This page focuses on highly purified prescription cannabidiol (CBD) oral solution, not dispensary products, artisanal cannabis oils, or THC-containing preparations. In modern epilepsy practice, CBD is best regarded as an adjunctive antiseizure medicine with the strongest evidence in Lennox-Gastaut syndrome (LGS), Dravet syndrome (DS), and tuberous sclerosis complex (TSC).1,8,9
Best-evidence indications: LGS, DS, TSC
Formulation used in trials: Oral solution 100 mg/mL
Typical starting dose: 2.5 mg/kg twice daily
Key safety issue:Transaminase elevation
Important: CBD is not interchangeable with over-the-counter “CBD oils”. The published evidence and regulatory approvals relate to a standardised, highly purified, prescription formulation.1
Label caveat: product indications differ by jurisdiction. In the United States, the licensed oral solution is approved for seizures associated with LGS, DS, or TSC in patients aged 1 year and older. In Europe, Epidyolex is licensed from 2 years, and for LGS/DS the indication is in conjunction with clobazam. Always prescribe to the local label and SmPC/PI. Overview
Cannabidiol is a non-intoxicating phytocannabinoid. In epilepsy practice, the clinically relevant issue is not whether it comes from Cannabis sativa, but whether it is delivered in a reliable, pharmaceutical-grade preparation with known concentration, manufacturing standards, interaction data, and trial evidence.1
In child neurology, CBD should usually be positioned as an adjunctive option for drug-resistant epilepsy, not as a replacement for syndrome-specific care, rational ASM sequencing, dietary therapy, surgery evaluation, or neurogenetic clarification.1,9,10
Where CBD fits well
Confirmed LGS, DS, or TSC with ongoing seizures despite appropriate prior therapies, especially when seizure burden remains clinically important and families can support careful monitoring.
Where caution is needed
Concomitant valproate, concomitant clobazam, pre-existing liver dysfunction, significant feeding difficulty, baseline somnolence, or poor capacity for laboratory follow-up.
Where expectations should be reset
Families seeking “natural” treatment without prescription-grade product, without measurable seizure targets, or with the hope that CBD will substitute for broader syndrome-directed management. Evidence base: what is established, and what is not
The strongest evidence for CBD comes from randomized controlled trials in DS, LGS, and TSC, followed by open-label extension studies showing sustained benefit in many patients over longer follow-up.2–9 Outside these syndromes, the literature is increasingly interesting but still largely observational and should not be described as evidence of the same grade.1,9,10
| Syndrome / setting | What the core studies showed | Practical interpretation |
|---|---|---|
| Dravet syndrome | In the pivotal placebo-controlled trial, add-on CBD 20 mg/kg/day reduced convulsive seizures more than placebo over 14 weeks and improved caregiver global impression, but adverse effects were more frequent.2 | Evidence is real and clinically meaningful, but benefit must be balanced against sedation, gastrointestinal intolerance, and liver monitoring. |
| Lennox-Gastaut syndrome | Two phase 3 trials showed that CBD reduced drop seizures more than placebo; one trial reported median reductions of about 44% with CBD versus 22% with placebo, and another showed about 42% and 37% reductions with 20 and 10 mg/kg/day versus 17% with placebo.3,4 | CBD is a reasonable adjunct for ongoing drop seizures. In day-to-day practice, many clinicians first aim for 10 mg/kg/day before deciding whether the extra benefit of a higher dose is worth the extra adverse effects. |
| Tuberous sclerosis complex | In TSC-associated seizures, CBD was superior to placebo. The 25 mg/kg/day regimen had a better safety profile than 50 mg/kg/day, which is why the lower regimen became the clinically preferred target.5 | For TSC, CBD is established therapy, but interaction with everolimus and closer nutritional/liver surveillance are especially relevant. |
| Long-term use | Open-label extension studies in LGS and TSC suggest sustained seizure reduction for many patients, with adverse effects broadly similar to those seen early in treatment.6,7 | Long-term continuation is reasonable when seizure burden, rescue use, severity, or day-to-day function has clearly improved. |
| Other epilepsies / off-label use | Real-world series suggest possible benefit in some refractory focal, genetic generalised, and unclassified epilepsies, but these data are retrospective and non-randomised.10 | Off-label use should be described as evidence-limited and preferably undertaken by specialist services using clear outcome measures and prospective review. |
Bottom line: if a child has LGS, DS, or TSC and remains significantly symptomatic after appropriate standard therapy, CBD is no longer fringe treatment. It is part of mainstream adjunctive care. Which children are good candidates?
A pragmatic candidate checklist is often more useful than a rigid algorithm. The best candidates usually have the following features:
- a reasonably secure epilepsy syndrome diagnosis;
- a persisting, countable seizure burden despite prior appropriate therapy;
- caregivers able to give a liquid medicine accurately and maintain a seizure diary;
- access to baseline and follow-up liver function testing; and
- a medication plan in which likely interactions have been reviewed in advance.
Especially reasonable to consider
- Ongoing drop seizures in LGS
- Ongoing convulsive seizures in DS
- Countable TSC-associated focal/generalised seizures
- Children already on a stable regimen where change can be judged clearly
Proceed more slowly or with more safeguards
- Valproate co-therapy
- Clobazam co-therapy with baseline drowsiness
- Moderate or severe hepatic impairment
- Low body weight, appetite concerns, or significant GI vulnerability
Poor starting point: using CBD before basic syndrome classification, before an adequate standard ASM trial sequence, or using a non-standard product with uncertain CBD/THC content. Dosing and prescribing
The prescription oral solution used in trials and licensing is 100 mg/mL. In practical terms, that means:
mL per dose = (weight in kg × mg/kg per dose) ÷ 100
Example: a 20 kg child prescribed 5 mg/kg twice daily needs 100 mg per dose, which equals 1.0 mL twice daily.
Standard titration
| Clinical setting | Week 1 | Week 2 | Further titration | Maximum usual dose |
|---|---|---|---|---|
| LGS / DS | 2.5 mg/kg twice daily (5 mg/kg/day) | 5 mg/kg twice daily (10 mg/kg/day) | Increase by 2.5 mg/kg twice daily at roughly weekly intervals if needed and tolerated | 10 mg/kg twice daily (20 mg/kg/day) |
| TSC | 2.5 mg/kg twice daily (5 mg/kg/day) | 5 mg/kg twice daily (10 mg/kg/day) | Increase by 2.5 mg/kg twice daily at roughly weekly intervals if needed and tolerated | 12.5 mg/kg twice daily (25 mg/kg/day) |
In LGS/DS, the lower target of 10 mg/kg/day often gives a useful balance between efficacy and tolerability. In TSC, the evidence-supported maintenance target is 25 mg/kg/day.
Per-dose volume examples (100 mg/mL solution)
| Weight | 2.5 mg/kg per dose | 5 mg/kg per dose | 10 mg/kg per dose | 12.5 mg/kg per dose |
|---|---|---|---|---|
| 10 kg | 25 mg = 0.25 mL | 50 mg = 0.50 mL | 100 mg = 1.00 mL | 125 mg = 1.25 mL |
| 20 kg | 50 mg = 0.50 mL | 100 mg = 1.00 mL | 200 mg = 2.00 mL | 250 mg = 2.50 mL |
| 30 kg | 75 mg = 0.75 mL | 150 mg = 1.50 mL | 300 mg = 3.00 mL | 375 mg = 3.75 mL |
| 40 kg | 100 mg = 1.00 mL | 200 mg = 2.00 mL | 400 mg = 4.00 mL | 500 mg = 5.00 mL |
Hepatic impairment
| Hepatic status | Practical message |
|---|---|
| Mild impairment (Child-Pugh A) | No label dose adjustment. |
| Moderate impairment (Child-Pugh B) | Start lower. For LGS/DS the labelled start is 1.25 mg/kg twice daily; for TSC, lower labelled targets also apply. Titrate more slowly and monitor more closely. |
| Severe impairment (Child-Pugh C) | Start very low. For LGS/DS the labelled start is 0.5 mg/kg twice daily. Use only when the benefit clearly justifies the complexity and risk. |
Practical prescribing pointers
- Give it consistently with respect to meals.
- Use the supplied oral syringe; do not use household spoons.
- In frail children, slower titration is often wiser than faster titration.
- Taper when stopping; avoid abrupt withdrawal when possible.
The two interactions that dominate routine practice are clobazam and valproate. In TSC, everolimus must also be reviewed carefully.
| What to monitor | Why it matters | Suggested practice |
|---|---|---|
| ALT, AST, total bilirubin | CBD can cause dose-related transaminase elevation, especially with valproate. | Check at baseline, then at 1, 3, and 6 months, and again within 1 month of dose increases or changes in liver-relevant co-medication. |
| Sleepiness / sedation | More common early in treatment; markedly more common with clobazam. | Ask specifically about daytime sleepiness, school participation, drooling, balance, and regression. Consider reducing clobazam before abandoning CBD. |
| Weight and appetite | Reduced appetite and weight loss are common enough to matter clinically. | Plot weight at each visit, especially in children with TSC, feeding issues, or higher doses. |
| Seizure diary / rescue medication use | Response is not only about raw count; severity and rescue burden matter. | Review countable seizures, clusters, injuries, rescue use, school attendance, and sleep. |
| Everolimus level (TSC) | CBD can increase orally administered everolimus exposure. | Use therapeutic drug monitoring and dose adjustment where relevant. |
Clobazam
CBD raises levels of N-desmethylclobazam, which is why somnolence often worsens after CBD starts. If the child becomes overly sleepy, think of reducing clobazam first rather than automatically abandoning CBD.
Valproate
The combination with valproate is the classic setup for liver enzyme elevation. Monitor more closely, and if transaminases rise, review whether valproate, CBD, or both need to be stepped back.
Everolimus
This matters especially in TSC. CBD can materially increase everolimus exposure, so level-guided dose adjustment is sensible rather than guesswork.
Strong enzyme inducers
Strong CYP3A4/CYP2C19 induction can lower CBD exposure. If such a combination is unavoidable, interpret “non-response” cautiously and follow the local label on dose adjustment.
When to stop for liver toxicity: discontinue CBD if transaminases rise to >3× ULN with bilirubin >2× ULN, or if there are sustained transaminase elevations >5× ULN.
One of the commonest real-world errors is forgetting that food materially changes CBD exposure. Pick a pattern — with meals or without — and keep it consistent. Adverse effects and troubleshooting
Across trials and clinical practice, the recurring adverse-effect pattern is familiar: somnolence, reduced appetite, weight loss, diarrhea, vomiting, fatigue, and elevated transaminases. In TSC, pyrexia and vomiting are particularly noticeable in some patients.5,7,9,10
| Problem | Likely explanation | What usually helps |
|---|---|---|
| Excessive sleepiness | Rapid titration or clobazam interaction | Hold the dose where it is, reduce clobazam if appropriate, and titrate more slowly. |
| Diarrhea / vomiting | Early dose-related GI intolerance | Slow titration, check meal consistency, maintain hydration, and reassess the dose ceiling. |
| Poor appetite / weight loss | Recognised treatment effect, often dose-related | Serial weights, dietetic input where needed, and consider stepping back the dose if clinically significant. |
| Rising ALT/AST | CBD effect, especially with valproate | Repeat promptly, review valproate/clobazam, and follow labelled interruption or discontinuation thresholds. |
| Rash / hypersensitivity | Drug intolerance | Assess severity promptly; significant hypersensitivity warrants stopping the drug. |
Practice Point: if the child is clearly sleepier but seizure burden is improving, the first question is often not “Should CBD stop?” but rather “Can clobazam come down?” Off-label use, what CBD is not, and unanswered questions
What CBD is not
- not a synonym for “medical cannabis”;
- not interchangeable with unregulated CBD oils;
- not first-line monotherapy for most childhood epilepsies;
- not free of important interactions; and
- not a substitute for syndrome-directed thinking.
Where evidence may grow
- other developmental and epileptic encephalopathies;
- refractory focal and genetic generalised epilepsies;
- long-term developmental and behavioural outcomes;
- better prediction of responders; and
- clearer interaction mapping across polytherapy.
The current off-label literature is promising enough to justify specialist interest, but not strong enough to justify casual overstatement. The most honest formulation for families is that CBD has high-grade evidence in selected syndromes and lower-grade evidence elsewhere.10
A good discipline for off-label use: define the seizure target in advance, review at a fixed interval, and record whether the child improved enough to justify continuation.
Take-home messages
- Think prescription purified CBD, not generic “cannabis oil”.
- Use it chiefly as adjunctive therapy in LGS, DS, and TSC.
- Start at 2.5 mg/kg twice daily and titrate methodically.
- In LGS/DS, many children do well enough at 10 mg/kg/day; higher doses bring more toxicity.
- In TSC, the evidence-based target is 25 mg/kg/day.
- Check ALT, AST, and bilirubin before treatment, then at 1, 3, and 6 months.
- Expect interaction trouble first with clobazam, valproate, and in TSC with everolimus.
- Keep dosing consistent with meals.
- Measure outcomes the way families live them: seizure count, clusters, rescue use, injuries, alertness, and participation.
- Off-label use can be reasonable, but it should be framed as evidence-limited.
References
- Arzimanoglou A, Brandl U, Cross JH, et al. Epilepsy and cannabidiol: a guide to treatment. Epileptic Disord. 2020;22(1):1-14. PubMed PMID: 32096470
- Devinsky O, Cross JH, Laux L, et al. Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome. N Engl J Med. 2017;376:2011-2020. PubMed PMID: 28538134
- Thiele EA, Marsh ED, French JA, et al. Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4): a randomised, double-blind, placebo-controlled phase 3 trial. Lancet. 2018;391:1085-1096. PubMed PMID: 29395273
- Devinsky O, Patel AD, Cross JH, et al. Effect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome. N Engl J Med. 2018;378:1888-1897. PubMed PMID: 29768152
- Thiele EA, Bebin EM, Bhathal H, et al. Add-on Cannabidiol Treatment for Drug-Resistant Seizures in Tuberous Sclerosis Complex: A Placebo-Controlled Randomized Clinical Trial. JAMA Netw Open. 2021;4:e2120703. PubMed PMID: 33346789
- Patel AD, Mazurkiewicz-Belzdinska M, Chin RFM, et al. Long-term safety and efficacy of add-on cannabidiol in patients with Lennox-Gastaut syndrome: Results of a long-term open-label extension trial. Epilepsia. 2021;62:2228-2239. PubMed PMID: 34287833
- Thiele EA, Bebin EM, Filloux F, et al. Long-term cannabidiol treatment for seizures in patients with tuberous sclerosis complex: An open-label extension trial. Epilepsia. 2022;63:426-439. PubMed PMID: 34957550
- Silvinato A, Floriano I, Bernardo WM. Use of cannabidiol in the treatment of epilepsy: Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex. Rev Assoc Med Bras (1992). 2022;68:1345-1357. PubMed PMID: 36417631
- Navarro CE. Practical approach to the safe use of cannabidiol in patients with refractory epilepsy: a mini review. Front Pharmacol. 2025. PubMed PMID: 40978479
- Hollander M, Mayer T, Klotz KA, et al. Use of cannabidiol for off-label treatment of patients with refractory focal, genetic generalised and other epilepsies. Neurol Res Pract. 2025;7:49. PubMed PMID: 40696489
Prescribing at a glance
Before starting: syndrome confirmed, baseline ALT/AST/bilirubin, medication interaction review.
Start: 2.5 mg/kg twice daily.
Common target: 5 mg/kg twice daily in LGS/DS; 12.5 mg/kg twice daily in TSC.
Watch closely: clobazam, valproate, everolimus, appetite, weight, sedation.
Recheck liver tests: 1, 3, and 6 months, then as clinically indicated and after relevant dose/medication changes.
Self-Assessment
Test your knowledge. Click your answer for each question, then submit to see your score and explanations.
