Postoperative reduction of anti-seizure medication is associated with cognitive improvement in seizure-free patients after anterior temporal lobectomy.
OBJECTIVE To determine whether postoperative reduction of anti-seizure medication (ASM) burden is associated with cognitive improvement after anterior temporal lobectomy (ATL) in seizure-free patients. METHODS We retrospectively analyzed data for adult patients with drug-resistant temporal lobe epilepsy who underwent unilateral ATL between March 2013 and January 2025. Analyses were restricted to patients achieving complete freedom from seizures (Engel Class I) at 1 year. ASM burden was quantified using the prescribed daily dose/defined daily dose ratio, and the reduction rate was calculated. Preoperative and 1-year postoperative neuropsychological outcomes, including full-scale intelligence quotient (FSIQ) and domain-specific indices, were assessed. The effect of ASM reduction on cognitive change was evaluated using a doubly robust estimation framework integrating inverse probability of treatment weighting with multivariable outcome regression. Robustness and potential dose-response relationships were assessed via sensitivity analyses, analysis of covariance (ANCOVA), and categorical modeling. RESULTS Fifty-seven seizure-free patients were included in the primary analysis (median age at surgery, 32 years; 33 female). Greater ASM reduction was significantly associated with improved working memory (WM) (β = .066, p = .001). A positive but non-significant trend was observed for FSIQ (β = .021, p = .062). A dose-response relationship was identified, with greater ASM reduction associated with progressively greater improvement in WM (p = .013). Findings were consistent across sensitivity and ANCOVA analyses. SIGNIFICANCE Postoperative ASM reduction was associated with domain-specific cognitive improvement following ATL, particularly in WM. These findings support ASM tapering as a potentially modifiable factor influencing postoperative cognitive outcomes.