Abstracts

Rationale: Epilepsy clinical trials collect prospective baseline seizure counts over a specified period, typically 4 to 8 weeks. When designing these trials, the investigators need to establish a stable baseline while taking into consideration the high seizure burdens in poorly controlled patients but also accounting for the variable nature of background seizure frequency and its impact on measurement of treatment effect. Here we report the results of a reanalysis of Fintepla Study 1, exploring the impact of shortening the baseline observation from 6 weeks, as the study was conducted, to 4 weeks in a population of patients with Dravet syndrome (DS). Methods: Subjects 2-18 years old with a diagnosis of DS and in whom convulsive seizures (CS) were not completely controlled by their current anti-epileptic drug regimen were enrolled. Subjects who had >=6 CS during the 6-week baseline period, and at least 2 seizures per 3 weeks, were randomized in a 1:1:1 ratio to ZX008 0.8 or 0.2 mg/kg/day (maximum 30 mg/day), or placebo administered BID. The primary efficacy endpoint was the change in mean monthly CS frequency (CSF) between ZX008 0.8 mg/kg/day and placebo during the 14-week treatment period compared to the 6-week baseline observation period. The re-analysis included subjects having at least 4 CS in the 4 weeks immediately preceding randomization, with at least 1 CS during the first 2-week period and during the last 2-week period. Results: In Study 1, a total of 119 subjects with DS were randomized to treatment (n=40, 0.8 mg/kg/day; n=39, 0.2 mg/kg/day; n=40, placebo) in the originally designed study. As shown in the table, there was no meaningful difference in baseline seizure frequency when comparing a 6-week to 4-week baseline period though one placebo patient who met baseline criteria in the study did not meet criteria using a 4-week baseline. For the primary endpoint as calculated originally per the analysis plan, ZX008 0.8 mg/kg/day showed a 62.3% reduction in mean monthly CSF vs placebo (P<0.001), whereas the reanalysis of the primary endpoint using a 4-week baseline ZX008 0.8 mg/kg/day showed a 59.7% improvement (