Retatrutide reduced knee pain more than placebo in a phase 3 trial of adults with excess weight and knee osteoarthritis. The findings highlighted in Lilly’s November 5 ACR preview are encouraging, but the sponsor’s “75.8%” pain-reduction figure needs context: it describes change from baseline in a post-hoc analysis, not a 75.8% advantage over placebo.
The more informative comparison is the difference between the randomized groups on the same pain scale. The published conference abstract supplies that comparison, while its accompanying safety table adds reasons to look beyond the headline. Retatrutide remains investigational; these findings do not create an approved treatment for knee osteoarthritis.
Who was studied, and what changed?
ACR abstract 0871 reports TRIUMPH-4, a 68-week, double-blind trial with 445 participants. Adults received weekly injections of retatrutide at research target doses of 9 mg or 12 mg, or placebo. These are experimental regimens, not dosing instructions.
The trial registry required a body mass index of at least 27, persistent knee pain and moderate osteoarthritis changes on an X-ray. It excluded diabetes. Participants’ average age was about 55 and average BMI about 40. The results therefore concern a defined population, not everyone with knee pain.
Researchers measured pain with the WOMAC pain subscale, scored from 0 to 10. The average starting score was 5.95. The abstract reports these modeled changes at week 68:
| Research group | Mean pain-score change | Mean body-weight change |
|---|---|---|
| Retatrutide 9 mg | −4.48 points | −26.4% |
| Retatrutide 12 mg | −4.41 points | −28.7% |
| Placebo | −2.43 points | −2.1% |
Lower pain scores represent improvement. Subtracting the placebo change gives an additional reduction of about 2 points with either retatrutide regimen. Both comparisons were statistically significant. The placebo group also improved substantially, which is why change from baseline alone cannot describe the treatment’s added benefit.
Why the percentage headline needs a footnote
Lilly explicitly states that percentage change in WOMAC pain was not prespecified and came from a post-hoc analysis. The registered co-primary endpoints were the change in pain-score points and percentage change in body weight. Those are different measurements, even though both can appear as percentages in a news summary.
A post-hoc calculation can help describe results, but it should not be presented as though it were the trial’s planned primary pain endpoint. Nor does an average percentage reduction mean that the same percentage of participants became pain-free.
The abstract’s figure and caption identify the tabulated figures as efficacy-estimand, model-based estimates. The curves show observed on-treatment measurements, with fewer contributors at later visits. These estimates are not simply the average of everyone’s recorded final measurement, and they should not be relabeled as treatment-regimen results. The complete statistical analysis plan was not reviewed here, so the detailed assumptions for treatment interruptions and missing data remain outside this report’s scope. Our estimand guide explains why that distinction matters.
The second knee study is a separate comparison
ACR abstract 2573 reports 574 participants in the TRIUMPH-1 osteoarthritis group, followed for 80 weeks. It also found greater pain improvement with retatrutide than placebo. That supports the broader research question, but the different trial population and follow-up should stay visible: an 80-week result cannot be pooled casually with TRIUMPH-4’s 68-week estimate.
Our earlier TRIUMPH-1 report covers the wider study. This conference update focuses on knee symptoms. Neither abstract supplies an active-drug comparison establishing superiority over another weight-management medicine. Pain improvement also does not, by itself, establish cartilage repair or prevention of knee replacement.
Tolerability changes the reading of the benefit
The TRIUMPH-4 safety table includes 444 participants who received at least one dose. Adverse events led to treatment discontinuation in 12.9% of the 9 mg group and 18.2% of the 12 mg group, versus 4.0% with placebo. Nausea and diarrhea were common; the abstract describes gastrointestinal events as mostly mild to moderate and concentrated during dose escalation.
Serious adverse events occurred in 9.5%, 10.8% and 2.0%, respectively. There was one death in each retatrutide group and none in the placebo group. These are reported events, not proof that treatment caused them; the table does not provide the clinical narratives needed to interpret individual cases. A conference summary cannot resolve rare or long-term risks. Our tolerability guide explains why discontinuations belong beside efficacy estimates.
What this means for readers now
The FDA’s current notice says retatrutide is not an ingredient in any FDA-approved drug and has not been found safe and effective for any condition. A conference abstract is a research disclosure, separate from a regulatory decision. The retatrutide profile explains the medicine’s development context.
The useful question for a care discussion is how weight, pain and daily function are being assessed together. Our joint-pain movement guide covers broader activity considerations. This update does not justify obtaining an unapproved product or changing treatment. Further full reports should clarify durability, treatment discontinuation and safety, alongside any evidence about joint structure.
