A once-daily injectable dual GLP-1 and GIP agonist, acmopatide (CT-868), reduced body weight by 3.3% to 6.7% over 16 weeks in adults with type 1 diabetes and overweight or obesity, against a 0.3% gain on placebo, according to a phase 2 trial published in Diabetes Care on 28 October 2026. Insulin doses fell and there were no episodes of diabetic ketoacidosis. No GLP-1 drug is approved for type 1 diabetes, so this is one of the few placebo-controlled looks at the drug class in that group.
What the trial measured
The study randomised 111 adults with type 1 diabetes, an HbA1c between 7.0% and 10.0% and a BMI of at least 27 equally to one of three acmopatide dose levels or a volume-matched placebo, given as daily injections under double-blind conditions at multiple US centres. Average baseline HbA1c was 7.5% and average BMI 34.3. The primary outcome was the change in HbA1c at week 16, with weight and insulin dose as secondary outcomes.
The glucose result was modest. HbA1c fell by 0.4 to 0.6 percentage points on acmopatide against 0.3 on placebo. Only the middle dose reached what the authors call nominal significance, a placebo-adjusted difference of 0.3 points with a 90% confidence interval of 0.1 to 0.6 points in favour of the drug. A 90% interval is narrower than the 95% interval most papers report, so this is weaker evidence than it first appears. On that dose, 55.6% reached an HbA1c below 7.0% against 33.3% on placebo.
Weight is the clearer signal. Subtracting the placebo change from each dose group’s mean gives a difference of roughly 3.6 to 7.0 percentage points, and every dose separated from placebo (P below 0.01). Total daily insulin fell by 3.4% to 14.9% more than placebo on the two higher doses. Systolic blood pressure fell by 3.1 to 8.4 mmHg, against 3.3 mmHg on placebo, so the lowest dose was no better than placebo on that measure. Side effects were mainly gastrointestinal, and there was no severe hypoglycaemia.
Where it fits
Acmopatide targets the same two receptors as tirzepatide, but it is designed as a signal-biased agonist, meaning it is engineered to favour some downstream receptor signalling pathways over others. Whether that design makes any practical difference for patients has not been shown in head-to-head studies. The drug came to Roche through its purchase of Carmot Therapeutics, and several authors are employees of Genentech and Roche.
The type 1 setting is the interesting part. Many people with type 1 diabetes now also live with overweight or obesity, and some clinicians already prescribe semaglutide or tirzepatide off-label for them. The worry with any drug that cuts appetite and insulin needs in type 1 diabetes is ketoacidosis, which can follow if insulin is reduced too far. The absence of any ketoacidosis here is reassuring, but in a study this small and short it cannot rule out an uncommon risk. Our GLP-1 overview covers how the approved drugs in this class compare.
What is not yet known
This is a phase 2 dose-finding trial: about 27 to 29 people per arm, followed for 16 weeks. Weight loss on GLP-1-based drugs usually keeps building well beyond four months, so these figures say little about where it would plateau. The HbA1c benefit was small and statistically fragile. The trial was industry-led, and a 16-week absence of ketoacidosis in about 80 treated people says little about how often it would occur across thousands of patients over years. Acmopatide is unapproved and is not available outside clinical trials, and the abstract does not say whether a phase 3 programme in type 1 diabetes is planned.
Sources: Klein KR, Pettus J, Sonnenberg GE, et al.; Acmopatide 004 Study Group. A phase 2, double-blind, randomized, placebo-controlled study to evaluate the efficacy and safety of acmopatide (CT-868), a fully biased GLP-1/GIP receptor agonist, in adults with type 1 diabetes and overweight or obesity. Diabetes Care, published online 28 October 2026. PMID 42949535. Abstract read via PubMed; full text not reviewed.
