Survodutide for Fatty Liver: What the Phase 3 Nature Medicine Trial Shows

Survodutide lowered liver fat and body weight in adults with obesity and MASLD. At week 48 of the phase 3 SYNCHRONIZE-MASLD trial, 68.5% of people assigned to 6.0 mg weekly had at least a 30% drop in liver fat on MRI, versus 28.6% on placebo, in the analysis that accounts for people who stopped treatment ^[1]. The trial reported imaging and blood markers, not follow-up biopsies or liver complications, and the paper still described survodutide as under investigation when it appeared in 2026 ^[1].

Who was in SYNCHRONIZE-MASLD

The trial treated 216 adults in the United States and Spain: 146 on once-weekly survodutide 6.0 mg and 70 on placebo, over a 24-week dose escalation and 24 weeks of maintenance ^[1]. Entry required a BMI of at least 30, or at least 27 with a metabolic complication, plus “at-risk MASLD”: liver fat of at least 8% on MRI-PDFF with either an elevated noninvasive test or biopsy-confirmed MASH in the previous 3 years ^[1]. On average, participants had a BMI of 39.6 and liver fat of 16.9%, and 38.4% had type 2 diabetes ^[1].

Only 9.7% had biopsy-confirmed MASH; the other 90.3% qualified by noninvasive tests, and the authors say most participants did not meet formal criteria for MASH ^[1]. They also describe baseline fibrosis as no more than mild to moderate for most people ^[1]. This was an earlier-disease group than the patients enrolled in biopsy-based MASH trials.

Boehringer Ingelheim funded the study, designed it with outside consultants, ran the data analysis and paid for writing support, and several authors are company employees ^[1]. An October 2026 author correction fixed investigator listings, a typo in the placebo discontinuation denominator and a figure caption; it did not change an efficacy result ^[2].

The two sets of numbers, and which to read first

Two analyses were prespecified. The treatment-regimen estimand, the primary one, reflects what happened with early discontinuation and prohibited obesity drugs included; for the liver-fat responder result, it used an imputation that assumed survodutide dropouts behaved like the placebo group ^[1]. The supplementary efficacy estimand asks what would have happened had everyone stayed on treatment ^[1]. The gap is large here because treatment discontinuation reached 41.1% with survodutide and 40.0% with placebo ^[1].

Under the treatment-regimen estimand, the odds ratio for the 30% liver-fat response was 5.9 (95% CI 2.9 to 11.9) and the estimated weight difference from placebo was −7.3 percentage points (95% CI −9.4 to −5.2) ^[1]. The deeper liver-fat results in the lower half of the table were reported only under the efficacy estimand, with no adjustment for multiple comparisons, so they describe a best case among people assumed to stay on the drug ^[1].

A weight comparison needs care. In SYNCHRONIZE-1, the 6.0 mg group changed −13.0% at 76 weeks by the treatment-regimen estimand, against −5.4% on placebo ^[10]. That figure is numerically larger than −8.7%, but it comes from a longer trial that excluded diabetes, while 38.4% of SYNCHRONIZE-MASLD participants had it ^[1]. Our survodutide explainer covers the wider development program.

Does less liver fat mean a healthier liver?

Probably to some degree, though the link is an association. In a meta-analysis of seven early-phase studies with 346 people, those reaching a 30% MRI-PDFF decline were more likely to have a histologic response (51% vs 14%) and NASH resolution (41% vs 7%) ^[4]. A retrospective analysis of 150 trial participants found that people with a cT1 reduction of at least 80 ms were about 5 times as likely to have a histologic response ^[5]. A secondary analysis of a pegozafermin trial, a different drug, tied a 50% liver-fat decline to fibrosis regression (adjusted odds ratio 3.08, 95% CI 1.32 to 7.19), while a 30% decline was not an independent predictor (2.10, 95% CI 0.85 to 5.16) ^[6].

Inflammation markers moved in the expected direction with survodutide. Under the efficacy estimand, a cT1 reduction of at least 80 ms occurred in 63.0% versus 21.4% on placebo, and ALT fell 36.8% versus 11.0% ^[1].

Scarring markers disagreed. Liver stiffness on vibration-controlled transient elastography (FibroScan) fell 28.7% versus 9.2% (estimated difference −19.6%, 95% CI −31.9 to −7.2), and the ELF score fell 0.34 versus 0.06 ^[1]. MR elastography, however, showed no difference (estimated difference −0.05, 95% CI −0.25 to 0.15) ^[1]. The authors suggest a floor effect from low baseline stiffness may explain that null result, and also note that high BMI may lead FibroScan readings to overestimate fibrosis ^[1].

What biopsies have shown so far

The biopsy evidence for survodutide comes from a phase 2 trial of 293 adults with MASH and F1 to F3 fibrosis. MASH improved without fibrosis worsening in 47%, 62% and 43% of the 2.4, 4.8 and 6.0 mg groups, versus 14% on placebo, and fibrosis improved by at least one stage in 34%, 36% and 34%, versus 22% ^[3]. The highest dose did not produce the highest response, and phase 3 SYNCHRONIZE-MASLD tested only 6.0 mg ^{[3] [1]}.

The drugs below were tested in separate company-funded trials with different entry criteria, durations and endpoint definitions (the phase 2 trial counted histologic improvement in MASH, the other two counted resolution), so the table does not rank them ^{[3] [7] [8]}. Semaglutide 2.4 mg at the 72-week interim of ESSENCE ^[7] and resmetirom at 52 weeks in MAESTRO-NASH ^[8] both reached their histology endpoints against placebo. The 2026 survodutide paper describes both as conditionally approved by the FDA for MASH with F2 or F3 fibrosis ^[1]. Our Wegovy for MASH review covers the semaglutide path in detail.

A network meta-analysis of 29 trials ranked survodutide second for MASH resolution without worsening fibrosis (SUCRA 90.87), just behind pegozafermin (91.75) and ahead of tirzepatide (84.70) ^[9]. A SUCRA score is a rank probability built on indirect comparisons, not a measured difference between drugs. Because the search ended in December 2024, its survodutide data predate the phase 3 program. Retatrutide, a triple agonist that also includes glucagon activity, has its own separate MASLD evidence.

Why a glucagon component might matter for the liver

The authors propose that glucagon-receptor activity raises fatty acid oxidation in the liver, which could speed fat clearance beyond what weight loss alone provides ^[1]. They also cite a conference abstract, from the phase 2 trial, suggesting that changes in liver markers were mostly independent of weight loss ^[1]. That idea remains a hypothesis: SYNCHRONIZE-MASLD did not compare survodutide with a GLP-1 drug, so it cannot show an added benefit from glucagon.

Side effects and who stopped

• Gastrointestinal events drove most of the burden. Nausea affected 56.2% on survodutide versus 14.3% on placebo, vomiting 42.5% versus 7.1%, diarrhea 29.5% versus 10.0% and constipation 28.1% versus 10.0% ^[1].
• Tolerability cost people the drug. Adverse events led to discontinuation in 23.3% versus 10.0%, including gastrointestinal events in 19.9% versus 4.3% ^[1]. The authors note that the protocol initially allowed limited flexibility in dose escalation ^[1].
• Serious events were not higher. They occurred in 9.6% on survodutide and 11.4% on placebo, with no adjudicated drug-induced liver injury or acute pancreatitis in the survodutide group ^[1].
• Smaller signals deserve follow-up. Heart rate rose 3.6 beats per minute versus 0.8 at week 52; major depressive disorder was reported in 2.7% versus 0% and diabetic retinopathy in 1.4% versus 0% ^[1]. With 216 people treated, the trial could not settle uncommon risks in either direction.

In efficacy or on-treatment analyses, waist circumference fell 11.1 cm versus 1.9 cm and HbA1c fell 0.6 versus 0.02 points; LDL and HDL changes did not differ significantly from placebo ^[1].

What this trial cannot answer

The authors name two limitations: 48 weeks is short, and recruitment in two countries, with mostly White participants, limits how widely the results apply ^[1]. Hard outcomes are the larger gap. No result here shows fewer cases of cirrhosis, liver failure or death.

The phase 3 LIVERAGE program is designed to test that. The main trial lists an estimated 1,800 people with MASH and moderate or advanced fibrosis, with week 52 histology endpoints and a composite clinical outcome followed for up to 7 years ^[11]. LIVERAGE-Cirrhosis lists an estimated 1,590 people and a composite outcome over up to 4.5 years ^[12]. Cirrhosis data so far come from a small, open-label, non-randomized phase 1 study, where the authors judged the drug generally tolerable; drug-related adverse events occurred in 87.5% of people with cirrhosis in the multiple-dose cohorts ^[13].

What to raise with a clinician

• Which stage of MASLD do your test results suggest, and would a biopsy or elastography change your options?
• Which currently approved treatments fit your fibrosis stage, and has their labeling changed since the trials above were published?
• If an investigational drug interests you, are you eligible for a registered trial, and what would you need to disclose about your current medicines?