Survodutide 10mg Peptide

Survodutide, known throughout its development as BI 456906, is a synthetic 29 amino acid peptide that activates two receptors at once: the glucagon receptor (GCGR) and the glucagon-like peptide-1 receptor (GLP-1R). It was engineered at Boehringer Ingelheim from chemistry licensed from Zealand Pharma, and it is built on a modified glucagon backbone rather than on the exendin-4 or oxyntomodulin scaffolds that most incretin analogues start from. The sequence is glucagon with positions 18, 20 and 23 exchanged for the corresponding GLP-1 residues and position 16 for the exendin-4 residue, which is how a glucagon peptide acquires GLP-1 receptor activity without becoming a GLP-1 analogue. A C18 diacid hangs off the lysine side chain at position 24 through a glycine-serine linker as the half-life extending principle, the non-coded residue Ac4c sits at position 2 to blunt DPP-4 cleavage, and the C-terminus is amidated.

The dual design addresses something a pure GLP-1 receptor agonist does not reach. GLP-1 receptor signalling lowers energy intake. Glucagon receptor signalling in hepatocytes raises energy expenditure and drives hepatic fatty acid oxidation directly, which is a different route to lower liver fat than the indirect one that follows from eating less. Pairing them lets the glucagon arm push energy out while the GLP-1 arm restrains intake and offsets glucagon's usual glycaemic effect. That is the whole thesis of the molecule, and it is why the survodutide programme has run liver endpoints alongside body weight endpoints from an early stage.

Survodutide is investigational in every jurisdiction. Its registered record on ClinicalTrials.gov runs to 31 studies as of August 2026, spanning phase 1 pharmacokinetics, phase 2 dose-finding in obesity and in type 2 diabetes, a biopsy-controlled phase 2 study in MASH, and the phase 3 SYNCHRONIZE and LIVERAGE programmes. Material supplied for laboratory work is intended for in vitro and preclinical research only and is not for human or veterinary use.

In CHO-K1 cells stably expressing the human receptors, survodutide raises intracellular cyclic AMP with an EC50 of 0.33 nM at GLP-1R and 0.52 nM at GCGR (Zimmermann et al., 2022). The ratio is close to balanced, with a bias of roughly 1.6-fold toward GLP-1R. That number understates how much work the glucagon arm does, because survodutide is itself 10 to 20 fold less potent at GLP-1R than semaglutide is, so at any given exposure a larger share of the total signal is glucagon receptor signal than the raw ratio suggests. Boehringer selected survodutide from a panel of 19 dual GCGR/GLP-1R agonists precisely on that balance point, using acute oral glucose tolerance in mice as the GLP-1R readout and hepatic NNMT mRNA plus plasma FGF21 as the GCGR readout (Thomas et al., 2024).

The GLP-1R arm reaches the brain without generally crossing the blood-brain barrier. Fluorophore-labelled survodutide given to mice accumulated in the circumventricular organs and the adjacent hypothalamic and hindbrain nuclei, and c-Fos labelling showed activation of multiple nuclei that govern food intake. Single-cell profiling in the same study found GLP1R expressed in the area postrema and the arcuate nucleus while GCGR was barely detectable in either, in mouse and in human tissue. A long-acting GCGR-selective agonist activated none of those satiety nuclei and did not reduce food intake, yet still lowered body weight (Zimmermann et al., 2026). Appetite suppression is therefore GLP-1R work; the glucagon contribution to weight arrives by another road.

That other road is largely hepatic. In lean mice, a single 100 nmol/kg dose raised liver NNMT mRNA 15 to 17 fold and plasma FGF21 up to sevenfold above vehicle. In diet-induced obese mice, survodutide at 30 nmol/kg once daily reduced body weight by about 32% over 28 days, against roughly 25% for semaglutide at 20 nmol/kg, and indirect calorimetry showed a dose-dependent rise in energy expenditure at 5, 10 and 20 nmol/kg. The acute anorectic effect faded by day 15 while weight loss continued, which is the clearest preclinical separation between the two arms. When the hepatic glucagon receptor was deleted in mice, a structurally related analogue lost its weight and metabolic advantage over GLP-1R agonism alone (Long et al., 2026), placing the extra efficacy in the liver rather than in adipose or muscle.

The glycaemic counterbalance is the reason the pairing is workable at all. Unopposed glucagon receptor agonism raises hepatic glucose output. In the phase 2 type 2 diabetes study, survodutide lowered HbA1c by 1.71 percentage points at the most effective weekly schedule over 16 weeks, which is not the direction an uncompensated glucagon agonist would move it (Blüher et al., 2024). Phase 1 target-engagement markers ran in both directions at once: plasma amino acids such as alanine and tyrosine fell, a glucagon receptor signature, while paracetamol absorption slowed, a GLP-1 receptor signature of delayed gastric emptying.

The single number that defines survodutide is 0.33 against 0.52. Those are the EC50 values for cyclic AMP accumulation at human GLP-1R and human GCGR in CHO-K1 cells, and they place the molecule near the middle of the dual-agonist spectrum rather than at either edge. For comparison, an oxyntomodulin-derived dual agonist typically sits far more heavily on the GLP-1 side, and a glucagon analogue with no incretin activity sits at the other end. Survodutide was chosen out of 19 candidate dual agonists on exactly this axis: the selection criterion was an improvement in acute oral glucose tolerance in mice comparable to semaglutide, combined with clear in vivo glucagon receptor engagement measured by hepatic NNMT mRNA and plasma FGF21.

Why does the glucagon receptor change the physiology rather than just adding to it? Because glucagon receptor agonism acts on a different side of the energy balance equation. GLP-1 receptor agonism lowers the intake side. Glucagon receptor agonism raises the expenditure side and, in the liver specifically, shifts substrate handling toward fatty acid oxidation and ketogenesis. In diet-induced obese mice this shows up as a dose-dependent rise in energy expenditure by indirect calorimetry, and as continued body weight loss after the anorectic effect has faded. In humans it shows up as plasma FGF21 movement and as a liver fat response that outruns what the body weight change alone would predict.

The receptor distribution study published in 2026 sharpened this. Single-cell analysis of human and mouse circumventricular organs found GLP1R transcripts in the area postrema and arcuate nucleus and essentially no GCGR. A fluorophore-tagged survodutide reached those regions and activated c-Fos in food-intake nuclei; a long-acting glucagon-selective agonist did neither, yet still lowered body weight in the same paradigm. The two arms are not redundant and they are not acting in the same tissue.

NCT04667377 enrolled 387 participants aged 18 to 75 with a BMI at or above 27 kg/m2 and without diabetes, at 43 centres across 12 countries between March and November 2021. Randomisation was 1:1:1:1:1 to once-weekly subcutaneous survodutide at 0.6, 2.4, 3.6 or 4.8 mg, or placebo, across 46 weeks split into a 20-week dose escalation phase and a 26-week maintenance phase. The primary endpoint was percentage change in body weight from baseline to week 46.

Mean changes in body weight under the planned-treatment analysis were -6.2% (95% CI -8.3 to -4.1) at 0.6 mg, -12.5% (-14.5 to -10.5) at 2.4 mg, -13.2% (-15.3 to -11.2) at 3.6 mg and -14.9% (-16.9 to -13.0) at 4.8 mg, against -2.8% (-4.9 to -0.7) on placebo. The dose-response curve had not plateaued at the top dose tested, which is the reason the phase 3 programme carried 6.0 mg forward.

Completion was the weak point of the study rather than efficacy. Only 233 of 386 treated participants finished the 46 weeks, 61% of the survodutide arms and 60% of placebo. Adverse events occurred in 91% of survodutide recipients against 75% on placebo, and were predominantly gastrointestinal in 75% against 42%. The subsequent phase 3 design used a slower escalation schedule in direct response, and a phase 1 analysis had already reported that dose-related gastrointestinal events could be mitigated by slower escalation.

The most distinctive part of the survodutide record is NCT04771273, a 48-week phase 2 trial in adults with biopsy-confirmed MASH and fibrosis stage F1 through F3, published in the New England Journal of Medicine in 2024. Participants were randomised 1:1:1:1 to once-weekly subcutaneous survodutide at 2.4, 4.8 or 6.0 mg or placebo, through a 24-week rapid dose escalation phase followed by 24 weeks of maintenance. A total of 293 randomised participants received at least one dose. The primary endpoint was histological improvement in MASH with no worsening of fibrosis on paired biopsy.

The primary endpoint was met in 47% of the 2.4 mg group, 62% of the 4.8 mg group and 43% of the 6.0 mg group, against 14% on placebo, with a quadratic dose-response curve as the best-fitting model. A liver fat reduction of at least 30% occurred in 63%, 67% and 57% against 14% on placebo. Fibrosis improvement of at least one stage occurred in 34%, 36% and 34% against 22%. Nausea was reported in 66% against 23%, diarrhoea in 49% against 23% and vomiting in 41% against 4%; serious adverse events were 8% against 7%.

Two features of that dataset deserve attention. First, the non-monotonic dose response, with 4.8 mg outperforming 6.0 mg on the histological endpoint, is unusual and is at least partly a tolerability artefact of the rapid escalation schedule. Second, a 2026 causal mediation analysis of the 170 participants with F2 to F3 fibrosis and paired biopsies decomposed the effect: 66.7% of the effect on MASH resolution and 71.8% on MASH improvement was mediated by weight reduction, but only 36.3% of the effect on fibrosis improvement was. For non-invasive markers of inflammation and fibrosis the weight-mediated share ran from 16.5% for AST to 38.6% for the Enhanced Liver Fibrosis score. Steatosis endpoints behaved the other way, at 58.2% for MRI-PDFF and 77.4% for FibroScan CAP. Read together, the steatosis signal tracks weight and the fibro-inflammatory signal does not, which is the strongest published argument that hepatic glucagon receptor agonism is doing something a pure incretin cannot.

A separate open-label phase 1 study extended the liver work into cirrhosis. In 41 people with Child-Pugh class A, B or C cirrhosis and matched healthy controls, single-dose AUC and Cmax were similar across groups, and a second cohort of 41 participants with overweight or obesity, with or without compensated or decompensated cirrhosis, was escalated from 0.3 mg to 6.0 mg over 24 weeks then maintained for 4 weeks. Liver fat content, liver stiffness, liver volume and body weight were generally reduced by week 28. Decompensated cirrhosis is a population usually excluded from investigational trials outright, so that dataset is unusual.

NCT04153929 randomised 413 adults with type 2 diabetes, HbA1c 7.0% to 10.0% and BMI 25 to 50 kg/m2, all on background metformin, to one of six survodutide schedules, placebo, or open-label semaglutide as an active reference. Survodutide arms escalated to 0.3, 0.9, 1.8 or 2.7 mg once weekly, or 1.2 or 1.8 mg twice weekly. The primary endpoint was absolute change in HbA1c at 16 weeks.

Adjusted mean HbA1c fell from a baseline of 8.07% by 0.91 points at the lowest schedule and by 1.71 points at the most effective, with the 0.9 mg weekly arm at 1.46 points against 1.47 points for semaglutide. Body weight fell dose-dependently to -8.7% at the highest twice-weekly schedule, against -5.3% for semaglutide. Adverse events were reported in 77.8% of survodutide participants, mainly gastrointestinal, against 52.0% for semaglutide and 52.5% for placebo.

The shape of that result is the practical summary of the dual mechanism at short exposure: matched glycaemic efficacy per unit of GLP-1 signal, more body weight movement, more gastrointestinal burden. A 2026 analysis of beta-cell function biomarkers across the diabetes and obesity cohorts reported improvements in insulin sensitivity indices alongside those changes.

SYNCHRONIZE-1 (NCT06066515) randomised 725 adults from 14 countries with BMI at or above 30 kg/m2, or at or above 27 kg/m2 with one obesity complication, and without type 2 diabetes, 1:1:1 to once-weekly subcutaneous survodutide escalated to 3.6 or 6.0 mg, or placebo, for 76 weeks. Baseline mean age was 47.1 years, BMI 37.9 kg/m2 and waist circumference 115.2 cm. Co-primary endpoints were percentage body weight change and the proportion achieving at least 5% reduction at week 76. The sponsor reported in April 2026 that both co-primary endpoints were met, with mean body weight change of up to -16.6% against -3.2% on placebo under the efficacy estimand, and 85.1% against 38.8% achieving at least 5% reduction. A pre-specified body composition analysis announced by the originator in June 2026 put visceral adipose tissue reduction at about 34% and liver fat reduction at about 63%, both far larger than the total body weight change, with lean mass loss described as minimised. Those figures come from a company announcement rather than a peer-reviewed report and should be treated as provisional until the full dataset is published.

SYNCHRONIZE-2 (NCT06066528) is the companion trial in 752 treated participants with obesity and type 2 diabetes across 133 sites in 19 countries, on the same 3.6 and 6.0 mg schedules and the same 76-week endpoints. Baseline HbA1c was 7.4%, 78.7% were on metformin and 34.2% on an SGLT2 inhibitor. SYNCHRONIZE-CVOT (NCT06077864) is an event-driven cardiovascular safety trial with a target enrolment of 4,935 adults with BMI at or above 27 kg/m2 and established cardiovascular or chronic kidney disease, or at least two weight-related risk factors, using time to first five-point major adverse cardiovascular event as the primary endpoint. Regional trials run alongside: SYNCHRONIZE-JP in Japan (NCT06176365) and a Chinese trial (NCT06214741).

On the hepatic side, SYNCHRONIZE-MASLD (NCT06309992) was the phase 3 imaging study of 216 adults, reported in Nature Medicine in 2026, and the LIVERAGE programme carries the biopsy work forward: LIVERAGE (NCT06632444) in MASH with moderate or advanced fibrosis, and LIVERAGE-Cirrhosis (NCT06632457) in MASH-related cirrhosis. Both were recruiting as of 2026. A phase 2 study in chronic kidney disease with albuminuria as the endpoint (NCT07206290) opened in 2026, extending the programme past liver and weight.

Retatrutide takes the same premise one receptor further. It is a triple agonist at GIP, GLP-1 and glucagon receptors, so it keeps survodutide's glucagon component and adds the GIP arm that tirzepatide uses. In its phase 2 obesity trial retatrutide reached about -24% body weight at 48 weeks at the highest dose, against survodutide's -14.9% at 46 weeks, though the two were never compared head to head and the escalation schedules differ. The interesting divergence is not the weight number but the endpoint each programme chose to chase: retatrutide's development has centred on the magnitude of weight reduction, while survodutide's has centred on histological liver disease, which is why survodutide has a paired-biopsy MASH dataset and a mediation analysis of it, and retatrutide does not.

Tirzepatide is the clean contrast because it has no glucagon receptor activity at all. It is worth being precise about what GIP actually contributes, since most comparisons treat it as a table cell rather than as pharmacology. The GIP receptor is an incretin receptor like GLP-1R: agonism at it amplifies glucose-dependent insulin secretion, acts on adipocytes to influence lipid storage and blood flow, and has a central component in the area postrema that appears to blunt nausea, which is part of why a GIP-containing agonist can be pushed to higher exposures than a GLP-1-only molecule. What GIP does not do is raise hepatic fatty acid oxidation. So a GIP/GLP-1 agonist lowers liver fat, but chiefly as a downstream consequence of eating less and weighing less. Survodutide's mediation analysis, showing that only about a third of its fibrosis effect tracks body weight, is the specific claim no GIP-based dual agonist has a mechanism to make. Leaving GIP out is therefore a design decision rather than a gap: it costs tolerability headroom and buys a direct hepatic lever.

Within the glucagon/GLP-1 class, mazdutide and cotadutide share survodutide's receptor pair but come from a different scaffold and a different ratio. Both are oxyntomodulin analogues, where survodutide is a glucagon analogue, and both are reported to sit further toward the glucagon receptor than survodutide does. Cotadutide is short-acting enough to require daily administration and its development has stalled. Mazdutide reached approval in China in June 2025, making it the first glucagon/GLP-1 dual agonist authorised anywhere, though only in that one market. Survodutide has the largest registered programme of the three and the only paired-biopsy liver dataset. A 2025 Bayesian network meta-analysis of 19 randomised trials in 29,506 adults grouped survodutide with tirzepatide as dual agonists and found the dual-agonist class and retatrutide at equivalent mean weight loss of about 11 kg against 9 kg for selective GLP-1 receptor agonists, with retatrutide ahead specifically on the proportion reaching at least 15% reduction and also carrying the highest adverse event risk. Class-level meta-analyses of that kind should be read carefully, since they pool molecules whose receptor ratios differ by an order of magnitude.

For comparative in vitro work, the useful pairing is survodutide against mazdutide at matched cyclic AMP potency to isolate scaffold effects at a shared receptor pair, or survodutide against retatrutide to isolate the GIP contribution while the glucagon contribution is held constant. Cagrilintide, an amylin analogue, sits outside this axis entirely and is a different comparison.

Survodutide is supplied as a lyophilised white powder. Like other C18-diacid acylated peptides in this class, the dry powder is stable for extended periods at -20 degrees Celsius protected from light, and tolerates ambient shipping temperatures for the few days a courier takes. Once in aqueous solution the molecule behaves like the rest of the acylated GLP-1 family: it is far less stable, should be held at 2 to 8 degrees Celsius, and is degraded by repeated freeze-thaw cycles and by agitation at the air-water interface. The C18 diacid tail makes the peptide surface-active, so vigorous shaking of a solution promotes aggregation and fibrillation at the meniscus rather than simple dissolution. Gentle swirling against the vial wall is the standard laboratory handling for this reason.

The Ac4c residue at position 2 exists to resist dipeptidyl peptidase-4, which cleaves the native His-Ser N-terminus of glucagon and GLP-1 within minutes. Combined with albumin binding through the C18 diacid, this pushes the terminal half-life past 100 hours in humans, which is what makes weekly administration possible and which also means steady state is not reached quickly: the phase 1 multiple-dose study noted that plasma exposure was still climbing at the end of the escalation period.

Analytical characterisation of a 29 amino acid acylated peptide of this size needs more than a single purity figure. Reversed-phase HPLC with a C18 column and a trifluoroacetic acid or formic acid modified gradient resolves the main peak from deletion sequences and from the des-acyl impurity, which is the one most worth watching because a missing lipid tail changes pharmacokinetics without changing receptor pharmacology much. Electrospray mass spectrometry should return a deconvoluted mass near 4231.7 Da for the intact molecule. Because the sequence carries multiple acidic residues, the observed charge envelope in positive mode is typically broad; deconvolution rather than raw m/z inspection is the reliable check. Functional confirmation is a cyclic AMP accumulation assay in cells expressing human GCGR and human GLP-1R separately, since a batch that has lost the correct ratio between the two is the specific failure mode that a purity chromatogram will not reveal.