Nobody knows yet, and the honest answer is that short. The first randomized trial of a GLP-1 drug in opioid use disorder has published its design and nothing else [1]. What exists around it is one completed trial in alcohol whose main endpoint missed, and one observational cohort. Neither licenses a claim. The pattern is the one set out in what an addiction trial of these drugs can conclude.
What is being tested
The trial is enrolling 310 adults with moderate to severe opioid use disorder. All of them have recently started buprenorphine. They are randomized one to one to tirzepatide or placebo, with weekly research visits for six months. Longer assessments fall at one, three and six months, and a safety visit at week 30.
Randomization is balanced by site and by buprenorphine formulation. Buprenorphine is the treatment here. Tirzepatide is the adjunct under test, and the authors call this the first randomized trial of it in this condition.
The nearest completed trial missed its endpoint
Fifty adults with moderate to severe alcohol use disorder took oral semaglutide or placebo for eight weeks [2]. The dose ran 3 mg daily for four weeks, then 7 mg daily for four. The prespecified primary outcome was lab cue-elicited craving at week 6, and it did not move. Drinks per day did not move either.
Several secondary outcomes did. Heavy drinking days fell, b −0.580, 95% CI −1.012 to −0.148. Drinks per drinking day fell, −1.177, interval −2.307 to −0.047. Craving reported in daily life fell, −2.195, interval −4.174 to −0.216. Three of those upper bounds stop just short of no effect, and with 50 people a handful of different participants would erase them.
A trial with many outcomes usually produces a significant one. That is arithmetic rather than evidence, and the primary endpoint exists to stop the arithmetic from talking. The authors conclude that continued development is warranted, which is not a conclusion that it works. The full reading sits in the alcohol trial that missed.
What the observational evidence adds
A Swedish national cohort followed 95,490 people with depression or anxiety who were taking an antidiabetic drug, of whom 22,480 used a GLP-1 [3]. Semaglutide was associated with less worsening substance use disorder, adjusted hazard ratio 0.53, 95% CI 0.35 to 0.80.
Three limits apply to that number. It is a secondary outcome in an observational study. It describes semaglutide, and the drug being tested in opioid use disorder is tirzepatide, which that cohort did not include. And the design compares periods of use against periods of non-use in the same person, which removes fixed confounders and not time-varying ones. The molecule-by-molecule split is set out in four drugs in one class.
What a reader should do with this
Nothing on this roster is sold for opioid use disorder. Nobody should obtain tirzepatide to treat it. Participants in the trial receive it alongside proper treatment under supervision, which is not a situation a telehealth purchase reproduces.
The rule worth carrying is narrower than the subject. A drug with a real effect on one addiction endpoint may do nothing for a neighboring one. Only a trial that named its endpoint in advance can tell the two apart. Every new indication eventually becomes a marketing claim somewhere, which is why the gates a seller has to clear matter more than the claim does.