Three hormones.
One shot a week.
Thirty percent.
The first obesity drug to post a number that used to belong to surgeons. Here is what it does, what the five Phase 3 trials actually showed, and what it means for the next three years.
INVESTIGATIONAL · EDUCATIONAL FIELD GUIDE
Trial doses and outcomes describe research, not personal treatment recommendations. Consult a licensed clinician about your care.
Chapter 1 · The number
Average body weight lost at 104 weeks on the 12 mg dose in TRIUMPH-1's extension cohort. That's 85 pounds off a 268-pound starting weight — for the average person in the group, not the best responder.
For a sense of scale: a placebo shot plus the same diet-and-exercise coaching produced 2.2% at 80 weeks. Nearly two in three people on 12 mg finished the trial with a BMI under 30 — clinically, no longer obese.
Where 30% sits · average total body weight loss
Semaglutide and tirzepatide figures are from their own pivotal obesity trials (STEP 1 at 68 weeks; SURMOUNT-1 at 72 weeks), not a head-to-head. Cross-trial comparisons are directional, not proof. The head-to-head against tirzepatide (TRIUMPH-5) hasn't read out yet.
Chapter 2 · Why three is different from two
One molecule, three receptors.
Think of it as three keys on one ring.
Semaglutide (Ozempic/Wegovy) turns one key. Tirzepatide (Mounjaro/Zepbound) turns two. Retatrutide turns all three — and the third key is the one nobody had dared to use before. Tap a ring to see what each one does.
Chapter 3 · The lineage
Each generation added a key.
Each key added roughly five to ten points.
Bars are scaled to 35%. Different trials, different populations, different durations — the direction is clear, the decimals are not comparable. Retatrutide's 80-week number is the primary endpoint; the 104-week 30.3% is from a heavier extension cohort (BMI ≥ 35).
Chapter 4 · Five Phase 3 trials, five wins
The TRIUMPH program didn't test one population. It tested the hard ones.
Over 5,800 people. Each trial picked a group where weight loss is usually harder or matters more. Every one of them hit its primary endpoint.
TRIUMPH-1 · Dose-response at 80 weeks · % body weight lost
Bars scaled to 35%. Efficacy estimand (what happens if you stay on the drug). The treatment-regimen estimand — everyone randomized, including those who stopped — was −25.0% at 12 mg, still well ahead of anything approved.
Chapter 5 · The furnace at work
The glucagon key shows up where you'd expect: the liver and the joints.
Liver fat
Relative reduction in liver fat at 48 weeks on 12 mg in a Phase 2a fatty-liver study. 93% of that group ended with normal liver fat (under 5%). Placebo: −4.6%.
Knee pain
Drop in WOMAC pain score at 68 weeks on 12 mg in TRIUMPH-4 (4.4 points on a 10-point scale). Placebo with the same coaching: −40%. One in eight finished pain-free.
Chapter 6 · The part the reps won't lead with
More keys, more torque.
The side-effect table scales with the dose.
Nothing new in kind — it's the same gut story as every GLP-1 — but the 12 mg arm sits at the high end of what's been seen in this class, and about one in nine people stopped because of it.
| TRIUMPH-1, 80 weeks | Placebo | 4 mg | 9 mg | 12 mg |
|---|---|---|---|---|
| Nausea | 14.8% | 28.6% | 38.4% | 42.4% |
| Diarrhea | 13.5% | 25.2% | 34.1% | 32.0% |
| Vomiting | 4.8% | 10.6% | 22.8% | 25.3% |
| DysesthesiaSkin tingling or hypersensitivity — new to this class, mostly mild, mostly resolved on treatment | 0.9% | 5.1% | 12.3% | 12.5% |
| Stopped because of side effects | 4.9% | 4.1% | 6.9% | 11.3% |
In TRIUMPH-4's smaller, heavier cohort, discontinuation reached 18.2% at 12 mg. The 9 mg dose captured most of the efficacy with a gentler curve — expect a lot of real-world patients to live there.
Chapter 7 · Where it is on the road
Not approved. Not close to a pharmacy shelf. And a fight over 41 amino acids.
Chapter 8 · Beyond the scale
Obesity isn't one disease. It's the engine behind six.
Turn the engine off and watch what happens downstream.
This is the part that matters more than any weight number. Fatty liver, kidney decline, sleep apnea, joint pain, type 2 diabetes and cardiovascular disease are largely the same disease wearing different uniforms. Here is what the trials have shown so far for each — and how strong each piece of evidence really is.
Type 2 diabetes
TRANSCEND-T2D-1 at 40 weeks; TRIUMPH-2 added −1.5 pts on top of 21% weight loss. For many patients that is a return to non-diabetic range — remission-level control, not a cure of the underlying biology.
Sleep apnea
In the TRIUMPH-1 OSA sub-study, breathing interruptions fell from 59 per hour to about 22 at 80 weeks — a 36-event drop. The average AHI moved from severe into the moderate range; this does not establish that prescribed CPAP can be stopped.
Knee osteoarthritis
TRIUMPH-4. The cartilage doesn't regrow, but 70 pounds off the joint plus lower inflammation produced pain relief that rivals what people report after a replacement.
Fatty liver (MASLD)
At 48 weeks on 12 mg, nearly everyone had normal liver fat. This is the strongest liver signal of any drug in the class. What's not yet shown: biopsy-proven reversal of scarring (MASH fibrosis). That trial is the next step.
Kidney disease
Protein leak into the urine — the earliest marker of kidney damage — fell by a third in a post-hoc look at the Phase 2 trials. Exploratory, not proof. TRANSCEND-CKD (146 people, measured GFR) and the 10,000-person TRIUMPH-Outcomes trial (2029) will answer whether it actually slows kidney decline.
Heart disease
TRIUMPH-3 moved every risk marker the right way: triglycerides −37%, systolic BP −9 mmHg, inflammation halved. Whether that translates into fewer heart attacks and strokes is exactly what TRIUMPH-Outcomes exists to prove. The in-trial event count leaned favorable but was too small to conclude.
Chapter 9 · Implications
What changes if a shot does what surgery did.
For patients
Obesity stops being a condition you manage and becomes one you can reverse — for the majority, not the lucky. And the diagnoses that ride with it — fatty liver, sleep apnea, prediabetes, aching knees — start coming off the chart with it.
The open questions are the ones this class always had: what happens when you stop, and how much of the loss is muscle. TRIUMPH-6 is the maintenance answer; the lean-mass data isn't public yet.
For clinicians
Dose titration becomes the skill. 4 mg is a serious drug on its own; 12 mg is a serious drug with a serious GI tax. Expect fatty liver and knee OA to become on-label reasons to prescribe, and expect referrals for bariatric surgery to keep falling.
The three-receptor mechanism also means new things to watch for — dysesthesia is a new line on the counseling sheet.
For the market
Analysts model roughly $15 billion a year at maturity — not the "trillion-dollar drug" of podcast lore, but the most valuable single pipeline asset in the industry. Novo's answer (CagriSema, amycretin) is behind; Lilly's own tirzepatide is the one it has to beat head-to-head.
The biologic-vs-drug ruling decides whether that revenue lasts five years or twelve before generics.
The one-line version
Two keys got us to twenty.
The third key gets us to thirty —
and into the liver, the kidneys,
the airway and the joints.
Filing early 2027. Watch the head-to-head, the maintenance trial, the 40-amino-acid ruling — and the 2029 outcomes trial that decides whether "risk markers" become "lives."
Sources
Lilly, TRIUMPH-1 topline, May 21 2026 · Lilly, TRIUMPH-2 & -3 topline, July 23 2026 · Lilly, TRIUMPH-4 topline, Dec 11 2025 · Sanyal et al., Nature Medicine 2024 (MASLD Phase 2a) · Jastreboff et al., NEJM 2023 (Phase 2) · AJMC on TRIUMPH-1 · AJMC, "trillion-dollar drug" context · BioSpace on the biologic classification dispute · GLP-3 Wiki trial tracker (TRANSCEND, TRIUMPH-5/6/Outcomes status) · Sleep Review on the TRIUMPH-1 OSA sub-study (ADA, June 2026) · Kidney parameters post-hoc analysis of Phase 2 (Kidney Int Rep 2025) · TRANSCEND-CKD design paper (NDT 2026). AASM: OSA severity definitions. Semaglutide 14.9% is STEP 1 (Wilding, NEJM 2021); tirzepatide 20.9% is SURMOUNT-1 (Jastreboff, NEJM 2022). Retatrutide is investigational and not approved anywhere. This is education, not medical advice.