Retatrutide

Status In clinical development, not approvedBest evidence E2Sources checked 2026-09-17

Retatrutide is an investigational triple receptor agonist that is not approved anywhere in the world and has no marketing application on file. Its most quoted weight loss figures come from an on-treatment analysis and from a selected extension subset, and every phase 3 obesity result in circulation is an unpublished press release. Testing of 18…

Identity

Sequence: Y-Aib-QGTFTSDYSI-(alpha-methyl-Leu)-LDKK-AQ-Aib-AFIEYLLEGGPSSGAPPPS-NH2, 39 residues, with a C20 fatty diacid on the lysine side chain via a gamma-glutamate linker and an AEEA spacer, and a C-terminal serinamide
CAS: 2381089-83-2

What it is

Retatrutide, code LY3437943, is a 39 amino acid synthetic peptide that activates three receptors: the glucose-dependent insulinotropic polypeptide receptor, the GLP-1 receptor and the glucagon receptor. A C20 fatty diacid on a lysine side chain binds albumin, giving a half life of about six days. Eli Lilly developed it. It is approved nowhere in the world.

Mechanism, and what the GIP arm actually does

In vitro the half maximal effective concentrations are 0.0643 nanomolar at the GIP receptor, 0.775 at the GLP-1 receptor and 5.79 at the glucagon receptor, roughly 12 times more potent at GIP than at GLP-1 and 90 times more than at glucagon. Against the native hormones it is 0.3 times as active as glucagon, 0.4 times as active as GLP-1 and 8.9 times as potent as native GIP.

Calling it a balanced triple agonist misreads the sponsor’s own sentence, which reads balanced GCGR and GLP-1R activity but more GIPR activity. Balanced covers two receptors and explicitly exempts the third.

The GIP arm is agonism, not blockade, and the most potent of the three. Why that helps is not known: Rosenkilde and colleagues report benefits from both GIP receptor antagonism and agonism, with the mechanism of the paradox unknown. No human study has isolated the GIP contribution inside retatrutide.

The 24.2 percent figure, and the word doing the work in it

It comes from a phase 2 trial of 338 adults with obesity and without type 2 diabetes, seven arms, 48 weeks. The 12 mg arm showed a least squares mean weight reduction of 24.2 percent against 2.1 percent on placebo, a secondary endpoint. The primary endpoint, at 24 weeks, was 17.5 percent.

It is also an efficacy estimand, which is where most weight loss claims come apart. That analysis keeps every randomised participant but discards data collected after a person permanently stops the drug, then models the rest as though they had stayed on. It is not a completer analysis, which restricts to those who finished, and not intention to treat, which keeps the post-discontinuation data. It answers how much weight people lose while taking the drug, not how much a population loses if you start them all on it.

Phase 3 quantifies the gap. In TRIUMPH-1 at week 80 the 12 mg arm was 28.3 percent on the efficacy estimand and 25.0 on the treatment regimen estimand, 3.3 points apart, while placebo moved the other way, 2.2 against 3.9 percent. The effect against placebo is therefore 26.1 points on one analysis and 21.1 on the other, about 5 points from the choice of analysis alone.

Both 28.3 and 30.3 percent circulate as the TRIUMPH-1 result. Both are real and describe different things. The 28.3 percent is the week 80 efficacy estimand for the 12 mg arm, all 2,339 randomised participants. The 30.3 percent is a week 104 figure from an extension of 532 people with a BMI of 35 or more at week zero who completed 80 weeks without discontinuing or reducing dose, a group selected for having tolerated the full dose that long.

The phase 3 programme, and what a press release is worth

The programme is larger than it looks: TRIUMPH-1 through TRIUMPH-9, plus TRIUMPH-Outcomes, a 10,000 participant cardiovascular and kidney outcomes trial reporting in 2029, plus a separate TRANSCEND programme in type 2 diabetes and chronic kidney disease. TRIUMPH-9, 600 participants over 104 weeks, exists solely to find the best dose escalation scheme and reports in October 2028.

That settles something. Every week by week titration table in circulation presents itself as a protocol, yet the sponsor, holding all of its own data, does not know which scheme is best and is running a two year trial to find out. All that is published about phase 3 escalation is that it is fixed, lasts 16 weeks and ends at 4, 9 or 12 mg.

Every phase 3 obesity figure in circulation is a press release figure. TRIUMPH-4, 445 people, knee osteoarthritis, 68 weeks, reported in December 2025: weight down 26.4 and 28.7 percent at 9 and 12 mg against 2.1 on placebo, and WOMAC knee pain down 4.5 and 4.4 points against 2.4, so placebo captured more than half the pain response. In July 2026 TRIUMPH-2 in type 2 diabetes reported 20.8 percent at 12 mg, TRIUMPH-3 in severe obesity with cardiovascular disease up to 22.6 percent. The TRIUMPH-1 sleep apnoea basket reported the apnoea-hypopnoea index falling by up to 36.1 events an hour, 60.6 percent from a baseline of 58.6, with no placebo figure in the announcement.

None of those four has a peer reviewed publication or posted registry results. The only peer reviewed phase 3 paper is TRANSCEND-T2D-1 in the Lancet, 537 people with type 2 diabetes over 40 weeks, 15.3 percent weight reduction at 12 mg on the treatment regimen estimand. A press release is not a publication: no tables, no confidence intervals, no adverse event list, no external review. The headline cannot be interrogated.

The plateau question is real. The phase 2 authors flagged that weight had not plateaued at 48 weeks, and week 80 to week 104 corroborates it, so the curve is still descending wherever it has been measured and the asymptote is unknown. That is not the same as continuing indefinitely. Hepatic fat is the contrast: it plateaus near maximally by week 24, at about 20 percent weight loss, on a stated floor effect.

The liver data, and its two ceilings

In the MASLD substudy, 98 people with liver fat of 10 percent or more by MRI, relative liver fat fell 82.4 percent at 12 mg at 24 weeks, a difference against placebo of 82.7 percent, 95 percent confidence interval 95.2 to 70.2. Liver fat below 5 percent was reached by 86 percent of that arm at 24 weeks and 93 percent at 48 weeks.

Two qualifiers matter more. No participant in any retatrutide trial has had a liver biopsy, so resolution of steatosis on imaging is not resolution of steatohepatitis, and 48 week MRI data were missing for 56.1 percent of the substudy. The authors call it hypothesis generating, not definitive.

Safety, with denominators

In TRIUMPH-1 at 12 mg, nausea occurred in 42.4 percent against 14.8 on placebo and vomiting in 25.3 against 4.8. Dysesthesia, an altered or unpleasant sensation, occurred in 5.1, 12.3 and 12.5 percent at 4, 9 and 12 mg against 0.9 percent on placebo. That is roughly fourteen times the placebo rate, it is absent from the phase 2 record, and no mechanism has been published for it.

Discontinuation because of adverse events tracks dose. TRIUMPH-1: 4.1, 6.9 and 11.3 percent against 4.9 on placebo. TRIUMPH-3: 9.8 and 13.5 against 4.8. TRIUMPH-4: 12.2 and 18.2 against 4.0, the highest in the programme, in an older population with joint disease and no 4 mg arm.

In phase 2, heart rate rose by up to 6.7 beats per minute, peaking at 24 weeks and declining afterwards. No heart rate data have been reported for any of the four phase 3 obesity trials, although pulse is collected across the programme per the design paper. The data exist and are unreported, which for a drug given 80 weeks is itself a finding.

Regulatory status

Retatrutide is not approved by any regulator anywhere, and no marketing application has been filed. The sponsor said in July 2026 that it plans to submit a Biologics License Application to FDA in the first quarter of 2027, a target that had already slipped because more manufacturing and quality control data were needed. A submission is not a review, and a review is not an approval.

No breakthrough therapy, fast track, priority review, orphan or PRIME designation is documented in any source located. That is a statement about the documents, not a claim that none exists.

A pre-approval expanded access programme exists and is narrow: physician initiated, for a BMI of 35 or more with two or more serious or life-threatening obesity-related complications, documented failure at the highest dose of an approved therapy, no accessible trial, and a prior discussion of all standard options including bariatric surgery. It is supply limited, and it is not availability.

What circulates, and why no numbers appear here

This site does not reproduce the circulating retatrutide figures, because the compound is unapproved, the harm sits in the escalation rather than the molecule, and the vials have been shown not to hold what the label says.

Product identity is the most useful fact here. Eighteen samples sold as retatrutide in Australia were tested in 2026. One matched its label. About 30 percent were more concentrated than labelled, about 30 percent less. One vial contained none of the peptide at all. The peptide typically made up only about 10 percent of the vial contents, leaving, in the analytical chemist’s words, another 90 per cent that is unaccounted for, which could contain a different toxin or something else unsafe. These figures come from the reported testing, because the primary publication was not retrievable.

Over-concentration was as common as dilution, which inverts the usual assumption that sellers cut to save money. For a molecule with a six day half life and dose-dependent adverse events, being unknowingly above the intended amount is the worse error. A purity certificate does not address it: purity and content are different measurements.

In a letter to the Federation of State Medical Boards dated 2 April 2025, FDA cautioned against unapproved products containing retatrutide that are being sold directly to consumers, often with false labeling and dosing instructions, and stated that compounded retatrutide products do not currently meet the requirements for exemptions under sections 503A or 503B of the Federal Food, Drug, and Cosmetic Act. Having no USP or NF monograph and not being a component of an approved drug product, it fails both pathways. A warning letter of 31 March 2026 established that a not-for-human-consumption disclaimer is no shield.

On the reported death, three clauses belong together or none do. One death has been reported in the United Kingdom, and the MHRA has logged 77 suspected adverse reaction reports. A suspected report is a suspicion, not a finding, and causation has not been established. And the exposure denominator is unquantified: given the testing above, nobody knows how many of those people received retatrutide, or how much.

What the record gets wrong

Three times as effective as semaglutide. No head to head trial against semaglutide in obesity exists. The only registered comparison, TRANSCEND-T2D-2, is in type 2 diabetes and open label. Every ratio in circulation is arithmetic across separate trials differing in duration, population and estimand. TRIUMPH-5, the only head to head against tirzepatide, reports in November 2026.

Retatrutide beat tirzepatide. Like for like on estimand, TRIUMPH-1 at 12 mg was 25.0 percent at week 80 and SURMOUNT-1 at 15 mg was 20.9 percent at week 72. The usual presentation, 28.3 against 20.9, sets an on-treatment analysis against an intention to treat one and makes the gap look about 80 percent larger than it is. It also omits the discontinuation figures, which run the other way.

Retatrutide is being studied in heart failure with preserved ejection fraction. There is no such trial. Heart failure appears only as one component of the TRIUMPH-Outcomes composite, as an adjudicated safety event and as an exclusion criterion. The claim was almost certainly imported from tirzepatide.

The problem is not confined to vendor pages. A 2025 review in Biomolecules says the phase 2 obesity trial used doses up to 8 mg; it used 12 mg, and the same paragraph then reports the 12 mg results correctly. A 2026 review in Cardiology in Review gives a 23.2 percent fat mass reduction as though it came from the obesity trial. It comes from the DXA substudy of the type 2 diabetes trial, it is not the maximum, because the pooled 8 mg arm reached 26.1 percent, and the same sentence calls it comparable to bariatric surgery with no comparator in the cited data. Go to the trial publication and the registry record, never to a review, for any number.

What is not known

Its molecular weight. Figures circulate in chemical catalogues, but no primary publication or regulatory document names one, so none is asserted here.

Whether it prevents anything. No cardiovascular or kidney outcome data exist. TRIUMPH-Outcomes, 10,000 participants across 743 sites, has a primary completion date of February 2029. Blood pressure, lipoproteins and inflammatory markers all move favourably and heart rate moves unfavourably, and none of those is an outcome.

What happens when it stops. The entire published record is one sentence in the MASLD substudy noting some weight regain at the safety follow-up four weeks after discontinuation, which for a molecule with a six day half life is less than five half lives. TRIUMPH-6, the randomised withdrawal trial, reports in April 2028. Until then, claims that losses are durable and claims that regain is rapid are equally unsupported for this molecule.

What the liver findings mean histologically. No participant in any retatrutide trial has had a liver biopsy. The hepatic data are MRI fat fraction and serum biomarkers, the substudy population was not enriched for steatohepatitis or significant fibrosis, and the phase 3 liver outcomes trial reports in August 2030.

What happens to muscle in people without diabetes. The only published body composition data come from a type 2 diabetes substudy in which 103 of 189 participants contributed paired scans, and the authors’ conclusion is about the proportion of lean mass lost, not absolute sparing. Because total weight loss is larger, a similar proportion means a larger absolute loss. The only DXA substudy in the phase 3 obesity programme sits inside TRIUMPH-3 and is unreported, and no muscle strength, grip, gait speed or physical function endpoint appears anywhere in the programme.

Anything about bone. There is no bone mineral density endpoint, no fracture endpoint and no bone turnover marker in any registered retatrutide trial, in any species. Native GIP induces osteoblast activity, which is a reason to expect benefit, and weight loss of 25 to 30 percent is itself associated with bone loss, which is a reason to expect harm. Which dominates has not been tested.

Why dysesthesia occurs in 12.5 percent of participants at 12 mg. No mechanism has been proposed in any source located, and the event is absent from the phase 2 record.

Anything in the populations every trial excluded: type 1 diabetes, anyone under 18, pregnancy and lactation, advanced fibrosis or cirrhosis, prior bariatric surgery, a history of pancreatitis, and anyone who had taken another weight loss drug within 90 days, which means no switching, add-on or sequencing data exist at all.

Doses used in published research

Amounts administered under a trial protocol. Not dosing, not a recommendation.
Trial	Population	Dose arms	Duration	What it measured
NCT04867785	281 adults with type 2 diabetes on metformin or on diet and exercise alone	Retatrutide 0.5, 4, 8 or 12 mg once weekly, dulaglutide 1.5 mg once weekly, or placebo; the 4 mg and 8 mg maintenance doses were each given twice, differing only in starting dose	36 weeks	Change in HbA1c at 24 weeks
NCT04881760	338 adults with obesity or overweight and a weight related condition, without type 2 diabetes	Retatrutide 1, 4, 8 or 12 mg once weekly, or placebo, across seven arms; the 4 mg and 8 mg arms each ran with a 2 mg or a 4 mg starting dose	48 weeks	Percent change in body weight at 24 weeks
TRIUMPH-1, NCT05929066	2,339 adults with obesity or overweight, without type 2 diabetes, including nested sleep apnoea and knee osteoarthritis baskets	Retatrutide 4, 9 or 12 mg once weekly after a 16 week fixed escalation, or placebo, 1:1:1:1	80 weeks, with an optional 24 week extension for about 500 participants	Percent change in body weight at week 80
TRANSCEND-CKD, NCT05936151	146 adults with overweight or obesity and chronic kidney disease, with or without type 2 diabetes	Retatrutide once weekly at the maximum tolerated dose up to 12 mg, or matched placebo, 1:1; not a fixed dose	24 weeks to the primary endpoint	Measured glomerular filtration rate by iohexol clearance
TRIUMPH-4, NCT05931367	445 adults with obesity or overweight and knee osteoarthritis, without type 2 diabetes	Retatrutide 9 or 12 mg once weekly after a fixed escalation, or placebo, 1:1:1	68 weeks	WOMAC pain subscale score and percent change in body weight, co-primary
TRANSCEND-T2D-1, NCT06354660	537 adults with type 2 diabetes inadequately controlled on diet and exercise alone	Retatrutide 4, 9 or 12 mg once weekly, or placebo, 1:1:1:1	40 weeks	Change in HbA1c at week 40
TRIUMPH-9, NCT07357415	600 adults with obesity or overweight, without type 2 diabetes	Different retatrutide dose escalation schemes, retatrutide only, no placebo; the schemes themselves are not published	104 weeks	Percent change in body weight at week 104

No dose of retatrutide has been established as safe or effective, because no regulator has reviewed the compound and no marketing application has been filed. What follows is a record of what trials administered, not an instruction.

The phase 2 obesity trial, NCT04881760, randomised 338 adults across seven arms to 1 mg, 4 mg, 8 mg or 12 mg once weekly for 48 weeks, or placebo. The 4 mg and 8 mg maintenance doses each appeared twice, differing only in whether the starting dose was 2 mg or 4 mg. The 2 mg start produced fewer gastrointestinal events at the same maintenance dose, and in the parallel phase 2 type 2 diabetes trial the fast escalation 8 mg arm had the highest gastrointestinal rate in the study at 50 percent while achieving no better result than the slow escalation arm.

The phase 3 TRIUMPH trials used maintenance doses of 4, 9 and 12 mg once weekly, or 9 and 12 mg only in TRIUMPH-3 and TRIUMPH-4, reached through what the design paper calls a fixed dose escalation regimen lasting 16 weeks, followed by 64 weeks of maintenance. The weekly step sizes inside that 16 weeks have never been published. TRANSCEND-CKD did not use a fixed dose at all, titrating to the maximum tolerated dose up to 12 mg.

The trials also attached things to the dose that no circulating schedule carries: permitted permanent dose reduction for gastrointestinal events after a de-escalation and re-escalation attempt, permitted reduction for anyone reaching a BMI of 22 or lower or who felt they had lost too much weight, and discontinuation at a BMI of 18.5 or lower.

Which escalation scheme is best is an open question the sponsor is still studying. TRIUMPH-9, 600 participants over 104 weeks, exists to compare escalation schemes and reports in October 2028.

Compiled from: Jastreboff 2023, New England Journal of Medicine; Rosenstock 2023, Lancet; NCT07357415, TRIUMPH-9; Heerspink 2026, Nephrology Dialysis Transplantation; Giblin 2025, Diabetes Obesity and Metabolism

No dose has been established as safe or effective for any indication.

Amounts above record what a named source reported administering in a study. They are not a recommendation, not a protocol, and not instructions.

Converting a vial and a syringe into a draw volume is a separate, mechanical question from what amount to use. The reconstitution calculator does that arithmetic for peptide, HCG, and HGH vials.