A weight-loss drug eased knee arthritis pain in two major trials

By Dr. Liji Thomas, MDReviewed by Lauren HardakerNov 10 2026

An investigational triple-action weight-loss drug significantly eased knee pain and improved mobility in two major trials, but higher-dose side effects and unanswered questions about long-term joint health raise concerns despite the promising results.

Study: Retatrutide Once Weekly in Adults with Knee Osteoarthritis Pain and Obesity. Image credit: Studio Romantic/Shutterstock.com

A recent paper published online in The New England Journal of Medicine reported on two phase 3 trials, TRIUMPH-4 and TRIUMPH-1-OA, which showed that retatrutide reduced knee pain and improved function significantly more than placebo and produced weight loss of about 17% to 25%.

Retatrutide targets three hormone receptors to promote weight loss

Obesity is a risk factor for knee osteoarthritis and increases the rate of progression primarily by increased joint loading, added to cartilage degeneration through increased joint stress, and obesity-related low-grade systemic inflammation. Thus, weight loss is an important component of knee osteoarthritis management in people with excess weight.

The glucagon-like peptide-1 receptor agonists (GLP-1RAs) like semaglutide have proven their ability to induce marked weight loss alongside significant metabolic improvements. A recent semaglutide trial showed the potential for reduced knee pain with GLP-1 RAs.

Retatrutide is an investigational once-weekly injection that activates two receptors in addition to GLP-1: glucose-dependent insulinotropic polypeptide (GIP) and glucagon. Earlier phase 2 trials with retatrutide showed weight loss of up to 24.2% at 48 weeks.

The current paper reported the results of two phase 3 trials aimed at assessing changes in knee pain associated with this weight loss.

Two phase 3 trials evaluate retatrutide for knee pain

Eli Lilly, which is developing retatrutide, funded the trials. Both trials were double-blind, randomized, and placebo-controlled.

They enrolled adults without diabetes who had chronic knee pain, X-ray evidence of osteoarthritis, and a body mass index (BMI) of at least 27 in TRIUMPH-4, while TRIUMPH-1-OA required a BMI of at least 30, or at least 27 with an obesity-related complication. TRIUMPH-4 had 445 participants, given retatrutide at dosages of 9 mg, 12 mg, or placebo for 68 weeks. TRIUMPH-1-OA had 574 participants, given the drug at 4 mg, 9 mg, 12 mg, or placebo, for 80 weeks.

In all intervention arms, the drug was started at 2 mg, and the dose increased every 4 weeks to the target dose. The primary outcomes were change in knee pain assessed by the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (both trials) and percent change in body weight (TRIUMPH-4). Pain-related outcomes were analyzed using a hybrid estimation method that uses a hypothetical strategy to analyze intercurrent events indicating treatment failure, while analyzing other events using a treatment-policy strategy. A pure intention-to-treat analysis was reported alongside for comparison. Weight-related outcomes were analyzed regardless of treatment discontinuation.

At the start of the trials, about 83% of participants had a BMI of at least 35, and the mean body weight was about 112 kg. At baseline, all participants registered 4 to 9 for knee pain on the WOMAC pain scale (an ascending pain scale from 0 to 10), with the mean being 6.

Retatrutide significantly reduces knee osteoarthritis pain in both trials

Both trials showed significant reductions in knee pain among participants receiving retatrutide, with improvements consistently exceeding those seen with placebo.

In TRIUMPH-4, participants receiving retatrutide reported reductions of approximately 3.7 to 4 points on the 10-point WOMAC pain scale after 68 weeks, compared with just over 2 points in the placebo group. This translated into an additional improvement of approximately 1.6 to 1.8 points over placebo, with both doses producing statistically significant benefits.

Similar findings emerged from TRIUMPH-1-OA, where all three retatrutide doses significantly reduced knee pain after 80 weeks. Participants improved by approximately 3.2 to 3.6 points, compared with 1.9 points among those receiving placebo. The additional pain reduction ranged from 1.35 to 1.77 points, with moderate to large treatment effects (Cohen's d of 0.55 to 0.79).

The intention-to-treat analyses showed somewhat smaller differences from placebo, but these remained statistically significant.

Physical function scores improve alongside reductions in knee pain

Physical function on the WOMAC scale showed the same direction of improvement, with differences of roughly 1.7 to 1.95 points from placebo.

Most retatrutide recipients achieve at least 50% pain reduction

In pooled data at week 68, an estimated 63% to 70% of retatrutide participants had at least a 50% drop in pain, compared to 38% on placebo. For physical function, the estimated figures were 63.5% to 64.6% against 33.1%.

Similar results were obtained following re-analysis including only participants with BMI of 35 or above.

Retatrutide produces up to 25% average weight loss

In TRIUMPH-4, body weight fell by 20.0% in the 9 mg group, and 23.7% with 12 mg retatrutide, compared to 4.6% with placebo. Moreover, an estimated 90% to 95% of retatrutide participants lost at least 5% of body weight.

In TRIUMPH-1-OA, weight reduced by 17.5%, 22.9%, and 24.8%, in the 4 mg, 9 mg, and 12 mg groups, respectively, against 4.3% with placebo. In TRIUMPH-4, an estimated 33.7% lost 25% or more of their baseline weight while on 9 mg retatrutide, and 45.1% with 12 mg, against 3.1% on placebo.

Higher retatrutide doses show increased serious adverse event rates

Although most participants completed the trials, remaining on the assigned treatment proved more challenging, particularly at higher retatrutide doses. Overall, approximately 85% of participants completed TRIUMPH-4 and 83% completed TRIUMPH-1-OA, but only 72% and 75%, respectively, completed their assigned course of retatrutide or placebo.

Safety findings also highlighted concerns with higher doses. Serious adverse events were generally more frequent in the higher-dose retatrutide groups than with placebo, affecting approximately 10% to 11% of participants receiving 9 mg or 12 mg in TRIUMPH-4, compared with 2% on placebo. In TRIUMPH-1-OA, rates ranged from 4.2% at 4 mg to 14.6% at 12 mg, compared with 6.9% in the placebo group.

Adverse events also contributed to treatment discontinuation, with rates generally increasing at higher doses. At 12 mg, approximately 18% of participants in TRIUMPH-4 and 15% in TRIUMPH-1-OA stopped treatment because of adverse events, compared with around 4% to 6% receiving placebo.

Gastrointestinal symptoms, particularly nausea, diarrhea, and constipation, were the most common adverse events. Although generally mild to moderate and concentrated during dose escalation, they accounted for approximately half of adverse event-related treatment discontinuations in TRIUMPH-4 and one-third in TRIUMPH-1-OA.

Another notable finding was dysesthesia, an abnormal skin sensation that was generally mild but considerably more frequent with retatrutide than placebo. It affected up to approximately 21% of participants in TRIUMPH-4 and 17% in TRIUMPH-1-OA, compared with 0.7% receiving placebo in each trial.

Two deaths occurred in TRIUMPH-4, one each in the 9 mg and 12 mg groups, although investigators considered neither related to treatment. No deaths were reported in TRIUMPH-1-OA.

Cancer events were also reported, although their frequency varied across treatment groups without a consistent dose-related pattern. In TRIUMPH-4, cancer was reported in 4.8% of participants receiving 9 mg, compared with 1.4% at 12 mg and 0.7% with placebo. The authors did not specifically discuss this finding, and the trials were not designed to determine whether retatrutide influences cancer risk.

Only two-thirds of participants maintained the highest retatrutide dose

Only about two-thirds of the 12 mg groups were still on the assigned dose at the end of the trials. Dose reductions were allowed for gastrointestinal effects, poor oral intake, or excessive weight loss. Conversely, 93.7% of participants on 4 mg were on the same dose by this point.

Mechanical and inflammatory explanations for pain relief are considered

According to the authors, the pain relief offered by retatrutide may compare favorably with existing treatments, with effect sizes of 0.37 for ibuprofen, 0.31 for tramadol, 0.61 for intra-articular steroid injections, and 0.50 for semaglutide. However, they could not perform direct comparisons due to differences in study design.

They propose that the pain benefit may result in part from reduced joint loading, with possible anti-inflammatory contributions, though the mechanism has not been established.

Trials cannot establish whether retatrutide prevents joint deterioration

The study has several limitations. The first pain assessment after baseline was at week 16, leaving the speed of onset of the drug’s action unknown. Inflammatory markers were measured only at the start and end of the trial.

The participants were overweight or obese with osteoarthritis, limiting the generalizability of the results. The trials were not designed to test whether retatrutide slows joint damage or prevents knee replacement. Whether improvements persist after treatment discontinuation also remains unknown. The dropout rate was substantial: only 72% completed assigned treatment in TRIUMPH-4 and 75% in TRIUMPH-1-OA. In TRIUMPH-1-OA, 62.5% of the placebo group completed treatment.

Retatrutide reduces knee pain while delivering substantial weight loss

In people with obesity and painful knee osteoarthritis, retatrutide reduced knee pain and improved function significantly more than placebo, and produced weight loss of about 17% to 25%. On the other hand, discontinuation because of adverse events reached 15% to 18% at the 12 mg dose, serious adverse events were more frequent in the higher-dose groups than with placebo, and long-term joint outcomes remain unknown.