The ERUK view of the current debate around cannabis-based products as a treatment for epilepsy

The ERUK view of the current debate around cannabis-based products as a treatment for epilepsy can be found below.

To summarise the statement, whilst there is evidence that medical-grade preparations of an active ingredient of cannabis, cannabidiol, can be beneficial in some types of epilepsy, cannabis oil itself cannot be considered a safe or effective treatment. This is due to the variety of compounds found in over-the-counter cannabis oils, some of which may have harmful effects. People with epilepsy and their families should consult with their epilepsy specialist doctor for advice as to treatment options.

Why the interest in cannabis?

The idea that cannabis might be helpful for some people with epilepsy has been around for thousands of years. However, it is only in the last few decades that we have started to understand why this might be. The brain makes chemicals (endocannabinoids) which are very similar in structure to some components of cannabis. These are involved in the regulation of how “excitable” the brain is, and so can also influence the liability to seizures. Cannabis contains hundreds of different chemicals, the two most important of which are Tetrahydrocannabidiol (THC) and Cannabidiol (CBD). Both THC and CBD seem to be effective against seizures [1]. THC is the part that makes people feel “high”, but it can also cause negative effects such as anxiety and paranoia [2]. THC has also been shown in some models of epilepsy to promote seizures. There is also a small risk of the brain getting used to THC, needing more over time (tolerance) and potentially leading to addiction. Taken regularly there is also significant evidence that THC, particularly in the developing brain (up to young adulthood), can sometimes cause permanent damage to IQ, mental health and brain structure. For these reasons doctors and scientists have focussed on CBD, which doesn’t seem to carry these risks.

Clinical Trials

The first good quality evidence appeared in 2016 [3], following which several well-controlled clinical trials of a near pure formulation of CBD (Epidiolex™, made by the UK Company GWPharma) were published [4-6]. EpidiolexTM contains less than 0.1% THC.
The best trials of new drugs include a period of time when patients and their doctors don’t know if the additional drug they are taking as part of the trial contains the drug being studied or not. There are strict entry criteria so that after allocating people randomly into taking the drug or the placebo, the two groups are otherwise well matched and the effect of the drug (as opposed to other factors) can be more clearly identified. These are called randomised blinded studies, and the pill or solution without the active drug is called the placebo. At the end of the blinded-placebo phase, everyone is then switched to the active drug. This is to measure the “placebo effect”: this refers to the known fact that some people will get better when they start a new pill or solution, even if it doesn’t’ actually contain the new drug under study.
The blinded placebo controlled studies of CBD have all been done in people with two particularly severe paediatric onset, drug resistant and often life limiting epilepsies: Dravet’s syndrome and Lennox Gastaut Syndrome. Most of the participants were children, all taking other antiepileptic drugs at the same time, and having at least 8 disabling motor seizures each month. The blinded phase of the studies lasts only a few months, so has to include people with very frequent seizures to measure change.

Observational studies

There have also been studies in over 3000 people with a range of different epilepsies taking CBD, with or without some THC, over periods of months to years, but without a blinded placebo arm (known as “Open studies”) [7]. Open studies are often much less well designed, with a higher risk of more positive reports of effectiveness than might reflect the true situation. That doesn’t mean people in these studies are lying, just that epilepsy is inherently variable and will sometimes improve over time without a change in drugs, so by chance alone this might sometimes be misinterpreted as a response to a new drug. There is also a lot of evidence for epilepsy that factors such as mood and stress can influence seizure frequency. Epilepsy trials, like trials in other conditions, show evidence of a placebo effect, and the placebo effect is known to be bigger where those taking what they think is a new drug have a strong belief that it will work. There has been intense social and media attention around cannabis in recent years. Together with a belief held by many that a natural product is inherently safer and more effective than a synthetic compound, this may mean a placebo effect is more likely with cannabis-derived drugs than some other new agents. Ultimately what this means is that observational data is useful in terms of detecting safety and side effects, but much less reliable as a measure of effectiveness.

Effectiveness, side effects and short term safety

Based on the blinded trials, and the better controlled open label studies, the best current estimates are that of every 8 people with drug resistant epilepsy treated with CBD, 1 will have a greater than 50% reduction in the frequency of disabling motor seizures, but only around 1 in 171 will achieve long term seizure freedom. Around 1 in 3 will have side effects, most commonly diarrhoea, drowsiness, reduced appetite, vomiting and fatigue. These are often mild, and improve over time or with a reduction in dose. Around 1 in 23 people will have serious side effects, usually meaning the drug has to be stopped. Interactions with other antiepileptic drugs, so far in particular with valproate and clobazam have also been identified, with dose adjustments sometimes needed [7].
Given the very severe nature of the epilepsies that have been studied in the trials, even a 50% reduction in disabling seizures can make a significant difference to the individual, and for the small number who have a dramatic response without side effects this can be life changing. It is this evidence that has resulted in the American licensing authorities approving Epidiolex™ for use in two severe drug-resistant paediatric onset epilepsies (Dravet and Lennox Gastaut Syndromes). European and UK authorities are expected to follow suit in 2019.

Comparison to other licensed antiepileptic drugs

There have been no studies comparing CBD, or other cannabis-based drugs to other licensed drugs. In broad terms, based on the number of people whose seizures improve and the risk of side effects, CBD looks similar to other newer drugs. However, we have much more data about safety and effectiveness of other drugs, especially taken in the long term, than we do about CBD. The possible benefits and risks of trying CBD in patients with drug-resistant epilepsies will vary considerably between individuals. Doctors specializing in epilepsy treatment will discuss this, and other treatment options, with patients and their families to help make the best decision for the individual.

Over the counter preparations.

There are a range of cannabis oils available in health food shops, over the internet, and in countries where “medicinal cannabis” is available. These products contain both CBD and THC in varying proportions. Products with <0.2% THC are legal in the UK, but are not manufactured under the same controlled conditions as pharmaceutical drugs are. They can contain other chemicals, including pesticides and impurities, including some components of the cannabis plant linked to cancers and other long term health problems [2]. The quantities of CBD and THC may not match what they say on the label and can vary from batch to batch. Doctors and scientists do not recommend use of these products. People with severe drug-resistant epilepsies should consult with their epilepsy specialist to explore treatment options, including whether there may be clinical trials they can take part in. With the impending change in licensing for cannabis-based products, and early promising evidence, it is very likely that there will be new trials recruiting in a broader range of epilepsy types in the coming months and years.

1. O’Connell, B.K., D. Gloss, and O. Devinsky, Cannabinoids in treatment-resistant epilepsy: A review. Epilepsy & Behavior, 2017. 70: p. 341-348.
2. Board, P.E.E. Cannabis and mental health. 2017 [cited 2018 13/07/2018]; E]. Available from:
3. Devinsky, O., et al., Cannabidiol in patients with treatment-resistant epilepsy: an open-label interventional trial. Lancet Neurology, 2016. 15(3): p. 270-278.
4. Devinsky, O., et al., Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome. New England Journal of Medicine, 2017. 376(21): p. 2011-2020.
5. Devinsky, O., et al., Effect of Cannabidiol on Drop Seizures in the Lennox–Gastaut Syndrome. New England Journal of Medicine, 2018. 378(20): p. 1888-1897.
6. Thiele, E.A., et al., Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4): a randomised, double-blind, placebo-controlled phase 3 trial. Lancet, 2018. 391(10125): p. 1085-1096.
7. Stockings, E., et al., Evidence for cannabis and cannabinoids for epilepsy: a systematic review of controlled and observational evidence. Journal of Neurology, Neurosurgery & Psychiatry, 2018. 89(7): p. 741-753.


« »