PHILADELPHIA — New open-label data from the expanded-access treatment program involving the cannabidiol Epidiolex (GW Pharma) show the median reduction in frequency of convulsive seizures after 3 months of treatment was 45% in all patients but higher in those with Dravet syndrome, among the most severe types of epilepsy.
The data are “very positive and promising,” said lead author Orrin Devinsky, MD, professor, neurology, neurosurgery and psychiatry, and director, New York University Comprehensive Epilepsy Center.
A “big caveat,” however, is that this was an open-label study, he said. Results of the first of four placebo-controlled trials of this cannabidiol product are expected early next year.
Dr Devinsky reported the findings here during the American Epilepsy (AES) 69th Annual Meeting.
Epidiolex is a purified cannabidiol compound in an oil base that is administered as a liquid.
The efficacy analysis included 261 patients with continuous exposure to the drug and on whom investigators had evaluable data. These patients ranged in age from 4 months to 41 years with a median of 11.8 years.
In these patients, 19% had a known cause of their epilepsy. The most common diagnoses were Dravet syndrome (17%), Lennox-Gastaut syndrome (LGS), or another rare type of epilepsy (15%).
The median number of convulsive seizures per 28 days in these patients was 31, so this was a group of children and young adults with “very frequent seizures,” commented Dr Devinsky.
Patients started cannabidiol at a dose of 2 to 5 mg/kg per day, which was gradually increased to a maximum dose of 25 mg/kg per day. They were also taking an average of three antiepileptic drugs (AEDs).
The primary outcome was change in the number of convulsive seizures, which are relatively easy to count, said Dr Devinsky.
“The seizure had to last at least 3 to 5 seconds with convulsive activity, cause the child to drop or fall to the floor or have a tonic-clonic convulsive event,” he noted. These are readily identifiable by most parents who tracked these events in seizure diaries, he said.
The median convulsive seizure frequency reduction was 45.1% in all patients and 62.7% in patients with Dravet syndrome. The median reduction of atonic seizures was 71.1% in patients with LGS.
In terms of achieving a greater than 50% reduction in seizures, this rate was 47% among all patients. During the last 4 weeks of the trial (between weeks 9 and 12), 9% of all patients and 13% of patients with Dravet syndrome were seizure-free.
“These are numbers that are much greater than what I would have predicted, and certainly would have ever predicted for a placebo response,” said Dr Devinsky.
He stressed that the study population was “an extremely treatment-refractory group.” Such patients are at high risk for sudden unexpected death from epilepsy and status epilepticus.
Safety data were available on 313 treated patients. The most common adverse effects were somnolence (23%), diarrhea (23%), fatigue (17%), decreased appetite (17%), convulsion (17%), and vomiting (10%).
Dr Devinsky stressed that only 4% of all patients stopped the study medication because of adverse effects. “For an antiepileptic drug trial, that is relatively low.”
About 12% of patients withdrew because of lack of effectiveness, which “again was relatively small” in this type of trial, said Dr Devinsky.
Serious adverse events were reported in 34% of patients and included seven deaths. “None of these were felt by investigators to be related to the study medicine,” said Dr Devinsky.
Although these new results “add more fuel to the fire showing that this drug is likely to be effective for some people,” without a double-blind study “we are essentially blind to know what the truth is,” said Dr Devinsky.
An interesting study finding was that patients who were taking the AED clobazam had a better overall rate of treatment response (57% with a 50% or greater seizure reduction compared with 39%). This may reflect elevations in the desmethyl clobazam metabolite, he said.
“We think that some of the efficacy — and some of the toxicity — in some patients may have been related to the clobazam,” said Dr Devinsky.
He noted, however, that there was “absolutely no difference” in the response rate between patients with Dravet syndrome and LGS taking clobazam and those not taking that drug.
Asked whether increasing the dose of clobazam in patients already receiving this drug might have resulted in the same improved treatment response, Kelly Knupp, MD, Children’s Hospital, Aurora, Colorado, said that’s “possible.”
Increased clobazam levels probably reflect “some drug-to-drug interactions that need to be sorted out,” said Dr Knupp, who treats children with epilepsy.
She, too, is keen to know what the placebo effect is. “The value of doing a double-blind randomized study is that it takes into account that placebo effect; we just don’t really understand it here.”
In the meantime, she said she’s “guardedly hopeful” about the new results.
The study was funded by GW Pharma. Dr Devinsky reports receiving research support from GW Pharma.
American Epilepsy Society (AES) 69th Annual Meeting. Abstract 3.397.