Safety and efficacy of perampanel in advanced Parkinson's disease: a randomized, placebo-controlled study.

Perampanel, a novel, noncompetitive, selective AMPA-receptor antagonist demonstrated evidence of efficacy in reducing motor symptoms in animal models of Parkinson's disease (PD). We assessed the safety and efficacy of perampanel for treatment of "wearing off" motor fluctuations in patients with PD. Patients (N = 263) were randomly assigned to once-daily add-on 0.5, 1, or 2 mg of perampanel or placebo. The primary objective was to determine whether there was a dose-response relationship for efficacy among the 3 perampanel doses and placebo. The primary efficacy endpoint for each treatment was measured as the least-squares (LS) mean change from baseline to week 12 in percent "off" time reduction during the waking day, as recorded by patient diaries. The primary efficacy analysis was a 1-sided Williams test for dose-response trend at the 0.025 level of significance. At week 12, dose-response trends, as determined by the Williams test, were not statistically significant for LS mean reduction in percent "off" time during the waking day (P = 0.061, with significance defined as P <or= 0.025). The 2 higher perampanel doses (ITT population; n = 258) produced nonsignificant reductions from baseline to week 12 in percent "off" time during the waking day versus placebo (7.59%, P= 0.421 [1 mg], 8.60%, P = 0.257 [2 mg] versus 5.05% [placebo]; significance for pairwise comparisons defined as P <or= 0.05). There were no significant changes in dyskinesia or cognitive function in any perampanel group versus placebo. Adverse events were similar across treatment groups. Perampanel treatment was well tolerated and safe, but failed to achieve statistical significance in primary and secondary endpoints.