Peptide & Experimental
MariTide (Maridebart Cafraglutide)
GIP receptor antagonist + GLP-1 agonist (peptide-antibody conjugate) · AMG 133
MariTide is Amgen's weight-loss candidate with the most unusual approach: it activates the GLP-1 receptor and at the same time blocks the GIP receptor, exactly the opposite of tirzepatide. And it is injected only once a month. It is not approved; phase 3 is under way.
In brief
MariTide, drug name maridebart cafraglutide, is a peptide-antibody conjugate: an antibody to which two peptides are coupled. This keeps it in the body for a long time, and it is injected every 4 weeks. In a double-blind phase 2 trial with 592 participants over 52 weeks, people without diabetes lost 12.3 to 16.2 percent of their weight depending on the group, versus 2.5 percent on placebo. Gastrointestinal complaints were common, and fat-free mass (measured by DXA) fell by 8.6 to 11.6 percent. The 20 percent passed around online is not in the publication. MariTide is an investigational drug; phase 3 is under way.
What MariTide is
MariTide is not a peptide in the usual sense but a peptide-antibody conjugate. Peptides are chemically attached to a monoclonal antibody. The antibody brings the long residence time, the peptides bring the effect. Because the peptides are made separately and then attached, building blocks that do not exist in natural proteins can also be incorporated, such as ring-shaped structures or artificial amino acids.
This results in the practical difference from everything approved today: the injection comes once a month instead of once a week. In the phase 2 trial, dosing every 8 weeks was also tested. That is not a marketing argument but follows from the design principle.
How it is supposed to work
The GLP-1 part works as with semaglutide: less appetite, a better insulin response, slower gastric emptying. The second part is the surprise. GIP, like GLP-1, is a gut hormone, and tirzepatide activates its receptor. MariTide blocks it.
Both work, and nobody knows exactly why. A whole chain of findings supports the blockade: mice without the GIP receptor remain protected from obesity and insulin resistance on a high-fat diet, even when the receptor is missing only in the central nervous system. Genome-wide studies in humans link variants with lower GIP receptor function to a lower BMI. In mice and monkeys, the blockade enhances weight loss from GLP-1. A phase 1 trial showed that this carries over to humans.
The specialist authors themselves call this state of affairs a paradox: blocking and activating the same receptor both have favorable effects, and the reasons are unknown. Anyone who sells the blockade as the superior principle goes beyond the data.
How strong the effect is
The phase 2 trial had 11 groups in 2 cohorts. In the obesity cohort with 465 participants, on average 47.9 years old and with a BMI of 37.9, weight loss after 52 weeks was between 12.3 and 16.2 percent, versus 2.5 percent on placebo. The individual groups reached 13.6, 15.5, 14.6, 12.3, 14.1 and 16.2 percent.
More revealing than the averages are the proportions: 80.4 to 85.7 percent of those treated lost at least 5 percent of their weight, versus 26.3 percent on placebo. 56.9 to 77.9 percent reached at least 10 percent, versus 11.8 percent on placebo. 45.5 to 62.3 percent managed at least 15 percent, versus 2.6 percent on placebo. And 23.4 to 38.5 percent of those treated reached at least 20 percent, while no one on placebo did.
In the second cohort with type 2 diabetes, 127 participants, weight loss was smaller, as is usual for this drug class: 8.4 to 12.3 percent versus 1.7 percent. The long-term blood sugar marker HbA1c fell by 1.2 to 1.6 percentage points, while on placebo it rose by 0.1 percentage points.
What is well supported
The primary endpoint of the phase 2 trial was met: the percentage change in weight after 52 weeks, in a double-blind, randomized, placebo-controlled design with 592 participants. For a drug at this stage of development, that is a solid basis, even if so far it rests on this one randomized trial.
Also supported is the effect on blood sugar in type 2 diabetes and, importantly for practice, that gastrointestinal complaints can be reduced by slow dose escalation and a lower starting dose. No unexpected safety signals appeared in the trial.
The concept itself also rests on more than an idea. The monthly dosing follows from the coupling to an antibody. And the GIP blockade is grounded in animal models, human genetics and a phase 1 trial, not merely asserted.
What the studies show
Phase 2 trial (NEJM 2025, 592 participants)
Double-blind, randomized, placebo-controlled, dose-finding with 11 groups in 2 cohorts over 52 weeks, funded by Amgen. The primary endpoint was the percentage change in weight after 52 weeks and was met: 12.3 to 16.2 percent loss versus 2.5 percent on placebo in obesity without diabetes. Dosing every 8 weeks and two dose escalation schemes were also tested.
Cohort with type 2 diabetes (same trial)
127 participants, on average 55.1 years old, BMI 36.5. Weight loss after 52 weeks 8.4 to 12.3 percent versus 1.7 percent on placebo. HbA1c fell by 1.2 to 1.6 percentage points, while on placebo it rose by 0.1 percentage points.
Body composition (review table 2026)
From a review that summarizes the trial data in tables: total fat mass fell by 26.2 to 36.8 percent versus 9.1 percent on placebo. Fat-free mass, measured by DXA, fell by 8.6 to 11.6 percent versus 2.1 percent on placebo. Fat-free mass includes muscle, but also water and organs. The loss is therefore substantial and part of the picture for this drug class.
Rationale for GIP blockade (Diabetes 2025)
A review on the question of why blocking the GIP receptor should help with weight loss. Mice without this receptor are protected from obesity on a high-fat diet, people with less active receptor variants have a lower BMI on average, and in mice and monkeys the blockade enhances the effect of GLP-1. The authors state that blockade and activation both have favorable effects and that the reasons are unknown.
Where the data stop
The number circulating online is not in the publication. There, for the estimator that includes all participants regardless of treatment discontinuation, 12.3 to 16.2 percent without diabetes and 8.4 to 12.3 percent with diabetes are reported. Values around 20 percent, as often quoted, do not come from this publication.
Two things are missing entirely. First, phase 3: no results are available yet, including from the large trial on cardiovascular events with 12,800 planned participants. Second, any direct comparison with semaglutide or tirzepatide. A trial on switching from a GLP-1 drug to MariTide has only just started.
Then there is the loss of fat-free mass: 8.6 to 11.6 percent, measured by DXA, versus 2.1 percent on placebo. What that means over years, nobody knows. We give no figure for discontinuation rates because of side effects, because the abstract does not report one and the full text was not accessible. And the mechanism remains a puzzle: that blocking and activating the same receptor both help is an open problem in the specialist literature, not a solved one. Incidentally, what is offered on the gray market as MariTide cannot be the original. A peptide-antibody conjugate cannot be replicated like a short peptide.
Status, approval and legal
MariTide is an investigational drug from Amgen and is not approved anywhere. In Germany, it is not a marketable medicine and is not available outside trials; we therefore give no dosage information. The phase 3 program is broad: among others, a trial on cardiovascular events with 12,800 planned participants, one in heart failure with preserved or mildly reduced ejection fraction with 5,056 participants, one in obesity without diabetes with 3,853 participants, one in type 2 diabetes with obesity with 1,105 participants, two trials in sleep apnea and two extension trials. In sport: the catch-all class S0 of the 2026 Prohibited List bans drugs in clinical development at all times. By this wording, MariTide falls under it.
Safety
The side effect picture is that of the class: gastrointestinal complaints were common in the phase 2 trial but occurred less often when the dose was escalated slowly and started at a lower level. There were no unexpected safety signals. The abstract does not give concrete discontinuation rates, and the full text was not retrievable, so no figure is given here. What must be stated clearly is the loss of fat-free mass, which includes muscle: 8.6 to 11.6 percent, measured by DXA, versus 2.1 percent on placebo. With a fat mass loss of 26.2 to 36.8 percent, that is part of the overall picture, but not a side issue. The longest published treatment duration is 52 weeks. According to the manufacturer (Amgen, January 2026), the large majority maintained their weight loss in the second trial year on a lower monthly or a quarterly dose; that is a manufacturer statement, and no peer-reviewed publication on it was available at the time of writing. Long-term data are lacking, and data on cardiovascular events are still being collected. There are no upper limits from EFSA, BfR or NIH, because this is not a nutrient. Pregnancy, breastfeeding and children and adolescents have not been studied. Outside a trial, MariTide is not available to anyone, and what is traded under this name is not the original molecule.
BK-Score Supported, with caveats
| Human evidence | 6 | |
|---|---|---|
| Mechanism | 8 | |
| Safety data | 5 | |
| Hype gap | 4 | |
| Track record of use | 3 |
The phase II trial (NEJM 2025, NCT05669599) included 592 participants in 11 groups over 52 weeks and met its primary endpoint: 12.3 to 16.2 percent weight reduction without diabetes versus 2.5 percent on placebo, with diabetes 8.4 to 12.3 percent versus 1.7 percent, HbA1c minus 1.2 to minus 1.6 percentage points. The often-quoted values around 20 percent do not appear in this form in the publication – 23.4 to 38.5 percent of those treated reached at least 20 percent weight loss, no one on placebo did. What is special is the monthly dosing, made possible by coupling peptide and antibody. Little advertised is the loss of fat-free mass: 8.6 to 11.6 percent, measured by DXA, versus 2.1 percent on placebo. Phase III is broad, including a cardiovascular outcome trial with 12,800 planned participants; published data beyond one year and any direct comparison with semaglutide or tirzepatide are lacking, and for the second trial year there are so far only manufacturer statements. The convenience argument is being promoted before the confirmatory trials are in.
The score rates the state of knowledge, not the substance. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)
Frequently asked questions about MariTide (Maridebart Cafraglutide)
How much weight do people lose with MariTide?
In the phase 2 trial over 52 weeks, it was 12.3 to 16.2 percent of body weight in obesity without diabetes, depending on the group, versus 2.5 percent on placebo. With type 2 diabetes, 8.4 to 12.3 percent versus 1.7 percent.
Is the 20 percent you read everywhere correct?
Not as a result of the publication. It reports 12.3 to 16.2 percent. However, 23.4 to 38.5 percent of participants in the drug groups achieved a loss of at least 20 percent, and no one on placebo did. That is something different from an average.
Is MariTide really injected only once a month?
Yes, every 4 weeks in the phase 2 trial, and in one group even every 8 weeks. This is made possible by coupling the peptides to an antibody, which keeps the drug in the body for a long time.
Why does MariTide block the GIP receptor while tirzepatide activates it?
That is the open paradox of this research. Both routes lead to weight loss, and the reasons are unknown. The blockade is supported by mice without the GIP receptor, people with less active receptor variants and a lower BMI, and animal experiments.
Do you lose muscle with MariTide?
In the trial data, fat-free mass, which includes muscle, decreased. Measured by DXA, it fell by 8.6 to 11.6 percent, versus 2.1 percent on placebo, with a fat mass loss of 26.2 to 36.8 percent. Strength training and enough protein are therefore not an afterthought with this drug class.
Is MariTide approved yet?
No. It is an investigational drug, and phase 3 is under way, including a trial on cardiovascular events with 12,800 planned participants. There is no approval anywhere, and in sport the drug is banned as a medicine in development.
Related
- Related topicMazdutide
- Related topicAmycretin
- Related topicOrforglipron (Foundayo)
- Related topicTirzepatide (Mounjaro / Zepbound)
Sources
- Jastreboff et al., N Engl J Med 2025 (phase 2 trial)
- Bassatne and Rizo, Adv Ther 2026 (review with trial table)
- Rosenkilde et al., Diabetes 2025 (rationale for GIP blockade)
- Yie et al., Antib Ther 2026 (peptide-antibody conjugates)
- ClinicalTrials.gov, NCT05669599 (phase 2 trial)
- ClinicalTrials.gov, NCT07037433 (cardiovascular outcome trial)
- Letters to the editor and authors' reply, N Engl J Med 2025
- WADA Prohibited List 2026, BGBl. III No. 219/2025
- Clinical Trials Arena, 2026-01-14 (Amgen statements on the second trial year, manufacturer statement)
Open in the database – with search, filters and comparison (German app)
Information only, not medical advice and no usage or dosing recommendation. Prescription-only and unapproved substances belong in medical hands. Last updated: 2026-10-04.