Focal epilepsy constitutes 60-70% of epilepsy, and up to half of patients do not achieve seizure freedom with their first antiseizure medication (ASM). When the first ASM fails, evidence guiding whether to switch or add-on another ASM and which ASMs to use is limited. This review synthesized evidence from randomized controlled trials (RCTs) on the comparative effectiveness of subsequent ASMs after first ASM failure in focal epilepsy. Following a pre-registered protocol (PROSPERO: CRD42025603003) and PRISMA guidelines, we included RCTs involving children or adults with focal epilepsy who failed first ASM therapy for any reason. The primary outcomes were seizure remission and responder rate (≥50% reduction in seizure frequency). Searches were conducted across major databases in February 2025. Risk ratios with 95% confidence intervals were calculated in R. Meta-analysis was not performed due to heterogeneity. Six RCTs (961 participants) published between 1998 and 2012 met inclusion criteria, assessing seven ASMs under add-on or switch strategies. Trial duration ranged from 12 to 52 weeks; most were judged high risk of bias (ROB-2). Seizure remission end-point was assessed at a short-term endpoint of 3 months. Remission rates ranged from 11 to 43% for add-on and 8-43% for switch trials; response rates were 34-63% and 38-76%, respectively. Valproate showed higher responder rates than primidone (RR 1.52, 95% CI 1.01-2.28) in an add-on trial, while lamotrigine had a lower responder rate than valproate (RR 0.74, 95% CI 0.55-0.99) in one trial. In a switch strategy, lamotrigine showed lower treatment failure rates than valproate (RR 0.61; 95% CI 0.45-0.83) and fewer adverse effects than carbamazepine and valproate. Evidence guiding treatment decisions after the first ASM failure remains limited and outdated. Seizure outcomes were modest and comparable across treatment strategies, with few significant differences between ASMs. The small number of trials, short follow-up for endpoints, and methodological heterogeneity across trials constrain interpretation and clinical applicability. Leveraging observational data with causal inference methods, alongside pragmatic trials evaluating machine-learning-guided ASM selections, is needed to address this evidence gap and improve seizure management.
Egesa et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: