Standardizing Cannabis Use Measures and Outcomes in Randomized Trials for Cannabis Use Disorder

1. Cannabis Use Measures and Outcomes Assessed in Randomized Controlled Trials for Cannabis Use Disorder Dustin C. Lee, Ph.D. Behavioral Pharmacology Research Unit Johns Hopkins University School of Medicine

2. Introduction Cannabis use disorder (CUD) is prevalent Cannabis frequently reported as primary substance in treatment admissions • third behind alcohol and opiates Evidence-based psychosocial treatments developed • relapse rates high • no approved pharmacotherapies Need for continued focus on improving treatment efficacy

3. Introduction What are the optimal clinical endpoints for CUD interventions? Consensus outcomes for tobacco and alcohol • Tobacco – prolonged abstinence • Alcohol – abstinence and reduction in use • elimination of heavy drinking days (i.e. 4/5 drinks per drinking day for females and males)

4. Introduction CUD intervention outcomes • abstinence widely accepted primary outcome • reduction in use might be meaningful alternative Challenges with measuring cannabis use in clinical trials • evolving cannabis landscape (potency, ROA, concentrates) • no standard “unit” assessment of cannabis quantity Important to standardize how cannabis use is measured to determine optimal outcome assessments

5. Objectives of Presentation Primary Purpose: summarize findings from systematic review to provide a platform for in-depth discussion focused on standardizing cannabis use measures and outcomes in CUD trials • Objectives for this presentation • describeapproaches for measuring cannabis use in existing randomized controlled trials for CUD • summarize cannabis use outcomes assessed across trials • provide brief overview of other outcome domains assessed in CUD trials

6. Search Strategy Overall aim was to maximize inclusion of all relevant randomized controlled trials for CUD Structured literature search in seven electronic databases • PubMed, Embase, Cochrane Central, Cochrane Reviews, Cochrane Other Reviews, CINAHL, and PsycINFO • developed in collaboration with medical librarian Terms specified population, intervention, and trial design characteristics relevant to CUD interventions Search included references cited in recent systematic reviews and CUD trials

7. Inclusion Criteria • Current cannabis users that were • seeking treatment for CUD and/or • met diagnostic criteria for CUD • Psychosocial and pharmacological interventions for CUD • Trials with any comparison condition • Cannabis use a primary/secondary outcome • Studies were required to be published in English in a peer-reviewed journal

8. Exclusion Criteria • Trials examining cannabinoids as therapeutic agents • Interventions not specific to cannabis use • Interventions in ambivalent/non-treatment seekers • Editorials, letters, case studies/series, commentaries, or conference abstracts

9. Study Characteristics 5,877 records identified, 188 full-text documents reviewed 58 randomized controlled trials included in review • 36 psychosocial Interventions (PSY) • 22 pharmacological Interventions (PHA) Majority of trials included adult participants • 78% included only adults (n=45; 26 PSY, 19 PHA) • 15% included only adolescents (n=9; all PSY) • 7% included adolescents and adults (n=4, 1 PSY, 3 PHA)

10. Cannabis Use Measures – Self-Report Self-reported cannabis use assessed in 53 trials Timeline Followback approach used in 43 trials Diaries or self-report calendars used in 8 studies GAIN/ASI used in two trials

11. Cannabis Use Measures – Toxicology Toxicology testing for THC/metabolites conducted in 46 trials • 26 PSY/20 PHA • Reported as a standalone outcome in 22 trials • 9 PSY/13 PHA • Used to confirm self-reported abstinence in 19 trials • 14 PSY/5 PHA • Both standalone outcome and confirmation of self-reported abstinence in 5 trials • 3 PSY/2 PHA

12. Cannabis Use Measures – Toxicology Urine tests most predominant biological assay • 45 trials (11 did not specify details) • 5 trials used blood/saliva/hair (4 in addition to urine) Qualitative “screening” tests used in 23 trials • 18 trials used recommended cutoff of 50 ng/mL • 4 trials used other cutoffs (100 ng/mL, 20 ng/mL) • multiple cutoffs (20, 50, 100 ng/mL) used in one trial Quantitative tests (GC/MS, LC/MS/MS) used in 11 trials • to confirm positive tests (6 trials) • only source of toxicology testing (5 trials)

13. Cannabis Use Measures – Toxicology Concordance between self-reported use and urine toxicology reported in 18 trials Concordance was high overall • 80 to 100% agreement • kappas ranged from 0.6 to 0.9

14. Summary of Cannabis Use Measures Cannabis use measures fairly consistent across trials Timeline Followback used in majority of studies • reliability demonstrated for in-person and self-administration via computer/phone Variability in urine toxicology testing methods • qualitative tests used most frequently • subset of trials used quantitative tests High concordance between self-report and toxicology

15. Cannabis Use Outcomes

16. Abstinence

17. Reduction in Frequency

18. Reduction in Quantity

19. Summary of Cannabis Use Outcomes Wide range of abstinence and reduction outcomes derived from self-report and toxicology measures Heterogeneity in time-points used to determine change in use both during and post-TX Optimal unit of measurement to assess change in quantity not clear • grams, joints most common

20. Other Outcomes – Withdrawal PHA trials reported withdrawal as an outcome measure Nine trials used self-report instruments • Marijuana Withdrawal Checklist • Cannabis Withdrawal Scale • Clinical Institute Withdrawal Assessment Scale Individual symptoms of withdrawal assessed in seven trials • self-report and objective measures of sleep, irritability

21. Other Outcome Domains ASI, Addiction Severity Index;BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; MPS, Marijuana Problem Scale; RTC, Readiness to Change Questionnaire; SDS, Severity of Dependence Scale; TLFB, Timeline Followback; WHOQOL-BREF-World Health Organization Quality of Life Assessment

22. Summary of Other Outcome Domains Secondary outcome domains driven by unique features of a given population/trial Several domains of interest across trials • e.g. withdrawal, severity of use/dependence, mood, psychosocial functioning, cannabis-related problems, readiness to change/self-efficacy, alcohol and other drug use No apparent consensus on optimal instruments for each domain

23. Limitations and Considerations Methodological quality of trials not assessed Trials evaluating brief interventions in non-dependent/non-treatment seekers excluded Quantitative comparisons of outcomes not included in review

24. Summary/Discussion Consistency in cannabis use self-report measures • TLFB, urine toxicology included in most studies Consensus needed on unit of measurement to assess reduction in quantity • is TLFB sufficient or are other measures needed? Urine toxicology widely-used objective measure • variability in use of qualitative vs. quantitative tests • standardized approach may improve consistency across trials

25. Summary/Discussion Cannabis use outcomes highly heterogeneous across trials • abstinence and reduction outcomes remain unclear • combining data across existing trials with common features may improve power to detect sensitive outcomes Wide-range of outcomes reported in other domains • future studies would benefit from a standardized “core” battery of measures in relevant outcome domains

26. Acknowledgements Thank you to ACTTION for providing the opportunity to conduct this systematic review Specific acknowledgements to co-authors Nicolas Schlienz, Erica Peters, Ryan Vandrey, Eric Strain, Robert Dworkin and Dennis Turk Thank you for your attention!