Study Finds Cannabidiol Reduces Frequency of Convulsive Seizures in Children and Young Adults with Dravet Syndrome

Dravet syndrome, a rare genetic childhood epilepsy associated with drug-resistant seizures and high mortality, may be amenable to treatment with cannabidiol (CBD), a constituent of cannabis (Cannabis sativa, Cannabaceae). While 4 small trials reported mixed results, in vitro and in vivo models show antiseizure activity. A standardized oral solution of CBD (prepared and provided by GW Pharmaceuticals; London, United Kingdom) has been tested for safety and efficacy in an open-label trial involving children and young adults with drug-resistant seizures. The authors conducted a double-blind, randomized controlled trial (RCT) of the same CBD solution in patients with Dravet syndrome.

Patients in this multinational, multicenter RCT were aged 2-18 years, diagnosed with Dravet syndrome, used ≥1 antiepileptic drugs, and had ≥4 convulsive seizures (CSs) in a 28-day baseline period. Baseline was followed by 14 weeks of treatment (2 weeks of dose escalation + 12 weeks of maintenance), a 10-day taper period, and a 4-week follow-up. Epilepsy drugs and interventions were stable for 4 weeks before screening and were to remain so during the RCT. Of 177 patients screened, 120 were randomly assigned on a 1:1 basis to receive placebo (a solution identical to the CBD solution except for the absence of CBD; prepared and provided by GW Pharmaceuticals) or CBD, plus their existing regimens. Both groups received doses 2x/day, with the CBD group totaling up to 20 mg CBD/kg body weight/day. The CBD oral solution contained 100 mg of CBD per mL. During the taper period, doses were reduced by 10% daily over a period of 10 days. Caregivers recorded CS data (daily number and type) on an interactive voice system. After completing the RCT, all patients could enter an open-label, long-term study if they chose. Clinical assessments were performed at baseline; 2, 4, 8, and 14 weeks of treatment; and the end of the taper period for those who withdrew after tapering the trial agent or who did not enter the open-label study.

Mean age of patients who were randomly assigned was 9.8 years (range, 2.3-18.4); 52% were male. Median baseline CS frequency was 13.0/month (range, 3.7-1717). A total of 108 patients (90%) completed the treatment phase; 52 of 61 (85%) in the CBD group and 56 of 59 (95%) in the placebo group. Twelve patients (10%) withdrew early (9 in the CBD group; 3, placebo). A total of 105 patients entered the open-label, long-term trial. Patients had previously used a median 4.0 antiepileptic drugs (range, 0-26) and were using a median 3.0 (range, 1-5). The most common type of convulsion was generalized tonic-clonic (94 patients, 78%), with secondary generalized tonic-clonic seizures in another 25 (21%); other types were less frequent. Non-convulsive seizures were reported by 37 patients in the CBD group (61%) and 41 in the placebo group (69%). Developmental delay in 114 of 118 patients with data available was severe or profound in 56 (48%).

The primary endpoint was the percentage change from baseline in CSs every 28 days during the 14-week treatment period among patients who received CBD compared to placebo. Numerous secondary endpoints included the Caregiver Global Impression of Change (CGIC); a number of patients whose seizures decreased by at least 25%, 50%, 75%, or 100%; sleep disruption; use of rescue medication; and more.

In the CBD group, CS frequency fell from a median 12.4/mo (range, 3.9-1717) at baseline to 5.9/mo (range, 0.0-2159) over the treatment period, representing a median change of −38.9%. In the placebo group, CS frequency fell from a median 14.9/mo (range, 3.7-718) to 14.1/mo (range, 0.9-709), a median change of −13.3%. The adjusted median difference in CS frequency between CBD and placebo groups was −22.8% (95% confidence interval [CI], −41.1 to −5.4; P = 0.01). Effects were seen in the first 4 weeks of treatment when the median number of CSs fell from 12.4 to 5.0/mo in the CBD group and from 14.9 to 13.0/mo in the placebo group (P = 0.002). A reduction in CS frequency by 50% or more was experienced by 43% of patients in the CBD group and 27% of those in the placebo group (odds ratio, 2.00; 95% CI, 0.93-4.30; P = 0.08). During the treatment phase, 3 patients in the CBD group and 0 in the placebo group were CS-free (P = 0.08). Rescue medications were used by 36 patients (59%) in the CBD group and 41 (69%) in the placebo group. On the CGIC, 37 of 60 caregivers (62%) rated their child’s condition improved in the CBD group compared to 20 of 58 (34%) in the placebo group (P = 0.02).

Adverse events (AEs) were seen in 93% of patients in the CBD group and 75% in the placebo group. While 89% of these (84% in the CBD group; 95%, placebo) were mild to moderate, 8 patients in the CBD group and 1 in the placebo group withdrew from the RCT due to AEs. AEs resolved completely in 8 of 10 patients in the CBD group after dose reductions; partially, in 1; but continued in 1 other. The most common AE was sleepiness, reported by 22 (36%) in the CBD group and 6 (10%) in the placebo group. Of those in the CBD group reporting somnolence, 18 of 22 were taking clobazam, as were 5 of 6 with somnolence in the placebo group. Ten serious AEs were reported in the CBD group and 3 in the placebo group, including status epilepticus in 3 patients in each group; however, these did not lead to withdrawal. Elevated liver enzymes led to the withdrawal of 3 in the CBD group and 1 in the placebo group; 9 others in the CBD group experienced elevated liver aminotransferase enzymes but did not withdraw; all 13 were taking a form of valproate. In the 9 patients with elevated liver enzymes who continued the RCT, levels returned to normal during CBD treatment. In the CBD group, other AEs reported include vomiting, fatigue and lethargy, upper respiratory infection, poor appetite, convulsion, and diarrhea. Diarrhea was experienced by 19 (31%) in the CBD group, the second most common AE in that group; among those in the placebo group, diarrhea equaled somnolence in frequency (n = 6, 10%).

CBD reduced the frequency of CSs in patients with Dravet syndrome in this RCT, but more studies are needed to determine its long-term safety, including drug interactions, and efficacy in this catastrophic disease affecting children and young adults.

The study was funded by GW Pharmaceuticals. GW Pharmaceuticals was responsible for the trial design (with input from investigators and other experts), trial management, site monitoring, trial pharmacovigilance, and data and statistical analyses. The findings reported in the study are specific to the GW Pharmaceuticals formulation of CBD and cannot be extrapolated to other CBD products.

Author S. Wright is employed by GW Pharmaceuticals and holds a pending patent on the use of cannabinoids in the treatment of epilepsy (WO2015193667) and a patent on the use of phytocannabinoids in the treatment of epilepsy (EP2448637). Author L. Laux reports having received grants from GW Pharmaceuticals during the conduct of the study. Author J.H. Cross reports receiving grant support from GW Pharmaceuticals, and other manufacturers of pharmaceutical drugs. Authors O. Devinsky, E. Marsh, I. Miller, I.E. Scheffer, and E.A. Thiele all report receiving grant support, consulting fees, honoraria, or travel support from several large pharmaceutical companies. Five members of the Cannabidiol in Dravet Syndrome Study Group are employed by GW Pharmaceuticals.



Devinsky O, Cross JH, Laux L, et al.; for the Cannabidiol in Dravet Syndrome Study Group. A trial of cannabidiol for drug-resistant seizures in the Dravet syndrome. N Engl J Med. May 25, 2017;376(21):2011-2020.