The bet that the obesity-drug class can treat addiction keeps moving down the list of substances. It started with alcohol, extended toward nicotine and opioids, and now reaches cannabis. A new multi-site randomized trial is testing tirzepatide, the GLP-1 and GIP dual agonist sold for metabolic disease, in moderate-to-severe cannabis use disorder, an indication that has no approved pharmacotherapy at all. For the cannabis industry, the same drug class arrives as both a threat and a possible cure.
The trial
The study is a randomized, double-blind, placebo-controlled, multi-site trial of roughly 100 patients with moderate-to-severe cannabis use disorder by DSM-5 criteria. Patients are assigned to weekly subcutaneous tirzepatide or matching placebo over 24 weeks, with the dose titrated up to a maximum of 15 milligrams. Its primary objective is twofold: to determine the maximum tolerated dose in this population and to evaluate whether the drug reduces cannabis use. Secondary endpoints include disorder severity, the share of cannabis-negative weekly urine screens, craving, withdrawal severity, retention in treatment, and quality of life. This is an early proof-of-concept and dose-finding study, not a pivotal trial. It sits within a federal addiction-research push that has been moving GLP-1 agonists across one substance-use disorder after another.
Why the indication matters
Cannabis use disorder has no FDA-approved medication. As potency climbs and legal availability widens, the population with problematic use grows, while the treatment toolkit remains essentially behavioral. In that vacuum, any credible pharmacological candidate is notable, and a novel mechanism more so. This is not a crowded field where a new entrant has to beat an incumbent; there is no incumbent.
The thesis underneath it
GLP-1 and GIP agonists appear to act as neuro-metabolic modulators of reward: they dampen craving and the reward response that drives compulsive use. The premise is that metabolic disease and addiction share overlapping reward biology. That is the same rationale the desk has tracked elsewhere. A semaglutide trial reduced drinking, and large observational analyses linked tirzepatide and semaglutide to lower rates of alcohol use disorder. Federal researchers have tested tirzepatide as an adjunct to buprenorphine for opioid use disorder. And Lilly’s purpose-built central-nervous-system successor, brenipatide, is already moving into alcohol and smoking programs. Cannabis is simply the next substance tested against a single reward mechanism. The bet is becoming addiction-wide: one class, many drugs of abuse.
The cannabis double bind
For the cannabis market specifically, the GLP-1 class is already a documented headwind. The industry has openly discussed weight-loss drugs as a demand disruption, because they blunt appetite, including the cannabis-driven appetite that fuels part of consumption. Now the same class is being tested to treat cannabis use disorder. So GLP-1s press on cannabis from two directions at once: trimming casual, appetite-linked use on one side, and potentially medicalizing the reduction of heavy use on the other. A drug class that simultaneously shrinks a market and treats that market’s disorder is an unusual form of competitive pressure, and cannabis is now squarely in its path.
The commercial shadow
Tirzepatide is Lilly’s. Lilly has already signaled it will pursue addiction with a purpose-built agent instead of leaving the field to academia, moving brenipatide into alcohol and smoking programs. So unlike the owner-less repurposing stories in this space, the GLP-1 and addiction field has a clear commercial owner positioned to act if the proof-of-concept signals hold. The cannabis trial is academic and early, but it sits in a field where the largest player has already committed capital and a molecule.
What This Trial Doesn’t Show Yet
The gaps are substantial. The trial is early: around 100 patients, a dose-finding objective, and no data yet. The cannabis-specific evidence is currently zero. The thesis rests instead on alcohol-trial and observational data, plus reward-biology reasoning extrapolated across substances. That cross-substance leap is unproven. Alcohol, opioids, nicotine, and cannabis engage different primary pharmacologies, and a reward-modulating effect that helps with one substance may not transfer equally to another. Cannabis is the least-evidenced target so far. Tirzepatide’s gastrointestinal side effects and substantial weight loss also raise suitability questions in a cannabis-use-disorder population that can include lean young adults, a group very different from the metabolic-disease patients the drug was built for. And this remains off-label academic use of an approved metabolic drug; no commercial cannabis-use-disorder program exists, and even a positive result is years from a labeled indication.
One Mechanism, an Expanding Addiction Bet
The GLP-1 and GIP class keeps enlarging its claim on addiction, and cannabis is the newest test. The trial matters for two reasons: cannabis use disorder has no approved drug, so a credible candidate is meaningful, and for the cannabis market the class is a double-edged force, trimming demand on one side and offering to treat problematic use on the other. The evidence is early, and the cross-substance leap is unproven. But the direction is the real story: a single reward mechanism being tested against the entire addiction map, with the industry’s largest pharmaceutical player already committed. There is something fitting about the obesity-drug era’s signature class being aimed at cannabis from both ends at once, as a reason people consume less and as a treatment for consuming too much. Whether tirzepatide reduces cannabis use is still unknown, and a small early trial will not settle it. But the fact that cannabis is now being tested at all shows how far the reward thesis has spread.