Opioids and Cannabis: An Uncertain Combination?

“[W]hile the rigorous study of Hamilton et al. advances scientific understanding of acute cannabinoid–opioid interactions, its findings—like much of the clinical trial literature—must be interpreted within the broader context of patient experience, real-world practice, and evolving policy.”

Image: Adobe Stock.

In a recent clinical trial in this issue of Anesthesiology, Hamilton et al.¹ conducted a randomized, crossover clinical trial comparing placebo, 10 mg dronabinol (synthetic Δ-9-tetrahydrocannabinol), 2 mg hydromorphone, and the combination of dronabinol + hydromorphone on pain severity, laboratory pain assessments, memory, and abuse liability among adults with knee osteoarthritis. Their findings did not indicate any synergistic analgesic effects when combining hydromorphone with dronabinol but did show increased working memory reaction time for this combination compared to placebo. Participants also reported more feelings of high and stronger drug effect with dronabinol than with hydromorphone. There was no reported risk for extramedical use for combined hydromorphone + dronabinol, but there was a higher risk of nausea with this combination than with any other group. This study followed a previous investigation by the same group that showed no benefits of combining 10 mg dronabinol with 4 mg hydromorphone compared with hydromorphone alone, dronabinol alone, or placebo among individuals with knee osteoarthritis.² Taken together, these studies make a thoughtful contribution to the body of experimental pharmacology and clinical trial research that points to lack of opioid–cannabinoid synergism in acute dosing paradigms among humans. Notably, these findings conflict with the consistent demonstration of opioid–cannabinoid synergism in preclinical models,³ highlighting the importance of conducting such studies in humans.

While the study of Hamilton et al.¹ was beautifully conducted and has internally valid results, it has limitations. Due to stringent exclusion criteria, no participants had a recent or significant history of cannabis use, nor were they allowed to be taking many common medications for pain (e.g., duloxetine, gabapentanoids). Thus, their results may only generalize to this rarified population of individuals who are trying a first dose of Δ-9-tetrahydrocannabinol with hydromorphone rather than to the complexity of real-world clinical application in chronic pain contexts in which participants often have substantial comorbidities, are taking a variety of medications, and take these medications over a long period of time. Further, fixed dosing models are difficult to translate to real-world practice, in which people often employ multiple different routes of administration (e.g., vaporizing, eating, topical products), cannabinoids, and doses—often in the same day.⁴

From a dosing perspective, 10 mg Δ-9-tetrahydrocannabinol is significantly higher than the 0.5- to 3-mg starting daily dose that expert guidance suggests for cannabis-naive individuals with chronic pain.⁵ Indeed, the starting dose of oral dronabinol for weight gain or as an antiemetic—the indications for which it is approved by the U.S. Food and Drug Administration (Silver Spring, Maryland)—is 2.5 mg.⁶ High doses among Δ-9-tetrahydrocannabinol–naive individuals can be disorienting and frightening, which may contribute to why a disproportionately high number of emergency room visits for cannabis in Colorado have been attributed to edibles.⁷ Further, Δ-9-tetrahydrocannabinol analgesia may follow a nonmonotonic pattern in acute contexts, in which too high of a dose may be associated with less pain relief than a lower dose, demonstrated by a recent study of dronabinol on laboratory pain among people receiving opioid agonists for opioid use disorder.⁸ This is why clinical trials of Δ-9-tetrahydrocannabinol–containing products for chronic pain typically employ a titration period to increase tolerability and dial in to a dose that minimizes side effects. For example, a recent phase III study of a novel, pharmaceutical grade, Δ-9-tetrahydrocannabinol–dominant product for low back pain involved a 3-week titration period. Participants started at doses of approximately 2.5 mg Δ-9-tetrahydrocannabinol twice a day, and many titrated up to mean total daily doses of 17.5 mg by end of the study treatment period.⁹ The weekly incidence of adverse events decreased once participants found a stable dose: from 15% during titration to 1.2% during the treatment phase.

When considering how to translate these results to clinical treatment of chronic knee osteoarthritis pain, the work of Hamilton et al.¹ suggests that adding oral Δ-9-tetrahydrocannabinol into a knee osteoarthritis treatment regimen that includes opioids may not cause enhanced analgesia in early stages. However, clinicians ought to be mindful of whether direct analgesia is the only desired effect of Δ-9-tetrahydrocannabinol–containing products for patients. As demonstrated by a recent survey of individuals with rheumatic conditions (253 of whom had knee osteoarthritis), many use cannabis products for numerous symptoms, including pain, sleep, mood, joint stiffness, and inflammation.¹⁰ Some symptoms, such as sleep and anxiety, reportedly improved more than pain itself. It is difficult to infer from the acute dosing results of Hamilton et al.¹ how these pain-adjacent symptoms may be affected when Δ-9-tetrahydrocannabinol products are taken daily over a long period of time (as is the case for chronic pain medications), especially with known bidirectional relationships between symptoms like sleep disturbance and pain.¹¹

While not conducted in knee osteoarthritis, two recent phase III trials of a Δ-9-tetrahydrocannabinol–dominant product for low back pain are instructive for this point.^9,12 In the first, participants who received active treatment reported significantly improved pain, sleep, physical health, disability, as well as substantially decreased use of rescue medication (ibuprofen) compared to placebo.¹² In the second, participants were randomized to receive tailored opioid treatment or the Δ-9-tetrahydrocannabinol product, with the same 3-week titration period as previously described.⁹ Those who received Δ-9-tetrahydrocannabinol had greater improvements in pain and sleep, as well as significantly outperforming opioids on constipation and other gastrointestinal-related side effects. These findings align with real-world data suggesting that often people substitute cannabis for opioids in the chronic pain context because of a favorable effect profile and better symptom management.¹³

So where do we go from here? Clinical trials remain the gold standard of research, and to effectively answer whether Δ-9-tetrahydrocannabinol might help synergize with opioid analgesia or be used in an opioid-tapering or sparing context, I believe we must be willing to expand the sorts of clinical trials we might employ to answer this question. For greater generalizability, such trials can and should use products from the legal cannabis marketplace. For example, a clinical trial (results not yet posted) was recently completed that investigated whether ad libitum medical cannabis in combination with a manualized prescription opioid taper support behavioral intervention resulted in more successful opioid tapering than the behavioral intervention alone.¹⁴ Bidwell et al.¹⁵ and Drennan et al.¹⁶ in Colorado have conducted novel clinical studies that carefully measure intake analgesic, intoxication, and other effects of commercially available products. Finally, my colleagues and I recently published a feasibility pilot¹⁷ of a novel intervention that provided tailored guidance on appropriate use of commercially available cannabis products to maximize benefit and minimize harm among veterans with chronic pain. While such trials do not offer the exacting stringency of classic Food and Drug Administration–regulated clinical trials when it comes to the drug product, they have two key advantages: (1) offering deeper understanding of how real-world use patterns and products may affect symptoms and (2) avoiding the substantial burden associated with storing, prescribing, and dispensing Δ-9-tetrahydrocannabinol–containing cannabis products—many of which remain Schedule I under federal law.

In summary, while the rigorous study of Hamilton et al.¹ advances scientific understanding of acute cannabinoid–opioid interactions, its findings—like much of the clinical trial literature—must be interpreted within the broader context of patient experience, real-world practice, and evolving policy. Doing so will help us move toward a pain care landscape that maximizes benefit, minimizes harm, and reflects the complexities and hopes of the communities we serve.

Acknowledgments

After the author drafted and revised this article, ChatGPT GPT-4o (Open AI, San Francisco, California) was used to check for grammatical errors and structural flow.

Research Support

Support was provided solely from institutional and/or departmental sources.

Competing Interests

Dr. Boehnke is supported by grant funding from Tryp Therapeutics (Victoria, Australia) for a clinical trial of psilocybin-assisted therapy and sat on a data safety and monitoring board for an ongoing clinical trial with Vireo Health (Minneapolis, Minnesota; unpaid). He has received grant funding from the National Institute on Drug Abuse, the National Center for Complementary and Integrative Health, and the National Institutes of Arthritis, Musculoskeletal, and Skin Diseases of the National Institutes of Health (Bethesda, Maryland). He has also received granting funding from the State of Michigan Veteran Marijuana Research Program (Lansing, Michigan) and has received speaking fees for lectures from the Southern Pain Society (Asheville, North Carolina) and the Michigan Center for Clinical Systems Improvement (Grand Rapids, Michigan). He received an honorarium for developing a podcast on fibromyalgia with Viatris Inc. (Canonsburg, Pennsylvania).