Abstracts

Rationale: Perampanel is a once-daily oral anti-seizure drug (ASD) for partial-onset seizures (POS) and primary generalized tonic-clonic (PGTC) seizures. There are limited data on real-world use of perampanel as an ASD in the US. PROVE (Study 506; NCT03208660) is a retrospective, multicenter, non-interventional Phase IV study of perampanel administered to patients (pts) with epilepsy during routine clinical care. Here, we report a post hoc analysis of the data from PROVE to assess whether regional differences existed in retention rate, safety, efficacy, and dosing experience between pts treated with perampanel at a single study site, Northeast Regional Epilepsy Group in Hackensack, New Jersey, US (known as Site #1009), and pts treated with perampanel across all other study sites. Methods: Data were obtained from medical records of pts who initiated perampanel treatment after January 1, 2014, at 38 sites across 20 states of the US. Follow-up was completed on March 15, 2019. Based on the Safety Analysis Set (SAS), the primary endpoint was retention rate (proportion of pts remaining on perampanel at 3, 6, 12, 18, and 24 months following treatment initiation). Dosing experience, efficacy, and safety (including monitoring of treatment-emergent adverse events [TEAEs]) were secondary objectives. Retention rates at 24 months and incidence of TEAEs between Site #1009 and all other sites were compared using χ2-test. Results: SAS included 70 pts at Site #1009 (mean [standard deviation (SD)] age, 27.0 [15.8] years; 62.9% female) and 1623 at all other sites (mean [SD] age, 28.6 [16.6] years; 52.2% female). Most common seizure types were complex POS (Site #1009, n=30 [42.9%]; other sites, n=968 [59.8%]) and PGTC (Site #1009, n=46 [65.7%]; other sites, n=812 [50.1%]). Most pts received 1-3 Baseline ASDs (Site #1009, 88.6%; other sites, 73.8%); 27.1% and 19.8% were on enzyme-inducing ASDs, respectively. At completion, 51.4% (n=36; Site #1009) and 50.7% (n=823; other sites) of pts remained on perampanel.Perampanel was uptitrated weekly (32.9% [Site #1009]; 18.1% [other sites]), biweekly (24.3%; 18.5%), every 3 weeks (0.0%; 1.7%), 'other' (14.3%; 53.7%), and 'unknown' (28.6%; 8.1%). Mean (SD, range) maximum daily perampanel doses were 5.9 (2.7, 2-12; Site #1009) and 6.7 (3.4, 0-52; other sites) mg; mean (SD, range) daily doses were 5.9 (2.4, 3.5-9.5) and 5.6 (3.0, 0.1-52.0) mg, respectively; mean (SD, range) cumulative duration of exposure was 13.8 (12.8, 0.0-45.1) and 17.3 (15.7, 0.0-77.1) months, respectively. Retention rates at 24 months were similar for Site #1009 and other sites (P=0.5), as were seizure outcomes based on investigator opinion (Figure 1). No seizure frequency data were recorded at Site #1009.TEAEs occurred in 25 (35.7%) pts at Site #1009 and 646 (39.8%) at other sites (P=0.5; Table 1). Conclusions: Despite some variability in titration rate, clinical outcomes, as measured by retention and side effect rates, were not significantly different between Site #1009 and all other sites. Additional analyses are planned to further investigate outcomes of perampanel use between specific PROVE study sites to establish whether regional differences do indeed exist. Funding: Eisai Inc.