Does Ecnoglutide's cAMP-Biased GLP-1 Receptor Signaling Translate Into Non-Inferior HbA1c Reduction Versus Dulaglutide With Comparable Tolerability in 2026?
Yes. The EECOH-2 phase 3 trial demonstrated that once-weekly ecnoglutide at both tested doses was non-inferior to dulaglutide in HbA1c reduction over 52 weeks, with the higher dose achieving superiority. The gastrointestinal adverse-event profile was mild-to-moderate and did not exceed dulaglutide rates.
What Is cAMP-Biased Agonism at the GLP-1 Receptor and Why Does It Matter Mechanistically?
cAMP-biased agonism at the GLP-1 receptor means a ligand preferentially activates the Gs-adenylyl cyclase-cAMP axis while minimising beta-arrestin recruitment. Because beta-arrestin drives receptor internalisation and desensitisation, a cAMP-biased agonist sustains insulin-secretory signalling longer per receptor occupancy event and may reduce emetic pathway activation linked to beta-arrestin-dependent brainstem circuits.
The GLP-1R is a class B GPCR that canonically couples to Gs, elevating intracellular cAMP and triggering protein kinase A-dependent insulin exocytosis in pancreatic beta-cells. Agonist binding simultaneously recruits beta-arrestin-1 and beta-arrestin-2, which uncouple the receptor from Gs and initiate clathrin-mediated endocytosis.
In diabetogenic conditions, beta-arrestin-2 expression in human islets is reduced, amplifying the relative importance of the cAMP arm for insulin release. This observation provides a pathophysiological rationale for why cAMP-biased agonism may be particularly advantageous in the diabetic milieu.
Preclinical work by Jones et al (2018, Nature Communications) established that beta-arrestin recruitment during sustained GLP-1R agonist exposure paradoxically attenuates insulin release by accelerating receptor internalisation. Genetic ablation or pharmacological reduction of beta-arrestin recruitment in mice improved anti-hyperglycaemic outcomes without proportional loss of efficacy.
A 2025 review by Douros et al in Frontiers in Endocrinology proposed that reduced beta-arrestin recruitment correlates with diminished activation of area postrema circuits mediating nausea and emesis. This offers a plausible pathway by which cAMP-biased agonists might improve GI tolerability relative to unbiased comparators.
How Was Ecnoglutide Engineered to Achieve cAMP Bias?
Ecnoglutide (XW003) is a 26-amino-acid GLP-1 analogue in which a valine substitution at a key receptor-contact position biases GLP-1R signalling toward cAMP generation over beta-arrestin recruitment. A C18 fatty-acid chain conjugated via a linker confers albumin binding, extending the half-life to approximately one week and enabling once-weekly subcutaneous dosing.
The discovery chemistry is described by Guo et al (2023, Cell Chemical Biology). In cell-based assays, ecnoglutide demonstrated potent cAMP induction with substantially reduced beta-arrestin-2 recruitment relative to native GLP-1 and semaglutide.
The valine substitution disrupts a hydrophobic contact that normally stabilises the beta-arrestin-receptor interface without materially altering Gs coupling geometry. The albumin-binding strategy mirrors the approach used in semaglutide but employs a distinct linker architecture.
In preclinical pharmacokinetic studies, ecnoglutide achieved a half-life consistent with once-weekly dosing in non-human primates. Head-to-head rodent comparisons showed more pronounced body-weight reduction with ecnoglutide than with semaglutide at equimolar doses, attributed to the sustained cAMP signal from reduced receptor internalisation.
What Did the EECOH-1 Placebo-Controlled Phase 3 Trial Show?
EECOH-1 was a randomised, double-blind, placebo-controlled phase 3 trial in adults with type 2 diabetes inadequately controlled by diet and exercise. At week 24, HbA1c reductions were minus 196 basis points with the low dose and minus 243 basis points with the high dose, versus minus 34 basis points with placebo. Both differences were statistically significant.
The trial enrolled patients across multiple Chinese centres with a mean baseline HbA1c of approximately 8 to 9 percent. The high dose achieved HbA1c below 7 percent in over 80 percent of participants at end of treatment, compared with approximately 9 percent on placebo.
Body-weight reductions of 5 percent or greater were recorded in one-third of the high-dose group versus 3 percent on placebo. Gastrointestinal adverse events were predominantly mild-to-moderate and concentrated in the dose-escalation phase.
No new safety signals were identified in EECOH-1. Hypoglycaemia rates were low in the absence of concomitant sulfonylurea or insulin, consistent with the glucose-dependent insulin-secretion mechanism of GLP-1R agonists.
What Did the EECOH-2 Active-Comparator Phase 3 Trial Demonstrate Against Dulaglutide?
EECOH-2 was a 52-week, multicentre, open-label, non-inferiority trial comparing ecnoglutide at two doses once weekly against dulaglutide as add-on to metformin. Both ecnoglutide doses met the non-inferiority margin for HbA1c reduction, and the higher dose additionally demonstrated superiority. Reductions were sustained through week 52.
The primary endpoint was change from baseline in HbA1c at week 32. The higher ecnoglutide dose achieved a reduction of approximately 191 basis points from baseline, numerically exceeding dulaglutide.
Secondary endpoints including fasting plasma glucose, 2-hour post-prandial glucose, body weight, and waist circumference all showed statistically significant improvements versus dulaglutide in the higher-dose arm. These results were published by He et al in The Lancet Diabetes and Endocrinology in 2025.
Serious adverse events were infrequent across all arms. Six patients (3 percent) in the low-dose group and eight patients (4 percent) in the high-dose group discontinued due to adverse events over 52 weeks.
Does the cAMP-Bias Hypothesis Translate Into a Measurably Better GI Tolerability Profile?
The EECOH-2 data show ecnoglutide's GI adverse-event incidence was comparable to dulaglutide but not statistically lower. The cAMP-bias hypothesis did not produce a detectable tolerability advantage in this active-comparator trial. The hypothesis remains mechanistically plausible but is not yet confirmed as a clinical differentiator in head-to-head data.
Several confounders limit the tolerability comparison. Dulaglutide is itself a relatively well-tolerated GLP-1R agonist, and the open-label design introduces ascertainment bias for subjective GI symptoms.
The trial was not powered to detect a tolerability difference as a primary endpoint; GI events were captured as secondary safety data. The most commonly reported gastrointestinal events for ecnoglutide were diarrhoea and nausea, consistent with the GLP-1R agonist class profile.
The Douros et al (2025) mechanistic framework predicts that tolerability advantages from cAMP bias may be most apparent at higher doses or against agents with heavier GI burdens. Ecnoglutide's obesity-indication phase 3 programme uses higher doses and may provide a more informative tolerability dataset once published.
What Is Ecnoglutide's Current Regulatory Status and What Does That Establish About the Evidence Base?
Ecnoglutide injection was approved by China's NMPA in January 2026 for adult type 2 diabetes and subsequently for chronic weight management. This makes it the world's first approved cAMP-biased GLP-1 receptor agonist. No regulatory filing has been submitted to the FDA or EMA as of mid-2026.
The NMPA approval was supported by the EECOH-1 and EECOH-2 phase 3 datasets, along with earlier phase 2 data published in Nature Communications by Zhu et al in 2024. Sciwind Biosciences has announced a commercial partnership with Pfizer China for domestic launch.
The absence of an FDA or EMA filing means the compound's evidence base has not been subjected to Western regulatory scrutiny. Whether the EECOH programme's Chinese-centre enrolment and open-label active-comparator design would satisfy FDA non-inferiority trial standards for a new molecular entity remains an open question.
What Are the Principal Evidence-Quality Limitations of the EECOH Phase 3 Programme?
The EECOH programme's principal limitations include single-country enrolment restricting generalisability, an open-label design in EECOH-2 introducing performance and detection bias, absence of cardiovascular outcomes data, and a comparator arm using dulaglutide rather than higher-efficacy agents. These factors constrain the strength of inference available from the non-inferiority result.
Non-inferiority trial design requires careful scrutiny of the margin selection. The pre-specified margin in EECOH-2 was 40 basis points in HbA1c, a standard choice in the field. The superiority result for the higher ecnoglutide dose partially addresses the concern that ecnoglutide might be meaningfully less effective than dulaglutide while still meeting the primary endpoint.
The 52-week duration is adequate for glycaemic endpoints but insufficient to characterise long-term durability, cardiovascular risk modification, or beta-cell preservation effects. The EECOH-2 publication calls for longer-duration comparative trials against semaglutide and tirzepatide.
What Is the Clinical Significance of Demonstrating Non-Inferiority via a Mechanistically Distinct Pathway?
Establishing non-inferiority through a structurally distinct signalling mechanism validates biased agonism as a clinically testable pharmacological strategy in diabetes. It opens a design space for next-generation GLP-1R agonists engineered around specific signalling arms, independent of whether ecnoglutide itself currently confers a measurable tolerability advantage over existing agents.
The EECOH programme provides the first phase 3 RCT evidence that deliberate signalling bias at a class B GPCR can be translated into a clinically approved compound without sacrificing efficacy. This proof-of-concept is relevant to ongoing programmes targeting beta-arrestin-biased or Gq-biased variants at GLP-1R, GIPR, and glucagon receptor.
For the diabetes pharmacology field, the data confirm that the cAMP arm of GLP-1R signalling is sufficient to drive HbA1c reductions required for regulatory approval. This mechanistic confirmation supports the hypothesis that receptor internalisation driven by beta-arrestin is not required for therapeutic glycaemic efficacy, and may in fact limit it. How Do GLP-1 Agonists and AOD-9604 Interact Mechanistically in a 2026 Weight-Loss Stack, and What Dosing Sequence Avoids Receptor Saturation? What Does the 2026 Rat Study Reveal About Semaglutide-Induced Prolonged GLP-1 Receptor Activation and Sodium Balance? Why Do GIP Receptor Agonists and Antagonists Both Produce Weight Loss in 2026 Obesity Trials?