Survodutide

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

Survodutide is an investigational glucagon and GLP-1 receptor dual agonist, built on the glucagon backbone rather than the GLP-1 one, and it is not approved anywhere. Almost every figure quoted for it, including 16.6 percent, 18.7 percent and 83 percent, is a secondary analysis rather than the trial primary result. Its phase 3 adverse event…

Identity

Sequence: H-Ac4c-QGTFTSDYSKYLDERAAKDFI-X-WLESA-NH2, where Ac4c is 1-aminocyclobutane-1-carboxylic acid at position 2 and X is a glycine serine linker at position 24 carrying a C18 diacid; the C terminus is amidated. The modifications are from the 2022 medicinal chemistry paper; the full residue string as printed comes from a supplier listing that was truncated at the linker, so treat the linker notation as approximate
Formula: C192H289N47O61, from a chemical supplier data sheet only, not confirmed against a peer-reviewed or regulatory source
Molar mass: 4231.7, from a chemical supplier data sheet only, not confirmed against a peer-reviewed or regulatory source
CAS: 2805997-46-8, from vendor catalogues, and salt-form assignments in those catalogues are not reliable

What it is

Survodutide is a synthetic 29 amino acid peptide built on the human glucagon backbone, not the GLP-1 backbone. It is a dual agonist of the glucagon and GLP-1 receptors, engineered to reproduce what the gut peptide oxyntomodulin does weakly. Boehringer Ingelheim develops it as BI 456906, licensed from Zealand Pharma. It has no GIP receptor activity, is not a triple agonist, and is approved nowhere.

Calling it a GLP-1 analogue, which happens constantly, gets the molecule backwards. Zimmermann and colleagues, in the 2022 discovery paper, document three modifications to the glucagon sequence: the non coded residue Ac4c, 1-aminocyclobutane-1-carboxylic acid, at position 2, blocking dipeptidyl peptidase-4 cleavage; a glycine serine linker at position 24 carrying a C18 diacid, which binds albumin and gives once weekly dosing; and C terminal amidation. It was picked from a screen of nineteen dual agonists, alongside BI 456908 and BI 456897, which are different molecules.

Mechanism, including the part nobody explains

Two potency figures circulate and both are correct, because they are different assays. Zimmermann measured potency in CHO-K1 cells at 0.33 nM for the GLP-1 receptor and 0.52 nM for the glucagon receptor, near balanced and about ten fold weaker than native GLP-1 at 60 pM and glucagon at 20 pM. In 100 percent human plasma the same molecule gave 1.0 nM and 8.3 nM. That second ratio is the source of the sponsor statement in Nature Medicine 2026, verbatim: survodutide has an eight-fold greater activation of the human GLP-1R than the glucagon receptor in vitro. The drug circulates bound to albumin, so the plasma figure is the relevant one.

Glucagon is the counter-regulatory hormone to insulin, so agonising its receptor in an insulin resistant person looks like adding to the problem. That resolves empirically. In diet induced obese mice, Zimmermann ran survodutide, semaglutide and a glucagon receptor only agonist at weight matched doses: the glucagon only agonist raised blood glucose, the other two lowered it. In the phase 2 type 2 diabetes trial, HbA1c fell by up to 18.72 mmol/mol, which is 1.71 percent, at 1.8 mg weekly, with plasma alanine and glucagon falling as target engagement.

The cleanest mechanistic statement available is a 2026 Molecular Metabolism paper. It found the glucagon receptor barely detectable in the area postrema and arcuate nucleus at single cell level while the GLP-1 receptor is expressed there, then ran the control, verbatim: a long-acting GCGR agonist did not induce neuronal activation in satiety-mediating regions, nor reduced food intake but showed reduction in body weight. Appetite is the GLP-1 arm. Weight loss beyond appetite is the glucagon arm. In mice, survodutide suppressed food intake less than semaglutide and lost more weight.

What the human trials found

The phase 2 obesity trial, NCT04667377, treated 386 adults at 0.6, 2.4, 3.6 or 4.8 mg weekly or placebo for 46 weeks, 20 of escalation and 26 of maintenance. By planned treatment, the primary analysis, weight changed minus 6.2, minus 12.5, minus 13.2 and minus 14.9 percent against minus 2.8 percent placebo. Completion was 60.4 percent.

SYNCHRONIZE-1, NCT06066515, randomised 725 adults with obesity and without diabetes to a dose adjusted up to 3.6 or 6.0 mg, or placebo, for 76 weeks. On the prespecified primary analysis, the treatment regimen estimand, mean weight change was minus 12.2 percent (95 percent confidence interval minus 13.6 to minus 10.8) at 3.6 mg, minus 13.0 percent (minus 14.4 to minus 11.6) at 6.0 mg and minus 5.4 percent (minus 6.9 to minus 4.0) on placebo, with at least 5 percent lost by 72.6, 71.9 and 46.3 percent.

SYNCHRONIZE-MASLD, NCT06309992, treated 216 people at 6.0 mg or placebo two to one for 48 weeks, co-primary endpoints liver fat responder rate and weight. On the treatment regimen estimand, 68.5 against 28.6 percent achieved at least a 30 percent liver fat reduction, and weight fell 8.7 against 1.4 percent. Both endpoints were met on both estimands. A Nature Medicine author correction appeared on 25 August 2026 and its content is not indexed, so figures from that paper need checking against it.

Why the number you have read is a secondary analysis

Every circulating survodutide headline is real, and every one is a secondary analysis. The 16.6 percent figure is the efficacy estimand at 6.0 mg in SYNCHRONIZE-1, where the prespecified primary is minus 13.0 percent against minus 5.4 percent placebo. The 18.7 percent figure is the actual treatment sensitivity analysis of the 4.8 mg phase 2 arm, where the primary is minus 14.9 percent.

The tell is the placebo. A treatment regimen estimand includes what happened after people stopped, escalated slowly or took prohibited weight medication. An efficacy estimand models the trial as if everyone had adhered and nobody used those medications, which lifts the drug arm and drops the placebo arm. So placebo moves in the opposite direction to the drug: 5.4 percent becomes 3.2 in SYNCHRONIZE-1, and 2.8 becomes 2.3 in phase 2. The placebo adjusted difference nearly doubles, from 7.6 percentage points to 13.4. Find the placebo figure first, ask which way it moved, and you can audit any weight loss claim.

The MASH result, as published

The phase 2 MASH trial, NCT04771273, randomised 293 treated adults with biopsy confirmed MASH and F1 to F3 fibrosis to 2.4, 4.8 or 6.0 mg or placebo for 48 weeks, including 24 weeks of what the authors call rapid escalation. In the New England Journal of Medicine report, MASH improvement without worsening of fibrosis occurred in 47, 62 and 43 percent against 14 percent placebo, p less than 0.001 for a quadratic dose response as best fitting model. Fibrosis improvement of at least one stage occurred in 34, 36 and 34 percent against 22 percent, with no p value, no interval and no claim of significance, and the conclusion names only the primary endpoint.

The 83 percent and 64.5 percent figures are not that result. They come from the sponsor release of 7 June 2024, a less conservative analysis of the same trial, and the giveaway is again the placebo rates: 18.2 percent for MASH and 25.9 percent for fibrosis in F2 to F3, against 14 and 22 percent in the publication. The dose response is also non monotonic. Response peaked at 4.8 mg and 6.0 mg did worse than 2.4 mg, which is why a quadratic was fitted; the likely reason is that 18 of 75 left the 6.0 mg arm during escalation and non completers counted as non responders. Quoting 62 percent alone presents the peak of that curve as the drug effect.

SYNCHRONIZE-MASLD does not fill that gap and does not show that survodutide treats MASH. It had no biopsy endpoint, 90.3 percent of participants lacked biopsy confirmed MASH, and liver stiffness by magnetic resonance elastography did not differ from placebo, p equals 0.6235. What does support a liver specific mechanism is a mediation analysis in 170 participants with paired biopsies, where only 36.3 percent of the fibrosis effect was mediated by weight loss.

The dose that is not the target dose

4.8 mg is not the target dose. It was a phase 2 arm and is a titration way-point in phase 3, which uses 3.6 and 6.0 mg in obesity and 6.0 mg in the liver trials. There is no label, so there is no established dosing, and any titration schedule circulating online did not come from a regulator.

Safety, including the correction that cuts the other way

Discontinuation for adverse events in phase 3 was roughly 24 percent at 3.6 mg and 25 percent at 6.0 mg against 5.4 percent placebo. Against the 2.6 to 7.1 percent range reported for approved agents in this class, that is three to nine times the rate, and it is the hardest single number against the compound. It stands, and it is usually quoted without its two qualifiers. Overall completion was effectively identical on drug and placebo in both programmes, 61 against 60 percent in phase 2 obesity and 41.1 against 40.0 percent discontinuing in SYNCHRONIZE-MASLD, so total attrition was not drug specific even though adverse event attrition was. And in the phase 2 obesity trial, where the breakdown was published, 74 percent of adverse event discontinuations, 56 of 76, happened during the 20 week rapid escalation phase, before anyone reached a maintenance dose. No equivalent breakdown has been published for either phase 3 trial.

Gastrointestinal adverse events in SYNCHRONIZE-1 reached 80.9 percent at 3.6 mg and 89.7 percent at 6.0 mg against 47.9 percent placebo. Serious adverse events were not increased, 8 against 7 percent in the phase 2 MASH trial. Heart rate rose 3.6 beats per minute against 0.8 at week 52 in SYNCHRONIZE-MASLD, and about 3.2 to 3.5 beats per minute at week 76 in SYNCHRONIZE-1. No deaths were reported.

Regulatory status

Survodutide is not approved by any regulator anywhere, for any indication, and no marketing application has been filed anywhere. Every expedited designation it holds is a liver pathway. The FDA granted Breakthrough Therapy designation on 8 October 2024, verbatim, for the treatment of adults with non-cirrhotic metabolic dysfunction-associated steatohepatitis (MASH) and moderate or advanced fibrosis (stages 2 or 3). It also holds FDA Fast Track, EMA PRIME, and breakthrough designations in China and Taiwan. Nothing covers obesity.

The 2024 to 2025 picture of a phase 2 asset with a liver signal is stale. SYNCHRONIZE-1 published in the New England Journal of Medicine and SYNCHRONIZE-MASLD in Nature Medicine, both June 2026. SYNCHRONIZE-2 has not reported and is scheduled for the EASD meeting in Milan, eleven days from the date on this page. SYNCHRONIZE-CVOT, 5,531 participants, completed on 1 June 2026 with no public results. LIVERAGE runs to December 2031.

What circulates, and why no numbers appear here

Survodutide is sold online as lyophilised powder in milligram labelled vials under research use framing, with titration charts, reconstitution guides and projected approval timelines presented as though a regulatory process were under way. None of it comes from a regulator.

This site does not reproduce the circulating figures for survodutide, because the compound is unapproved, the escalation is where the harm sits, and material sold under this name has no established relationship between what the label says and what the vial contains. In the trials, escalation ran 20 to 24 weeks with four weekly clinical contact and permitted dose reduction, and a quarter of participants left anyway. A web page table is not that.

What the record gets wrong

One aggregator page is the origin of most of the structural and pharmacokinetic errors in circulation. It places the modifications at positions 18, 20 and 23, where the documented sites are position 2, position 24 and the C terminus. It gives survodutide GLP-1 receptor potency as 20 pM and glucagon potency as 108 pM against native GLP-1 at 5 pM, where the primary source gives 0.33 nM, 0.52 nM and 60 pM; its 20 pM is in fact the Zimmermann figure for native glucagon. It gives a human half life of 109 to 115 hours, uncited, and no human terminal half life could be sourced to a primary publication for this entry, so none is stated here. It also dates the Breakthrough Therapy designation to September 2024.

The question nobody is asking

In SYNCHRONIZE-1, going from 3.6 mg to 6.0 mg bought 0.8 percentage points of weight loss, minus 12.2 against minus 13.0 percent, with responder rates effectively identical at 72.6 and 71.9 percent. Over the same step, gastrointestinal adverse events rose from 80.9 to 89.7 percent and gastrointestinal discontinuation from 17.8 to 20.2 percent. And 6.0 mg is the dose carried into LIVERAGE, the trial that has to prove the liver indication, even though 4.8 mg outperformed it on phase 2 histology. Whether 6.0 mg is the right dose for either indication is the sharpest open question about this compound, and no published source addresses it.

What is not known

Whether survodutide raises energy expenditure in a human being. That is the central mechanistic claim made for the compound, it rests on rodent indirect calorimetry that reached significance at one dose on two days with no change in core temperature or activity, and it has never been measured in a published human study. Two trials will report: NCT06745284, energy expenditure and fatty acid oxidation against semaglutide in 64 people, primary completion 8 January 2027, and NCT05202353, hepatic glucagon receptor and pancreatic GLP-1 receptor occupancy by PET and MRI, primary completion 6 January 2027.

Whether it changes MASH histology or fibrosis. The phase 2 primary endpoint was MASH improvement, a two point NAFLD activity score reduction, not resolution. The fibrosis secondary endpoint was numerically favourable with no significance claimed. The one phase 3 liver trial that has reported had no biopsy endpoint. MASH resolution and one stage fibrosis improvement are co-primary endpoints of LIVERAGE, primary completion 27 December 2031, and liver related events and mortality are the endpoints of LIVERAGE-Cirrhosis, 5 June 2029.

Cardiovascular outcomes. SYNCHRONIZE-CVOT, 5,531 participants with an adjudicated five point major adverse cardiovascular event primary endpoint, completed on 1 June 2026 and its results are not public. The dedicated thorough QT study, NCT06200467, completed on 22 October 2025 and is also unpublished. Until both report, the cardiovascular profile rests on a heart rate increase of about 3.2 to 3.6 beats per minute, blood pressure reductions, and no confirmed events in a 725 person trial that was not designed to detect them.

What happens after 76 weeks, and after stopping. Seventy-six weeks is the longest published exposure. No survodutide trial reports an off-treatment follow-up period, weight regain has not been studied, and the post hoc beta cell analysis states directly that the durability of those improvements after treatment cessation remains unknown.

The human terminal half life. No primary source giving a value in hours was located for this entry, so none is stated. Circulating figures for it are uncited.

What the loss of lean tissue amounts to. The SYNCHRONIZE-1 imaging substudy reported lean body volume decline of 9.8 percent, and sponsor communications describe lean mass loss as no more than 10.8 percent and no more than 11.3 percent of the change in total tissue mass, which are not the same figure. Set against the documented fall in circulating amino acids, no published study measured muscle strength, physical function, nitrogen balance or protein turnover on this drug.

How it compares with the drugs people compare it to. No head-to-head randomised trial against tirzepatide, retatrutide or any approved obesity agent exists or is registered. The oral contraceptive interaction study NCT05896384 and the renal impairment study NCT06352411 are both complete and both unpublished, and anti-drug antibody data appear nowhere in the published record.

Doses used in published research

Amounts administered under a trial protocol. Not dosing, not a recommendation.
Trial	Population	Dose arms	Duration	What it measured
NCT04153929	413 adults with type 2 diabetes on metformin, HbA1c 7.0 to 10.0 percent, BMI 25 to 50	Survodutide up to 0.3, 0.9, 1.8 or 2.7 mg once weekly, or 1.2 or 1.8 mg twice weekly, against placebo and open label semaglutide up to 1.0 mg once weekly	16 weeks	Absolute change in HbA1c at 16 weeks
NCT04667377	386 treated adults with BMI 27 or above, without diabetes	Survodutide 0.6, 2.4, 3.6 or 4.8 mg once weekly, or placebo, 1:1:1:1:1	46 weeks, 20 weeks of escalation then 26 weeks of maintenance	Percentage change in body weight at week 46
NCT04771273	293 treated adults with biopsy confirmed MASH and fibrosis stage F1 to F3	Survodutide 2.4, 4.8 or 6.0 mg once weekly, or placebo, 1:1:1:1	48 weeks, 24 weeks of rapid escalation then 24 weeks of maintenance	Histological MASH improvement with no worsening of fibrosis
NCT05296733	82 adults with Child-Pugh A, B or C cirrhosis and matched healthy or overweight controls; 41 in the multiple dose part	Single subcutaneous 0.3 mg in the single dose part; in the multiple dose part once weekly doses escalated from 0.3 mg to 6.0 mg over 24 weeks then maintained	28 weeks in the multiple dose part	AUC and maximum plasma concentration after a single dose, and drug related adverse events on multiple dosing
SYNCHRONIZE-1, NCT06066515	725 adults with BMI 30 or above, or 27 or above with an obesity related complication, excluding diabetes	Survodutide once weekly at a dose adjusted up to 3.6 mg or up to 6.0 mg, or placebo, 1:1:1	76 weeks	Percent change in body weight and weight reduction of at least 5 percent, co-primary
SYNCHRONIZE-MASLD, NCT06309992	216 treated adults with obesity and at risk MASLD	Survodutide 6.0 mg once weekly or placebo, 2:1; dose reduction permitted, with 2.4 mg the lowest maintenance dose allowed	48 weeks, 24 weeks of escalation then 24 weeks of maintenance	Liver fat reduction of at least 30 percent on MRI-PDFF and percent change in body weight, co-primary
LIVERAGE, NCT06632444	Approximately 1,800 adults with non-cirrhotic MASH and F2 to F3 fibrosis	Weekly survodutide injections reaching a maximum of 6 mg, or placebo	52 weeks in part one, then approximately seven years in part two	MASH resolution without worsening of fibrosis, and at least one stage fibrosis improvement without worsening of MASH, co-primary in part one

No dose of survodutide has been established as safe or effective, because the compound is not approved in any country and has no label anywhere. What follows is a record of what named trials administered, in the past tense, and nothing in it is a regimen.

Phase 3 obesity, SYNCHRONIZE-1 and SYNCHRONIZE-2, NCT06066515 and NCT06066528: once weekly subcutaneous injection at a dose adjusted up to 3.6 mg or up to 6.0 mg, or placebo, over 76 weeks. SYNCHRONIZE-JP used the same two doses. SYNCHRONIZE-CN used 3.6 mg or 4.8 mg with its primary endpoint at week 52, and is the only phase 3 trial in the programme with a 4.8 mg maintenance arm.

Phase 3 liver trials: SYNCHRONIZE-MASLD, NCT06309992, escalated to 6.0 mg once weekly over 24 weeks then maintained for 24 weeks, with 2.4 mg the lowest maintenance dose permitted. LIVERAGE and LIVERAGE-Cirrhosis, NCT06632444 and NCT06632457, use weekly injections reaching a maximum of 6 mg, and both are still recruiting.

Phase 2: the obesity trial NCT04667377 used 0.6, 2.4, 3.6 or 4.8 mg once weekly across 20 weeks of escalation and 26 of maintenance. The MASH trial NCT04771273 used 2.4, 4.8 or 6.0 mg once weekly across 24 weeks of rapid escalation and 24 of maintenance. The type 2 diabetes trial NCT04153929 used up to 0.3, 0.9, 1.8 or 2.7 mg once weekly and 1.2 or 1.8 mg twice weekly over 16 weeks, and its maximum once weekly dose was well below anything used in the obesity and liver programmes.

The only per-step escalation scheme documented in a primary publication is from the phase 1 cirrhosis study NCT05296733: 0.3 mg escalated to 6.0 mg over 24 weeks, then four weeks of maintenance. Every trial escalation ran with four weekly clinical contact, standardised diet and activity counselling and permitted dose reduction, and roughly a quarter of participants still discontinued for adverse events.

Compiled from: Bluher 2024, Diabetologia, phase 2 type 2 diabetes; le Roux 2024, Lancet Diabetes and Endocrinology, phase 2 obesity; Sanyal 2024, New England Journal of Medicine, phase 2 MASH; Lawitz 2024, Journal of Hepatology, cirrhosis; Diabetes Obesity and Metabolism 2025, phase 2 subgroup and discontinuation analysis; NCT06632444, LIVERAGE; SYNCHRONIZE-MASLD, Nature Medicine 2026; SYNCHRONIZE-1, New England Journal of Medicine 2026; Ji 2026, Diabetes Therapy

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.