Reviews/Epilepsy Management/Cannabidiol in Lennox-Gastaut syndrome
Novel Mechanisms · EP-NOV-02 · v2
Across this corpus, cannabidiol added to existing medicines is consistently linked to fewer drop seizures in people with Lennox-Gastaut syndrome. The strongest evidence comes from re-analyses of two placebo-controlled trials, which show that patients on cannabidiol gained more seizure-free days than those on placebo and that the benefit appeared within the first week [10][6]. A long-term follow-up study and a real-world chart review point the same way, with drop-seizure reductions sustained over many months [2][9]. The panel noted that the one pivotal trial abstract in the corpus states its conclusion in the title but reports no actual figures [13], and that most other studies had no comparison group. Because the placebo-controlled signal is clear and every other source agrees on direction, the panel treated the basic claim that cannabidiol reduces drop seizures as well supported, while being careful not to over-read the exact size of the effect.
Post hoc analyses of the two phase 3 LGS trials show cannabidiol produces more drop-seizure-free days than placebo, with least-squares mean differences of 2.83 (95% CI .98-4.68) and 3.64 (95% CI 2.18-5.10) days per 28 days [10], and drop-seizure separation from placebo becoming nominally significant by Day 6 (p = .008) [6]. The open-label extension reports median drop-seizure frequency reductions of 48% to 60% through week 48 [2]. An independent real-world cohort without concomitant clobazam reports ≥50% drop-seizure reduction in 19% to 30% of LGS patients [9]. The pivotal GWPCARE4 RCT [13] names this result in its title. Direction is consistent across RCT-derived and real-world sources. The contested point: The division is not over direction but over how directly the corpus demonstrates the primary placebo-controlled effect: the pivotal GWPCARE4 abstract [13] carries no effect size, confidence interval or p-value, so the causal signal rests on post hoc analyses [10][6] rather than a primary endpoint in the corpus. Most remaining studies ([1][2][5][7][8][9][11][12][14][15][16]) are uncontrolled, and several report motor, major-motor, total or tonic-clonic seizures rather than adjudicated drop seizures ([1][15][12][16]).
All 17 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Epilepsy Management, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The pivotal GWPCARE4 RCT [13] states the drop-seizure conclusion only in its title; the corpus abstract body reports no effect size, CI or p-value, so its primary numbers cannot be weighed directly — the placebo-controlled signal rests on post-hoc analyses [6] and [10].
Most included papers ([1], [2], [5], [7], [8], [9], [11], [12], [14], [15], [16]) are open-label, single-arm, retrospective or caregiver-reported with no placebo control, leaving them exposed to regression-to-the-mean, expectation effects and diary-reporting bias.
Where it remains uncertain. The pivotal GWPCARE4 abstract reports no numbers [13], so the placebo-controlled effect depends on post hoc analyses [10][6], while most other included studies are uncontrolled and some measure non-drop seizure types ([1][15][12][16]).
Aleatory
Several real-world cohorts and EAP series carry modest or shrinking denominators (e.g. drop-seizure responder analysis at n = 69 then n = 53 [9]; UK series n = 26 [15]), so point estimates are subject to sampling variability.
Epistemic
The GWPCARE4 abstract [13] reports no effect size, CI or p-value, so the primary placebo-controlled drop-seizure result cannot be weighed directly from the corpus and must be inferred from post hoc analyses [10][6].
Inferential
Several strong-relevance papers report major-motor, motor, total or generalised tonic-clonic seizures ([1][15][12][16]) rather than adjudicated drop seizures, so applying their reductions to the exact claim requires an inferential step.
The load-bearing papers from a corpus of 17, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Shows sustained, directly measured drop-seizure reductions in LGS over 48 weeks (open-label extension).
“Median reduction from baseline in drop seizure frequency (quantified monthly over 12-week periods) ranged from 48% to 60% through week 48.”
Panel auditInternal validity: low · scope of inference: Long-term uncontrolled drop-seizure reduction in LGS patients who continued CBD from two prior RCTs.Key limitations: open-label, no control · interim analysis · selected RCT completers
Confirms placebo-controlled drop-seizure separation in the pivotal LGS RCTs.
“Differences in drop seizure reduction between placebo and CBD emerged during the titration period and became nominally significant by Day 6 (p = .008) for pooled CBD treatment groups.”
Panel auditInternal validity: high · scope of inference: Timing of drop-seizure treatment effect of CBD vs placebo in LGS.Key limitations: post hoc analysis · 'nominally' significant · focused on time-to-onset not primary endpoint
Independent real-world drop-seizure responder data in LGS, and without clobazam.
“Achievement of ≥50% reduction in drop (LGS) or convulsive (DS) seizures at 3 and 12 months: LGS, 19% (n = 69) and 30% (n = 53)”
Panel auditInternal validity: low · scope of inference: Real-world drop-seizure reduction and retention of CBD without clobazam in LGS.Key limitations: retrospective, no control · responder rates are modest · denominator attrition over time
The abstract reports major motor and total seizure reductions in a pooled LGS/DS cohort, never a drop-seizure outcome isolated to LGS, so the direction for this specific claim is not determinable.
Panel auditInternal validity: low · scope of inference: Long-term tolerability and uncontrolled seizure reduction in treatment-resistant LGS/DS patients on add-on CBD.Key limitations: no control group · LGS and DS pooled · outcome is major motor/total, not drop seizures
4 of 17 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Convergent placebo-controlled RCT-derived drop-seizure data and consistent real-world cohorts carry the directional claim at the Established tier, trimmed slightly because the pivotal trial's primary numbers appear only via post-hoc analysis and a numberless abstract.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Primary placebo-controlled drop-seizure effect size in the corpus. The pivotal GWPCARE4 result exists only as a title-level claim [13]; the primary endpoint numbers are not weighable from the corpus and are carried by post hoc analyses [10][6]. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Include the full GWPCARE3/GWPCARE4 primary-endpoint publications so the placebo-controlled drop-seizure reduction and its CI can be read directly rather than reconstructed.
Open question: What is the primary placebo-controlled drop-seizure effect size and confidence interval from the pivotal LGS trials? Why it is still open: The only pivotal RCT abstract in the corpus [13] reports no figures, and the effect is inferred from post hoc analyses [10][6].
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19