Four Phase 3 TRIUMPH trials of retatrutide have reported, and two now have full papers. In The Lancet on 29 September 2026, TRIUMPH-2 (type 2 diabetes) reports −18.8% body weight at week 80 on 12 mg against −5.1% on placebo, counting every randomised participant. Lilly's headline 20.8% uses the efficacy estimand, which assumes everyone stayed on treatment. TRIUMPH-1 (obesity without diabetes) reports −25.0% in NEJM on the same basis, 28.3% on Lilly's. TRIUMPH-4 showed 28.7% and TRIUMPH-3 22.6% in toplines. Lilly plans to file for FDA approval in Q1 2027.
Four Phase 3 trials of retatrutide have reported, and two now have full peer-reviewed papers. TRIUMPH-2, in people with type 2 diabetes, reports −18.8% body weight at 80 weeks on 12 mg against −5.1% on placebo in The Lancet's primary analysis. Lilly's headline figure for the same arm is 20.8%. Both numbers are correct; they come from different estimands, explained below. Lilly plans to file its Biologics License Application with the FDA in Q1 2027.
Retatrutide is not approved anywhere. Everything below is trial data.
What did TRIUMPH-2 show in The Lancet?
The Lancet paper (Bellido et al., 29 September 2026, doi 10.1016/S0140-6736(26)01861-1) randomised 1,152 adults with a BMI of 27 or more and type 2 diabetes across 92 sites in eight countries. Mean baseline BMI was 38.2 and mean HbA1c 7.71%. Participants took placebo or retatrutide 4, 9 or 12 mg weekly for 80 weeks, and 84% completed study drug.
| Week 80, body-weight change | 4 mg | 9 mg | 12 mg | Placebo |
|---|---|---|---|---|
| Treatment-regimen estimand (Lancet primary) | −11.9% | −16.8% | −18.8% | −5.1% |
| Difference vs placebo | −6.9 pts | −11.8 pts | −13.8 pts | |
| Efficacy estimand (Lilly headline) | −12.7% | −19.1% | −20.8% | −4.0% |
| HbA1c change, points (Lilly) | −1.4 | −1.6 | −1.5 | −0.2 |
| Lost ≥20% of body weight (Lilly) | 24.0% | 45.3% | 52.0% | 1.5% |
| Lost ≥25% of body weight (Lilly) | 10.8% | 30.6% | 34.9% | 0.8% |
Rows marked "Lilly" come from Lilly's TRIUMPH-2 release of the same day and use the efficacy estimand. In participants who started with a BMI of 35 or more, the 12 mg arm reached −23.4% on that basis.
Two details stand out. HbA1c fell about the same on 9 mg and 12 mg, so the top dose bought weight, not further glucose control. And the step from 9 mg to 12 mg added 2.0 percentage points of weight loss on the primary estimand, while the step from 4 mg to 9 mg added 4.9.
Why does Lilly report 20.8% when the paper says 18.8%?
The Lancet paper reports −18.8% on 12 mg at week 80 under the treatment-regimen estimand, which counts every randomised participant whether or not they stayed on the drug. Lilly's headline 20.8% uses the efficacy estimand, which estimates the effect had everyone stayed on treatment.
Neither number is wrong. The treatment-regimen figure is closer to what a population of real patients would see, because some people stop. The efficacy figure describes what the drug does in people who keep taking it. Lilly led with the efficacy number in its TRIUMPH-1 and TRIUMPH-2 releases, and news coverage that repeats it rarely names the estimand.
The same split applies to TRIUMPH-1. The NEJM paper (Jastreboff et al., 29 September 2026) reports −17.6%, −23.7% and −25.0% on 4, 9 and 12 mg against −3.9% on placebo under the treatment-regimen estimand. Lilly's May topline gave 28.3% for 12 mg on the efficacy estimand. When two sources disagree about retatrutide by two to three points, the estimand almost always explains it.
What is the TRIUMPH program?
TRIUMPH is Eli Lilly's Phase 3 program for retatrutide, a triple agonist of the GLP-1, GIP and glucagon receptors. The trials differ by population rather than by drug, so reading them together shows how the same doses behave in different patients.
| Trial | Population | Participants | Doses | Duration | Reported |
|---|---|---|---|---|---|
| TRIUMPH-4 | obesity/overweight + knee osteoarthritis, no diabetes | 445 | 9, 12 mg | 68 weeks | 11 Dec 2025 (topline) |
| TRIUMPH-1 | obesity/overweight, no diabetes; knee OA and sleep apnea subgroups | 2,339 (+532 extension) | 4, 9, 12 mg | 80 weeks (+24-week extension) | 21 May 2026 (topline); NEJM 29 Sep 2026 |
| TRIUMPH-2 | type 2 diabetes + obesity/overweight | 1,152 | 4, 9, 12 mg | 80 weeks | 23 Jul 2026 (topline); Lancet 29 Sep 2026 |
| TRIUMPH-3 | severe obesity (BMI ≥35) + established cardiovascular disease | 1,949 | 9, 12 mg | 80 weeks | 23 Jul 2026 (topline) |
What did each trial show?
| Trial | 12 mg, Lilly headline (efficacy estimand) | 12 mg, paper primary (treatment-regimen) | Other result |
|---|---|---|---|
| TRIUMPH-4 | 28.7% at 68 weeks | no paper yet | WOMAC knee pain −74.3% vs −40.3% placebo |
| TRIUMPH-1 | 28.3% at 80 weeks | 25.0% (NEJM) | 30.3% at 104 weeks, BMI ≥35 extension |
| TRIUMPH-2 | 20.8% at 80 weeks | 18.8% (Lancet) | HbA1c −1.5 points on 12 mg |
| TRIUMPH-3 | 22.6% at 80 weeks | no paper yet | primary endpoint met |
TRIUMPH-4 and TRIUMPH-3 figures are Lilly toplines. Compare like with like: a paper's primary figure next to another trial's press-release figure overstates the gap by two to three points.
Why is weight loss lower in the diabetes trial?
TRIUMPH-1 and TRIUMPH-2 ran the same doses for the same 80 weeks and differ mainly by whether participants had type 2 diabetes. On 12 mg:
- Efficacy estimand (Lilly): 28.3% without diabetes, 20.8% with.
- Treatment-regimen estimand (papers): 25.0% without diabetes, 18.8% with.
The gap is six to seven and a half points whichever basis you use. Semaglutide and tirzepatide show the same direction: tirzepatide's SURMOUNT-2, in type 2 diabetes, reported −14.7% on 15 mg at 72 weeks (The Lancet, 2023), well below its result in people without diabetes. Retatrutide behaving the same way is consistent with the drug class, and it means a single headline percentage without a population attached is misleading.
What about the side effects?
The TRIUMPH-2 paper gives per-dose rates in the safety population:
| TRIUMPH-2 event | 4 mg | 9 mg | 12 mg | Placebo |
|---|---|---|---|---|
| Diarrhea | 27% | 34% | 34% | 13% |
| Nausea | 14% | 21% | 28% | 8% |
| Dysesthesia | 4% | 6% | 7% | 1% |
| Hypotension | 1% | 5% | 6% | under 1% |
Dysesthesia, an abnormal skin sensation, and hypotension are the two signals the full papers flag beyond GI events. The NEJM TRIUMPH-1 paper notes that hypotension was more common in participants already on blood-pressure medication.
Discontinuation is the number that carries the most information:
| Stopped for side effects | 4 mg | 9 mg | 12 mg | Placebo | Source |
|---|---|---|---|---|---|
| TRIUMPH-1 | 4.1% | 6.9% | 11.3% | 4.9% | Lilly topline |
| TRIUMPH-2 (includes deaths) | 4% | 12% | 8% | 5% | Lancet |
| TRIUMPH-3 | n/a | 9.8% | 13.5% | 4.8% | Lilly topline |
TRIUMPH-2 breaks the pattern of the other two trials: its 9 mg arm lost more participants than its 12 mg arm. Lilly's decimals show the same order (11.6% vs 7.7%). The arms held about 285 people each, so the gap is about eleven participants. In TRIUMPH-1 and TRIUMPH-3, dropout rose steadily with dose and reached one in eight or more on 12 mg.
Seven participants died in TRIUMPH-2 (two, three and one on 4, 9 and 12 mg; one on placebo). Investigators judged all seven deaths unrelated to the study drug. The full breakdown, including heart rate, is on our retatrutide side effects page, and the dose ladder each arm followed is in retatrutide dosing protocols.
When could retatrutide actually be approved?
Lilly's 29 September 2026 release restates that it "plans to submit a Biologics License Application (BLA) in Q1 2027 to the U.S. Food and Drug Administration." The July release had tied that timing to completing chemistry, manufacturing and controls work.
A Q1 2027 filing under standard review puts an FDA decision in late 2027 or 2028. Any timeline claiming a 2026 approval or a 2026 submission is out of date.
Until then there is no legal prescription channel. Retatrutide is not eligible for 503A compounding, explained in our 503A and PCAC piece. Lilly runs a narrow expanded-access program and is suing online sellers; both are covered in where to buy retatrutide.
What the program has not answered
Sleep apnea in TRIUMPH-2. The trial included a basket for moderate to severe obstructive sleep apnea. Lilly says those results will be presented later.
No head-to-head trial against tirzepatide or semaglutide. Every comparison, including ours in retatrutide vs tirzepatide, is cross-trial.
No cardiovascular outcomes result. TRIUMPH-3 enrolled people with established cardiovascular disease and met its weight endpoint, but that is not a trial of heart attacks, strokes or cardiovascular deaths.
Body composition. How much of the loss is fat versus lean mass is not in the abstracts. Our muscle loss on GLP-1s piece covers what is known for the class.
Nothing beyond 104 weeks. The longest data is the TRIUMPH-1 extension.
How it ranks against Lilly's own next drugs. Lilly's amylin agonist eloralintide, alone and paired with tirzepatide, reported Phase 2 results in 2025 and 2026; we line those numbers up against TRIUMPH, with the cross-trial caveats, in eloralintide vs retatrutide.
Correction, 6 August 2026
The earlier version of this article was wrong in ways worth stating plainly.
It reported TRIUMPH-4 as showing 24% weight loss at 48 weeks in an obesity-plus-comorbidities population enriched for cardiovascular risk. TRIUMPH-4 ran 68 weeks, in obesity with knee osteoarthritis, and showed 28.7% at 12 mg. The article had TRIUMPH-3 and TRIUMPH-4 populations transposed: TRIUMPH-3 is the severe-obesity-plus-cardiovascular-disease trial.
It also listed all of TRIUMPH-1 through TRIUMPH-4 as published as of May 2026. TRIUMPH-2 and TRIUMPH-3 did not report until 23 July 2026. And it described a "2026 FDA submission target," which Lilly has since specified as a Q1 2027 BLA.
The program roster in the earlier version, showing five trials with a TRIUMPH-5 cardiovascular outcomes study, could not be verified against Lilly's own announcements and has been removed rather than rewritten. Logged in our corrections record.
Sources
- Bellido et al., TRIUMPH-2, The Lancet, 29 September 2026 (doi 10.1016/S0140-6736(26)01861-1)
- Lilly - TRIUMPH-2 full results and Q1 2027 BLA statement, 29 September 2026
- Jastreboff et al., TRIUMPH-1, NEJM, 29 September 2026 (doi 10.1056/NEJMoa2604169)
- Lilly - retatrutide successful in two additional Phase 3 obesity trials (TRIUMPH-2, TRIUMPH-3) - 23 July 2026
- Lilly - TRIUMPH-1 topline results - 21 May 2026
- Lilly - retatrutide weight loss up to 71.2 lbs with osteoarthritis pain relief (TRIUMPH-4) - 11 December 2025
- Garvey et al., SURMOUNT-2, The Lancet, 2023
- Triple-Hormone-Receptor Agonist Retatrutide for Obesity - NEJM Phase 2
- Internal: retatrutide dosing protocols, retatrutide side effects, retatrutide vs tirzepatide, where to buy retatrutide
Frequently asked
What did TRIUMPH-2 show in The Lancet?
TRIUMPH-2 randomised 1,152 adults with obesity and type 2 diabetes to retatrutide 4, 9 or 12 mg or placebo for 80 weeks. On The Lancet's primary analysis, weight fell 11.9%, 16.8% and 18.8% against 5.1% on placebo. HbA1c fell 1.4 to 1.6 points against 0.2 (Lilly). The paper was published on 29 September 2026.
Why does Lilly say 20.8% when The Lancet says 18.8%?
They answer different questions. The Lancet's 18.8% uses the treatment-regimen estimand, which counts every randomised participant, including people who stopped the drug. Lilly's 20.8% uses the efficacy estimand, which estimates the result had everyone stayed on treatment. Placebo reads 5.1% and 4.0% on the two bases.
How much weight did people lose on retatrutide in Phase 3?
It depends on the population and the estimand. On 12 mg: TRIUMPH-1, obesity without diabetes, 28.3% (Lilly) or 25.0% (NEJM primary) at 80 weeks. TRIUMPH-2, with type 2 diabetes, 20.8% (Lilly) or 18.8% (Lancet primary). TRIUMPH-4, knee osteoarthritis, 28.7% at 68 weeks. TRIUMPH-3, cardiovascular disease, 22.6%.
When will retatrutide be FDA approved?
No approval date exists. Lilly's 29 September 2026 release says it "plans to submit a Biologics License Application (BLA) in Q1 2027." A standard review after a Q1 2027 filing would place a decision in late 2027 or 2028. Retatrutide is not approved anywhere today.
What share of people lost 25% of their weight on retatrutide?
In TRIUMPH-2, on 12 mg, 34.9% of participants lost at least 25% of body weight and 52.0% lost at least 20%, against 0.8% and 1.5% on placebo. Those figures come from Lilly's release and use the efficacy estimand. Among participants with a starting BMI of 35 or more, 12 mg reached 23.4%.
What are the side effects of retatrutide in the Phase 3 trials?
Mostly gastrointestinal. In TRIUMPH-2, nausea affected 14%, 21% and 28% of the 4, 9 and 12 mg arms against 8% on placebo, and diarrhea 27% to 34% against 13%. The Lancet paper also flags dysesthesia (up to 7%) and hypotension (up to 6%). Our side-effects page has the full table.
Is retatrutide better than tirzepatide?
The weight numbers are larger, but no head-to-head trial has been run. In people with type 2 diabetes, TRIUMPH-2 reported −18.8% at 80 weeks on 12 mg, and tirzepatide's SURMOUNT-2 reported −14.7% at 72 weeks on 15 mg, both on the treatment-regimen estimand. Different trials, different populations, so treat the gap as indicative.