CBD and epilepsy: Epidyolex, the only authorised medicine

Epidyolex is an oral solution of highly purified cannabidiol, authorised by the European Commission on the opinion of the European Medicines Agency. The EMA's EPAR page sets out the indication: adjunctive therapy of seizures associated with Lennox-Gastaut syndrome or Dravet syndrome, in combination with clobazam, in patients aged 2 years and above. An extension of indication covers tuberous sclerosis complex, likewise from 2 years, in combination with other antiepileptics.

Three elements are worth retaining. The indication covers rare, severe and drug-resistant epilepsies, not epilepsy in general. It is prescribed in combination with another antiepileptic, not as a replacement. And the substance is a purified pharmaceutical cannabidiol, dosed and controlled to the standards of a medicine.

In France, the Haute Autorité de santé has issued its transparency opinions on this medicine: substantial actual benefit and a minor improvement in actual benefit, at level IV, in Lennox-Gastaut syndrome and Dravet syndrome, in an opinion of 13 May 2020. The medicine is on list I, its prescription is restricted to specialists and to departments of neurology and paediatric neurology, and it is reimbursed at 65 %. The commission also flagged a risk of dosing error linked to the packaging: the posology is expressed in milligrams, but the administration devices are graduated in millilitres.

Two trials published in the New England Journal of Medicine underpinned the authorisation.

The first, Devinsky et al., 2017, randomised 120 children and young adults with Dravet syndrome, to 20 mg per kilogram per day of cannabidiol or a placebo, on top of their usual antiepileptic regimen. The median frequency of convulsive seizures went from 12.4 to 5.9 per month on cannabidiol, against 14.9 to 14.1 on placebo. The proportion of patients with at least a 50 % reduction in seizures was 43 % on cannabidiol against 27 % on placebo, a difference that did not reach the significance threshold used.

The second, Devinsky et al., 2018, enrolled 225 patients aged 2 to 55 with Lennox-Gastaut syndrome. The median reduction in drop seizures was 41.9 % at 20 mg per kilogram and 37.2 % at 10 mg per kilogram, against 17.2 % on placebo.

These results are real and clinically useful for patients who have run out of options. They are also modest in absolute terms: the gap against placebo is 20 to 25 points on the median reduction in seizures in Lennox-Gastaut syndrome, and 16 points on the proportion of responders in Dravet syndrome, because the control group improves too in this kind of protocol.

Most mainstream content glosses over this point. A child weighing 20 kilos receiving 20 mg per kilogram per day receives 400 mg of cannabidiol a day. An adult of 70 kilos on the same posology would be at 1 400 mg a day.

By way of comparison, EFSA has adopted a far lower provisional daily reference value for dietary exposure, detailed in our article on the daily CBD dose. The posologies used in the trials therefore belong to a prescribing setting with laboratory monitoring, and have no equivalent in consumer use. Transposing a trial conclusion to an over-the-counter product is an error of reasoning.

The summary of product characteristics puts figures on the most frequent adverse effects at doses of 10 to 25 mg per kilogram per day: somnolence in 23 % of patients, decreased appetite in 21 %, diarrhoea in 20 %, pyrexia in 16 %, vomiting in 12 %.

The liver signal is the most important. A rise in transaminases beyond three times the upper limit of normal was observed in 12 % of patients on cannabidiol, against under 1 % on placebo. That proportion depends heavily on the medicines taken alongside: 23 % where valproate and clobazam are combined, 19 % with valproate alone, 3 % with clobazam alone and 3 % with neither. It also depends on the dose: 15 % at 20 or 25 mg per kilogram, against 3 % at 10 mg. A check on transaminases and bilirubin is required at one, three and six months.

Cannabidiol interacts with several cytochrome P450 enzymes, which alters the blood concentration of other medicines, starting with clobazam. Through inhibition of CYP2C19, it multiplies the levels of its active metabolite three- to fourfold; somnolence then concerns 43 % of patients, against 14 % without clobazam. These mechanisms are set out in our articles on CBD and liver enzymes and on the grapefruit effect. Anyone on a course of medicines should raise it with their doctor or pharmacist before adding a product containing CBD.

The boundary is clear and deliberately strict. A medicine has a marketing authorisation, a clinical dossier, pharmacovigilance and a prescription. A consumer CBD product has none of that, and can claim no effect on epilepsy nor on any other condition: doing so would be enough to reclassify it as a medicinal product by presentation.

The ANSM has devoted an alert to this very subject, calling on patients with epilepsy and the parents of children with epilepsy not to use cannabidiol products sold outside the legal circuit. The agency notes that these products are sometimes accompanied by claims of a health benefit aimed in particular at epilepsy. It also notes that CBD can be toxic to the liver and can raise the concentration of other medicines, antiepileptics among them, and that neither the actual quantity of CBD nor the presence of other compounds is guaranteed outside the legal circuit.

That separation works in the other direction too: the existence of Epidyolex does not make an over-the-counter oil comparable to a prescribed course, and access to cannabis for medicinal use falls under a distinct scheme, described in our review of the general roll-out of cannabis for medicinal use in France. On what the other published trials show, and do not show, our summary of the clinical studies on CBD takes stock.