Lilly's triple-hormone obesity drug cuts weight 22.6% — and lands six points below its own earlier trials

Eli Lilly's retatrutide produced 22.6% weight loss in a 1,949-patient phase 3 trial of people with severe obesity and established cardiovascular disease, hitting its primary endpoint at 80 weeks on the 12-milligram dose. The result lands about six points below the same molecule's earlier readouts in lower-acuity populations. That compression, paired with an unclear cardiovascular-outcome signal, complicates the case for the drug's glucagon arm in sicker patients.

Retatrutide is a single molecule engineered to activate three hormone receptors at once: GIP and GLP-1, the same targets as Lilly's tirzepatide, and glucagon, which is the new piece. The glucagon arm is the differentiator — it accelerates energy expenditure and liver-fat burning on top of the appetite suppression GIP and GLP-1 already deliver. Lilly's bet has been that adding glucagon would push weight loss past the company's own tirzepatide, which targets GIP and GLP-1, and Novo Nordisk's semaglutide, which targets GLP-1 alone.

In May, the TRIUMPH-1 trial reported 28.3% weight loss at 80 weeks in patients with obesity but no diabetes and no established cardiovascular disease. In December, TRIUMPH-4 reported 28.7% at 68 weeks in patients with obesity and knee osteoarthritis — also no diabetes. Both readouts came in roughly where Lilly's mechanism story predicted.

Thursday's two new trials, per Lilly's investor release, test the molecule in sicker cohorts. TRIUMPH-2 enrolled 1,152 patients with both type 2 diabetes and overweight or obesity and produced 20.8% weight loss at 80 weeks on the 12-milligram dose. TRIUMPH-3 enrolled 1,949 patients with severe obesity and established cardiovascular disease and produced 22.6%. Both had placebo arms at 3.2%. All of the trial percentages here reflect the "efficacy estimand" — the count of treatment-adherent patients without rescue therapy — which the regulator's intent-to-treat analysis will treat more conservatively at filing.

The 5-to-6-point compression is the story. Patients with type 2 diabetes tend to lose less weight on GLP-1-based drugs than patients without, a pattern seen across the class. The 22.6% in cardiovascular-disease patients is harder to map directly, because the cohort is older and on more background medications than the earlier readouts enrolled. The shape of the result, though, is consistent with the broader GLP-1 pattern: as patient complexity rises, the magnitude compresses.

TRIUMPH-3 was also the first cardiovascular-outcomes test of any GIP/GLP-1/glucagon molecule. The class needed a clean answer: GLP-1 drugs have already shown reductions in major adverse cardiovascular events (MACE — heart attack, stroke, and cardiovascular death) in diabetes and high-risk patients, and that win is the basis for premium pricing in obesity. Retatrutide's readout did not deliver one. The FierceBiotech coverage cut off mid-sentence at the MACE count, and Lilly's top-line described the result as "less clear-cut" rather than a positive readout. Without a statistically clear cardiovascular-outcomes win, retatrutide's label and pricing math look more like the base obesity case than the premium CV-expanded case.

In TRIUMPH-2, 7.7% of patients on retatrutide discontinued because of adverse events. In TRIUMPH-3, discontinuation climbed to 11.6% on the 9-milligram arm and 13.5% on the 12-milligram arm, against 11.3% in the lower-acuity TRIUMPH-1. The gastrointestinal profile on the 12-milligram dose was heavy: diarrhea in 33.6% of TRIUMPH-2 patients and 24.4% of TRIUMPH-3; nausea in 38% and 22.4%; constipation in 16.8% and 15.7%. These are not new safety signals for the class, but they are the rates real prescribers will see.

Citi analysts, writing in May around the TRIUMPH-1 readout, called that result nominally cleared against a roughly 25% placebo-adjusted bar but said it fell short of some expectations. The same framing now applies to TRIUMPH-3 — the bar was set by GLP-1 CV wins in adjacent populations, and the molecule did not clear it on the first try.

Lilly plans to file retatrutide for obesity in 2027, framing the two new readouts as supportive of that base case. The bar to watch is not the next percentage point. It is whether other obesity-drug readouts in the same patient populations — type 2 diabetes, established cardiovascular disease — show similar magnitude compression, or whether the triple-hormone mechanism delivers more in sicker patients than the GLP-1 backbone alone. If competing drugs compress the same way, the glucagon-marginality thesis softens: the third receptor may add less in sicker patients than the early readouts suggested. If they do not, retatrutide's 22.6% in this population is the most direct read on what the class can deliver where the cardiovascular stakes are real.