Abstracts

Rationale: A phase 3 study (GWPCARE1; NCT02091375) in patients with Dravet syndrome inadequately managed by =1 antiepileptic drug (AED), demonstrated that add-on treatment with a pharmaceutical formulation of highly purified CBD significantly reduced monthly convulsive seizure frequency over a 14-week treatment period by a median of 39% for CBD vs 13% for placebo. At baseline, most patients in the study reported having had some response to prior AEDs; however, a small cohort reported never having a response on any AED. This subgroup may be more refractory to next treatment and/or less prone to the placebo effect. Methods: GWPCARE1 Part B enrolled 120 patients (61 CBD; 59 placebo). Patients were asked at baseline “has any medication reduced seizure frequency?” Baseline characteristics, reductions in seizure frequency, adverse events (AEs), and caregiver global impression of change (CGIC) were assessed and analyzed in those patients who answered no to this baseline question. Results: Of the 120 patients in the study, 18 (15%) reported having no response to any prior AED; 9 treated with add-on CBD and 9 with add-on placebo. The mean age was 9.1 years, with one third less than 6 years. Patients tried a median of 6 AEDs (range, 4 to 9) for CBD vs 7 AEDs (5 to 17) for placebo. Use of clobazam was identical between groups, with more placebo patients taking valproate and stiripentol. At baseline, patients on CBD vs placebo had a median monthly seizure frequency of 18.0 (5.4 to 196) vs 7.7 (3.7 to 718) convulsive seizures and 22.6 (11.0 to 1705) vs 14.0 (5.6 to 725) total seizures. During the treatment period, the median reduction in monthly seizure frequency for patients on CBD vs placebo was 70% (interquartile range [IQR], 44 to 91) vs -11% (an increase; IQR, -36 to 27) for convulsive seizures, with =50% reduction in 6/9 vs 1/9 patients; and was 70% (IQR, 28 to 91) vs -20% (an increase; IQR, -36 to 33) for total seizures, with =50% reduction in 5/9 vs 2/9 patients. AEs occurred in 15/18 patients; most events were of mild/moderate severity with no serious AEs. The most common AEs were diarrhea (3 on CBD, 1 on placebo) and somnolence (3 on CBD, 0 on placebo). Two patients treated with CBD (both on valproate) had transaminase elevations >3 × ULN; both resolved spontaneously. Improvement in overall condition at last visit was reported for 6/9 CBD vs 3/9 placebo patients. Conclusions: A marked treatment effect of CBD vs placebo was observed in this small subgroup of patients reporting no response to any prior AED. The incidence and type of AEs were similar to the full study population, with AEs occurring more frequently in the CBD vs placebo group. Although this post hoc analysis is limited by the small sample size and non-rigorous selection criteria, results suggest that meaningful seizure reduction is possible with CBD, even in patients who have previously had no response to multiple AEDs. Funding: GW Research Ltd