Medical Cannabis for NonCancer Chronic Pain: Systematic
Review
Mary Butler, PhD, MBA
Co-Director, Minnesota EPC
Assistant Professor, UMN SPH HP&M
Review Team
•
•
•
•
•
•
Mary Butler
Erin Krebs, MD, MPH
Michelle Brasure, PhD, MLIS
Ben Sunderlin, MPH
Victoria Nelson, MS
Robert Kane, MD
PICOS
PICOS
Populations
Inclusion
Children or adults experiencing
chronic non-cancer pain
Interventions Unmodifed whole plant material
Whole Plant Extracts
Nabiximols (Sativex®)
Dronabinol (Marinol®)
Nabilone (Cesamet®) (synthetic)
Comparators Placebo
Active pain treatment (1)
Outcomes
Exclusion
Acute pain
Animal studies
Studies will not be
excluded for comparator;
however, a comparator
arm must be present to
assess benefits
Pain measures (ex: visual analog Intermediate outcomes
scales, McGill Pain Scale)
such as lab values
PICOS (cont.)
PICOS
Settings
Inclusion
Outpatient
Study
Designs
Benefits: Randomized controlled
trials, controlled trials, prospective
or retrospective cohort with
comparators
Harms: case control, case series
(at least 10 participants) for
potential serious harms
(hospitalizable events)
No date limitations
Other
limitations
Exclusion
Inpatient (hospital
treatment in response
to acute episode)
Results – Literature Flow
Results – Summary of Study Characteristics
Characteristic category
Countries in which studies were conducted
Funding Source
Study Design
Number randomized (or enrolled for observational)
Number of studies (unless otherwise noted)
Austria (1)
Belgium, (4)
Canada (8)
Czech Republic (4)
Denmark (1)
France, (2)
Israel (1)
Italy (1)
Romania, (2)
Spain, (1)
UK (11)
US (1)
Industry (17)
Non-governmental (2)
Not reported (2)
No funding (1)
Multisite parallel arm RCT (4)
Single-site parallel arm RCT (4)
Crossover RCT (6)
Open-label extension of RCT (4)
Open-label extension with randomized withdrawal (1)
Case Series (2)
RCTs – median 42 (range 13-339)
Open-label extensions – mean 104 (range 28-234)
Case series – mean 17 range (13-21)
Results – Summary of Study Characteristics
Characteristic category
Populations
Interventions
Treatment Duration
Study Risk of Bias
Number of studies (unless otherwise noted)
Pain related to MS (6)
Fibromyalgia (2)
Rheumatoid arthritis (1)
Mixed populations with chronic pain (3)
Unilateral peripheral neuropathic pain and allodynia (2)
Central neuropathic pain (2)
Brachial plexus injury (1)
Neuropathic pain (1)
Allodynia or diabetic neuropathy (1)
Diabetic neuropathy (1)
Medication overuse headache (1)
Chronic upper motor neuron syndrome (1)
Sativex (11)
Nabilone (7)
Dronabinol (2)
Delta-9-THC suspended in olive oil (1)
RCTs – 2 weeks to 14 weeks
Open-label extensions – 4 weeks to 124 weeks (31 months)
Case series – up to 36 months to up to over 48 months
RCT High risk of bias (7)
RCT Moderate risk of bias (6)
RCT Low risk of bias (1)
Open-label and case series High risk of bias (7)
Results –Comparative Effectiveness
• Nabilone vs dihydrocodeine (adjunct) for
neuropathic pain
– Small effect favors dihydrocodeine, AEs similar
• Nabilone vs amitriptyline for chronic insomnia
in people with fibromyalgia
– No difference for pain outcomes
Insufficient to draw conclusion for both.
Results - Efficacy
• For people with MS and central neuropathic
pain
– Sativex not different from placebo for pain
responders (low strength)
• For people with peripheral neuropathic pain
with allodynia
– Sativex favored for pain responders (low
strength)
Results – Efficacy (cont.)
• Evidence is insufficient to address patients with
rheumatoid arthritis, fibromyalgia, brachial plexus
injury, medication overuse headache, unilateral
peripheral neuropathic pain, diabetic neuropathy,
chronic upper motor neuron syndrome with
spasticity-related pain.
• No evidence available for other pain conditions
Results – Overall Efficacy Patterns
• Signal for hypothesis that medical cannabis
is beneficial for some patients.
• Effects are small (which could signal better
question is “for whom”)
• Consistent with growing basic science for
pain mechanisms and treatment pathways.
Results - Harms
• Medical Cannabis (compared to placebo) is
associated with greater risk of any AE,
serious AEs, trial withdrawal due to AEs,
other specified AEs.
• No literature (beyond 1 small study for
dihydrocodeine) for comparative harms with
opioids or other analgesics.
Results – JAMA
• Removed studies that were cancer pain,
unpublished studies, and shorter than 2
weeks (original pooled studies still >2wks)
• Only speaks to Sativex
• Evidence grade dropped (1 outcome became
insufficient)
• Numerical Rating Scale outcome became
insufficient
Applicability of Findings
• Studies provide indirect evidence for whole
plant extract medical cannabis
• Treatment durations too short for likely longterm use
– Benefits diminish over time??
– Rare harms??