Abstracts

Rationale: Eslicarbazepine acetate (ESL) is a novel once-daily anticonvulsant, extensively converted after oral administration to eslicarbazepine, which is thought to block voltage-gated sodium channels. In three phase III, randomized, double-blind trials, subjects treated with once-daily ESL 800 mg or 1200 mg as adjunctive therapy had significantly greater seizure frequency reduction than placebo-treated subjects. This analysis examined the association between seizure reduction and baseline depression status. Depressive symptoms are common in patients with epilepsy, and have been shown to be associated with a poorer response to AE therapy. (Miller JM et al. Drugs 2008) Methods: Depressive symptoms at baseline for subjects in the pooled per-protocol population (N=1006) were assessed using the Emotional Well Being (EWB) domain of the Quality of Life in Epilepsy Inventory-31 (QOLIE-31). EWB is equivalent to the Mental Health Index-5 of the 36-item Short Form health survey. Presence of depression symptoms was defined two ways: EWB ≤ 52 (100-point scale) based on published literature (Kelly MJ et al. BMC Psychiatry 2008), and as quartiles based on the study distribution. Response was defined two ways: ≥ 50% and ≥ 75% reduction in seizure frequency during maintenance therapy (12 weeks). Logistic regression models examined depressive symptoms as a response predictor as well as modification of treatment effects by depression status (4 models). Results: Evaluable baseline EWB scores were available for all subjects; average age of 37.8 years; average duration of epilepsy of 20.9 years; 77.7% Caucasian; and 50.3% female. Using EWB ≤ 52, 339 (33.7%) subjects were classified as having depressive symptoms. Depressive symptom quartiles were 0‑48 (n=263), 49‑64 (n=318), 65‑76 (n=201), and 77‑100 (n=224). Using EWB ≤ 52 as a cutoff value, depressive symptom were not associated with ≥ 50% response (p=0.849) or ≥ 75% response (p=0.546). Response rates are presented in Table 1. Similarly, using quartiles, depression status was not associated with response (≥50%: p=0.155; ≥75%: p=0.573). Upon examination of the interaction terms of treatment and depression status, baseline depression status did not modify the effect of treatment in any of the 4 models (p-values ranged from 0.136 to 0.834). Conclusions: In this post-hoc analysis of the pooled phase III clinical trials for ESL, depressive symptoms at baseline, as assessed by the QOLIE-31 EWB domain, were not predictive of seizure response, and did not modify the association between ESL treatment and seizure frequency response.