3 Severe cutaneous adverse reactions time latency between beginning of drug use and onset of reaction

Abstract

Background Toxic epidermal necrolysis (TEN) and Stevens–Johnson syndrome (SJS) are acute life-threatening severe cutaneous adverse reactions (SCAR) with an unclear pathogenesis mainly caused by drugs. Allopurinol and trimethoprim/sulphamethoxazole (TMS) are both well known to be associated with these conditions. In addition, TMS is known to induce generalized bullous fixed drug eruption (GBFDE), a less severe condition with a very short induction period that is clinically often confused with SJS or TEN. Aim We want to further investigate the risk profile of allopurinol and TMS for inducing SCAR, as the hazard functions of these substances are different. Furthermore, the re-review of cases using more specific criteria to differentiate between SCAR and GBFDE should allow us to detect misclassification of cases. Methods 984 cases of SJS, SJS/TEN overlap and TEN were ascertained by a population-based registry between 1990 and 1999. The following analysis is based on a random sample of 115 cases earlier accepted as SJS or TEN, which were exposed to either allopurinol or TMS, and 38 cases excluded in the previous review. An independent expert committee blinded for possible causes re-reviewed these cases in clinical terms, as the original review process took place over a period of 10 years. In this analysis special emphasis is given to the time latency between beginning of drug use and onset of SCAR. Results Before re-review 162/984 patients with SCAR reported the use of allopurinol and 131/984 the use of TMS within 2 weeks prior to the onset of the adverse reaction. After the re-review the percentage of doubtful cases was higher for TMS (28/57) than for allopurinol (30/83). For definite cases of SJS or TEN the range between the lower and upper quartile of the time latency between beginning of drug use and onset of SCAR was 14–34 days for allopurinol, in contrast to 5–15 days for TMS. The time latency for doubtful and excluded cases after the use of TMS was much shorter (2.5 and 2 days, respectively). Conclusions The high numberof doubtful cases after the re-review reveals the difficulty of applying approved detailed definitions to the variety of clinical patterns of cutaneous adverse reactions. We could confirm a high correlation of time latency between beginning of drug use and onset of SCAR and GBFDE for allopurinol and TMS, which may have an important impact on the risk profile of these and other suspected drugs, as well as on pathogenetic and therapeutic considerations of severe adverse events. When drug exposure occurs outside the relevant interval of time latency for SJS and TEN, other risk factors and/or differential diagnoses such as GBFDE should be considered.