Epidiolex (Cannabidiol)

Plain-language summary

Epidiolex is a prescription medicine made of highly purified cannabidiol (CBD) extracted from the cannabis plant. It was approved by the U.S. Food and Drug Administration in June 2018 — the first cannabis-derived drug to clear that bar ^[1]. It is approved for seizures associated with three conditions: Dravet syndrome, Lennox-Gastaut syndrome (LGS), and tuberous sclerosis complex (TSC), in patients one year of age and older ^[1][2].

It is taken by mouth as an oil. Doses used in trials and clinical practice (10-20 mg/kg/day) are far higher than the amounts in over-the-counter CBD products — a 30 kg child on 15 mg/kg/day takes 450 mg of CBD daily, more than most retail bottles contain in total.

This article is not medical advice. Epidiolex is a prescription drug with real interactions and side effects. Decisions about starting, stopping, or dosing it should be made with a neurologist or epileptologist.

What probably works (strong evidence)

Three indications are supported by randomized, double-blind, placebo-controlled Phase III trials:

• Dravet syndrome. In a pivotal trial, Epidiolex at 20 mg/kg/day reduced convulsive seizure frequency by a median of 38.9%, versus 13.3% for placebo ^[3]. Strong evidence
• Lennox-Gastaut syndrome. Two Phase III trials showed reductions in drop seizure frequency of roughly 40-44% at 20 mg/kg/day, versus 17-22% on placebo ^[4][5]. Strong evidence
• Tuberous sclerosis complex. A Phase III trial showed a 48.6% reduction in TSC-associated seizures at 25 mg/kg/day, versus 26.5% for placebo ^[6]. Strong evidence

These are meaningful effects in patients who have typically failed multiple other antiseizure drugs. They are not cures — most patients still have seizures — but the reductions are clinically significant.

What might work (weak or emerging evidence)

Off-label use of Epidiolex or purified CBD has been explored for:

• Other treatment-resistant epilepsies (e.g., CDKL5 deficiency disorder, Aicardi syndrome, Doose syndrome). Open-label expanded-access data suggest some benefit, but without placebo controls the effect size is uncertain ^[7]. Weak / limited
• Schizophrenia adjunct therapy. A small RCT of 1000 mg/day CBD (not Epidiolex specifically) showed modest symptom improvement ^[8]. Not replicated at scale. Weak / limited
• Anxiety disorders. Some small trials of high-dose oral CBD show acute anxiolytic effects in induced-anxiety paradigms, but evidence for chronic clinical anxiety remains preliminary ^[9]. Weak / limited

None of these are FDA-approved uses.

What doesn't work or has weak evidence

Several popular claims about CBD do not transfer to Epidiolex-style evidence:

• Low-dose 'wellness' CBD for seizures. The doses that work in trials are 10-25 mg/kg/day. A 70 kg adult would need 700-1750 mg/day. A typical 25 mg gummy is not a substitute. No data
• CBD for chronic pain as monotherapy. Systematic reviews find inconsistent or null effects for pure CBD in chronic pain at studied doses ^[10]. Weak / limited
• CBD for sleep. Limited and mixed trial data; effects may reflect anxiety reduction rather than direct hypnotic action. Weak / limited
• CBD 'cures' cancer, autism, Alzheimer's. No Phase III human evidence supports these claims. No data

What we don't know

• Long-term safety beyond ~4 years. Open-label extensions look reasonable, but multi-decade pediatric data don't exist yet.
• Mechanism. CBD's antiseizure mechanism is not fully understood. It is not primarily a CB1/CB2 agonist; proposed targets include GPR55, TRPV1, and adenosine reuptake ^[11]. Disputed
• Who responds. No reliable biomarker predicts which patients will benefit.
• Whether the entourage effect matters here. Purified CBD works in these syndromes. Whether whole-plant extracts work better, worse, or the same is not established by head-to-head RCTs.

Comparison with standard treatments

For Dravet syndrome, first-line agents include valproate and clobazam; stiripentol, fenfluramine, and Epidiolex are added for refractory cases. Fenfluramine showed larger seizure reductions (~62-74% vs placebo ~9%) in its trials, though direct head-to-head comparisons with Epidiolex are limited ^[12].

For LGS, standard options include valproate, lamotrigine, rufinamide, clobazam, topiramate, and the ketogenic diet. Epidiolex is typically added when seizures remain uncontrolled.

For TSC, everolimus (an mTOR inhibitor) is also approved for associated seizures and addresses the underlying biology more directly; Epidiolex is an alternative or adjunct.

In practice, neurologists layer these drugs. Epidiolex is rarely a first choice and usually added to an existing regimen.

Risks, side effects, and interactions

Common adverse effects in trials ^[3][4][5][6]:

• Somnolence and sedation (especially with concomitant clobazam)
• Decreased appetite
• Diarrhea
• Fatigue
• Elevated liver transaminases — dose-dependent, more common with concomitant valproate. The label requires baseline and periodic liver function testing ^[2]. Strong evidence

Drug interactions that matter:

• Clobazam. CBD inhibits CYP2C19, raising levels of N-desmethylclobazam (an active metabolite) several-fold. Much of the sedation — and possibly some of the seizure benefit — in trials reflects this interaction ^[13]. Strong evidence
• Valproate. Combination increases risk of hepatotoxicity.
• CYP3A4 and CYP2C19 substrates broadly: CBD can shift levels of many co-administered drugs.

Withdrawal: As with any antiseizure medication, abrupt discontinuation can precipitate seizures or status epilepticus. Taper under medical supervision.

Not medical advice. This article summarizes published evidence. It is not a substitute for evaluation by a qualified clinician. Do not start, stop, or change doses of Epidiolex — or substitute over-the-counter CBD for it — without consulting a neurologist.

Sources

How this page was made