[Drugs associated with acute generalized exanthematic pustulosis]

Acute generalized exanthematous pustulosis is a severe eruption which is usually drug related. If the causative drug is discontinued, acute generalized exanthematous pustulosis resolves spontaneously in ten days. The aim of this study was to compare drugs suspected of causing acute generalized exanthematous pustulosis reported to French Pharmacovigilance centres and those reported in the literature.

All cases of "pustular eruption" qualified as "serious" reported to the French Pharmacovigilance Centers between January 1985 and December 2001 were analyzed. Cases for which the diagnosis of acute generalized exanthematous pustulosis was not clearly identified were reviewed by a dermatologist. The relationship between acute generalized exanthematous pustulosis and drug exposure was re-examined by one of us. An exhaustive review of the literature was also performed.

Review of the data base revealed 207 cases of serious acute generalized exanthematous pustulosis leading to death in 4 cases (2%). Of these cases of acute generalized exanthematous pustulosis, only one drug was suspected in 107 cases (51.6%). The main drugs involved were: pristinamycin (18 cases), amoxicillin (+/- clavulanic acid) (16 cases), hydroxychloroquine (8 cases) and a combination of spiramycin + metronidazole (5 cases).

The most frequent causal drugs in our study and in the literature are: amoxicillin +/- clavulanic acid, pristinamycin, hydroxychloroquine, ampicillin, diltiazem, co-trimoxazole, terbinafine, carbamazepine and spiramycin +/- metronidazole. Only pristinamycin and diltiazem have information in their summary of product characteristics regarding the risk of acute generalized exanthematous pustulosis. Because it is essential to discontinue the causative drug as soon as possible if a pustular eruption occurs, physicians must be informed of the risk, which should be added to the "adverse events", and "warnings" sections of the summary of product characteristics of the drugs concerned.

Our results show the relevance of notification of side effects by physicians to pharmacovigilance centres, leading to the identification of a signal and public health dissemination of warnings.