Antibiotic use in children with penicillin allergy

April 2005

Penicillin is the most frequent cause of adverse allergic drug reactions and anaphylaxis. Approximately 10% of the population claims to have a history of penicillin allergy. However, several published studies have revealed that when these people are skin tested for penicillin allergy, less than 20% react positively (ie, they do not have penicillin-specific IgE antibodies).

While a good history and description of past potential allergic reactions to penicillin or other ß-lactam antibiotics may be helpful, history alone is insufficient in making an accurate diagnosis of penicillin allergy. Although pediatricians can choose from numerous other antibiotics when treating a child with a history of a penicillin allergy, the usefulness of penicillin for some important infectious diseases, and the chemical similarity of penicillin with other ß-lactam antibiotics (eg, amoxicillin, cephalosporins), causes the accuracy of a labeling of “penicillin allergy” to become important. This issue is also timely, as the commercial product for evaluating penicillin allergy by skin testing, PrePen, is not currently available. Production of this product was halted by the FDA over concerns of inappropriate manufacturing techniques. The American Academy of Allergy, Asthma and Immunology is coordinating efforts with potential manufacturers to make available a new commercial product for evaluating penicillin allergy (see Web site, in reference listing). However, when, and if, a new product suitable for adequately testing penicillin allergy becomes available is not currently known.

Drug allergy

The Gell and Coombs classification system is used to classify drug allergies, and Type I – IgE-mediated – is likely the allergic reaction most concerning to clinicians, as this allergic reaction can be fatal. Type I, anaphylactic, reactions result in urticaria, pruritis, laryngeal edema, bronchospasm, cardiovascular collapse, and potentially death. Fortunately, anaphylaxis from penicillin is rare – 0.01% - 0.05% incidence for each course of penicillin. Non-Type I reactions, characterized by symptoms such as maculopapular, polymorphous rash, arthralgia, or emesis, are more common. The challenge arises, however, when a child’s caregiver states that their child is allergic to penicillin, and their history is either vague or not convincing of a Type I, or IgE-mediated, reaction. Such a history may include the presence of a rash which developed several days after penicillin or Augmentin therapy, and when asked, the caregiver describes what seems to be a generalized, maculopapular rash, without hives or systemic symptoms indicative of anaphylaxis. Should the clinician avoid amoxicillin or Augmentin, or a cephalosporin, for the child’s next acute otitis media? While the use of an alternative antibiotic may be a relatively easy option, such antibiotics may be less effective, more costly, more likely to induce bacterial antimicrobial resistance, or more likely to cause significant adverse effects.

The best method for adequately evaluating the potential for an IgE-mediated reaction to penicillin is immediate hypersensitivity skin testing. When Pre-Pen was available, this was possible. Pre-Pen, however, only tested for allergic reaction to the penicillin “major determinants” (penicilloyl polylysine), the major metabolite of penicillin. Additional products of penicillin metabolism, the “minor determinants,” are more likely responsible for anaphylactic reactions, and up to 20% of those with documented anaphylaxis do not react to testing with the major determinants only. The minor determinants have never been commercially available, although additionally skin testing with a diluted penicillin G solution is usually an adequate substitute. Skin testing for penicillin allergy is predictive only for IgE-mediated hypersensitivity reactions, and does not predict the likelihood of the patient developing other significant, non-IgE immune reactions, such as glomerulonephritis, vasculitis, hemolytic anemia, erythema mulitforme, or Stevens-Johnson syndrome. Unfortunately, there is no reliable means to predict the occurrence of these adverse effects, and patients with a history of such a reaction with penicillin should not receive penicillin again.

Amoxicillin and antibiotics

Amoxicillin may be the most useful antibiotic to the practicing pediatrician because of its good activity toward Streptococcus pneumoniae, its effectiveness for a wide array of infectious diseases, its low cost and adverse effect profile, and its favorable pharmaceutical profile (good taste of oral suspension and numerous dosage forms). Amoxicillin or ampicillin may result in a unique rash which is not indicative of an IgE-mediated penicillin allergy and which does not require skin testing. This maculopapular rash occurs in approximately 5%-10% of patients given amoxicillin or ampicillin. Upon readministration of amoxicillin or ampicillin, this rash may reoccur, although systemic symptoms rarely occur. For the child who suffers a reaction more indicative of an IgE-mediated reaction, such as an urticarial rash, penicillin skin testing should be utilized.

Many commercially available antibiotics and classes of antibiotics share a common chemical structure, the ß-lactam ring, and these antibiotics are referred to as ß-lactam antibiotics. Penicillin (and the penicillins {eg, piperacillin}), amoxicillin, the cephalosporin class and carbapenem (eg, imipenem, meropenem) antibiotics are all classified as ß-lactam antibiotics, and all share the potential for cross-reactivity in a patient with a history of an IgE-mediated hypersensitivity reaction to penicillin. The cephalosporins are likely the most utilized class of non-penicillin ß-lactam antibiotics prescribed by office-based pediatricians. These cephalosporins include cephalexin (Keflex, Advancis), cefuroxime (Ceftin, GlaxoSmithKline), cefpodoxime (Vantin), cefdinir (Omnicef, Abbott), and ceftriaxone (Rocephin, Roche). The risk of a child, who has reacted positively to penicillin skin testing, suffering an allergic reaction to a cephalosporin antibiotic is quite low – <2%. It is believed that the 1st-generation cephalosporins (eg, cephalexin) are more likely than 2nd-generation (eg, cefuroxime) or 3rd-generation (cefpodoxime) cephalosporins to be cross-reactive. This is due to the chemical similarity of side-chains of the ß-lactam ring between penicillin and 1st-generation cephalosporins. If a child with a history suggestive of an IgE-mediated allergic reaction to penicillin (eg, urticarial rash) requires treatment with a cephalosporin, skin testing for penicillin therapy should be done. If skin testing for penicillin allergy is positive, therapeutic options include avoiding use of cephalosporin antibiotics, or graded challenges or desensitization with a cephalosporin (the latter are best preformed by a clinician trained in allergy-immunology). While the risk of cross-reactivity between the penicillin and cephalosporin class is relatively low, anaphylactic deaths have been documented in penicillin skin test-positive patients who have received cephalosporins. Children with a history of an IgE-mediated allergic reaction to a cephalosporin who require penicillin or amoxicillin should be skin tested for penicillin allergy. In the child with a history of penicillin allergy who has recently safely been treated with a cephalosporin, cephalosporin skin testing is not necessary. Commercially available cephalosporin skin testing reagents are not available, and although protocols for cephalosporin skin testing with diluted cephalosporin antibiotics are available, their use has not been standardized and the negative predictive value is not known.

Studies

Perhaps one of the most intriguing and practical questions confronting clinicians treating a child with a vague history of an allergy to penicillin or amoxicillin (eg, a mild, maculopapular rash) is, “Can I use amoxicillin in this child who now has an acute otitis media?” Several studies have evaluated this premise.

Mendelson evaluated 240 children younger than 20 who had a history of allergy to penicillin, amoxicillin, or a cephalosporin antibiotic. All children were skin tested with penicillin major and minor determinants, and then challenged with a 10-day oral penicillin course. Four weeks later, they were retested to determine if they had been resensitized. Of the 240 subjects tested, 21 (8.75%) had one or more positive skin tests. Of 40 subjects with a history of allergy not suggestive of an IgE-mediated reaction (maculopapular eruption, joint pains, fever), five (12.5%) reacted positively to skin testing. Of 24 subjects with a history of a non-IgE-mediated reaction to amoxicillin, two (8%) reacted positively to skin testing.

Kalogeromitros studied 638 patients (mostly adults) in Greece who had an indication for penicillin therapy, and they were evaluated for their history of previous allergic reactions and their true allergic status. Patients were skin tested with both penicillin major and minor determinants. Of 96 patients with a vague history of a Type I penicillin or ß-lactam allergy (non-anaphylactic or non-urticarial reactions), 14 (14.6%) reacted positively to skin testing. Of the 18 patients with a convincing history of a previous Type I allergic reaction, 13 (72.2%) reacted positively to skin testing, and five of 18 (27.8%) reacted negatively.

In a similar study, Pichichero evaluated 247 children (6 month–18 years of age) with a history of an allergic reaction to penicillin, amoxicillin, or a cephalosporin. These children were skin tested with penicillin (major and minor determinants), ampicillin, cefazolin, cefuroxime, or ceftriaxone. It is important to consider, however, that skin testing for IgE-mediated reactions to ampicillin or cephalosporins has not been standardized. Oral challenge, repeat testing, and follow-up were also evaluated. Of the 247 children tested, 84 (34%) reacted positively to skin testing or oral challenge. For specific antibiotics, 32%, 34%, and 50% of penicillin, amoxicillin, and cephalosporin reactions, respectively, were positive and thus indicative of an IgE-mediated mechanism. Thus, of the 247 evaluated children with a history of a suspected allergic reaction, 163 (66%) did not react to skin testing, and thus could received future courses of the antibiotic evaluated. Of particular interest is the finding that of 68 children with a history of a pruritic, polymorphous rash (maculopapular, erythematous) from either penicillin, amoxicillin, or a cephalosporin, 15 (22%) reacted positively to skin testing. Of 54 children with a previous allergic reaction to amoxicillin and who reacted positively to skin testing with ampicillin, 15 (28%) had a previous history of a non-IgE-mediated reaction. Similarly, 19% of children with a previous history of a non-IgE-mediated reaction to penicillin reacted positively to skin testing.

Solensky completed a literature review of published studies evaluating children and adults with a history of penicillin allergy who were subsequently skin tested. Thirty studies (published between 1966-1998) were assessed. Allergic histories were classified as “convincing” (more indicative of an IgE-mediated mechanism – anaphylaxis, urticaria, angioedema, bronchospasm, or pruritic rash) or “vague” (maculopapular rash, gastrointestinal symptoms, or unknown reaction). Of a total of 1,063 people who reacted positively to skin testing, 347 (33%) had a “vague” history of a previous allergic reaction to penicillin. The skin testing procedures in these studies, however, were not similar, and thus these results maybe somewhat inaccurate.

Conclusions

Published studies evaluating the accuracy of a patient history of penicillin allergy indicate that the majority (80% or more) does not possess an IgE-mediated true allergic response to penicillin or perhaps other ß-lactam antibiotics. While it may be tempting to rule-out an allergy to penicillin by history alone, especially when the history is vague or not convincing, several studies have shown that this is not always reliable. In the studies described above, 12.5%-33% of subjects with a vague history of penicillin allergy reacted positively to skin testing. On the contrary, one study demonstrated that more than 25% of subjects with a “convincing history” of penicillin allergy reacted negatively to skin testing (the number of subjects tested were small in this study, however). Thus, the most reliable means to assess for IgE-mediated allergy to penicillin, and possibly to other ß-lactam antibiotics, is skin testing. When properly completed by a clinician trained in skin testing techniques, a negative skin test reaction to penicillin suggests that 97% to 99% of patients can safely be given penicillin. Unfortunately, the current status of the commercial availability of a penicillin skin-testing kit is not known. It is hoped that an appropriate manufacturer for a product containing both the major and minor penicillin determinants will soon be located. Do all patients with a history of “penicillin allergy” require referral for skin testing? Perhaps not. Better candidates for referral may include children with infectious diseases for which penicillin is the superior, or only, effective treatment. Younger children or infants with risk factors for frequent infection (eg, frequent bouts of AOM) may also be good candidates. It is best that children be skin tested for penicillin allergy prior to pending use for a specific infectious disease. While other classes of antibiotics (eg, macrolide or azalide antibiotics) may frequently be used in such children, their antimicrobial activity toward Streptococcus pneumoniae or other important pathogens is likely to be less than amoxicillin. Cephalosporin antibiotics may often be substituted in such children as well, even though they are also classified as ß-lactam antibiotics. The risk of an adverse allergic reaction to a cephalosporin in a child with suspected penicillin allergy is small, although it can be significant. Skin testing for penicillin allergy remains the most reliable means to accurately evaluate such children.

For more information:

• American Academy of Allergy, Asthma, and Immunology. PrePen initiative make progress. www.aaaai.org/members/academynews/2004/11/prepen.stm.
• Bernstein IL. Disease management of drug hypersensitivity, a practice parameter. Ann Allergy, Asthma, Immunol. 1999;83:665-700.
• Mendelson LM. Routine elective penicillin allergy skin testing in children and adolescents: study of sensitization. J Allergy and Clin Immunol. 1984;73:76-81.
• Kalogeromitros D. Penicillin hypersensitivity: value of clinical history and skin testing in daily practice. Allergy Asthma Proc. 2004;25:157-60.
• Pichichero ME. Diagnosis of penicillin, amoxicillin, and cephalosporin allergy: reliability of examination assessed by skin testing and oral challenge. J Pediatr. 1998;132:137-43.
• Solensky R. Penicillin allergy: prevalence of vague history in skin test-positive patients. Ann Allergy Asthma Immunol. 2000;85:195-9

• Edward Bell, PharmD, BCPS, is an associate professor of pharmacy practice at Drake University College of Pharmacy, and a clinical specialist at Blank Children’s Hospital, Des Moines, Iowa.