CT-868 (acmopatide)
Carmot Therapeutics (Roche)
Small A1c effect, big weight and insulin-dose effect. Early days.
An investigational once-daily dual GLP-1/GIP receptor agonist — and, per the American Diabetes Association, the first drug of its class developed specifically for type 1 diabetes rather than borrowed from type 2. In a 111-person Phase 2 it lowered A1c modestly, cut daily insulin dose by up to 15%, and drove up to about 7% weight loss, with no diabetic ketoacidosis and no severe hypoglycemia over 16 weeks. Roche moved acmopatide into its phase-3 T1D pipeline in July 2026, although no pivotal-trial registry identifier or readout date has been disclosed.
The scorecard
In type 1 diabetes, A1c fell 0.34% from a baseline of 7.5% at the 4.1 mg dose, and 56% of participants reached an A1c below 7%. That is a real effect, but a modest one — the weight and insulin-dose results are the bigger story. Continuous-glucose time-in-range figures have not been published.[1]
Generally well tolerated, with no diabetic ketoacidosis and no level 3 (severe) hypoglycemia reported over 16 weeks. Time spent below 54 mg/dL (3.0 mmol/L) rose numerically on the drug but stayed under the recommended 1% ceiling in every group. Held back from a higher score because this is conference-only data — 111 people over four months cannot rule out rarer harms, and no peer-reviewed type 1 publication exists yet.[3]
The strongest results in the trial: dose-dependent weight loss of up to about 7% versus placebo, and daily insulin dose reduced by up to 15%. Participants started at a mean BMI of 34, so this targets a real and common problem. No blood-pressure, cardiovascular or kidney outcomes have been reported.[1]
One randomized, placebo-controlled Phase 2 in 111 adults, 16 weeks long, presented at a conference and not yet peer-reviewed. Roche's 23 July 2026 pipeline update lists acmopatide as new to phase 3 for T1D, but provides no pivotal-trial identifier, design, start date, or readout timing. It remains unapproved anywhere for any indication.[5]
Access convention (cheaper and more available is better): purely investigational. Unlike most adjuncts in this category, it cannot even be obtained off-label, because it is not a marketed drug. Available only inside a trial.[1]
These are scored as add-ons to insulin in Type 1 diabetes, never on their Type 2 diabetes evidence — a drug with a large Type 2 trial base but only small Type 1 studies scores low on maturity. Safety carries heavy weight here because the best-known adjunct risk, diabetic ketoacidosis at normal glucose levels, is potentially fatal and specific to Type 1. Most of this class is used off-label; that is reflected in access, not hidden.
Editor’s take
Read the numbers carefully, because the headline and the substance point in different directions. The A1c move — 0.34% from a starting point of 7.5% — is small. The weight loss (up to about 7%) and the insulin-dose reduction (up to 15%) are not. This is best understood as a metabolic-health drug for the large minority of adults with type 1 diabetes who also carry excess weight and insulin resistance, rather than a glucose-control breakthrough. Reassuringly, there was no diabetic ketoacidosis and no severe hypoglycemia over 16 weeks — but 111 people for four months is not where you learn about rare harms, and the results have only been presented at a conference, not published. The genuinely novel thing here is who it was built for: this is an incretin drug designed for type 1 diabetes from the start, not one borrowed from the type 2 shelf and hoped for the best. Roche's phase-3 advancement is a real programme milestone, but until the pivotal protocol appears it is a sponsor-pipeline status, not evidence that a phase-3 participant has been dosed.
The full picture
CT-868, also called acmopatide, is a once-daily injected drug that activates two gut-hormone receptors at once — GLP-1 and GIP. Drugs in this family are familiar from type 2 diabetes and obesity, where they lower glucose, slow the stomach's emptying, and cause substantial weight loss. What makes CT-868 different is intent. According to the American Diabetes Association, it is the first dual GLP-1/GIP agonist developed specifically for type 1 diabetes, rather than a type 2 drug that people with T1D and their doctors have tried off-label and hoped would translate.
The evidence is one trial. It was randomized and placebo-controlled, which matters, and it enrolled 111 adults who had type 1 diabetes and a BMI of at least 27 kg/m² — mean age 41, mean BMI 34, mean starting A1c 7.5%. Participants took acmopatide at 1.8, 4.1 or 6.6 mg once daily, or placebo, for 16 weeks. Everyone stayed on insulin; this is an add-on, not a replacement, and nothing about it changes the fact that a person with type 1 diabetes needs insulin to live.
At the 4.1 mg dose, A1c fell by 0.34% from baseline, and 56% of participants got below 7%. Weight loss was dose-dependent and reached about 7% versus placebo. Daily insulin dose dropped by up to 15% versus placebo.
Those three results are not equally impressive, and it is worth being blunt about which is which. A 0.34% A1c reduction from a baseline of 7.5% is a modest glycemic effect — real, but not transformative, and the trial did not release continuous-glucose time-in-range figures, which is the number most people would actually want. The weight and insulin-dose effects are the substantial ones. For the sizeable group of adults with type 1 diabetes who also carry excess weight and insulin resistance — sometimes called "double diabetes" — that combination is the point of the drug.
Safety is the question that decides whether any adjunct is worth taking in type 1 diabetes, and here the early news is reassuring. Over the 16 weeks there were no cases of diabetic ketoacidosis and no cases of level 3 (severe) hypoglycemia. Time spent below 54 mg/dL (3.0 mmol/L) did rise numerically in the groups taking the drug, but every group stayed under the recommended ceiling of 1% of the day. The drug was described as generally well tolerated. Set against that: this class slows gastric emptying and reduces food intake, and the trial cut people's insulin doses — a combination that demands care with dose timing, and one that 111 people followed for four months cannot fully characterise.
Two caveats deserve equal weight with the results. First, this is conference data, not a publication. The type 1 findings were presented at the ADA's 86th Scientific Sessions in June 2026 and released by press release; they have not been peer-reviewed. There is a peer-reviewed Phase 2 of CT-868 — but it is in type 2 diabetes, and type 2 evidence does not establish anything about safety or benefit in type 1. Second, the development path has advanced but remains under-described. Carmot Therapeutics, which started the trial, was acquired by Roche while it was running. Roche's 23 July 2026 pipeline now lists RG6641/acmopatide as new to phase 3 for T1D.1 Roche has not disclosed the pivotal protocol, trial identifier, start date, readout timing, or filing plan. A sponsor pipeline move is a licence to run a phase 3, not a licence to sell a drug.
One quieter reason this belongs on the artificial-pancreas map as well as the pharmacy shelf: an incretin drug blunts the post-meal glucose spike at its source, by slowing how fast food arrives. A closed loop struggles most with meals it was not told about. A drug that flattens the meal curve makes the loop's hardest guess an easier one — the same logic behind pairing insulin with amylin in fully-closed-loop research. That is a hypothesis, not a finding; nobody has yet tested CT-868 inside an automated insulin delivery system.
References
-
Roche. Half-year 2026 pharmaceuticals pipeline (status 23 July 2026). https://www.roche.com/pharmahy26.pdf ↩
Coming soon
ETA · Phase 2 in type 1 diabetes is complete and was reported at the ADA's 86th Scientific Sessions in June 2026. Roche moved acmopatide into its phase-3 T1D pipeline in July, but has not published the pivotal design, trial identifier, readout date, or filing timeline. A T1D licence therefore remains years away even if phase 3 succeeds.
- →Full peer-reviewed publication of the type 1 Phase 2 results — the numbers below all come from a conference presentation and press release
- →Roche disclosure and registration of the phase-3 T1D protocol, including endpoints, enrollment, and timing
- →Continuous-glucose data (time in range) from the Phase 2, which have not been released
Sources
- [1]First Dual GLP-1/GIP Receptor Agonist Developed for Type 1 Diabetes Shows Promising Benefits in Blood Glucose Control · conference — Presented at the ADA's 86th Scientific Sessions, June 2026. Randomized, placebo-controlled Phase 2: 111 adults with type 1 diabetes and BMI at or above 27 kg/m2 (mean age 41, mean BMI 34, mean baseline A1c 7.5%), four arms (1.8, 4.1 or 6.6 mg once daily, or placebo), 16 weeks to the primary endpoint. At 4.1 mg, A1c fell 0.34% and 56% reached below 7%; dose-dependent weight loss up to about 7%; insulin dose reduced up to 15% versus placebo.
- [2]First Dual GLP-1/GIP Receptor Agonist Developed for Type 1 Diabetes Shows Promising Benefits in Blood Glucose Control (ADA newsroom) · conference — ADA newsroom release accompanying the June 2026 Scientific Sessions presentation.
- [3]Once-daily GLP-1/GIP dual agonist lowers HbA1c, weight in type 1 diabetes · news · 2026-06-06 — Conference coverage reporting the safety detail the headline release did not: no cases of level 3 hypoglycemia and no cases of diabetic ketoacidosis during the trial. Time below 54 mg/dL (3.0 mmol/L) increased numerically in the acmopatide arms but all groups stayed below the recommended 1% threshold.
- [4]Efficacy and safety of CT-868, a fully biased dual GLP-1/GIP receptor agonist, in type 2 diabetes: a phase 2 trial · peer-reviewed — Published 2026 in Diabetes, Obesity and Metabolism. The peer-reviewed *type 2* Phase 2. Useful for background pharmacology only — it is not evidence for use in type 1 diabetes and this item is not scored on it.
- [5]Roche half-year 2026 pharmaceuticals pipeline · manufacturer · 2026-07-23 — Roche's pipeline table, status 23 July 2026, lists “RG6641 acmopatide — T1D” among Q2 changes new to phase III. It does not identify the pivotal protocol or give a start or readout date.