CT-388: Roche's Dual GLP-1/GIP Agonist With 22.5% Weight Loss
Roche's CT-388 delivered 22.5% placebo-adjusted weight loss in Phase 2 and resolved obesity in 54% of patients at the top dose. The signaling-biased science explained.
Medically Reviewed
Reviewed by Dr. James Chen, MD, PhD, FACE on July 17, 2026
Our medical review process ensures clinical accuracy and patient safety.
CT-388 is Roche's once-weekly dual GLP-1/GIP receptor agonist, and in January 2026 it posted the strongest Phase 2 weight-loss result the obesity field has seen from a two-hormone drug: 22.5% placebo-adjusted weight loss at 48 weeks. That figure puts CT-388 in territory previously reserved for triple agonists and bariatric surgery, and it arrived from a molecule engineered around a mechanistic bet that most of its competitors did not make — deliberate signaling bias at both receptors. Roche has since designated CT-388 a fast-track asset and moved it into a Phase 3 program.
Evidence: "CT-388 is a unimolecular peptide-based dual GLP-1R/GIPR agonist that is cAMP signal-biased at both receptors, with minimal receptor internalization vs their native ligands." — Chakravarthy MV, Rodriguez R, Hergarden A, Elliott MA, et al. Molecular Metabolism. 2025. DOI: 10.1016/j.molmet.2025.102291
The drug engages the same two receptors as tirzepatide — the approved GLP-1/GIP dual agonist behind Mounjaro and Zepbound — but the way it engages them is not the same, and that difference is the entire thesis behind the molecule.
Why Signaling Bias Is CT-388's Central Design Choice
When a peptide binds the GLP-1 receptor, two things happen. The receptor triggers cAMP signaling, which is the pathway that drives insulin secretion, appetite suppression, and the metabolic effects patients want. But the receptor also recruits a protein called β-arrestin, which tags the receptor for internalization — physically pulling it inside the cell, where it can no longer respond to the drug.
Internalization is a self-limiting brake. The more aggressively an agonist drives β-arrestin recruitment, the faster it strips its own receptors off the cell surface, and the more the tissue desensitizes to continued exposure. A biased agonist is one engineered to activate the useful pathway while largely skipping the one that shuts it down.
Evidence: "Reduced GLP-1 receptor internalization and recycling, achieved through biased agonism, results in greater sustained insulin release and improved glycaemic efficacy in vivo compared with agonists that drive rapid receptor endocytosis." — Jones B, Bloom SR, Buenaventura T, et al. Nature Communications. 2018;9:1602. DOI: 10.1038/s41467-018-03941-2
CT-388 takes that principle to an unusual extreme. In the receptor pharmacology published in Molecular Metabolism, CT-388 produced essentially no β-arrestin-2 recruitment at the human GLP-1 receptor — an Emax of 4%, which is background noise — and internalized the receptor at only 6% of the efficacy of native GLP-1. GIP receptor internalization was characterized as very low by the same measures. The molecule activates cAMP at both receptors and then, functionally, leaves the receptors where they are.
The receptor math
| Property | CT-388 | Implication |
|---|---|---|
| GLP-1R potency (EC₅₀) | 0.087 nM | High-potency GLP-1 arm |
| GIPR potency (EC₅₀) | 2.47 nM | ~28× weaker than GLP-1R |
| β-arrestin-2 at GLP-1R | Emax 4% | Essentially no recruitment |
| GLP-1R internalization | 6% of native GLP-1 | Receptors stay surface-exposed |
| Dosing | Once weekly, SC | Half-life supports weekly |
The potency ratio matters as much as the bias. CT-388 is roughly 28-fold more potent at GLP-1R than at GIPR, making it a GLP-1-dominant molecule with a GIP component rather than a balanced 50/50 dual agonist. That asymmetry is a design decision, not an accident, and it places CT-388 in a different pharmacological position than the balanced dual agonists in development.
What GIP Adds to a GLP-1 Backbone
GIP was the incretin the obesity field spent years arguing about. Blocking GIP produces weight loss in animal models; so does agonizing it. Amgen's MariTide is built on the antagonist hypothesis, while tirzepatide and CT-388 bet on agonism — and both approaches produce weight loss in humans, which is one of the stranger findings in metabolic pharmacology.
Evidence: "GIP receptor agonism enhances the weight-lowering efficacy of GLP-1 through complementary actions on adipose tissue function, nutrient partitioning, and central pathways regulating energy balance, and may also improve the gastrointestinal tolerability of GLP-1 receptor agonism." — Samms RJ, Coghlan MP, Sloop KW. Trends in Endocrinology & Metabolism. 2020;31(6):410-421. DOI: 10.1016/j.tem.2020.02.006
The tolerability argument is the practically important one. Preclinical work suggests GIP receptor activation in the area postrema — the brainstem region that generates nausea — dampens the emetic signal that GLP-1 agonism produces. If that translates, adding GIP lets you push the GLP-1 arm harder before nausea forces you to stop escalating. Since dose ceilings in this class are set almost entirely by gastrointestinal tolerability, a mechanism that raises the ceiling raises the achievable weight loss.
The clinical proof of concept for GIP agonism came from tirzepatide, which established that two receptors beat one.
Evidence: "The mean percentage change in weight at week 72 was −15.0% with 5 mg weekly tirzepatide, −19.5% with 10 mg, and −20.9% with 15 mg, as compared with −3.1% with placebo. 57% of participants in the 15-mg group had a reduction in body weight of 20% or more, as compared with 3% with placebo." — Jastreboff AM, Aronne LJ, Ahmad NN, et al. New England Journal of Medicine. 2022;387(3):205-216. DOI: 10.1056/NEJMoa2206038
The CT-388 Phase 2 Results: 22.5% at 48 Weeks
CT388-103 (NCT06525935) randomized 469 adults with obesity, or overweight with at least one weight-related comorbidity, to once-weekly subcutaneous CT-388 across several dose-escalation regimens topping out at 24 mg, or placebo, for 48 weeks.
The trial met its primary endpoint with a placebo-adjusted mean weight loss of 22.5% at the 24 mg dose (p < 0.001) under the efficacy estimand — which measures the effect in patients who actually stay on the drug — and 18.3% under the treatment-regimen estimand, which counts everyone regardless of discontinuation. Both figures are reported because they answer different questions: the first asks how well the drug works when taken, the second asks how well it works when prescribed. Serious comparisons across trials should use matching estimands, a caveat worth holding onto when CT-388's 22.5% gets set against tirzepatide's 20.9%.
Two details from the readout carry more weight than the headline number.
The curve had not plateaued. At week 48, weight loss at 24 mg was still descending with a clear dose-response relationship. Every drug in this class eventually flattens as intake and expenditure re-equilibrate; a trial that ends before the plateau has not found the molecule's ceiling. Whether the true ceiling sits at 25% or 30% is a Phase 3 question.
More than half the top-dose patients stopped having obesity. 54% of participants on 24 mg reached a BMI under 30 kg/m², versus 13% on placebo. Reframing the endpoint from "percent lost" to "disease resolved" is the shift this generation of drugs is forcing on the field.
Responder rates at 24 mg (week 48)
| Threshold | CT-388 24 mg |
|---|---|
| ≥5% weight loss | 95.7% |
| ≥10% weight loss | 87.0% |
| ≥20% weight loss | 47.8% |
| ≥30% weight loss | 26.1% |
| BMI <30 (obesity resolution) | 54% |
A 26.1% rate of ≥30% weight loss is the number that separates CT-388 from the approved drugs. Losing 30% of body weight was, until recently, a bariatric surgery outcome; the comparison between the two approaches is one we cover in our analysis of bariatric surgery versus GLP-1 medications.
Glycemic results tracked the weight: 73% of participants who entered with prediabetes returned to normal blood glucose, against 7.5% on placebo.
Safety and Tolerability
CT-388's adverse event profile was consistent with the incretin class — predominantly gastrointestinal, predominantly mild to moderate, front-loaded during dose escalation. Roche reported no new or unexpected safety signals.
The tolerability number worth isolating is discontinuation: 5.9% on CT-388 versus 1.3% on placebo. For a drug pushed to 24 mg with weight loss still accelerating at week 48, a sub-6% dropout rate is a favorable result, and it is the strongest available evidence that the signaling-bias engineering is doing something clinically real rather than merely something interesting in a cell assay.
The Phase 1 program had already set that expectation. Across single-ascending-dose and multiple-ascending-dose cohorts in 64 adults with overweight or obesity, mean weight change by day 29 ran from −4.7% to −8.0% versus −0.5% on placebo. Gastrointestinal events occurred in 40–100% of participants depending on cohort, and one serious adverse event — intractable vomiting in the single-dose portion — was judged probably drug-related. No participant discontinued for safety reasons in the multiple-dose portion.
That single serious event is worth stating plainly rather than burying: the bias engineering reduces the GI burden of this class, it does not abolish it. Patients starting CT-388, if it reaches approval, should expect the same nausea that accompanies every drug in this family.
How CT-388 Compares With the Field
| Drug | Receptors | Best weight loss | Status |
|---|---|---|---|
| Semaglutide 2.4 mg | GLP-1 | ~14.9% at 68 wks | Approved |
| Tirzepatide 15 mg | GLP-1 + GIP | ~20.9% at 72 wks | Approved |
| CT-388 24 mg | GLP-1 + GIP (biased) | 22.5% placebo-adj. at 48 wks (Ph 2) | Entering Phase 3 |
| Retatrutide 12 mg | GLP-1 + GIP + glucagon | ~24.2% at 48 wks (Ph 2) | Phase 3 |
The comparison that defines CT-388's commercial case is against retatrutide, Lilly's triple agonist. Retatrutide reaches roughly 24% by recruiting a third receptor — glucagon — which buys additional energy expenditure at the cost of a more complex safety surface, including heart-rate effects. CT-388 is arguing it can approach the same efficacy with two receptors and better-behaved pharmacology, purely by engineering how it engages them. If that argument survives Phase 3, signaling bias becomes a design principle the whole field has to answer, not a Roche curiosity.
The benchmarks in that table come from separate trials with different populations, durations, and estimands, so the ordering is indicative rather than definitive. Only a head-to-head trial settles it, and none is currently running.
Roche has launched the Phase 3 program under the names Enith1 and Enith2. Registrational obesity trials of this size typically read out over roughly two years, which places CT-388's pivotal data in the 2028 range and any approval decision after that.
Key Takeaways
CT-388 delivered 22.5% placebo-adjusted weight loss at 48 weeks in Phase 2, resolved obesity in 54% of top-dose patients, normalized glucose in 73% of those with prediabetes, and did it with a 5.9% discontinuation rate — the best efficacy-tolerability combination reported for a dual agonist to date.
The mechanistic story is the part worth following. CT-388 is not a stronger version of tirzepatide; it is a differently-engineered one, built to activate cAMP at both incretin receptors while refusing to recruit β-arrestin or internalize the receptor it depends on. The Phase 2 data are consistent with that engineering paying off, though consistency is not proof, and receptor-level elegance has failed to survive Phase 3 before.
What to watch: whether the weight-loss curve finds its plateau in Enith1 and Enith2 or keeps descending; whether the tolerability advantage holds in thousands of patients rather than hundreds; and whether the drug retains weight loss after discontinuation any better than its predecessors, which it almost certainly will not — weight regain after stopping remains this entire class's unsolved problem, and no amount of receptor bias changes the fact that obesity is a chronic disease requiring chronic treatment.
References
- Chakravarthy MV, Rodriguez R, Hergarden A, Elliott MA, et al. Effects of CT-388, a once-weekly signaling-biased dual GLP-1/GIP receptor agonist, on weight loss and glycemic control in preclinical models and participants with obesity. Molecular Metabolism. 2025;103:102291. DOI: 10.1016/j.molmet.2025.102291
- Jones B, Bloom SR, Buenaventura T, Tomas A, Rutter GA, et al. Targeting GLP-1 receptor trafficking to improve agonist efficacy. Nature Communications. 2018;9:1602. DOI: 10.1038/s41467-018-03941-2
- Samms RJ, Coghlan MP, Sloop KW. How may GIP enhance the therapeutic efficacy of GLP-1? Trends in Endocrinology & Metabolism. 2020;31(6):410-421. DOI: 10.1016/j.tem.2020.02.006
- Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity (SURMOUNT-1). New England Journal of Medicine. 2022;387(3):205-216. DOI: 10.1056/NEJMoa2206038
- Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity (STEP 1). New England Journal of Medicine. 2021;384(11):989-1002. DOI: 10.1056/NEJMoa2032183
- Jastreboff AM, Kaplan LM, Frías JP, et al. Triple-hormone-receptor agonist retatrutide for obesity — a phase 2 trial. New England Journal of Medicine. 2023;389(6):514-526. DOI: 10.1056/NEJMoa2301972
- Roche. Roche announces positive Phase II results for its dual GLP-1/GIP receptor agonist CT-388 in people living with obesity. Ad hoc announcement pursuant to Art. 53 LR, 27 January 2026. Company release (CT388-103, NCT06525935)
Last updated: 2026-07-17 Medical review: Dr. James Chen, MD, PhD, FACE
Tags
Written By
Dr. Sarah Mitchell
Medical Director, MD, FACP
Dr. Sarah Mitchell is a board-certified internist specializing in metabolic medicine and weight management. With over 15 years of clinical experience, she has helped thousands of patients achieve sustainable weight loss through evidence-based approaches.
Medical Reviewer
Dr. James Chen
Endocrinologist, MD, PhD, FACE
Dr. James Chen is a fellowship-trained endocrinologist with expertise in diabetes, metabolism, and hormone-related weight disorders. His research on GLP-1 receptor agonists has been published in leading medical journals.
Editorial Standards
This article follows our strict editorial guidelines. All content is based on peer-reviewed research and reviewed by medical professionals. This information is for educational purposes only — always consult your healthcare provider before making medical decisions.