Cannabidiol Reduces Seizure Frequency in Dravet Syndrome: Phase 3 Trial
Abstract
In this landmark Phase 3 RCT, CBD reduced convulsive seizure frequency by 38.9% vs. 13.3% with placebo in children with Dravet syndrome. This trial supported FDA approval of Epidiolex. Adverse events included somnolence and elevated liver enzymes.
Study Summary
This landmark Phase 3, double-blind, placebo-controlled RCT enrolled 120 children and young adults (ages 2–18) with Dravet syndrome — a severe, treatment-resistant epilepsy caused by SCN1A mutations. Participants were randomized 1:1 to receive CBD (Epidiolex, 20mg/kg/day) or placebo as add-on therapy to existing antiepileptic drugs for 14 weeks. The primary endpoint was change in convulsive seizure frequency from baseline. CBD reduced convulsive seizure frequency by 38.9% vs. 13.3% with placebo (p=0.01). The proportion of patients with ≥50% seizure reduction was 43% with CBD vs. 27% with placebo. Five patients in the CBD group became seizure-free vs. none in the placebo group. Adverse events were more common with CBD: somnolence (36% vs. 10%), decreased appetite (28% vs. 5%), and elevated liver aminotransferases (20% vs. 3%, particularly in patients on valproate). This trial directly supported the FDA approval of Epidiolex in June 2018 — the first FDA-approved cannabis-derived medicine.
Key Findings
- 1CBD reduced convulsive seizure frequency by 38.9% vs. 13.3% placebo (p=0.01)
- 243% of CBD patients achieved ≥50% seizure reduction vs. 27% placebo
- 35 patients became seizure-free with CBD vs. 0 with placebo
- 4Liver enzyme elevations in 20% of CBD patients — highest risk with concurrent valproate
- 5This trial directly supported FDA approval of Epidiolex in June 2018
Clinical Implications
- CBD (Epidiolex) is FDA-approved and evidence-based for Dravet syndrome seizure management
- Liver function monitoring is required, especially in patients on valproate
- Clobazam dose may need adjustment — CBD inhibits CYP2C19, increasing N-desmethylclobazam
- The 20mg/kg/day dose used in this trial is the FDA-approved starting dose
Study Limitations
- 14-week trial duration — long-term efficacy and safety require open-label extension data
- Add-on design — efficacy as monotherapy not established
- Blinding integrity: somnolence and appetite changes may have unblinded some participants
- Generalizability limited to Dravet syndrome — other epilepsy types require separate evidence