MANAGEMENT OF THE PATIENT WITH SUSPECTED b-LACTAM ALLERGY
Despite more than 40 years of clinical use, penicillin remains an important agent for many types of infection. Currently, parenteral b-lactam antibiotics (penicillins, cephalosporins, cephamycins, carbapenems, and monobactams) comprise the largest group of parenteral antimicrobials administered to hospitalized patients. Similarly, oral b-lactams are frequently used for outpatients with a wide variety of presumably treatable infections, and they have more recently been extensively employed as step-down therapy for patients being switched from parenteral antibiotics in preparation for hospital discharge. They remain popular because of reasonable safety, dosing flexibility, ease of administration (few drug-drug interactions), and an excellent spectrum of activity with proven efficacy in many clinical situations.
Allergy is a major problem associated with the use of b-lactams, and these agents may be the most common cause of drug allergy. Although rates of immediate allergy among penicillins have historically been felt to be similar, a recent investigation suggests that allergy to ampicillin may be the most common. Allergy to other cephalosporins, carbapenems, and carbacephems has been less well studied. Up to 20% of hospitalized patients claim to be penicillin-allergic. This rate is much higher than what has been demonstrated in clinical studies and is substantially higher than the rates associated with most other antimicrobials, with the possible exception of the sulfonamides. Because of similarities in structure among various types of b-lactams, it is not surprising that cross-allergenicity has been described. This issue was initially complicated because early cephalosporins were probably contaminated by penicillin. Some data have suggested that allergy may be targeted to specific side chains, and thus allergy to amoxicillin but not to penicillin has been reported. However, a recent investigation assess-ing the risk for cross-allergenicity between amoxicillin and cefadroxil (which share side chains) demonstrated that almost 90% of persons allergic to the former tolerated the latter.
Rates of allergy to cephalosporins in patients with and without penicillin allergy are thought to be 5% to 16% and 1% to 3%, respectively. Regarding penicillin, early investigations suggested that 0.7% to 10% of patients receiving this agent had allergic reactions, with a fatality rate below 0.002%. The first reported case of anaphylaxis was in 1946 and the first reported death in 1949. Administration of parenteral penicillin is more likely to be associated with fatalities, probably because of the higher doses employed. However, more recent data demonstrate no difference in rate of allergy between high oral doses and similar doses administered parenterally. Additionally, the risk for penicillin allergy is higher in persons with a history of previous reaction. Allergy may nonetheless develop in persons without documented prior b-lactam exposure, for reasons that include prior exposure through foods and other environmental factors. Risk is lower in children and the elderly, with fatal outcomes more common in the latter group. A history of atopy is not an independent risk factor for penicillin allergy. Signs of allergy may persist for weeks after discontinuation of the offending agent.
Anaphylaxis to penicillin occurs in 0.01% of those who receive this agent, with a mortality rate of 9%. Alternatively, urticaria can be noted in up to 5% of patients receiving penicillin, even in the absence of known allergy. Neither penicillins nor cephalosporins cause significant cross-allergenicity with monobactams (i.e., aztreonam). This fact provides de facto data that the allergenic moiety does not reside within the b-lactam ring itself.
Many patients carry the label “penicillin-allergic” for obscure reasons. The problem is confounded because many adverse reactions to these compounds are not associated with true allergic mechanisms (although they may be reproducible), and such reactions are associated with little further risk on repeated administration. Examples include the maculopapular rash that occurs many days into the course of b-lactam administration, or the severe maculopapular rash that develops in patients with mononucleosis on administration of ampicillin. Table 56-1 summarizes known allergic mechanisms and their associated clinical manifestations. Problems associated with overemphasis of allergy include the use of secondary antibiotics that may be more expensive, toxic, and less reliable.
Table 56-1. Mechanisms of antimicrobial allergy
The role of skin testing for b-lactam allergy is controversial, with the problem complicated by a lack of available antigens for the cephalosporins and the commercial availability of only the major determinant for penicillin. Skin testing is indicated for patients who may need penicillin yet whose history of immediate reaction is uncertain. Only about 10% of patients labeled as penicillin-allergic will react to skin tests.
After administration, penicillin is broken down into the penicilloyl group (major determinant) and other derivatives (minor determinants). Most immediate reactions are the result of IgE antibodies reacting against the latter. Alternatively, later urticarial reactions usually are associated with IgE antibody activity against the major determinant. In patients without a reaction to major or minor determinants, IgE-mediated penicillin allergies are unlikely to develop. However, these tests have no value in determining the likelihood of other reactions.
The clinician who is considering the use of a b-lactam antimicrobial in a patient with presumed allergy needs to determine (a) the likelihood of true allergy and the conditions surrounding the initial adverse reaction, (b) the possibility of using an alternative b-lactam, (c) the availability of suitable alternative agents if a b-lactam is contraindicated, and (d) the role of desensitization if a b-lactam must be employed despite immediate allergy. It is incumbent on the clinician to assess the true history of presumed b-lactam allergy and to document the reaction and its relationship to the product.
True allergy must be differentiated from adverse reaction and toxicity. Stories such as, “My mother was allergic, and she told me that I could be allergic,” and “My stomach became upset when I took penicillin,” and “I got diarrhea when I took ampicillin,” are examples of clinical situations where there is not a need for alternative agents. Indeed, the true clinical significance of many of the untoward reactions associated with unknown or idiopathic mechanisms is undetermined, and many such patients can receive the offending agent. The problem of defining true allergy is further confounded by the fact that selected adverse reactions may occur simultaneously with the administration of a b-lactam but are not related to the product itself. Examples include procaine or vasovagal reaction, pain at the site of injection, and the Jarisch-Herxheimer reaction. Procaine-induced adverse reactions rapidly follow accidental IV injection of procaine penicillin and can include agitation, hallucinations, seizures, and hypertension. Problems tend to subside within minutes. The Jarisch-Herxheimer reaction is most often noted following the administration of penicillin for spirochetal disease (most commonly syphilis) and is related to the release of endotoxin from dying organisms. Manifestations include fever, rigors, and other constitutional symptoms and generally begin several hours after the first dose of drug.
The majority of patients with reactions to b-lactams have late maculopapular rashes associated with unknown mechanisms. These are unlikely to cause severe reactions on readministration of a b-lactam product, and therefore an alternative b-lactam can be safely utilized. As an example, the patient in whom a maculopapular reaction developed on day 5 of parenteral nafcillin can be safely treated with cefazolin, and the outpatient with a similar history of allergy to dicloxacillin can be managed with an oral cephalosporin. Although data are scarcer, patients with most other forms of non–IgE-mediated reactions can probably also be managed with alternative b-lactams. Examples include serum sickness, interstitial nephritis, and drug fever.
Patients with a history of an immediate reaction to a b-lactam antimicrobial should, if possible, not be treated with any of these products. Fortunately, the current antimicrobial armamentarium allows alternative strategies for most infections. Table 56-2 summarizes alternative agents for selected infections when a b-lactam is contraindicated. The route, dose, and duration of therapy depend on the site and severity of the infection.
Table 56-2. Alternative antimicrobials when b-lactams are contraindicated
Rarely, a patient who has had an immediate reaction to a b-lactam requires treatment with the offending agent, and in these instances desensitization should be performed. Examples include gonococcal endocarditis, enterococcal endocarditis (vancomycin-resistant organism), and selected forms of syphilis and Lyme borreliosis during pregnancy. Desensitization appears to work by binding IgE antibodies so that interaction with mast cells is prevented. Many authorities recommend that desensitization be performed following informed consent and in a carefully controlled environment, such as an ICU. Desensitization consists of administering progressively larger doses of penicillin (or another agent). There is no need for antihistamine-corticosteroid premedication. Table 56-3 provides an oral desensitization regimen that can be employed so long as the patient can reliably absorb oral medications. The typical interval between doses is at least 20 minutes; intradermal injections should be carefully observed for wheal/flare reactions. Once a final dose has been reached, it must be administered according to standard dosing regimens without doses being missed. The best data demonstrate that approximately one third of persons undergoing desensitization have mild reactions, either during desensitization or during active treatment. The risk for IgE-mediated reactions during desensitization is low. Mild allergic reactions seen during desensitization require that the offending dose be repeated until tolerated. More severe reactions require active management, and the dose of the offending product should be decreased by at least 90%.
Table 56-3. Oral desensitization of b-lactam-allergic patients
In summary, the initial management of a patient with a suspected allergy to b-lactam antimicrobials requires a careful evaluation of the type of reaction and its immunologic basis. Many patients with late (generally on a poorly defined immunologic basis) reactions can be successfully treated with alternative b-lactams or even with the offending antibiotic. Those with immediate or rapidly progressive reactions can generally be managed with alternative antimicrobials. For patients with defined IgE-mediated reactions who must be treated with the offending agent, desensitization is usually successful. (R.B.B.)
Anne S, Reisman RE. Risk of administering cephalosporin antibiotics to patients with histories of penicillin allergy. Ann Allergy 1995;74:167–170.
Available data were reviewed to assess the risks of administering cephalosporins to patients allergic to penicillin. Few data demonstrated a likelihood of cross-allergenicity. The authors are not convinced that skin testing provides any useful data, and conclude that penicillin allergy does not predict allergy to cephalosporins.
Chisholm CA, et al. Penicillin desensitization in the treatment of syphilis during pregnancy. Am J Perinatol 1997;14:553–554.
The authors conducted a retrospective study of 16 patients who required penicillin desensitization for syphilis during pregnancy. The oral regimen was equally effective, and the cost was approximately 40% of the cost of parenteral desensitization. The authors conclude that oral desensitization is as effective as parenteral desensitization and less expensive.
Pichichero ME, Pichichero DM. Diagnosis of penicillin, amoxicillin and cephalospo-rin allergy: reliability of examination assessed by skin testing and oral challenge. J Pediatr 1998;132:137–143.
Approximately 250 children or adolescents with clinically documented allergies to b-lactams were studied prospectively by skin testing and oral rechallenge. Of these, only one third demonstrated an IgE-mediated reaction on skin testing or oral challenge. None of the patients with negative test results demonstrated a significant reaction on oral rechallenge. The authors conclude that true allergy is overdiagnosed based on prior history, and that skin testing is indicated to rule out allergy. With negative test results, oral rechallenge is safe.
Sastre J, et al. Clinical cross-reactivity between amoxicillin and cefadroxil in patients allergic to amoxicillin and with good tolerance of penicillin. Allergy 1996;51:383–386.
The authors investigated 16 patients with allergy to amoxicillin but without clinical intolerance to penicillin to assess the role of the side chain on the former as a moiety for allergy. Patients with amoxicillin allergy were given cefadroxil, which contains the same side chain. Only 12% demonstrated immediate reactions, which shows that the side chain is not the target of hypersensitivity.
Saxon A, et al. Immediate hypersensitivity reactions to b-lactam antibiotics. Ann Intern Med 1987;107:204–215.
This conference discusses the types of allergic reactions and provides an excellent overview of the chemistry of the b-lactams and the relationship to allergy and skin testing. It is the best recent review of the topic.
Weiss ME, Adkinson NF. b-Lactam allergy. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and practice of infectious diseases, 4th ed. New York: Churchill Livingstone, 1995.
This excellent summary of issues related to b-lactam antimicrobial allergy provides historic material, a sound rationale for penicillin testing, and several tables for penicillin desensitization, including one for oral desensitization.