Medical Cannabis for Epilepsy: Epidyolex, Unlicensed Medicines and What NICE Says

In September 2019, the National Institute for Health and Care Excellence issued Technology Appraisal Guidance 614 — a document that, in its quiet regulatory language, marked a significant turning point. Epidyolex, a pharmaceutical-grade oral solution of cannabidiol manufactured by GW Pharma, became the first cannabis-derived medicine to receive a positive NICE recommendation for use on the NHS. For the families who had spent years importing unlicensed cannabis oils across borders, fighting clinical panels, and submitting parliamentary petitions, it was partial — but real — vindication.

Epilepsy had always been the condition most likely to force the UK’s hand on medical cannabis. The political pressure was not abstract: Billy Caldwell, a boy from County Fermanagh whose mother Charlotte purchased CBD oil in Canada and had it confiscated at Heathrow in 2018, collapsed into life-threatening seizures within days of his supply being cut. The Home Secretary at the time, Sajid Javid, personally intervened to secure emergency access. Six months later, cannabis-based medicinal products were rescheduled from Schedule 1 to Schedule 2 under the Misuse of Drugs Regulations — legally available, for the first time, on prescription.

That decision, in November 2018, was the regulatory breakthrough. What followed was considerably more complicated.

Epidyolex: Britain’s First NICE-Approved Cannabis Medicine

Epidyolex is not cannabis in any conventional sense. It contains no THC — the psychoactive compound associated with recreational use — and is delivered as a precisely dosed oral solution standardised to 100mg/ml of cannabidiol. It is manufactured under pharmaceutical-grade conditions, subject to MHRA licensing, and prescribed within clearly defined clinical criteria.

NICE approved Epidyolex in October 2019 as an adjunct therapy for seizures associated with two rare and severe childhood epilepsy syndromes: Dravet syndrome and Lennox-Gastaut syndrome, in patients aged two years and over. A subsequent 2021 guidance update extended recommendations to Tuberous Sclerosis Complex (TSC), following the GWPCARE6 trial results.

The clinical bar for approval was high. GW Pharma’s GWPCARE trial programme — six randomised, placebo-controlled studies — provided the evidence base. GWPCARE1 and GWPCARE2, published in the New England Journal of Medicine in 2017 and 2018, demonstrated statistically significant reductions in convulsive seizure frequency in Dravet syndrome patients: 39% median reduction with cannabidiol versus 13% with placebo in the landmark Devinsky et al. 2017 study. GWPCARE3 and GWPCARE4 addressed Lennox-Gastaut syndrome, showing comparable efficacy on drop seizures. The data was, by the standards of a notoriously treatment-resistant patient population, persuasive.

NHS England commissioning followed the NICE guidance. Epidyolex is now available through specialist paediatric neurology services, subject to prescribing criteria. The annual treatment cost — approximately £10,000 to £25,000 per patient depending on dose and weight — is covered for eligible patients, though the funding pathway involves NHS England’s Highly Specialised Technology evaluation processes.

Beyond Epidyolex — Unlicensed Medical Cannabis for Treatment-Resistant Epilepsy

Epidyolex covers a narrow population. For the broader group of patients with treatment-resistant epilepsy who do not have Dravet, Lennox-Gastaut or TSC, access to cannabis-based medicines remains through unlicensed channels — a pathway that exists, but is encumbered with obstacles that make routine clinical use exceptional rather than standard.

Since the November 2018 rescheduling, specialist doctors — those on the specialist register — may legally prescribe unlicensed cannabis-based medicinal products. These include full-spectrum cannabis flower and oil preparations containing THC, CBD or combinations of both. They are dispensed through a small number of specialist pharmacies and, increasingly, through dedicated medical cannabis clinics that have established supply chains from licensed importers.

The clinical rationale for considering THC-containing preparations in epilepsy is scientifically credible, even if the evidence base is thinner than for Epidyolex. THC acts on CB1 receptors in the brain in ways that may modulate neuronal excitability; some clinicians and patients report meaningful seizure reduction with preparations that Epidyolex, as a pure-CBD product, does not provide. However, the absence of large-scale randomised trial data for THC in epilepsy creates a regulatory and clinical vacuum that individual prescribers navigate differently.

For paediatric patients specifically, prescribing THC-containing preparations requires a level of clinical confidence and regulatory comfort that most NHS paediatric neurologists currently lack. The result is a two-tier landscape: Epidyolex-eligible patients receiving NHS-funded, evidence-based treatment; everyone else in a private, expensive, and clinically inconsistent grey zone.

Dravet Syndrome and Lennox-Gastaut: The Evidence Base

Dravet syndrome is a rare genetic epilepsy, typically caused by mutations in the SCN1A gene, characterised by prolonged febrile seizures in infancy and evolving into a lifelong, drug-resistant condition. Patients frequently fail multiple antiepileptic drugs. Stiripentol, clobazam and valproate have partial efficacy; fenfluramine, approved by MHRA in 2020, offers additional benefit in some cases. Cannabidiol fits into this algorithm as an adjunct, not a replacement.

The Devinsky et al. New England Journal of Medicine trial, published in May 2017, enrolled 120 patients with Dravet syndrome. Those receiving cannabidiol (20mg/kg/day) had a median reduction in convulsive seizures of 38.9% versus 13.3% in the placebo group. The responder rate — patients with at least a 50% reduction in seizures — was 43% with cannabidiol versus 27% with placebo. Adverse events, including somnolence, diarrhoea and elevated liver enzymes (particularly in patients also taking valproate), were more common in the active group.

For Lennox-Gastaut syndrome, a similarly refractory condition characterised by multiple seizure types and cognitive impairment, the GWPCARE3 and GWPCARE4 trials showed reductions in drop seizure frequency of 41.9% with high-dose cannabidiol (20mg/kg/day) and 37.2% with low-dose (10mg/kg/day), compared with 17.2% for placebo. These figures, published in The Lancet in 2018, provided the clinical data on which NICE based its positive recommendation.

Real-world evidence has broadly supported these trial findings. The UK Medical Cannabis Registry — operated by Drug Science, a scientific research organisation founded by Professor David Nutt — has collected observational data from several hundred epilepsy patients accessing unlicensed products. A 2021 analysis of 88 patients with treatment-resistant epilepsy showed statistically significant improvements in seizure frequency and patient-reported quality of life outcomes at three and six months. The limitations are those inherent to observational data — no control arm, self-selection bias, variable product consistency — but the directional consistency with trial evidence is notable.

How UK Paediatric Neurologists Are Approaching Cannabis Prescriptions

The British Paediatric Neurology Association issued a position statement in 2019, updated subsequently, that has shaped clinical practice cautiously. The BPNA position endorses the use of Epidyolex within its licensed indications, while expressing significant reservations about the broader prescribing of unlicensed cannabis-based medicines for children with epilepsy outside of clinical trial settings.

The BPNA’s concerns are not ideological but practical. Unlicensed cannabis preparations vary in cannabinoid content, purity and consistency in ways that pharmaceutical products do not. A preparation purchased through a private medical cannabis clinic may contain the advertised quantity of cannabidiol — or it may not. Batch-to-batch variation in imported flower products adds further uncertainty. For a child with severe, medically complex epilepsy, this inconsistency is clinically meaningful: a change in cannabinoid dose, even unintentional, can trigger breakthrough seizures.

Within NHS paediatric neurology, the practical outcome is that Epidyolex prescriptions happen within the system; everything else is largely privatised. Many paediatric neurologists decline to prescribe unlicensed cannabis preparations, directing families toward private clinics — a referral that comes with costs of £150 to £400 per consultation and ongoing medicine costs that the NHS does not cover.

A minority of specialist centres — Great Ormond Street Hospital among them — have established more structured approaches to evaluating individual patients for whom cannabis-based medicines outside the Epidyolex indication may be appropriate. But these are exceptions. The BPNA’s guidance, while scientifically defensible, has in practice made NHS access to anything beyond Epidyolex extremely rare.

The Access Problem: Why Getting a Cannabis Prescription for Epilepsy Is Still Difficult

Seven years after the 2018 rescheduling, the gap between legal availability and practical access remains substantial. Multiple structural barriers persist.

First, the NHS prescribing barrier. General practitioners cannot prescribe cannabis-based medicines; only specialists on the specialist register may do so. For epilepsy, this means paediatric or adult neurologists. Most NHS neurologists, conscious of BPNA guidance and NHS England commissioning restrictions, will prescribe Epidyolex within its licensed scope and nothing further. Patients seeking THC-containing preparations or cannabidiol outside the licensed indications are directed, in effect, to the private sector.

Second, cost. A private cannabis prescription for epilepsy typically costs £400 to £600 per month in medicine alone, plus clinic fees. Medical cannabis is not reimbursable on the NHS for indications outside NICE guidance, and private health insurers routinely exclude it. The financial burden falls entirely on families, many of whom are already navigating the substantial costs associated with caring for a severely disabled child.

Third, the pharmacy network. Not all pharmacies dispense cannabis-based medicines. A specialist dispensing pharmacy with appropriate Schedule 2 storage and import licences is required. The network is growing — from fewer than 20 in 2019 to several hundred today — but geographical distribution is uneven and waiting times for product availability can run to weeks.

Fourth, stigma. Families report that GPs respond to private cannabis prescriptions with scepticism, reluctance to co-prescribe or monitor, and occasional active discouragement. For parents who believe cannabis-based medicines are helping their child, navigating a medical establishment that treats their treatment choice as peripheral to mainstream care is exhausting.

What the Data Shows — Clinical Trials, NICE Reviews, Real-World Evidence

The evidence landscape for cannabis in epilepsy is, by the standards of the field, relatively robust — particularly for Epidyolex within its licensed indications. The Cochrane Review on cannabinoids for epilepsy, last updated in 2018, found moderate-quality evidence that cannabidiol reduces seizure frequency in Dravet syndrome and Lennox-Gastaut syndrome, while noting that evidence for other epilepsy types is insufficient to draw conclusions.

The NICE 2019 evidence review was thorough. Technology Appraisal TA614 examined the GWPCARE programme data, health economic modelling (including quality-adjusted life year calculations), and patient and carer testimony. The committee concluded that, for eligible patients, Epidyolex provided clinically meaningful benefit at a cost the NHS could accept relative to the existing standard of care.

For unlicensed preparations, the evidentiary picture is messier. The Drug Science UK Medical Cannabis Registry provides the most substantial UK real-world dataset, but observational registries cannot substitute for randomised controlled trials in terms of regulatory credibility. BPNA and NHS England both note that the heterogeneity of unlicensed products — ranging from imported pharmaceutical-grade flower to artisanal CBD oils — makes population-level conclusions difficult.

There are ongoing trials. The MHRA has granted clinical trial authorisation for studies examining THC:CBD combinations in paediatric epilepsy; results from several are expected in the next two to three years. Until those data mature, the evidence gap for non-Epidyolex cannabis medicines in epilepsy will remain a legitimate clinical concern rather than merely a bureaucratic obstruction.

Risks, Monitoring and Long-Term Safety

Cannabidiol is not without adverse effects. The GWPCARE trials documented somnolence (in up to 36% of patients), decreased appetite, diarrhoea and elevated hepatic transaminase levels. The liver enzyme elevations — potentially clinically significant in a minority of patients — are most common when Epidyolex is co-administered with sodium valproate. NICE guidance and the prescribing information both mandate baseline liver function testing and monitoring at regular intervals, with dose reduction or discontinuation if transaminases rise significantly.

Drug-drug interactions are a particular concern in a patient population already taking multiple antiepileptic drugs. Cannabidiol inhibits CYP2C19 — the enzyme responsible for metabolising several common antiepileptics including clobazam, stiripentol and to some extent phenobarbital. This can raise plasma levels of co-administered drugs, occasionally to toxic concentrations. Clinical monitoring and dose adjustment of existing medications is essential when initiating Epidyolex.

For unlicensed THC-containing preparations, the risk profile is less well characterised in paediatric epilepsy specifically. THC has known psychoactive effects that are dose-dependent and potentially problematic in children with existing neurodevelopmental vulnerabilities. Long-term effects of THC on the developing brain — while the subject of ongoing research — have not been definitively characterised in this clinical context. This uncertainty underpins the BPNA’s caution and, in the view of many paediatric neurologists, represents a legitimate reason to restrict THC-containing prescriptions in children to exceptional cases with robust clinical justification and monitoring.

For adult patients with treatment-resistant epilepsy, the benefit-risk calculation may shift. Adults whose seizures remain uncontrolled despite multiple antiepileptic drugs — a group in which quality of life is profoundly impaired and medication side-effect burden is already high — may have more to gain from a trial of cannabis-based medicine and more tolerance of its risk profile. Clinical experience from private cannabis clinics suggests that some adult epilepsy patients report meaningful improvement; systematic evidence in this population specifically remains underdeveloped.

Parents, Patients and Campaigners: How Families Fought for Access

The political history of medical cannabis in the UK is, more than in most countries, a story of parental advocacy. Charlotte Caldwell and Hannah Deacon — whose son Alfie became the second high-profile case in 2018, with a different clinical picture but equally compelling urgency — demonstrated that individual cases could, under sufficient media and political pressure, shift government policy.

The 2018 rescheduling was the direct result. Sajid Javid, Home Secretary at the time, later described his decision as responding both to the specific medical cases and to the broader evidence that restricting specialist access to cannabis-derived medicines served no clinical or public protection purpose. The rescheduling placed cannabis-based medicinal products in Schedule 2 — the same schedule as other controlled drugs with recognised medicinal value, such as morphine and fentanyl.

Campaigners had expected faster NHS implementation to follow. It did not. The access problems that families documented in 2019 and 2020 — inability to find NHS specialists willing to prescribe, prohibitive private costs, inconsistent pharmacy availability — were substantially the same ones documented in 2023 and 2024. The All-Party Parliamentary Group on Medical Cannabis has produced multiple reports identifying the same systemic failures; NHS England has acknowledged them; tangible change has been slow.

The Medical Cannabis Clinicians Society and patient advocacy organisations including End Our Pain and the United Patients Alliance continue to press for systemic reform: NHS prescribing pathways for unlicensed medicines, reimbursement mechanisms, and mandatory prescribing training for neurology trainees. The direction of travel is favourable; the pace is not.

For families currently seeking access, the practical advice from advocacy groups is consistent: pursue an Epidyolex prescription through NHS paediatric neurology if eligible; seek a second specialist opinion if initially refused; access private clinics with caution and due diligence regarding product quality standards; and engage with patient registries such as the Drug Science UK Medical Cannabis Registry where possible, to contribute to the evidence base that may eventually widen NHS access.

The NHS epilepsy landscape for cannabis medicines in 2026 looks meaningfully different from 2018 — Epidyolex is now routinely prescribed within its licensed indications, the pharmacy network has expanded, and clinical familiarity with cannabis-based medicines has grown across the neurology community. But for the substantial population of treatment-resistant epilepsy patients who sit outside the NICE-approved categories, the gap between what is legally possible and what is practically accessible remains a defining failure of post-rescheduling implementation.

Medically reviewed by the Cannamedical Britannia Clinical Team, May 2026. This article is for informational purposes only and does not constitute medical advice. Patients should consult a qualified specialist before making any changes to their treatment. For information on how to access medical cannabis through a licensed UK prescriber, visit our Find a Doctor page or speak to your neurologist. Clinicians seeking prescribing information can visit our For Clinicians section. For quality and product information, see our Quality Standards page. A comprehensive overview of the UK medical cannabis landscape is available in our Medical Cannabis UK Guide.