Abstracts

Rationale: PER is the first AMPA-receptor antagonist approved for the adjunctive treatment of POS. BRV is structurally similar to levetiracetam (LEV) but has demonstrated a higher selectivity and affinity for the SV2A protein. Efficacy and safety between PER and BRV for the adjunctive treatment of POS are compared using ITC. Methods: Phase III randomized controlled trials (RCTs) of PER (Study 304, 305 & 306; N=1,298) and BRV (N01253, N01252, N01358 and N01254; N=1,742) were included. Data were pooled across trials of the same AED to generate one value per efficacy or safety variable per AED. The Bucher method was utilized to perform an ITC of the relative risks (RRs) for responder rate, seizure freedom, any adverse event (AE), serious AE, severe AE, and discontinuation due to AE. Patient characteristics were similar between the PER and BRV trials with the exception of the distribution of LEV use [Prior LEV, Concomitant LEV, No LEV] and the total number of baseline AEDs. ITC results are presented for the overall population and subgroups of LEV use as BRV efficacy values varied according to concurrent LEV exposure. Results: In the overall ITC analysis, PER compared to BRV demonstrated similar efficacy on 50% responder rate (RR 1.0, 95% CI 0.8-1.4) while no statistical difference was observed with seizure freedom (RR 0.4, 95% CI 0.1-2.4). PER patients were more likely to experience ?-50% reduction in seizure frequency compared to BRV patients in those taking concomitant LEV (RR 2.5, 95% CI 1.1-5.7). The responder rate was similar between the two treatments in patients with prior LEV use (RR 1.7, 95% CI 0.5-5.7) and no LEV use (RR 0.9, 95% CI 0.6-1.2). Differences in the rates of adverse events between the PER and BRV groups were not statistically significant [any AE (RR 1.1, 95% CI 1.0-1.3); serious AE (RR 1.7, 95% CI 0.7-3.8); severe AE (RR 1.5, 95% CI 0.8-3.0); discontinuation due to AE (RR 1.1, 95% CI 0.6-2.2)]. Conclusions: PER was not different from BRV on 50% responder rate for the overall population. PER demonstrated a higher 50% responder rate than BRV in patients taking concomitant LEV. These differences may be due to the similarity in the mechanism of action between BRV and LEV. Funding: This research was funded by Eisai Inc. G. Tremblay and V. Patel worked on this project while employees of Eisai Inc.