Pediatric Vulnerability to Anti-Seizure Medication-Induced SJS, TEN, and DRESS: A Pharmacovigilance Analysis of the AEMS Database

Background: Anti-Seizure Medications (ASMs) are associated with an elevated risk of Stevens-Johnson Syndrome (SJS), Toxic Epidermal Necrolysis (TEN), and Drug Reaction with Eosinophilia and Severe Symptoms (DRESS). These events are grouped as Severe Cutaneous Adverse Reactions (SCARs), which can be fatal. The pediatric population represents a uniquely vulnerable group, yet safety data specifically focusing on children remain limited. This analysis investigates which Anti-Seizure Medications generate the strongest signals for SJS, TEN, and DRESS in the pediatric population using the FDA Adverse Event Monitoring System (AEMS) database.

Objective: To determine which commonly prescribed ASMs generate the strongest disproportionality signals for Stevens-Johnson Syndrome, Toxic Epidermal Necrolysis and Drug Reaction with Eosinophilia and Severe Symptoms in pediatric AEMS reports.

Methods: This retrospective pharmacovigilance study analyzed spontaneous adverse event reports involving pediatric patients treated with Anti-Seizure Medications (ASMs). Data were extracted between June 2026 and July 2026 from the FDA Adverse Events Monitoring System (AEMS) for 13 ASMs: cannabidiol, cenobamate, clobazam, clonazepam, carbamazepine, gabapentin, lacosamide, lamotrigine, levetiracetam, oxcarbazepine, phenytoin, perampanel, topiramate, and valproic acid. Fenfluramine was identified as a medication of interest, but excluded as there were no AEMS reports identifying SJS/TEN/DRESS events. Reports were filtered to include cases where the ASM was prescribed for seizure and/or epilepsy where skin and subcutaneous tissue disorders were reported, specifically, specifically SJS, TEN, and DRESS. The dataset was restricted to a ten year period ranging from 2016 through 2026. The pediatric cohort was defined as patients aged 17 years or younger. Each case outcome (death, hospitalization, life-threatening event, or disability) and treatment regimen (monotherapy vs. polytherapy with other ASMs) were recorded. Duplicate reports were identified and removed. After duplicate removal, the analytic dataset included 94,961 unique AEMS reports. Disproportionality was calculated using reporting odds ratios (RORs) and 95% confidence intervals (CIs) for each medication. Additionally, primary disproportionality and sensitivity analysis on monotherapy was performed

Results: Out of 94,961 unique AEMS reports, 13,420 met pediatric eligibility criteria, including 233 cases of SJS, 154 of TEN, and 390 of DRESS. Lamotrigine generated the largest and most consistent signals across all three severe skin conditions in both primary analysis (SJS: ROR=6.34 ,CI: 4.86–8.26; TEN: ROR=4.05, CI: 2.94–5.57; DRESS: ROR=3.61, CI: 2.94–4.42) and monotherapy analysis, where its effect sizes notably increased (SJS: ROR increased to 9.29, TEN: ROR increased to 5.26, DRESS: ROR increased to 5.77). Carbamazepine and phenytoin demonstrated a similar significant rise in ROR within the monotherapy cohort for DRESS and TEN. Conversely, primary signals for valproic acid and oxcarbazepine with SJS had initial elevations of ROR but failed to persist during monotherapy analysis. Clobazam demonstrated signaling for DRESS in monotherapy analysis only.

Conclusion: Based on the pharmacovigilance analysis, lamotrigine demonstrated consistent disproportionality signals with significant findings across the three SCARs of interest. These risk signals intensified in monotherapy settings, suggesting that polytherapy diluted the signal, and lamotrigine’s risk for SCARs is highest when used alone. Carbamazepine and phenytoin exhibited similar patterns with significant ROR increases in monotherapy for DRESS and TEN. The lack of elevation of ROR in monotherapy analysis with valproic acid and oxcarbazepine with SJS suggests potential confounding polytherapy. The clobazam DRESS signal appeared only in monotherapy analysis and should be considered a secondary finding. These results are consistent with established literature identifying these medications as high risk agents for SCARs and applies this information to a pediatric cohort. Our findings reinforce the importance of clinical monitoring during the initiation of ASMs, particularly lamotrigine, phenytoin, and carbamazepine in children, particularly when used as monotherapy.