Background: beta-Lactams are the drugs most frequently involved in hypersensitivity reactions mediated by immunoglobulin (Ig) E.
Objective: To evaluate a population of patients with suspected B-lactam allergy using a validated algorithm that includes specific IgE antibodies, skin testing, and/or a drug provocation test.
Methods: A total of 1032 patients with symptoms compatible with B-lactam allergy were evaluated by means of their clinical history, specific immunoglobulin (Ig) E antibody determinations (benzylpenicillin, ampicillin, and amoxicillin), and skin tests with major determinants (penicilloyl-polylysine) and minor determinants (minor determinant mixture) of benzylpenicillin, penicillin G, ampicillin, and amoxicillin-clavulanic acid. Patients whose skin test results were negative were challenged with amoxicillin-clavulanic acid. Only immediate hypersensitivity reactions were evaluated. All patients with negative study results and for whom a reaction occurred more than 1 year before were retested using the same protocol.
Results: A total of 170 patients (16.4%) were finally confirmed as having immediate allergic reactions to beta-lactams (62.3% by skin testing, 16.5% by specific IgE, and 21.2% by drug provocation test). The mean age of these patients was 43.3 years, and the drug most frequently involved in the reaction was amoxicillin (41.1%), followed by the combination amoxicillin-clavulanic acid (36.4%). In the remaining 22.5%, different beta-lactams were involved or the culprit drug was not known. Only mild reactions were observed after the drug provocation test. A retest was required in 23% of patients in order to confirm their hypersensitivity.
Conclusions: These results indicate that a diagnostic protocol based on the combination of skin testing and in vitro determination of specific IgE antibodies plus, if required, drug provocation testing is an appropriate procedure for evaluating immediate hypersensitivity reactions to beta-lactams. Because the sensitivity of skin testing and in vitro IgE assays is not optimal and a considerable proportion of patients are tolerant, drug provocation tests are necessary to achieve the diagnosis or confirm tolerance. A large percentage of patients (23%) were diagnosed using retest.