Key messages

• It is not clear whether cannabinoids reduce the frequency of seizures because the certainty of the evidence is very low.
• Cannabinoids are probably associated with transient but frequent adverse effects.

What is the evidence.
See evidence matrix  in Epistemonikos later

We found five systematic reviews [1],[2],[3],[4],[5] including six primary studies [6],[7],[8],[9],[10],[11], of which four [6],[7],[8],[9] correspond to randomized controlled trials.This table and the summary in general are based on the latter.

What types of patients were included

All trials [6],[7],[8],[9] included patients with epilepsy, however only one trial [9] specified the type of epilepsy (secondarily generalized).
Two trials [8],[9] included refractory epilepsy. Only one trial [9] reported seizure frequency (one per week) and one trial [7] included patients with mental retardation.
No trial reported the age of the participants, but two trials mentioned adult patients were included [8],[9].

What types of interventions were included

All trials [6],[7],[8],[9] compared oral cannabidiol in different doses and different periods against placebo.
The regime used was 200 to 300 mg of cannabidiol daily for 18 weeks. In one trial [9], 100 mg cannabidiol daily for one week, and then 200 mg daily for another three weeks in one trial [7], 200 mg of cannabidiol for three months in one trial [8] and one trial used placebo for six months and then 300 mg of cannabidiol [6].

What types of outcomes
were measured

The systematic reviews identified [1],[2],[3],[4],[5] grouped outcomes in the following way:
Seizure frequency in a specific period (the duration of the intervention), and adverse effects associated with the use of cannabidiol.
One of the reviews [1] states none of the trials reported seizure-free period for 12 months.

To whom this evidence does and does not apply

• The evidence presented in this summary applies to all patients with epilepsy, especially those refractory to treatment with antiepileptics.
• The trials analyzed include oral cannabidiol in different doses. They do not address the effect of smoked or vaporized cannabis, neither the effect of tetrahydrocannabinol. One of the reviews [4] advises against the use of smoked tetrahydrocannabinol, because of its possible proconvulsant effect.

• The outcomes selected for this summary are those critical for decision-making according to the opinion of the authors. They coincide with those presented in the systematic reviews identified.
• Although the trials analyzed report minimal adverse effects associated with the use of cannabidiol, higher quality evidence from other conditions [12] reported more frequent adverse effects. There are no good reasons for not expecting a similar rate of adverse effects in this population.

• The evidence about benefits is of very low certainty, and adverse effects are frequent. The benefit/risk balance is probably unfavorable.

• Most patients and doctors should lean against the use of this intervention based on the existing evidence.
• However, some patients putting higher value in an uncertain benefit might consider its use, especially in the context of preconceptions they might have about the particular.

• Commercial formulations of cannabinoids are generally expensive. As there is no certainty about a possible benefit, it is not possible to estimate a proper cost/benefit balance.

Differences between this summary and other sources

• The key messages of this summary are consistent with the findings of the systematic reviews identified.
• This summary also coincides with the Position on Medical Marijuana of the American Epilepsy Society [13], it does not recommend its use for the control of epilepsy, because safety and efficacy of this intervention are unknown. However, it strongly supports clinical research to determine its real value.

• The probability that future evidence change the conclusions of this summary about the benefits of cannabinoids in epilepsy is high, because there is high uncertainty. Regarding adverse effects, the probability is low.
• There are several ongoing trials, according to the International Clinical Trials Registry Platform of the World Health Organization, which could provide relevant information in the future.