Long‐term safety and treatment effects of cannabidiol in children and adults with treatment‐resistant epilepsies: Expanded access program results

From Wiki Journal Club
Jump to: navigation, search
Szaflarski JP, et al. "Long-term safety and treatment effects of cannabidiol in children and adults with treatment-resistant epilepsies: Expanded access program results". Epilepsia. 2018. 59(8):1540-1548.
PubMedFull text

Clinical Question

In patients with treatment resistant epilepsy of multiple etiologies, does the addition of cannabidiol to an antiepileptic regimen reduce seizure frequency and is it safe/tolerable to utilize?

Bottom Line

The addition of cannabidiol to an antiepileptic regimen can reduce seizure frequency in multiple epilepsy etiologies and is generally well tolerated with the most common adverse effects including somnolence and diarrhea.

Major Points

The use of antiepileptic treatment regimens in patients with treatment resistant epilepsies may provide only partial seizure reduction with severe adverse effects. Cannabidiol, a highly purified pharmaceutical product has shown an adequate safety profile, however the long term effects are unknown. Phase 3 clinical trials have shown that adjunct therapy with cannabidiol was efficacious for seizures associated with Dravet syndrome and Lennox-Gastaut syndrome. The open label expanded access program was approved to assess the tolerability and safety as well as efficacy of cannabidiol as add on therapy to an antiepileptic regimen in all treatment resistant epilepsies.

Frequency of median monthly convulsive seizures reduced by 51% and frequency of median monthly total seizures reduced by 48% after 12 weeks of treatment suggesting cannabidiol can reduce seizure frequency in multiple seizure epilepsy etiologies.

Overall, cannabidiol was generally well tolerated as 76% of patients remained on treatment over a median of a 48 week time period from baseline.

The most common adverse effects were diarrhea (29%) and somnolence (22%) which increased with increasing doses suggesting it may be a dose related effect.


As of October 2018, no guidelines have been published in regards to this trial. However, the American Academy of Neurology supports rescheduling of medical marijuana from schedule I to schedule II to encourage research for evaluation of long term safety and efficacy.


  • Open-label expanded-access program or state initiated Investigational New Drug (IND) applications
  • Descriptive purposes only
  • N=607
    • Safety Analysis (N=607)
    • Efficacy analysis (N=580)
  • Setting: 25 USA based epilepsy centers
  • Enrollment: January 15, 2014 through December 16, 2016
  • Mean follow-up: 12, 24, 48, 72, and 96 week visits
  • Analysis: Descriptive data for assessing the efficacy analysis and last observation carried forward analysis (LCOF) for missing data
  • Primary outcome:
    • Establish the safety and tolerability profile of cannabidiol


Inclusion Criteria

  • Eligibility criteria and endpoints varied by site-specific protocols
  • Treatment resistant epilepsy
  • Stable doses of antiepileptic drugs for ≥ 4 weeks prior to enrollment
  • Safety analysis set consisted of patients with ≥ 1 dose of cannabidiol and ≥1 postbaseline evaluation
  • Efficacy analysis set consisted of patients from the safety analysis set who had >0 seizures at baseline and ≥1 postbaseline visit

Exclusion Criteria

  • Eligibility criteria varied by site-specific protocols

Baseline Characteristics

Researchers conducted both a safety and efficacy analysis. Each arm used a different cohort of patients, but the demographics remained similar. The results below are demographics from the safety analysis.

  • Mean age: 13.2 years
  • Male: 52%
  • Female: 48%
  • Seizure Type
    • Lennox-Gastaut syndrome: 15%
    • Dravet Syndrome: 10%
    • Tuberous sclerosis complex: 4%
    • Aicardi syndrome: 3%
    • CDKL5: 3%
    • Doose, Dup15q, or febrile infection-related epilepsy syndromes: 4%
    • Other: 40%
    • Unknown: 20%
  • Concomitant AEDs taken at baseline, median: 3


  • Patients or caregivers kept diaries during the 4 week baseline period to collect data on the amount of convulsive and total seizures that occured
    • Convulsive seizures are defined as tonic, clonic, tonic-clonic, atonic, or secondary generalized
    • Nonconvulsive seizures were defined as myoclonic, absence, myoclonic-absence, and focal seizures with or without impaired consciousness
  • Antiepileptic drugs were documented at baseline and dose modifications were allowed at follow up
  • Patients received a purified CBD 100mg/ml oral solution (Epidiolex) with gradual titration starting at 2-10 mg/kg/day up to a maximum dose of 25-50 mg/kg/day based on the site protocol
  • Follow up occurred every 2-4 weeks for the first 16 weeks and then every 2-12 weeks after
  • Efficacy outcomes were assessed for the 12, 24, 48, 72, and 96 week visit periods
  • Weekly seizure frequency was converted to the frequency per 28 days (weekly frequency x 4)
  • Percentage change in seizure frequency was calculated by ([seizure frequency per 28 days] - [seizure frequency at baseline]) / [seizure frequency at baseline] x 100
  • The median percentage change in seizure frequency was calculated due to interpatient variability
  • Response rates were calculated by the proportion of patients that had at least 50%, 75%, and 100% reduction in monthly convulsive seizures and the total seizure frequency compared to baseline
  • The Medical Dictionary for Regulatory Activities (MedDRA) was utilized to classify and monitor treatment emergent adverse events


Primary Outcomes

Percentage of patients who tolerated and remained on treatment (median duration of 48 weeks)
76% (461/607)
Percentage of patients with treatment emergent adverse effects and serious adverse effects during median treatment duration of 48 weeks
88% and 33% respectively
Most common adverse effects
diarrhea (29%)
somnolence (22%)
convulsion (17%)
Most common all cause serious adverse effects
convulsion (9%)
status epilepticus (7%)
pneumonia (5%)
emesis (3%)

Secondary Outcomes

Median monthly frequency of convulsive seizures and total seizures after 12 weeks of treatment compared to baseline
convulsive seizures reduced by 51%, total seizures reduced by 48%
Percentage of patients with ≥ 50%, ≥ 75%, and 100% reduction in convulsive seizures from baseline at week 12
52%, 31%, 11% respectively
Percentage of patients with ≥ 50%, ≥ 75%, and 100% reduction in total seizures from baseline at week 12
49%, 30%, and 6% respectively

Subgroup Analysis

Median dose of cannabidiol between weeks 12 and 96
25 mg/kg/day
% of patients that had a dose reduction of cannabidiol during follow up
50% (330/605)
Patients taking concomitant clobazam with a dose reduction from baseline
Patients taking concomitant valproate with a dose reduction from baseline

Adverse Events

Abnormal liver AE (AST and ALT >3x upper limit)
10% (67/607) of all patients
75% (46/607) of patients were on valproate
≥ 1 of the MedDRA preferred terms for pneumonia
10% (67/607) patients
Somnolence of patients taking concomitant clobazam compared to patients not taking clobazam
38% (123/320) vs. 14% (40/287) respectively


  • Lack of a controlled study
  • Investigators and patients are not blinded to treatment (recall bias)
  • Variability in reporting methods per each site (reporting bias)
  • Certain site protocols based enrollment off of predetermined seizure frequency which was made known to possible participants, risk of over reporting may have occurred to enter the study
  • Site specific protocol may vary regarding patient eligibility and endpoints (selection bias)
  • Certain seizure types may be difficult to count (ex. absence) and may alter the total seizure frequency data


  • GW Research Ltd and the Epilepsy Therapy Project of the Epilepsy foundation funded the expanded-access program

Further Reading

1. Kanner A., Ashman E., et al. Practice guideline update summary: Efficacy and tolerability of the new antiepileptic drugs II: Treatment-resistant epilepsy Neurology Jul 2018, 91 (2) 82-90; DOI: 10.1212/WNL.0000000000005756

2. Brust J., Fee D., et al. Position Statement: Use of Medical Marijuana for Neurologic Disorders. The American Academy of Neurology (AAN). Feb 2018.