By Dr. Benjamin Caplan, MD | Board-Certified Family Physician, CMO at CED Clinic | Evidence Watch
A 2025 systematic review of 22 randomized controlled trials found that most studies reported statistically significant pain reductions with cannabis-based medicines for neuropathic pain, but small sample sizes, short study durations, and compromised blinding from psychoactive effects prevent firm clinical recommendations. Patients and clinicians should view these findings as cautiously encouraging rather than definitive, with larger and more rigorous trials still needed.
Cannabis Medicines Show Modest Promise for Nerve Pain, But Evidence Remains Inconclusive
A new systematic review of 22 clinical trials spanning two decades of research finds more positive than negative results for cannabis-based medicines in neuropathic pain, yet pervasive methodological weaknesses, including compromised blinding, small samples, and short treatment durations, prevent firm treatment recommendations and underscore the urgent need for better-designed studies.
#72
Strong Clinical Relevance
Directly addresses the most common clinical question about cannabinoids and neuropathic pain, though the absence of meta-analysis and formal quality appraisal tempers the strength of its conclusions.
Cannabis-Based Medicines
Systematic Review
THC and CBD
Randomized Controlled Trials
Neuropathic pain affects an estimated 7 to 10 percent of the general population and remains one of the most treatment-resistant chronic pain conditions. First-line pharmacotherapies, including gabapentinoids, tricyclic antidepressants, and serotonin-norepinephrine reuptake inhibitors, provide meaningful relief for only a fraction of patients, and many who do respond experience dose-limiting side effects. Cannabis-based medicines have been proposed as alternative or adjunctive treatments, but clinicians lack clear guidance on whether the existing trial evidence justifies their use. A rigorous synthesis of the randomized controlled trial literature is essential for any informed prescribing decision in this space.
This PRISMA-guided systematic review searched Web of Science, PubMed, and MEDLINE for randomized controlled trials published between January 2003 and December 2024 evaluating cannabis-based medicines for neuropathic pain. From 5,397 database records and 24 hand-searched citations, 22 RCTs met inclusion criteria. The trials examined five distinct cannabinoid types, including CT-3 (ajulemic acid), delta-9-THC, CBD, THC-CBD combinations, and cannabidivarin (CBDV), across nine neuropathic pain etiologies spanning multiple sclerosis, spinal cord injury, diabetic neuropathy, postherpetic neuralgia, HIV-associated neuropathy, complex regional pain syndrome, peripheral neuropathic pain, chronic radicular pain, and lower extremity neuropathy. Many trials employed individualized dose-titration strategies, reflecting real-world clinical practice patterns.
Fifteen of the 22 trials reported statistically significant reductions in pain scores with cannabinoid treatment versus placebo, while seven did not, though some negative trials noted secondary improvements in sleep or mood. Importantly, the review did not conduct a meta-analysis and did not apply a formal risk-of-bias tool such as Cochrane’s RoB 2, meaning this 15-of-22 ratio is a raw count that does not account for study size, quality, or effect magnitude. The authors identify several shared limitations across the trial literature: psychoactive side effects that compromise participant blinding, small sample sizes, short treatment durations that limit extrapolation to chronic conditions, and high placebo response rates. The review concludes that cannabis-based medicines show potential but that benefits must be carefully balanced against adverse effects and response variability, and that larger, longer, and more rigorously designed trials are needed before definitive recommendations can be made.
This review does a reasonable job of capturing the state of the cannabinoid literature for neuropathic pain, and its honest acknowledgment of the blinding problem is refreshing. In nearly every cannabis trial, participants can usually tell whether they received the active drug or placebo, and that single methodological vulnerability casts a shadow over the entire evidence base. The 15-of-22 positive ratio sounds encouraging, but without pooled effect sizes or quality weighting, it is impossible to know whether we are looking at a robust therapeutic signal or a collection of underpowered trials amplified by expectation bias.
In my practice, I see neuropathic pain patients who have already cycled through gabapentin, pregabalin, duloxetine, and sometimes opioids without adequate relief. For those patients, I do consider cannabinoid therapies, but I frame them as a careful, individualized trial rather than an evidence-backed recommendation. I typically start with low-dose THC-CBD combinations, use slow titration, set clear outcome benchmarks at four to six weeks, and remain transparent with patients that the evidence, while suggestive, is not yet conclusive. The most important thing we can offer right now is structured monitoring and honest communication about what we know and what we do not.
This review sits in a familiar position within the cannabinoid research arc: enough positive signal to sustain clinical interest, but not enough methodological rigor to shift treatment guidelines. The absence of a meta-analysis is a meaningful gap, because the heterogeneity of cannabinoid formulations, dosing regimens, and pain conditions across these 22 trials demands quantitative tools to discern which combinations and contexts drive the positive results. Clinicians should note that the positive trials are not evenly distributed across conditions or cannabinoid types, and that drawing condition-specific conclusions from this synthesis requires reading the individual trial data rather than relying on the aggregate count.
From a pharmacological standpoint, prescribers must remain attentive to the drug interaction profile of THC and CBD, particularly CYP
