Amycretin, now called zenagamtide, remains an investigational medicine as of the documentary check on 17 October 2026. The trials show signals of weight loss and improved blood glucose. They do not allow you to choose a personal dose or conclude that a vial sold for research matches the product studied.
The striking figure often quoted comes from an early injectable trial: an estimated −24.3% change in body weight at thirty-six weeks, in a particular cohort. It is neither the average for all participants in the program, nor a result obtained with the tablet, nor a direct comparison with tirzepatide.
Here, I separate routes, populations and analysis methods. The 2026 data in people with diabetes add to the first publications from 2025: stopping at the older announcements would give an incomplete picture of the evidence.
Amycretin and zenagamtide: one molecule, several targets
Amycretin is the development name to look for in the early papers. The 2026 publications use zenagamtide, previously amycretin. The code NNC0487-0111 also appears in registries. These names refer to the same candidate, not three available treatments.
GLP-1 and amylin are two hormonal signals involved in regulating meals and metabolism. The aim is to combine their actions in a single molecule. This differs from CagriSema, which combines two molecules, cagrilintide and semaglutide. The cagrilintide guide · FR covers that other program.
Mechanism under study
Two highlighted pathways, one molecule
- GLP-1 pathwayA signal involved in regulating appetite and blood glucose.
- Amylin pathwayA signal involved in satiety and regulating food intake.
The shorthand “dual agonist” therefore describes the two main pathways highlighted, without capturing all the reported pharmacology. The phase 2 publications explicitly mention GLP-1, amylin and calcitonin. Adding a target to a description does not let you calculate an extra percentage of efficacy.
Oral trials: twelve weeks do not become thirty-six
The data below come from verified publication abstracts, registries and, where specified, a sponsor communication. The full texts and appendices of all trials could not be consulted. “Not quantified in the abstract” never means “no withdrawals.”
Oral route table
| Study and population | Sample size and duration | Reported efficacy | Available tolerability and withdrawal data |
|---|---|---|---|
| Phase 1, Gasiorek and colleagues, 2025. Adults aged 18 to 55, with overweight or obesity | 144 in total, across several parts. Only 60 in the twelve-week parts | Up to −13.1% body weight versus −1.2% with placebo at twelve weeks, according to the press release accompanying the publication | Mainly gastrointestinal events, mild or moderate in the abstract. Discontinuation rate not quantified in this abstract |
| Phase 2, Mora and colleagues, 2026. Type 2 diabetes treated with metformin, with or without an SGLT2 inhibitor | 186 participants: 156 receiving the molecule, 30 receiving placebo. Thirty-six weeks | Primary endpoint: HbA1c. Estimated mean reduction of 0.9–1.4 percentage points depending on the group, with significant differences versus placebo | Gastrointestinal events: 26–47% depending on the active-treatment group, 23% with placebo. Discontinuation rate not quantified in the abstract |
In phase 1, the primary endpoint concerned adverse events. Weight was exploratory. Not all 144 participants therefore received twelve weeks of treatment: some parts tested a single dose or short exposure. Presenting “144 people, −13.1%” erases that distinction.
The oral weight figure comes from the Novo Nordisk press release of 20 June 2025, because it is not detailed in the scientific abstract consulted. I identify it as such. The phase 2 publication adds a population with diabetes and a longer duration, with HbA1c as the primary endpoint: it is not a repeat of the first trial.
HbA1c reflects blood glucose exposure over several weeks. A reduction expressed in HbA1c percentage points is not weight loss expressed as a percentage. Putting these numbers in the same column labelled “efficacy” without specifying the endpoint creates a comparison that does not exist.
Subcutaneous trials: distinguish the cohorts
Injectable route table
| Study and population | Sample size and duration | Reported efficacy | Available tolerability and withdrawal data |
|---|---|---|---|
| Phase 1b/2a, Dahl and colleagues, 2025. Adults aged 18 to 55, BMI 27–39.9 | 125 in total: 101 receiving the molecule, 24 receiving placebo. Several parts and durations | Escalation cohort: −24.3% versus −1.1% at thirty-six weeks | Many withdrawals reported. A substantial proportion is attributed to reasons other than adverse events. Exact rate not provided in the abstract |
| Other cohorts in the same trial, not new independent trials | Twenty, twenty-eight or thirty-six weeks depending on the cohort | Respectively −9.7%, −16.2% and −22.0%, versus +2.0%, +2.3% and +1.9% with the corresponding placebos | Frequent gastrointestinal events, mostly mild or moderate. Durations and exposures differ |
| Phase 2, Mora and colleagues, 2026. Type 2 diabetes, background treatment comparable to the oral part | 262 randomized, 261 exposed. Thirty-six weeks | HbA1c: estimated reduction of 0.9–1.7 percentage points depending on the group. Weight: up to −14.6% according to the sponsor’s ADA 2026 presentation | Mainly gastrointestinal events. Twenty-one of 261 participants reported serious events, including three receiving placebo. Discontinuation rate not quantified in the abstract |
The 2025 results are estimated mean changes by cohort, not the proportion of people reaching a target. The publication press release specifies the assumption that all participants had followed the planned treatment through to the end. This assumption matters particularly when withdrawals are numerous.
The −14.6% figure in people with diabetes is attributed to the ADA 2026 communication. It must not be used to declare the injectable less effective than in the previous trial: the population, exposures and methods are not identical.
The two oral and injectable phase 2 papers come from the same program registered under NCT06542874, with 448 participants in total. Counting them as two fully independent replications would overstate the evidence base. The trials cited are funded by Novo Nordisk, with several authors employed by the sponsor.
Tolerability: withdrawals are part of the result
An adverse event is a problem that occurred during follow-up. It is not automatically caused by the molecule. Likewise, “mild or moderate” describes intensity, whereas a “serious” event meets other criteria, such as hospitalization. These terms are not interchangeable.
For the early injectable data, the authors report many withdrawals, a substantial proportion for reasons other than adverse events. This rules out two shortcuts: attributing every withdrawal to intolerance, or treating missing data as unimportant.
The oral phase 2 abstract also reports seven participants with serious events in the active-treatment groups, none receiving placebo, and no deaths in the trial. This is not enough to attribute these events to the medicine. Small sample sizes and limited follow-up leave uncertainties about rare and late risks.
When reading a full publication, you need several distinct denominators:
- The number of people randomized and the number actually exposed.
- Treatment discontinuations and study withdrawals, which may differ.
- Discontinuations attributed to adverse events and those for other reasons.
- The number of measurements available at the stated endpoint and how missing data were handled.
This guide therefore does not provide a numerical ranking of withdrawal rates. The openly available abstracts do not allow one to be constructed correctly for every formulation. Filling missing cells with zero would make a neat table and bad information.
Risks that remain poorly characterized
The trials mainly report gastrointestinal problems. They do not establish safety for everyone or a complete interaction profile. Tolerability in frail older people, during pregnancy, with certain diseases or with combinations of treatments cannot be inferred from the first cohorts.
Consult a doctor or pharmacist about decisions concerning your weight or diabetes, particularly if you are taking medication, are pregnant or breastfeeding, or have kidney or liver disease. A study in selected adults does not justify extending use to those under 18. Do not replace a prescribed treatment with an experimental product.
After exposure to a non-prescribed product, severe or persistent abdominal pain, repeated vomiting or an inability to drink require prompt medical advice. Feeling seriously unwell, difficulty breathing or facial swelling warrant urgent help, by calling 15 or 112 in France. Keep the product information for the healthcare professional, without taking another dose to “check” the reaction.
Why comparisons with tirzepatide are limited
The semaglutide and tirzepatide comparison · FR distinguishes their respective trials and indications. You cannot add amycretin as a new contender on the podium based on −24.3% alone.
An interpretable comparison must align several parameters:
- Population with or without diabetes, BMI and background treatments.
- Total duration, exposure-escalation period and time spent on the maintenance regimen.
- Route and formulation, with their absorption constraints.
- Endpoint studied, comparator and analysis method after discontinuations.
An estimand is the precise question the analysis seeks to answer. “What would happen if everyone followed the treatment?” differs from “What is the result after treatment assignment, accounting for interruptions?” Two accurate percentages can answer two different questions.
The phase 3 trial AMAZE 7 is registered for a comparison with semaglutide. As of the check on 17 October 2026, it is recruiting and has no posted results. A study of switching from tirzepatide also appears in the registry, but this type of tolerability study does not replace a randomized superiority trial. No superiority over tirzepatide is established by the publications presented here.
Status in France: documentary check on 17 October 2026
No authorization entry found under amycretin, amycrétine or zenagamtide in the data consulted. The check covers the EMA medicines export and the medicinal product and composition files from the French Public Medicines Database, dated 29 September 2026.
This conclusion documents the search, subject to each database’s update delays. It does not replace a fresh check at the time of any later publication. I am not announcing a French sale date, price, reimbursement or unverified early access.
The registries do, however, confirm phase 3 development: AMAZE 1 for the injectable and AMAZE 9 for the tablets are recruiting as of the check. Their results have not yet been posted. Authorizing a clinical trial and authorizing marketing are two different decisions.
A research product does not inherit a trial’s results
Moving from a studied peptide to a medicine involves a controlled formulation, controlled production and clinical monitoring. An “amycretin” or “research use only” listing does not document this chain. Claimed purity is not enough to establish identity, actual quantity, contaminants, sterility or stability.
The EMA and national authorities warn about fake GLP-1 medicines sold online. They may not contain the stated substance or may contain other harmful substances. The ANSM has also published measures against illegal sales, a document consulted in the copy distributed by the Île-de-France pharmacovigilance centers. These alerts do not prove the composition of any particular amycretin vial, but they explain why a label is not enough.
The Toll, here, is the hidden cost of a shortcut: you leave behind the conditions that make trials interpretable, without obtaining their quality and monitoring safeguards. To prepare for a consultation, bring your goal, your treatments and your medical history instead. Do not use this guide to convert between tablets, injections and other medicines.
