Medical Cannabis: Limited Evidence for Specific Conditions with Significant Safety Concerns
Medical cannabis has insufficient evidence for chemotherapy-induced nausea/vomiting and should not replace FDA-approved cannabinoids (dronabinol, nabilone), while for chronic pain it may provide marginal benefit only as a third-line option after failure of gabapentin, duloxetine, and combination therapy, with a number needed to treat of 20 and substantial neuropsychiatric risks. 1, 2
Chemotherapy-Induced Nausea and Vomiting
Evidence remains insufficient for recommending medical marijuana for prevention or treatment of chemotherapy-induced nausea and vomiting. 1
- ASCO guidelines (2020) explicitly state that evidence is insufficient for medical marijuana in chemotherapy-induced nausea/vomiting, and insufficient for using medical marijuana in place of FDA-approved cannabinoids dronabinol and nabilone 1
- First-line antiemetics for high-emetogenic chemotherapy remain NK1 receptor antagonists, 5-HT3 receptor antagonists, and dexamethasone 1
- FDA-approved cannabinoids (dronabinol and nabilone) are available for refractory nausea/vomiting but are not first-line agents 1
Critical caveat: Cannabis itself can paradoxically cause cannabinoid hyperemesis syndrome, requiring treatment with capsaicin cream, benzodiazepines, and haloperidol—not more cannabinoids 3
Chronic Neuropathic Pain
Medical cannabis should only be considered as third-line therapy after documented failure of gabapentin/pregabalin and duloxetine at therapeutic doses for at least 4 weeks each. 2
First-Line Therapies (Try These First)
- Gabapentin: Start 100-300 mg at bedtime, titrate to 2400 mg/day in divided doses over 2-4 weeks 2
- Pregabalin: Start 150 mg/day in 2-3 divided doses, increase to 300 mg/day after 1-2 weeks (maximum 600 mg/day) 2
- Duloxetine: 60 mg once daily, can increase to 120 mg/day if needed 2
- Topical capsaicin 8%: Single 30-minute application provides relief for at least 12 weeks for localized peripheral neuropathic pain 2
Second-Line Therapies (If First-Line Fails)
- Combination therapy: Gabapentin/pregabalin plus duloxetine or tricyclic antidepressant targets different neurotransmitter systems 2
- Tricyclic antidepressants: Nortriptyline or desipramine 10-25 mg at bedtime, titrate to 75-150 mg/day (requires ECG screening in patients over 40 years) 2
Third-Line: Medical Cannabis (Only After Above Failures)
- Cannabis may increase patients achieving 50% pain relief compared to placebo, but this represents marginal benefit with number needed to treat of 20 2
- Systematic reviews consistently rate evidence as low to very low quality 2
- Do not use cannabis for chemotherapy-induced neuropathy—randomized trials showed no improvement with more toxicity than placebo 2
Multiple Sclerosis Spasticity
- Nabiximols (THC/cannabidiol oromucosal spray, not FDA-approved in US) showed high-quality evidence for reducing spasticity on numerical rating scales and visual analogue scales in multiple sclerosis 4, 5
- Moderate-quality evidence supports cannabinoids for spasticity reduction (Ashworth scale weighted mean difference -0.36) 5
- FDA-approved oral cannabinoids (dronabinol, nabilone) have very limited data for spasticity 1
HIV/AIDS-Related Cachexia
- Dronabinol is FDA-approved specifically for anorexia and weight loss related to AIDS 1
- Low-quality evidence suggests cannabinoids may be associated with weight gain in HIV infection 5
Drug-Resistant Epilepsy
- Cannabidiol (CBD) is FDA-approved for seizures associated with rare forms of severe epilepsy (Lennox-Gastaut syndrome, Dravet syndrome) 1
- This represents the strongest FDA indication for a cannabinoid product 1
Musculoskeletal Pain
Acetaminophen and NSAIDs remain first-line for musculoskeletal pain, not cannabis. 6
- No high-quality guideline evidence supports cannabis for musculoskeletal pain 6
- Gabapentin 2400 mg/day in divided doses is recommended for neuropathic pain components 6
- Physical therapy, cognitive behavioral therapy, and yoga have stronger evidence than cannabis 6
Significant Safety Concerns Across All Indications
High Withdrawal Rates Due to Adverse Events
- 10% of cannabis users withdraw due to adverse events versus 5% with placebo (number needed to harm = 25) 2
Neuropsychiatric Effects Are Common
- Nervous system adverse events occur in 61% of cannabis users versus 29% with placebo (number needed to harm = 3) 2
- Psychiatric disorders occur in 17% versus 5% with placebo (number needed to harm = 10) 2
- Common adverse effects include somnolence, fatigue, dizziness, confusion, nausea, dry mouth, hypotension, disorientation, drowsiness, loss of balance, and hallucination 1, 5
Route-Specific Risks
- Edible cannabis products accounted for only 0.32% of sales but 10.7% of emergency department visits 1
- Delayed onset with edibles leads to repeat dosing and delayed higher plasma concentrations 1
- Avoid smoked cannabis in patients with preexisting severe lung disease 2
Absolute Contraindications
- Active psychiatric disorders 2
- Substance use disorder history 2
- Cognitive impairment 2
- Severe lung disease (for smoked forms) 2
Regulatory and Legal Context
- Medical marijuana is not FDA-approved for any indication despite state-level legalization 1
- DEA classifies marijuana as Schedule I (high abuse potential, no accepted medical use) 1
- 24-40% of cancer patients in the US use marijuana despite lack of FDA approval 1
- Providers should assess for cannabinoid use and provide education on state and federal regulations 1
Clinical Decision Algorithm
- Document failure of evidence-based first-line therapies at therapeutic doses for at least 4 weeks 2
- Document failure of second-line combination therapy 2
- Screen for absolute contraindications (psychiatric disorders, substance use history, cognitive impairment, severe lung disease) 2
- Consider patient factors: Prior cannabis use history may predict better response 2
- If proceeding, use FDA-approved cannabinoids (dronabinol, nabilone, cannabidiol) rather than medical marijuana when possible 1
- Monitor closely for neuropsychiatric adverse effects and discontinue if they occur 2
Common pitfall: Prescribing cannabis before exhausting evidence-based therapies with stronger efficacy and safety profiles. The marginal benefit (NNTB = 20) does not justify bypassing gabapentin, pregabalin, duloxetine, or combination therapy, which have superior evidence and safety profiles. 2