For families managing Dravet syndrome — one of the most treatment-resistant childhood epilepsies — every incremental reduction in seizure burden carries profound consequences for quality of life and neurodevelopmental trajectory. A new phase 3 trial testing soticlestat, a novel cholesterol 24-hydroxylase (CH24H) inhibitor, offers a mixed but instructive picture for the epilepsy field.
The SKYLINE trial randomized 144 patients aged 2–21 years with Dravet syndrome and uncontrolled convulsive seizures (minimum four per month despite existing therapy) to receive weight-adjusted soticlestat 300 mg twice daily or placebo over 16 weeks. Critically, 81% of participants were already on three or more antiseizure medications, underscoring the severity and pharmacoresistance of the enrolled population. The primary endpoint — median reduction in monthly convulsive seizure frequency across the full titration and maintenance period — reached −22.16% in the soticlestat group versus −8.64% with placebo, a difference of approximately 15.6 percentage points that narrowly missed statistical significance (p = .061). The maintenance-period analysis similarly fell short (p = .089). The proportion achieving a clinically meaningful ≥50% seizure reduction was also assessed as a key secondary endpoint, though results beyond the excerpt were not available for full interpretation.
The failure to cross the p < .05 threshold despite a directionally promising absolute difference raises important considerations. Soticlestat's mechanism — reducing neuronal oxysterol levels by inhibiting CH24H — is biologically plausible in hyperexcitable networks, and earlier phase 2 signals in both Dravet and Lennox-Gastaut syndromes generated genuine optimism. However, the relatively modest effect size and a pronounced placebo response (nearly 9% reduction) in a highly medicated cohort are consistent with challenges that have plagued Dravet trials historically. The 16-week duration also limits conclusions about sustained efficacy. This result is more setback than refutation — the mechanistic rationale remains intact — but the path to regulatory approval will require rethinking trial design or patient stratification.