乌鸦传媒

SEATTLE -- Epileptic patients and especially parents of children with the condition believe that the cannabidiol component of marijuana was effective, but there remains little solid evidence that it was not just a placebo effect, researchers said here.

Two surveys of parents of children with severe seizure disorders, presented at the American Epilepsy Society annual meeting, found high rates of perceived efficacy, particularly for Lennox-Gastaut syndrome and infantile spasms.

But the authors of those studies said the results were, in the words of one, "too good to be true" and in direct conflict with their own experiences in seeing patients taking cannabidiol-based products.

That is not to say the drug is definitely or even probably ineffective, however. Researchers were adamant that no firm conclusions could be reached until a randomized, controlled trial is conducted -- and such studies are launching shortly.

Cannabidiol is one of the many dozens of compounds present in marijuana. It is not psychoactive but interacts with cannabinoid and other pathways. Animal data have indicated that it has anti-seizure properties and epilepsy is now a major focus for medical marijuana use.

What most people use is some "artisanal" product grown or refined to emphasize cannabidiol over tetrahydrocannabinol (THC), the main psychoactive component responsible for the recreational "high." For pediatric treatment, these are formulated as oral concentrates which can be taken either straight or with food.

However, the AES meeting also had some early reports on a pharmaceutical cannabidiol formulation called Epidiolex, which is under development by the same company (GW Research) that sells the mixed cannabidiol-THC oral spray product called Sativex.

Parent Report Studies

, of Children's Hospital of Colorado in Denver, said epilepsy is the most commonly sought indication for medical marijuana use in his state.

Chapman was senior investigator on one of the parental-report studies, targeting parents of children using oral medical cannabinoids in Colorado. It was actually a medical chart review of 75 children treated at his center and not a survey per se. But all information about clinical responses came from parental reports, making the results indistinguishable from survey data.

The 75 patients were those whose charts indicated use of medical cannabinoids, including 41 who were established Colorado residents and 34 whose parents had moved to the state in order to obtain medical marijuana.

Mean patient age was 7.3 when starting on the cannabis products and the duration of clinician observation was 5.6 months. Diagnoses were mixed: absence seizures, focal seizures, generalized tonic-clonic seizures, myclonic seizures, epileptic spasms, Doose syndrome, Lennox-Gastaut syndrome, and Dravet syndrome were all represented.

Overall, 33% of the children were reported by parents to have shown at least 50% reductions in seizure frequency, the standard definition of a treatment response in epilepsy. There were no obvious differences in responder rates among the different seizure types.

But there was one finding that Chapman said was very revealing: responder rates were three times as high among patients whose families had just moved to Colorado as compared with the established residents. He said this was a fairly clear indication of parental bias in reporting responses to treatment.

Chapman said he had seen children seizing in his office despite parents' assurance that they were now seizure-free. And in the chart review, parent-reported responders with interictal EEG data available showed no objective improvement.

, of Mattel Children's Hospital at the University of California Los Angeles, who led the second survey study reported here, noted that parents giving cannabinoid products to their children tend to be "highly motivated and invested" in the drugs' efficacy. Obtaining these products is often illegal and, even in states with medical marijuana laws, they are expensive and not always easy to find. Parents have typically exhausted conventional treatments and are desperate for something that works.

Consequently, Hussain said, parents really want to believe that the cannabinoid products are helping their children.

The study by Hussain and colleagues was an online survey promoted to parents of children with infantile spasms, Lennox-Gastaut syndrome, and Dravet syndrome -- all devastating seizure disorders that develop in early childhood -- through Internet forums. Of 200 total respondents, 117 said they had used cannabidiol-type products with their children. Respondents were then asked to report previous treatments and to categorize the children's treatment responses to cannabinoid therapy as seizure-free, fewer seizures, no change, or more seizures.

According to the survey responses, the children were indeed a very severely afflicted and heavily pretreated group. Median time from symptom onset to starting on cannabinoids was 5 years. A median of eight medications had been tried unsuccessfully, 45% had failed a ketogenic diet, 13% had failed surgery, and 21% had failed vagal nerve stimulation.

Parental reports of cannabinoid efficacy were highly favorable. From 60% to 79% said their child had fewer seizures after starting the cannabinoid treatment, and 13% to 15% reported complete seizure freedom.

These response rates are "astronomical," Hussain said -- so much so that they are "too good to be true."

What the data from both studies show, is that well-controlled randomized trials are desperately needed, Hussain and Chapman said.

Epidiolex

That point was echoed by , of New York University's Epilepsy Center in New York City, who presented data from compassionate-use, open-label treatment of various seizure disorders with Epidiolex in 151 children and young adults.

Devinsky said the long-awaited randomized trial data on cannabinoids should be forthcoming with respect to Epidiolex. GW has listed a series of placebo-controlled phase III trials in Lennox-Gastaux and Dravet syndromes on Clinicaltrials.gov, and recruitment is expected to open shortly.

Mean age was 11, and patients were taking a mean of 2.7 other anti-epileptic drugs (AEDs).

Devinsky reported the mean reduction in seizure frequency at 40% for all patients. Dravet patients appeared to derive the most benefit, with a 72% mean reduction in seizure frequency and complete seizure freedom in 22% of the group (seven of 32).

He called the reported improvements "meaningful" but cautioned that, like in the other studies, they were ultimately based on unchecked parent reports.

Also, one concerning finding from the Epidiolex study was that the drug appears to affect blood levels of other AEDs. Cannabidiol is an inhibitor of CYP2C and CYP3A groups of drug-metabolizing enzymes, which are also important for some AEDs.

Findings from patients treated with Epidiolex suggested that clobazam is one of those most strongly affected -- blood levels of its active metabolite N-desmethylclobazam rose in all patients taking it, by more than threefold in one case.

The concern about AED blood level effects from cannabidiol was corroborated in an otherwise encouraging case report from a group led by , of the University of California San Diego. A child with Doose syndrome was started on an artisanal cannabidiol product of unknown provenance while continuing on valproate, on which the patient initially had no spells for 2 weeks but then relapsed.

Adding the cannabis product initially worsened the spells, but a check on valproate blood levels indicated that they had increased 50%. When the valproate dosage was adjusted downward and normal therapeutic levels were reestablished, the child became seizure-free and remained so for 5 months, with normal-appearing EEG readings.

Gold said these findings point to the importance of checking blood levels of conventional AEDs, which typically have very narrow therapeutic ranges, in patients starting on cannabinoid products.

That turn makes it critical to have a trusting relationship with patients and their parents, he said, because cannabinoids are still illegal in much of the country and parents may fear being reported to child services or the police.

Chapman agreed, and noted that the problem may not be confined to AEDs. Interactions with anesthesia agents are possible as well, and children receiving cannabinoid drugs tend to be very sick with frequent hospital stays and status epilepticus episodes, which are typically treated with anesthetics.

Comment