The risk of Stevens-Johnson syndrome and toxic epidermal necrolysis in new users of antiepileptic drugs

Epilepsia. 2017 Dec;58(12):2178-2185. doi: 10.1111/epi.13925. Epub 2017 Oct 13.


Objective: Older antiepileptic drugs (AEDs) are known to cause Stevens-Johnson syndrome and toxic epidermal necrolysis (SJS/TEN). However, evidence for newer AED is sparse. We quantified risks of SJS/TEN in association with use of all AEDs in the United Kingdom.

Methods: In a matched case-control study of 480 previously validated SJS/TEN cases (1995-2013) we used conditional logistic regression to calculate odds ratios (ORs) with 95% confidence intervals (CIs), and calculated absolute risks of SJS/TEN within separate cohorts of new users of 28 AEDs. We assessed causality between drugs and SJS/TEN in each exposed case, using an adapted version of the algorithm of drug causality for epidermal necrolysis (ALDEN) score.

Results: We observed a strong association between SJS/TEN and new use of carbamazepine (OR 92.57, 95% CI 19.89-∞), phenytoin (OR 49.96, 95% CI 10.13-∞), and lamotrigine (OR 26.90, 95% CI 4.88-∞), where causality, according to the ALDEN score, was very probable or probable for most exposed cases. Absolute risks for SJS/TEN were highest for phenytoin (45.86 cases/100,000 exposed), lamotrigine (44.17 cases/100,000 exposed), and carbamazepine (20.38 cases/100,000 exposed). Despite increased ORs for valproate (40,941 exposed), gabapentin (116,037 exposed), pregabalin (59,967 exposed), and clobazam (4,300 exposed), ALDEN suggested no causal association. There were no observed cases of SJS/TEN among new users of levetiracetam (n = 96,77), clonazepam (n = 18,075), or topiramate (n = 11,307).

Significance: The results of our study are consistent with those of previous studies of SJS/TEN, which found increased risks of SJS/TEN in new use of carbamazepine, phenytoin, and lamotrigine. Despite frequent use, no ALDEN-score confirmed cases were observed in new users of valproate, gabapentin, pregabalin, levetiracetam, topiramate, or clonazepam.

Keywords: Associations; Clinical practice research datalink; Culprit drugs.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Anticonvulsants / adverse effects*
  • Case-Control Studies
  • Cohort Studies
  • Databases, Factual
  • Drug Eruptions / epidemiology*
  • Female
  • Humans
  • Male
  • Risk Assessment
  • Risk Factors
  • Stevens-Johnson Syndrome / epidemiology*


  • Anticonvulsants