A research team scanning ClinicalTrials.gov for next-generation obesity compounds in 2023 would have found something unusual: a Phase 3 program for a GLP-1/glucagon dual agonist running entirely at Chinese sites, with no parallel U.S. arm and no FDA IND announcement attached. That compound is mazdutide (laboratory code IBI362), developed by Suzhou-based Innovent Biologics. It is not a me-too semaglutide analog — it pairs GLP-1 receptor agonism with glucagon receptor agonism, a mechanism shared by only a handful of clinical-stage molecules worldwide. For clinicians and researchers tracking the post-tirzepatide landscape, mazdutide is a useful case study in how obesity and diabetes drug development is increasingly proceeding through non-U.S. regulatory pathways first, and what that means for the evidence base available before a compound reaches Western prescribing decisions.
What Is Mazdutide? A Dual GLP-1/Glucagon Receptor Agonist
Mazdutide is a 39-amino-acid peptide engineered from oxyntomodulin, a native gut hormone that weakly activates both the GLP-1 receptor and the glucagon receptor (GCGR). Innovent's engineering work rebalanced potency at the two receptors and added fatty-acid acylation — the same half-life-extension strategy used in liraglutide and semaglutide — to permit once-weekly subcutaneous dosing rather than the daily injections oxyntomodulin itself would require.
The pharmacologic logic is additive rather than redundant. GLP-1 receptor activation slows gastric emptying, increases satiety signaling, and supports glucose-dependent insulin secretion, the same pathway described in detail in coverage of GLP-1 receptor mechanism of action and tolerability. Glucagon receptor activation, by contrast, is catabolic at the liver — it increases hepatic glucose output but also drives fatty-acid oxidation and resting energy expenditure. The hypothesis behind dual GLP-1/glucagon agonists, also under investigation in cotadutide and survodutide, is that pairing the two offsets the metabolic-adaptation slowdown seen with GLP-1 monotherapy alone, though this remains a mechanistic hypothesis rather than a settled clinical finding.
From Oxyntomodulin to IBI362: Development History
Mazdutide's Chinese trials were run with Innovent Biologics: the Phase 2 trial was funded by Innovent and several of its authors were Innovent employees (Ji L, et al. Nat Commun. 2023; PMID 38092790). A separate U.S.-based Phase 2 trial was funded by Eli Lilly and Company (Hsia SH, et al. Lancet Diabetes Endocrinol. 2026; PMID 42628555). The Chinese program reflects a broader pattern among Chinese biopharma companies: rather than waiting for a multinational partner to lead a global trial program, firms like Innovent are running full Phase 1 through Phase 3 development domestically and filing with the National Medical Products Administration (NMPA) first.
That sequencing matters for how the evidence base accumulates. A compound reviewed by the NMPA before the FDA generates its pivotal safety and efficacy dataset in a single-ethnicity population, on a China-specific dosing and monitoring framework, well before Western clinicians have peer-reviewed data to evaluate. The Phase 1 dose-escalation and Phase 2 dose-ranging studies for mazdutide followed a standard obesity-drug template — single ascending dose, then multiple ascending dose with weekly titration — before the compound advanced into the confirmatory Phase 3 program described below.
Phase 2 Trial Data: What the Dose-Ranging Study Showed
The published Phase 2 trial was a randomized, double-blind, placebo-controlled, multi-arm study in Chinese adults with overweight or obesity, testing multiple mazdutide doses against placebo over a 24-week treatment period that included gradual up-titration to reduce GI intolerance. Dose arms spanned a range from a low-dose comparator through the maximum tested dose, structured to define a dose-response curve rather than to serve as the confirmatory trial.
The published results followed the pattern seen across the GLP-1 class. Among 248 randomized participants, mean body-weight change at week 24 was −6.7% with 3 mg, −10.4% with 4.5 mg and −11.3% with 6 mg, versus +1.0% with placebo, a placebo-adjusted difference of 7.7 to 12.3 percentage points (all p<0.0001) (Ji L, et al. Nat Commun. 2023; PMID 38092790). This dose-response shape is consistent with what was later reported for the triple agonist retatrutide, covered in detail in the retatrutide Phase 2 trial results at 48 weeks, where higher-dose arms similarly produced the largest separation from placebo. The key limitation of any Phase 2 dataset, mazdutide's included, is duration and sample size: 24 weeks is sufficient to characterize early dose-response and tolerability but not durability, and Phase 2 cohorts are typically too small to detect low-frequency safety signals.
Phase 3 GLORY-1: Trial Design and Study Population
Mazdutide's Phase 3 program splits into two indications. The weight-management arm, registered under the GLORY-1 designation, enrolls Chinese adults meeting local BMI thresholds for obesity or overweight with a weight-related comorbidity. A separate DREAMS series of trials addresses type 2 diabetes, evaluating mazdutide's glycemic effects either alone or against active comparators. GLORY-1 treated participants for 48 weeks, with its two primary end points (percent change in body weight and the share reaching at least 5% weight reduction) assessed at week 32 (Ji L, et al. N Engl J Med. 2025; PMID 40421736), which allows for assessment of weight-loss plateau timing and longer-duration tolerability.
Primary and secondary endpoints follow the structure now standard across the obesity-trial field: percent change in body weight from baseline as the primary endpoint, with secondary endpoints covering waist circumference, the proportion of participants reaching weight-loss thresholds of 5%, 10%, and 15%, and cardiometabolic labs including lipid panels and fasting glucose. Registration details are searchable directly on ClinicalTrials.gov's mazdutide trial listings. A structural difference from the pivotal STEP and SURMOUNT programs is enrollment: GLORY-1 and the DREAMS trials recruit exclusively from Chinese sites, which narrows the population the data can be generalized to until a multi-region trial or bridging study is conducted.
Phase 3 Efficacy Signals: Weight Loss and Glycemic Outcomes
GLORY-1, published in 2025, randomized 610 Chinese adults with obesity or overweight to mazdutide 4 mg, mazdutide 6 mg or placebo. At week 48, mean body-weight change was −11.00% with 4 mg and −14.01% with 6 mg, versus +0.30% with placebo, and 35.7% and 49.5% of participants on the two doses lost at least 15% of body weight, versus 2.0% on placebo (P<0.001 for all comparisons with placebo) (Ji L, et al. N Engl J Med. 2025; PMID 40421736). A second trial, GLORY-2, tested a 9 mg dose in Chinese adults with obesity and reported a mean change of −16.65% at week 60, versus −1.50% with placebo (Gao L, et al. JAMA. 2026; PMID 42251595).
On the diabetes side, the DREAMS trials evaluate HbA1c reduction alongside body-weight change, reflecting the dual metabolic rationale of a GLP-1/glucagon mechanism. A randomized Phase 3 trial published in Nature assigned 731 Chinese adults with type 2 diabetes on background oral therapy to mazdutide 4 mg, mazdutide 6 mg or dulaglutide 1.5 mg for 28 weeks. Mazdutide lowered HbA1c more than dulaglutide, by a least-squares mean difference of 0.24 percentage points with 4 mg and 0.30 points with 6 mg, and lowered body weight more, by 3.78% and 5.76%; gastrointestinal adverse events were more frequent with mazdutide (Guo L, et al. Nature. 2026; PMID 41407860). Figures for trials that have not yet been published, circulating from investor calls or press releases, should still be treated as preliminary rather than clinical-decision-grade evidence.
Safety and Tolerability Profile
The adverse-event pattern reported across mazdutide's program to date is GI-predominant and concentrated during dose titration, mirroring the class-wide mechanism described in GLP-1 receptor agonists and gastric emptying. Commonly reported events include:
- Nausea, typically most frequent during the first weeks of a titration step
- Vomiting and diarrhea, generally described as mild to moderate
- Decreased appetite, an expected on-target effect rather than an unintended adverse event
- Injection-site reactions, consistent with subcutaneous peptide administration
Because mazdutide also activates the glucagon receptor, a monitoring point distinct from GLP-1-only agents is heart rate. The published trial abstracts do not quantify a mazdutide-specific heart-rate change, so this remains a parameter to read from the full trial reports. Discontinuation because of adverse events was low in GLORY-1: 1.5% with 4 mg, 0.5% with 6 mg and 1.0% with placebo (Ji L, et al. N Engl J Med. 2025; PMID 40421736). In the U.S. Phase 2 trial it was highest with the 16 mg dose, at 20%, mainly because of gastrointestinal disorders (Hsia SH, et al. Lancet Diabetes Endocrinol. 2026; PMID 42628555).
Why China First — Regulatory and Market Pathway
Running the full clinical program through the NMPA before any FDA interaction is a deliberate commercial and regulatory strategy, not an accident of trial logistics. China's obesity and type 2 diabetes markets are large and growing, and the NMPA has streamlined priority-review pathways for domestically developed innovative drugs, shortening the time from Phase 3 completion to conditional or full approval relative to some historical FDA timelines for novel mechanisms.
The tradeoff is an access lag for patients and clinicians outside Greater China. Even if mazdutide receives NMPA approval, availability in the United States or European Union would require a separate regulatory submission — either a new bridging trial in a non-Chinese population or a full independent trial program, similar to how the FDA required extensive outcomes evidence such as the four-year cardiovascular follow-up data summarized in the SELECT trial's cardiovascular outcomes review before granting broader label claims for existing GLP-1 agents. Until that submission occurs, clinicians outside China evaluating mazdutide are working from trials published in Chinese populations and a single U.S. Phase 2 trial rather than an FDA-reviewed label, a distinction worth stating plainly to any patient asking about it.
Mazdutide vs Tirzepatide, Retatrutide, and Other Multi-Agonists
Mazdutide sits in a specific niche within the expanding multi-receptor agonist field. Tirzepatide (Mounjaro/Zepbound) pairs GIP and GLP-1 receptor agonism; head-to-head data against semaglutide is detailed in the SURMOUNT-5 tirzepatide vs semaglutide trial analysis. Retatrutide adds a third receptor, combining GLP-1, GIP, and glucagon agonism in a single molecule. CagriSema, covered in the REDEFINE-1 Phase 3 data review, takes a different combination approach entirely, pairing a GLP-1 agonist with an amylin analog rather than adding receptor targets to a single peptide.
Mazdutide's dual GLP-1/glucagon mechanism, without a GIP component, positions it closest to cotadutide and survodutide rather than to tirzepatide or retatrutide. Whether that two-receptor combination produces efficacy closer to GLP-1 monotherapy or closer to the triple-agonist class is an empirical question no published head-to-head trial has yet answered. Cross-trial comparisons of percent weight change between differently designed studies, in different populations, with different titration schedules, are informative for hypothesis generation but not a substitute for a randomized head-to-head design — a caveat that applies to every comparison in this section.
Open Questions and Research Gaps
Several gaps limit how far the current mazdutide evidence base can be extrapolated. First, no long-term cardiovascular outcomes trial comparable in scale or duration to SELECT exists yet for mazdutide, so its effect on major adverse cardiovascular events remains unestablished. Second, every published Phase 3 trial to date has enrolled Chinese participants, and the only U.S. data come from a Phase 2 trial (Hsia SH, et al. Lancet Diabetes Endocrinol. 2026; PMID 42628555), which raises open questions about generalizability given known differences in obesity-related BMI thresholds and metabolic phenotype across populations — a multi-region trial or formal bridging study would be needed before extrapolating dosing or efficacy expectations elsewhere.
Third, durability after treatment discontinuation has not been reported for mazdutide. The pattern seen with other GLP-1-class agents, including the weight regain documented in the STEP 4 long-term discontinuation follow-up, suggests discontinuation-related regain is a class-wide phenomenon worth tracking specifically for dual agonists, where the glucagon-driven energy-expenditure component could theoretically behave differently on withdrawal, though this is speculative absent data. Finally, GLORY-1, GLORY-2 and two type 2 diabetes trials, one against dulaglutide (Guo L, et al. Nature. 2026; PMID 41407860) and one against placebo (Zhu D, et al. Nature. 2026; PMID 41407859), are now peer-reviewed. The semaglutide-comparator DREAMS-3 trial has so far been published only as a design and baseline report (Luo Y, et al. Contemp Clin Trials. 2026; PMID 41260459), so its results should be read from the full publication when it appears rather than from sponsor topline disclosures.
For clinicians and researchers monitoring this compound, the practical next step is to track the ClinicalTrials.gov registry entries and PubMed indexing for mazdutide directly rather than relying on secondary press coverage, and to treat any specific percentage or comparator claim as provisional until it appears in a peer-reviewed manuscript. Patients asking about mazdutide access outside China should be directed to discuss its unapproved, investigational status with a licensed prescriber before considering any import or off-label use.
This article summarizes research and does not constitute medical advice. Consult a licensed clinician for diagnosis, treatment, or any decisions about medications or supplements.