Randomized controlled trial (Phase 3, double-blind, placebo-controlled)Strong — gold-standard RCT design published in The Lancet2018

The Second Confirmation: CBD Works for Lennox-Gastaut Seizures Too

Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4): a randomised, double-blind, placebo-controlled phase 3 trial

Thiele EA, Marsh ED, French JA, Mazurkiewicz-Beldzinska M, Benbadis SR, et al.·Lancet·PubMed

A landmark RCT in The Lancet confirmed CBD reduces the most dangerous seizure type (drop seizures) in Lennox-Gastaut syndrome by 43.9% vs 21.8% placebo — proving CBD's anticonvulsant effect isn't limited to Dravet and completing the evidence for Epidiolex's FDA approval.

If Dravet syndrome is the disease that launched the CBD epilepsy movement, Lennox-Gastaut syndrome is the one that confirmed it wasn't a fluke.

LGS is rarer than Dravet, harder to treat, and in many ways more devastating. It doesn't just cause seizures — it causes multiple seizure types simultaneously, including "drop seizures" that send children crashing to the ground without warning. More than 90% of patients develop moderate to severe intellectual disability. Five percent die in childhood. At least 90% of adults with LGS still have seizures decades later.

The Devinsky Dravet trial had proven CBD could reduce seizures in one catastrophic childhood epilepsy. The question was whether it would work in a second, pharmacologically distinct form. If CBD's anticonvulsant mechanism was specific to Dravet's sodium channel dysfunction, it might not generalize. If it was broader — acting through mechanisms relevant to multiple seizure types — it would work here too.

It worked here too.

The Disease

The Trial

The Results

43.9% vs. 21.8%

median reduction in monthly drop seizure frequency — CBD versus placebo. The between-group difference was 17.21 percentage points (95% CI: -30.32 to -4.09, p = 0.0135). In a patient population where nothing else had provided adequate control, CBD produced a clinically meaningful reduction in the most dangerous seizure type.

The placebo response (21.8%) is notable — placebo effects in epilepsy trials are well-documented and substantial. This is exactly why the Dravet trial used a double-blind design and why uncontrolled anecdotes about CBD for seizures, however compelling, can't substitute for RCTs.

Thiele et al. (2018), Lancet 391:1085-1096

Why This Trial Mattered Beyond Dravet

The Dravet result could have been a one-off. Dravet syndrome involves a specific sodium channel mutation (SCN1A). Maybe CBD's anticonvulsant mechanism was specific to that ion channel defect. If so, it wouldn't generalize.

Lennox-Gastaut syndrome has completely different and heterogeneous underlying causes. The fact that CBD worked here too meant its anticonvulsant mechanism was broader than any single ion channel. Whatever CBD does to reduce seizures, it works across different types of epilepsy with different etiologies.

This had two implications:

The Safety Profile

Myth vs. Reality

✕Myth

CBD is a natural supplement with no significant side effects.

✓Reality

At therapeutic doses for epilepsy (20 mg/kg/day), CBD has a real drug safety profile. In this trial, 86% of patients experienced adverse events. The most common were diarrhea, somnolence, fever, decreased appetite, and vomiting. Elevated liver enzymes occurred, particularly in patients taking valproate concurrently — requiring monitoring. 14% of patients discontinued due to adverse events. One patient died (deemed unrelated). CBD at these doses is a pharmaceutical with pharmaceutical risks — not a gentle supplement.

The Evidence

Thiele et al. (2018): 86% adverse event rate, 14% discontinuation, elevated liver enzymes requiring monitoring, one death. Devinsky et al. (2017): 93% adverse event rate in the CBD group of the Dravet trial.

Thiele et al. (2018), Lancet 391:1085-1096; Devinsky et al. (2017), NEJM 376:2011-2020

The liver enzyme elevation is worth understanding. CBD inhibits cytochrome P450 enzymes (particularly CYP3A4 and CYP2C19), which are the same enzymes that metabolize many anti-epileptic drugs. When CBD is added to valproate — a standard first-line drug for LGS — the combination can elevate liver transaminases. This isn't a reason to avoid CBD, but it's a reason it needs to be prescribed and monitored by a neurologist, not self-administered from a dispensary product.

The Path to Dual Approval

Together with the Dravet trial, this study completed the evidence package that the FDA required. In June 2018, Epidiolex was approved for both Dravet syndrome and Lennox-Gastaut syndrome in patients aged 2 and older — the first cannabis-derived drug ever approved by the FDA.

The approval was notable for what it wasn't: it wasn't for "epilepsy" broadly. It was for two specific, catastrophic, treatment-resistant syndromes where rigorous RCTs had demonstrated efficacy. The FDA approved the evidence, not the enthusiasm.

Frequently Asked Questions

Lennox-Gastaut syndrome (LGS) is a severe form of epilepsy that typically begins between ages 3-5. It's characterized by multiple seizure types (especially "drop seizures" that cause sudden falls), a distinctive EEG pattern, and progressive cognitive impairment. More than 90% of patients develop intellectual disability. At least 90% still have seizures as adults. It's caused by diverse factors including brain malformations, injuries, and genetic conditions — unlike Dravet syndrome, which has one primary genetic cause. Treatment options before Epidiolex were limited and often inadequate.

Different disease, different seizure type measured, same result. The Dravet trial measured convulsive seizures in patients with a sodium channel mutation. This trial measured drop seizures in patients with heterogeneous causes. The fact that CBD worked in both — reducing seizures by ~39-44% versus placebo — proved its anticonvulsant mechanism is broad, not disease-specific. Both trials used the same drug (Epidiolex), same dose (20 mg/kg/day), and same gold-standard design (double-blind, placebo-controlled).

Placebo responses in epilepsy trials are notoriously high — typically 10-25%. Several factors contribute: natural fluctuation in seizure frequency (regression to the mean), the psychological effect of receiving treatment, increased compliance with existing medications during a trial, and caregiver reporting bias. This is exactly why placebo-controlled trials are essential for evaluating seizure medications. Without the placebo comparison, the 43.9% reduction in the CBD group would look impressive — but you'd have no way to know how much was drug effect versus context effect.

Epilepsy CBD clinical-trial lennox-gastaut epidiolex

Cite this study

Thiele EA, Marsh ED, French JA, Mazurkiewicz-Beldzinska M, Benbadis SR, et al.. (2018). Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4): a randomised, double-blind, placebo-controlled phase 3 trial. Lancet. https://doi.org/10.1016/S0140-6736(18)30136-3

View on PubMed

More from these researchers

Thiele EA(2)Marsh ED(1)French JA(1)