Effectiveness, safety and therapeutic potential of medical cannabis across treatment-resistant neuropsychiatric disorders: systematic review and meta-analysis

Why clinicians should know about it

• Picked for Neurology (clinical) (paper of the day, 8 November 2026): Clinical meta-analysis includes drug-resistant epilepsy and multiple sclerosis spasticity

Abstract

BACKGROUND: Medical cannabis has been suggested to have a therapeutic role in selected treatment-resistant neuropsychiatric disorders. We analyzed the available evidence to assess the effectiveness and safety of these interventions in treatment-resistant neuropsychiatric disorders. METHODS: For this systematic review and meta-analysis, we searched Web of Science, Cochrane, Embase, EBSCO, and PubMed for studies published between January 1, 1970, and April 25, 2026. We considered studies examining any type or formulation of medical cannabis in treatment-resistant neuropsychiatric disorders. Primary outcomes were disorder-specific clinical improvement measures. Safety outcomes were adverse events, serious adverse events, and withdrawals due to adverse events. Evidence from studies was synthesized as equivalent odds ratios (eORs) for comparative outcomes and as pooled proportions for single-arm safety outcomes, via random-effects meta-analyses. Risk of bias was assessed using RoB 2 for randomized trials, ROBINS-I for non-randomized studies, and structured quality assessment for case series and case reports, and certainty of evidence was assessed with GRADE. This study is registered with PROSPERO, CRD420251148868. FINDINGS: 127 eligible studies were included, of which 32 studies involving 3645 participants contributed to meta-analysis. Meta-analysis showed improvement in drug-resistant epilepsy (eOR 0.50, 95% confidence interval (CI) 0.34-0.72), refractory pain (eOR = 0.48, 95% CI = 0.30-0.78), and refractory multiple sclerosis spasticity (eOR = 0.22, 95% CI = 0.07-0.75). Evidence for psychiatric disorders was scarce. Small randomized trials reported improvement in PTSD-related nightmares and cannabis dependence after failed cessation. Uncontrolled studies reported improvements in hyperarousal, residual obsessive-compulsive symptoms, and refractory Tourette syndrome. Findings for treatment-resistant anxiety were inconsistent, and schizophrenia did not show a reliable therapeutic effect. Controlled analyses showed higher odds of any adverse events in drug-resistant epilepsy (eOR = 2.22, 95% CI = 1.52-3.23) and refractory multiple sclerosis (eOR = 1.89, 95% CI = 1.01-3.52), with higher odds of serious adverse events and withdrawals due to adverse events also observed in drug-resistant epilepsy. In single-arm and real-world studies, pooled proportions of any adverse events ranged from 15.5% in multiple sclerosis to 66.0% in Tourette syndrome, with substantial heterogeneity. INTERPRETATION: Adjunctive pharmaceutical cannabidiol has the most consistent comparative evidence for reducing seizure frequency in selected drug-resistant epilepsy. Evidence for refractory pain and multiple sclerosis is heterogeneous and varies by clinical indication, cannabinoid formulation, and outcome definition. Evidence in treatment-resistant psychiatric disorders remains limited and is insufficient to support routine clinical recommendations. Further high-quality studies directly examining diagnosis-specific and formulation-specific effects, safety, and tolerability are needed. FUNDING: STI2030-Major Projects and the National Natural Science Foundation of China.

Abstract as published, via PubMed.