Long-term tolerability, safety and efficacy of adjunctive perampanel in the open-label, dose-ascending Study 231 and extension Study 233 in Japanese patients with epilepsy

Abstract

PurposeTo evaluate long-term tolerability, safety and efficacy of adjunctive perampanel in a Phase II, multicentre, open-label, dose-ascending Study 231 (NCT00849212) and its extension (Study 233; NCT00903786) in Japanese patients with refractory partial-onset seizures (POS), with/without secondarily generalised seizures. MethodsIn Study 231, patients received adjunctive perampanel ≤12 mg/day during a 10-week treatment period. Patients completing Study 231 could enter Study 233 (≤316-week treatment period). Assessments included monitoring of treatment-related treatment-emergent adverse events (TEAEs), median percent change in seizure frequency per 28 days, 50% responder and seizure-freedom rates. During Study 231, a pharmacokinetic analysis assessed the effects of enzyme-inducing antiepileptic drugs. ResultsOverall, 23/30 (76.7%) patients completed Study 231; 21/30 (70.0%) received perampanel ≥8 mg/day and 10/30 (33.3%) achieved a maximum tolerated dose of 12 mg/day. Median percent change in seizure frequency per 28 days was –35.0%. 50% responder rate was 37.0%; 4 (13.3%) patients achieved seizure freedom. Twenty-one patients entered Study 233. Mean duration of exposure was 195 weeks; 9 (42.9%) patients received perampanel for ≤208 weeks. Seizure control was sustained for 316 weeks in 3/21 (14.3%) patients; 2 achieved seizure freedom. Treatment-related TEAEs were tolerable; the most common was dizziness (Study 231, 53.3%; Study 233, 14.3%). Mean perampanel plasma concentrations were lower with concomitant carbamazepine vs non-inducers (152.7 ng/mL vs 389.4 ng/mL across perampanel groups); small patient numbers for non-inducers (n = 2) should be considered when interpreting these data. ConclusionAdjunctive perampanel demonstrated a favourable safety profile and long-term tolerability in Japanese patients with refractory POS for ≤316 weeks.