Cannabis for the Treatment of Pediatric Epilepsy The Case for a Living Systematic Review Jesse Elliott, Deirdre DeJean, Tammy Clifford, Doug Coyle, Beth Potter, Becky Skidmore, Christine Alexander, Alexander Repetski, Bláthnaid McCoy, George A. Wells Faculté de médecine | Faculty of Medicine École d’épidémiologie et santé publique uOttawa.ca School of Epidemiology and Public Health uOttawa.ca
2 Disclosures • I have no conflicts • Dr. Bláthnaid McCoy is principal investigator in a study of cannabinoids for Dravet syndrome • None declared by any other author uOttawa.ca
Acknowledgements • Alexander Repetski , Christine Alexander: Patient family representatives • George Wells, Doug Coyle, Tammy Clifford, Beth Potter: School of Epidemiology and Public health, University of Ottawa • Bláthnaid McCoy: Division of Neurology, The Hospital for Sick Children Toronto • Becky Skidmore: independent information specialist • Deirdre DeJean: Centre for Health Law, Policy and Ethics, University of Ottawa uOttawa.ca
Goals • Introduce the concept of living systematic reviews • Present an example living systematic review • Cannabis for the treatment of epilepsy in children uOttawa.ca
2010: 75 RCTs per day 2018: 165 RCTs per day uOttawa.ca Bastian et al. PLoS Med 2010;7:e1000326
Median time from primary study publication to inclusion in a published systematic review: 2.5 to 6.5 years uOttawa.ca Elliott et al. PLoS Med 2014;11(2)
Living systematic review • A systematic review that is continually updated , incorporating new evidence as it becomes available Key concepts • Core systematic review methods • A priori defined update plan and intervals uOttawa.ca
When is a living systematic review appropriate? • High-priority question for decision making • Important uncertainty in the evidence base • Emerging evidence uOttawa.ca
A living systematic review: CANNABIS FOR THE TREATMENT OF PEDIATRIC EPILEPSY uOttawa.ca
Research question • What are the benefits and harms of cannabis-based treatments for pediatric epilepsy? Objective • To provide a comprehensive, up-to-date overview of the evidence in order to inform decision-making uOttawa.ca
Why is a living systematic review needed? High-priority question for decision making o Parents of affected children o Clinical practice o Policy Important uncertainty in the evidence base o Few studies, different methodologies, discrepant findings Emerging evidence o 2015: 4 studies o 2018: 30 studies, with 35 studies registered in ClinicalTrials.gov uOttawa.ca
Living systematic review plan Baseline review Search for Summarize Disseminate Evaluate studies findings findings eligibility (up to Apr. 2018) Update 1 Search for studies New (Apr. to Oct. 2018) studies 6-month update frequency: Update X Update 2: April 2019 Search for Update 3: October 2019 studies No new Update 4: April 2020 (6 months from studies found … last update) uOttawa.ca
PROSPERO: CRD42018084755 Elliott et al. Systematic Reviews 2018;7:95 Methods • A broad systematic review of the available randomized and non-randomized evidence in the published and grey literature PICOS criteria Population Children (<19 yr) with epilepsy Intervention Any type of cannabis-based product Comparator Pharmacologic or non-pharmacologic treatments, placebo, usual care, no treatment • Primary outcome Seizure freedom Seizure frequency (total, 50% reduction) • • Secondary Status epilepticus • • outcomes Quality of life Deaths • • Sleep Gastrointestinal adverse events • Study designs Randomized controlled trials (RCT) • Non-randomized studies (NRS) uOttawa.ca
Elliott et al. Epilepsia 2019;60(1):6 – 19. SUMMARY OF FINDINGS BASELINE REVIEW uOttawa.ca
Baseline review: April 2018 • 23 published studies, 1665 participants – 4 Randomized controlled trials (RCTs) – 19 Non-randomized studies (NRS) • Risk of bias – RCTs: Low – NRS: High (e.g., selection, ascertainment, performance) • 33 studies registered in ClinicalTrials.gov uOttawa.ca
Baseline review: summary of findings (RCTs) Outcome Finding* BASELINE REVIEW GRADE assessment No significant Risk difference: Seizure freedom Low certainty difference 5% (95%CI – 1% to 11%) Reduced with Median difference: Seizure frequency Moderate certainty cannabidiol – 20% (95%CI – 27% to – 13%) No significant Mean difference: Quality of life Moderate certainty difference 0.6 (95%CI – 2.6 to 3.9) Increased with Relative risk: Diarrhea Low certainty cannabidiol 2.25 (95%CI 1.38 to 3.68) *Versus placebo. uOttawa.ca AE = adverse event, CI = confidence interval
Baseline review: summary of findings (NRS*) Pooled % of participants Outcome BASELINE REVIEW GRADE assessment Seizure freedom 3% (95%CI 0% to 6%) Very low certainty Treatment response 44% (95%CI 36% to 56%) Moderate certainty (>50% seizure reduction) Quality of life Improved QoL Very low certainty Diarrhea 9% (95%CI 3% to 16%) Low certainty uOttawa.ca *Prospective before – after cohort studies; AE = adverse event, CI = confidence interval, QoL = quality of life
SUMMARY OF FINDINGS UPDATE 1: OCTOBER 2018 uOttawa.ca
UPDATE 1: October 2018 • 7 published studies, 812 participants – 0 RCTs – 7 NRS • High risk of bias (e.g., selection, ascertainment, performance) • An additional 2 studies registered in ClinicalTrials.gov uOttawa.ca
What is new after UPDATE 1? (RCTs) Outcome BASELINE REVIEW UPDATE 1 GRADE assessment Seizure freedom No significant difference Low certainty Reduced with Seizure frequency Moderate certainty cannabidiol No change (no new RCTs) Quality of life No significant difference Moderate certainty Increased with Diarrhea Low certainty cannabidiol NOTE: Of the 35 studies registered in ClinicalTrials.gov, 6 are RCTs uOttawa.ca AE = adverse event
What is new after UPDATE 1? (NRS*) Pooled % of participants Outcome BASELINE REVIEW UPDATE 1 GRADE assessment Seizure freedom 3% (95%CI 0% to 6%) 5% (95%CI 1% to 9%) Very low certainty Treatment response 44% (95%CI 36% to 56%) 50% (95%CI 43% to 58%) Moderate certainty (>50% seizure reduction) Quality of life Improved QoL Improved QoL Very low certainty Diarrhea 9% (95%CI 3% to 16%) 23% (95%CI 12% to 33%) Low certainty uOttawa.ca *Prospective before – after cohort studies; AE = adverse event, CI = confidence interval, QoL = quality of life
State of the evidence (October 2018) • Cannabidiol probably reduces seizures but increases the risk of diarrhea • Certainty is low to moderate and may change with the publication of ongoing studies • Evidence related to quality of life is less clear uOttawa.ca
Limitations • Certainty of the available evidence is very low to moderate, and there is high risk of bias among the non-randomized studies • Heterogeneity in terms of interventions and duration of treatment • Most available evidence is related to the use of cannabidiol; effect of other cannabinoids is unclear uOttawa.ca
CHALLENGES AND CONSIDERATIONS uOttawa.ca
• Time and resource implications • Consistency of judgements over time (e.g., study selection, risk of bias) • File management and storage • Changes to authorship over time • Dissemination of findings – cannabisandepilepsy.blogspot.com uOttawa.ca
jcrai065@uottawa.ca cannabisandepilepsy.blogspot.com uOttawa.ca
Recommend
More recommend
