What antibiotics should be avoided by a patient who developed anaphylaxis when treated with Kåvepen

Fråga: What antibiotics should be avoided by a patient who developed anaphylaxis when treated with Kåvepenin (phenoxymethylpenicillin, PcV), and where serum antibodies against PcV and PcG (bensylpenicillin) have been detected?

Sammanfattning: The general recommendations set by the Drug Information Committee in 1988 are still valid. In a case of previous anaphylaxis due to penicillin allergy, no other penicillins should be tried, unless there are very strong indications and that high security arrangements are considered for test dosing. The same is true for cephalosporins, even though the risk of cross-reactivity is likely to be low.

Svar: In 1988, The Drug Information Committee of the National Board of Wealth and Welfare in Sweden, published a summary of a Workshop on Penicillin Allergy (1). Some of the main conclusions were that patients with a history of an anaphylactic reaction to penicillin, should never again be exposed to any type of penicillin (including semisynthetic penicillins), and a warning label was required in their medical records. Semisynthetic penicillins on the Swedish market include ampicillin, pivampicillin, bakampicillin, amoxicillin, isoxazolylpenicillins (cloxa-, dicloxa-, and flucloxacillin), mecillinam, pivmecillinam and piperacillin.

In the Workshop recommendations, there is however no specific mention of piperacillin and mecilllinam (1). Documentation of cross-reactivity with penicillin is very limited (mecillinam) or not available (piperacillin). One report on mecillinam-allergic workers in a mecillinam/amoxicillin-production plant suggested cross-reactivity to penicillin to be less than 20 percent (2). No conclusive cases of severe hypersensitivity reactions triggered by mecillinam were found in the Swedish Drug Information System, Swedis (3). In contrast, several reports have described severe hypersensitivity reactions strongly associated with piperacillin (4, 5), including 10 cases of anaphylaxis and two severe skin reactions listed in Swedis (3). Both mecillinam and piperacillin show a great structural similarity with amoxicillin, and all three agents are contra-indicated in Pc-allergy in the Swedish Catalogue of Approved Medical Products (FASS) (6).

Among other types of betalactam antibiotics, carbapenems (for example imipenem on the Swedish market), are not recommended to patients with a history of Pc anaphylaxis because of a high extent of cross reactivity with penicillin (around 50 percent) (1, 7). In contrast, the monobactam antibiotic aztreonam is poorly immunogenic and considered as a safe alternative in Pc allergic patients (1, 8), even though one case of cross-reactivity has been described (9).

Concerning cephalosporins, the group of betalactams with the highest number of different products, the general recommendation to a Pc allergic patient has been to avoid these because of cross-reactivity (1), initially estimated to around 10 percent (10). However, this has been a matter of some controversy (10-12), and cross-reactivity against more modern, third-generation cephalosporins might be as low as 1-3 percent (10, 11). The immunogenic epitopes on cephalosporins are predominantly localised in the bulky side chains, and not the betalactam core of the molecule, thereby reducing the risk of cross-reactivity with PcV or PcG. However, if the cephalosporin side chains are similar to a particular semisynthetic penicillin, a high risk of cross reactivity exists in a patient with an allergy against the penicillin. This has been demonstrated in patients with amoxicillin allergy, who also react to cephadroxil, a molecule with an identical side chain (13). Taken together when a strong indication for the use of cephalosporins, eg cystic fibrosis (10, 11) exists, a suitable cephalosporin might be tested under strict safety measurements to meet possible anaphylactic reactions, in cases of Pc allergy.

In sensitised patients, allergic reactions against penicillins appear to decline over time, and less than 20 percent of patients with a previous history of an allergic reaction show a positive skin reaction 10 years later. Skin reactions (prick testing) appear to be the most sensitive and reliable test of existing Pc allergy, where a positive reaction means a high risk for anaphylaxis if the patient is challenged with Pc, and the opposite is true for negative reactions (8, 10 and references therein). However, in some hypersensitive individuals even the testing itself might trigger a generalised allergic reaction. Concerning analysis of anti-Pc-IgE-antibodies by RAST (standard serum testing in Sweden), it should be mentioned that antibody titers might decrease to undetectable levels as soon as six months after the allergic reaction, even if the risk for serious allergy remains. Moreover, RAST analysis could be false negative if high levels of anti-Pc-IgG are expressed in serum, and false positive in the presence of high levels of non-specific IgE (1, 10).