| Trial results
Two retatrutide knee osteoarthritis abstracts, both from Lilly trials, are now posted on the American College of Rheumatology meeting site (ACR Convergence 2026, Orlando). TRIUMPH-4 was scheduled as a plenary talk on November 8 at 11:15 AM ET, and the New England Journal of Medicine published the TRIUMPH-4 paper the same day. The knee osteoarthritis part of TRIUMPH-1 has a poster scheduled for November 10. Retatrutide is investigational and is not approved for any use.
TRIUMPH-4 in short
TRIUMPH-4 randomized 445 adults with painful knee osteoarthritis and a body mass index (BMI) of 27 or higher to retatrutide 9 mg, retatrutide 12 mg or placebo, injected once a week for 68 weeks. At the start, knee pain averaged 5.95 on a 0 to 10 scale, body weight averaged 112.7 kg and BMI averaged 40.4.
At week 68, counting only people still on treatment (the efficacy estimand, which estimates the effect while on the drug), knee pain fell by 4.48 points on 9 mg and 4.41 points on 12 mg, against 2.43 points on placebo. Body weight fell 26.4% and 28.7%, against 2.1% on placebo.
Counting everyone who was randomized, whether or not they stayed on treatment (the treatment-regimen estimand, from Lilly's December 2025 topline release), pain fell 4.0 and 3.7 points against 2.1 on placebo, and weight fell 20.0% and 23.7% against 4.6%.
Knee pain, points on a 0 to 10 scale
Efficacy estimand, people still on treatment
Treatment-regimen estimand, everyone randomized
Body weight, percent
Efficacy estimand, people still on treatment
Treatment-regimen estimand, everyone randomized
Change at week 68 in TRIUMPH-4. Longer bars mean a bigger fall. Both charts start at zero. Sources: ACR abstract 0871 and Lilly's December 11, 2025 release.
Side by side with the TRIUMPH-1 knee study
TRIUMPH-1 and TRIUMPH-4 are different trials. The TRIUMPH-1 knee study is the knee osteoarthritis group inside TRIUMPH-1, the larger weight-loss program reported in the TRIUMPH-1 NEJM report. The table does not combine the two.
| Measure | TRIUMPH-4 | TRIUMPH-1 knee study |
|---|---|---|
| Registry number | NCT05931367 | NCT05929066 (the knee osteoarthritis group inside TRIUMPH-1) |
| Design | Standalone phase 3, double-blind, placebo-controlled | Knee osteoarthritis group within the TRIUMPH-1 weight program, double-blind, placebo-controlled |
| People | 445 adults with painful knee osteoarthritis, BMI 27 or higher | 574 adults with painful knee osteoarthritis, BMI 27 or higher |
| Arms | Retatrutide 9 mg, 12 mg, placebo | Retatrutide 4 mg, 9 mg, 12 mg, placebo |
| Length | 68 weeks | 80 weeks |
| Average starting knee pain (0 to 10) | 5.95 | 5.97 |
| Average starting weight | 112.7 kg (BMI 40.4) | 112.5 kg |
| Pain change in points, efficacy estimand | 9 mg: −4.48. 12 mg: −4.41. Placebo: −2.43 | 4 mg: −3.66. 9 mg: −4.19. 12 mg: −4.30. Placebo: −2.24 |
| Pain change in points, treatment-regimen estimand | 9 mg: −4.0. 12 mg: −3.7. Placebo: −2.1 | 4 mg: −3.22. 9 mg: −3.48. 12 mg: −3.64. Placebo: −1.87 (NEJM hybrid-treatment estimand) |
| Weight change, efficacy estimand | 9 mg: −26.4%. 12 mg: −28.7%. Placebo: −2.1% | −18.6% to −28.7% across retatrutide arms. Placebo: −1.5% |
| Weight change, treatment-regimen estimand | 9 mg: −20.0%. 12 mg: −23.7%. Placebo: −4.6% | Not in the abstract |
| Gap to placebo in pain points (Atlas arithmetic, efficacy estimand) | 9 mg: 2.05. 12 mg: 1.98 | 4 mg: 1.42. 9 mg: 1.95. 12 mg: 2.06 |
| Most common side effects | Nausea, diarrhea, constipation, vomiting, decreased appetite | Mild to moderate stomach and gut effects. The 4 mg group had fewer of them and fewer stops because of side effects |
| Where the data stand | Published in NEJM on November 8 (the Atlas has read the abstract on PubMed). Lilly released topline results on December 11, 2025 | Published in NEJM on November 8 with TRIUMPH-4 (abstract read on PubMed). Poster scheduled for November 10 |
What the numbers show, and what they do not
Placebo improved too. Knee pain fell by 2.43 points on placebo in TRIUMPH-4, more than half the 4.48-point fall on 9 mg. The extra fall with retatrutide is about 2 points on the 0 to 10 scale on the efficacy count (the Atlas calculated this from the reported averages) and 1.6 to 1.9 points on the treatment-regimen count.
Pain and weight moved together. The abstract reports no analysis that separates the two, so the trial cannot say how much of the pain relief came from weight loss and how much from anything else the drug does.
The pain score is a questionnaire. WOMAC pain is filled in by patients. The abstract lists no X-ray, MRI or cartilage measure, so nothing in it shows cartilage regrowth, slower joint damage or a cure for arthritis.
Most people started with a high BMI. The average was 40.4, and Lilly said 84.0% of participants started at 35 or above. The results may not carry over to people with a lower BMI.
On side effects, Lilly's topline reported nausea in 38.1% on 9 mg and 43.2% on 12 mg, against 10.7% on placebo. Dysesthesia (unusual skin sensations such as tingling) affected 8.8% and 20.9%, against 0.7%. Stopping treatment because of side effects happened in 12.2% and 18.2%, against 4.0%. Lilly said some of those stops were for perceived excessive weight loss.
A note on percentages: Lilly's percent improvements in pain (for example 75.8%) were not planned in advance and come from a post hoc analysis. This page leads with points on the 0 to 10 scale.
What comes next
Update, November 14, 2026: TRIUMPH-4 and the TRIUMPH-1 knee study were published together in the New England Journal of Medicine on November 8 (Schnitzer et al., DOI 10.1056/NEJMoa2610676). The Atlas has now read the abstract on PubMed. It has not read the full paper, so the side-effect rates on this page still come from the December 2025 Lilly topline release. The paper main pain figures use what it calls a hybrid-treatment estimand: treatment failures are handled with a hypothetical strategy and everyone else is counted as randomized. In TRIUMPH-4 at week 68, knee pain fell 3.96 points on 9 mg and 3.74 on 12 mg against 2.12 on placebo (differences from placebo −1.83 and −1.61). Counting everyone as randomized, the differences from placebo were −1.70 and −1.38. Body weight fell 20.0% and 23.7% against 4.6% on placebo. In the TRIUMPH-1 knee study (574 adults, 80 weeks), pain fell 3.22, 3.48 and 3.64 points on 4, 9 and 12 mg against 1.87 on placebo (differences −1.35, −1.61 and −1.77; counting everyone as randomized, −0.96, −1.41 and −1.64). The abstract says serious adverse events were more common with retatrutide than with placebo in both trials, and the most common side effects were gastrointestinal. Funded by Eli Lilly.
Worth knowing: The tables above use the company topline release and the conference abstracts. The Atlas has read the NEJM abstract but not the full paper. Placebo improved too: in the paper, pain fell 2.12 points on placebo in TRIUMPH-4 and 1.87 in the TRIUMPH-1 knee study, so the added effect on the paper main count is about 1.4 to 1.8 points. Retatrutide is investigational and is not approved for any use.
Follow retatrutide on the tracker. Lilly plans to file for approval in Q1 2027 (company statement). The next date in the trial registry is a Phase 3 trial completion in Nov 2026. Every dated trial completion, filing and FDA decision for retatrutide is on the pipeline tracker, each with its source.
Related: Retatrutide medicine page | TRIUMPH-1 in NEJM | Lilly tracker | Drug pipeline
Source: ACR Convergence 2026, abstract 0871 (Conaghan et al.), retatrutide once weekly for knee osteoarthritis pain and obesity Conference abstract, not peer reviewed. ACR Convergence 2026, abstract 2573 (Hunter et al.), on the ACR abstract site Conference abstract, not peer reviewed. Eli Lilly and Company, TRIUMPH-4 topline release, December 11, 2025 Company announcement, not peer reviewed. Eli Lilly and Company, ACR Convergence 2026 preview, November 5, 2026 Company announcement, not peer reviewed. ClinicalTrials.gov, TRIUMPH-4 (NCT05931367) and TRIUMPH-1 (NCT05929066) Trial registry. Giblin K, et al., TRIUMPH rationale and design, Diabetes, Obesity and Metabolism, 2025 Journal article. New England Journal of Medicine, TRIUMPH-4 paper, November 8, 2026 Peer-reviewed journal.