jones B. Hypersensitivity reac-
vitro assessment of risk. J Clin Invest 1988;82:1826-32. tions to teniposide (VM-26): an analysis. J Clin Oncol 1986;4:1262-9.
54. Bachot N, Roujeau JC. Physiopathology and treatment of severe drug 67. Kellie S, Crist W, Pui C, Crone M, Fairclough D, Rodman J, et al. Hyper-
eruptions. Curr Opin Allergy Clin Immunol 2001;1:293-8. sensitivity reactions to epiphodophyllotoxins in children with acute lym-
55. Knowles S, Shapiro L, Shear N. Anticonvulsant hypersensitivity syndrome: phoblastic leukemia. Cancer 1991;67:1070-5.
incidence, prevention and management. Drug Saf 1999;21:489-501. 68. MacLean J, Moscicki R, Block K. Adverse reactions to heparin. Ann
56. Eisenhauer EA, ten Bokkel Huinink WW, Swenerton KD, Gianni L, Allergy 1990;65:254-9.
Myles J, van der Burg ME, et al. European-Canadian randomized trial of 69. Harenberg J, Huhle G, Wang L, Hoffmann U, Bayerl C, Kerowgan M.
paclitaxel in relapsed ovarian cancer: high-dose versus low-dose and long Association of heparin-induced skin lesions, intracutaneous tests, and
versus short infusion. J Clin Oncol 1994;12:2654-66. heparin-induced IgG. Allergy 1999;54:473-7.
57. Lorenz W, Reimann H, Schmal A, Dormann P, Schwarz B, Neugebauer 70. Grassegger A, Fritsch P, Reider N. Delayed-type hypersensitivity and
E, et al. Histamine release in dogs by Cremophor EL and its derivatives: cross-reactivity to heparins and danaparoid: a prospective study. Derma-
oxethylated oleic acid is the most effective constituent. Agents Actions tol Surg 2001;27:47-53.
1977;7:63-7. 71. Aijaz A, Nelson J, Naseer N. Management of heparin allergy in pregnan-
58. Markman M, Kennedy A, Webster K, Kulp B, Peterson G, Belinson J. cy. Am J Hematol 2001;67:268-9.
Paclitaxel-associated hypersensitivity reactions: experience of the gyne- 72. Pratt E, Miles P, Kerr D. Localized insulin allergy treated with continu-
cologic oncology program of the Cleveland Clinic Cancer Center. J Clin ous subcutaneous insulin. Diabet Med 2001;18:515-6.
Oncol 2000;18:102-5. 73. Wessbecher R, Kiehn M, Stoffel E, Moll I. Management of insulin aller-
59. Zanotti K, Rybicki L, Kennedy A, Belinson J, Webster K, Kulp B, et al. gy. Allergy 2001;56:919-20.
Carboplatin skin testing: a skin-testing protocol for predicting hypersen- 74. Dykewicz M, Kim H, Orfan N, Yoo T, Lieberman P. Immunologic analy-
sitivity to carboplatin chemotherapy. J Clin Oncol 2001;19:3126-9. sis of anaphylaxis to protamine component in neutral protamine Hage-
60. Broome C, Schiff R, Friedman H. Successful desensitization to carbo- dorn human insulin. J Allergy Clin Immunol 1994;93:117-25.
platin in patients with systemic hypersensitivity reactions. Med Pediatr 75. Zeltser R, Valle L, Tanck C, Holyst M, Bitchlin C, Gaspari A. Clinical,
Oncol 1996;26:105-10. histological, and immunophenotypic characteristics of injection site reac-
61. Rose P, Fusco N, Fluellen L, Rodriguez M. Carboplatin hypersensitivity tions associated with etanercept. Arch Dermatol 2001;137:893-9.
reactions in patients with ovarian and peritoneal carcinoma. Int J Gynecol 76. Stevenson, DD, Simon, RA. Lack of cross-reactivity between rofecoxib
Cancer 1998;8:365-6. and aspirin in aspirin-sensitive patients with asthma. J Allergy Clin
62. Billet A, Carls A, Belber R, Sallan S. Allergic reactions to Erwinia Immunol 2001;108:47-51.
asparaginase in children with acute lymphoblastic leukemia who had pre- 77. Stevenson DD, Simon RA. Sensitivity to aspirin and nonsteroidal antiin-
vious allergic reactions to Escherichia coli asparaginase. Cancer flammatory drugs. In: Middleton E Jr, Reed CE, Ellis EF, Adkinson NF,
1992;70:201-6. Yunginger JW, Busse WW, editors. Allergy: principles and practice, vol
63. Woo M, Hak L, Storm M, Evans W, Sandlund J, Rivera G, et al. Anti- 2. 5th ed. St Louis: Mosby; 1998. p. 1225-34.
asparaginase antibodies following E. coli asparaginase therapy in pedi-
atric acute lymphoblastic leukemia. Leukemia 1998;12:1527-33.