MariTide (Maridebart Cafraglutide): What the Evidence Shows

MariTide (maridebart cafraglutide) is Amgen’s experimental obesity drug, an injection given once a month or less often. It switches on the GLP-1 receptor and blocks the GIP receptor, which is the opposite of what tirzepatide does to GIP. In its 52-week phase 2 trial, people with obesity lost 12.3% to 16.2% of their body weight depending on the dose, against 2.5% on placebo; people who also had type 2 diabetes lost 8.4% to 12.3%.^[1] Nausea and vomiting were common, mostly around the first doses, and slower dose increases reduced them. Phase 3, called MARITIME, is under way, with the first main weight-loss trials expected to finish in early 2027.^[8] MariTide is not approved anywhere, Amgen has given no release date, and it cannot legally be bought.

What MariTide is

MariTide is built differently from today’s weight-loss shots. Semaglutide and tirzepatide are modified peptides. MariTide is a fully human antibody that blocks the GIP receptor, with two GLP-1-like peptides attached to it by short amino-acid linkers.^[4] Amgen calls it an antibody-peptide conjugate. Its code name during early development was AMG 133, and many older papers use that name.

The antibody part is what makes monthly dosing possible. Antibodies stay in the body for weeks, so MariTide lasts far longer than a weekly peptide. Amgen says its design “supports starting with monthly dosing, and staying on MariTide with as few as 4 or 6 doses per year.”^[6] Amgen calls it “an investigational medicine,” meaning it is still being tested and has not been approved.^[7]

Why blocking GIP is the interesting part

GIP and GLP-1 are both gut hormones released after eating. Tirzepatide, sold as Zepbound and Mounjaro, activates both receptors: its label says “Tirzepatide is a GIP receptor and GLP-1 receptor agonist.”^[11] MariTide does the reverse on GIP. It activates GLP-1 but blocks GIP.^[1]

Two drugs pushing the same receptor in opposite directions, and both causing weight loss, is a real puzzle in obesity research. A 2026 review describes it as “a fundamental paradox: both pharmacological activation and blockade of GIP receptor (GIPR) signaling reduce body weight and enhance the efficacy of GLP-1-based therapies.”^[12] The leading ideas are that switching GIP on may calm nausea and act on brain appetite circuits, while blocking it may strengthen GLP-1 signaling and limit fat storage. Neither explanation is settled. Amgen says its choice grew out of human genetics work at its deCODE unit in Iceland, which pointed toward blocking the GIP receptor.^[7] Our explainer on what GIP does on its own covers the hormone in more detail.

How much weight people lost in the trials

Phase 1: the first signal

The first human study was small and short. At the highest repeated dose, 420 mg every four weeks for three doses, average weight fell 14.5% by day 85, while the placebo group gained 1.5%. Weight stayed down by as much as 11.2% some 150 days after the last dose,^[4] which is an early hint that the drug lingers. A phase 1 study cannot show how well a drug works in practice; it mainly tells researchers which doses are worth testing next.

Phase 2: the main evidence so far

The phase 2 trial, which the New England Journal of Medicine carried in 2025, is the best evidence to date. It enrolled 592 adults in two groups: 465 with obesity but no diabetes, and 127 with obesity and type 2 diabetes. Participants received 140, 280 or 420 mg every four weeks, 420 mg every eight weeks, or placebo, for 52 weeks. Some arms started at full dose; others built up to it.^[1]

You will often see a bigger figure quoted: “up to about 20%.” That comes from Amgen, and it uses a different analysis. The efficacy estimand estimates what would have happened if everyone had stayed on the drug for the full year. By that measure, Amgen reports 16.3% to 19.9% weight loss without diabetes (2.6% on placebo) and 12.1% to 17.0% with diabetes (1.4% on placebo).^[2] Both numbers are real; they answer different questions. The lower one, which counts people who dropped out, is closer to what a typical patient group might see. Amgen also says weight loss “had not plateaued by 52 weeks,” meaning the curve was still heading down when the trial stopped.^[2]

People with type 2 diabetes usually lose less weight on GLP-1 drugs than people without it, and MariTide followed that pattern. Their HbA1c, a measure of average blood sugar, fell by 1.2 to 1.6 percentage points in the published analysis.^[1] Amgen’s own releases, using a different analysis, put the best result at up to 2.2 points.^[3]

The second year

People who lost at least 15% of their weight and were still on treatment at week 52 could join a second year, in which they were re-assigned to lower monthly doses, a dose every 12 weeks, or placebo.^[2] More than 90% of eligible participants signed up.^[3] Amgen said in February 2026 that “the large majority of participants maintained the weight loss achieved in Part 1 for an additional 52 weeks on a lower monthly dose or quarterly dose,” with “a very low incidence of nausea and vomiting.”^[5] That is encouraging, but it is a company summary without published numbers, and it covers only people who had already done well in year one.

Side effects: what the trials found

The main problem was the stomach. The published paper says gastrointestinal side effects “were common with maridebart cafraglutide, although less frequent with dose escalation and a lower starting dose,” and that “no unexpected safety signals emerged.”^[1] Amgen’s early summary named nausea, vomiting and constipation as the most common side effects. It said nausea and vomiting were mostly mild, short-lived and “primarily associated with the first dose.” In the arms that built up gradually, nausea typically cleared within about six days and vomiting within one to two days.^[3]

Starting at a high dose was the issue. Some phase 2 arms began at the full dose with no build-up, and that is where stomach symptoms were worst. Amgen reports that in the arms that did build up gradually, about 11% of people stopped because of any side effect, and up to 7.8% because of stomach side effects, fewer than in the arms without a build-up.^[2][3] These dropout figures are company-reported.

Amgen then ran a separate study of gentler starting doses. Of 121 people, those who began at 21 mg had a 24.4% rate of vomiting over the first six weeks, and those who began at 35 mg had 22.5%, with nobody stopping because of stomach symptoms.^[2] That is still roughly one person in four or five vomiting at least once. It shaped phase 3: everyone there starts at 21 mg, then 35 mg, then 70 mg, over an eight-week build-up before moving to one of three target doses.^[2] Whether that schedule fixes the problem at scale is one of the things phase 3 will show.

Amgen also reported no link between MariTide and changes in bone mineral density in phase 2.^[3] Long-term safety, including rare events, can only come from the much larger phase 3 trials.

The MARITIME phase 3 program

MARITIME is Amgen’s phase 3 program. It is broad: weight loss with and without diabetes, heart outcomes, heart failure, sleep apnea, switching from weekly drugs, and long-term follow-on studies.^[6] The two trials most likely to support a first approval for weight loss are MARITIME-1 and MARITIME-2, each measuring weight change at 72 weeks.^[8][9]

Amgen also says three phase 3 trials in people with type 2 diabetes will start in 2026, and a phase 2b study is testing MariTide on liver fat.^[6] The full list is on ClinicalTrials.gov, which showed 26 registered studies of the drug in October 2026.^[10] Note that the extension and switch trials test dosing every eight or twelve weeks. Whether people can keep their weight off on four to six shots a year is central to how Amgen hopes to set MariTide apart.

When MariTide might be available

Amgen has not announced a filing date, an approval target, a launch date or a price. Its most recent updates describe the MARITIME trials as ongoing or enrolling.^[6] The registry lists January 2027 as the estimated completion date for the main results of MARITIME-1 and MARITIME-2.^[8][9] After results come analysis, an application to FDA and a review, and Amgen has not said how quickly it will move through those steps. Any site naming a launch year is guessing.

For comparison with other drugs in the same queue, see retatrutide’s evidence, CagriSema’s phase 3 results and amycretin’s early trial. Trial lengths, doses and populations differ, so the headline percentages do not line up neatly. Our GLP-1 pipeline tracker keeps all of them on one page.

Can you buy MariTide now?

No. MariTide has no FDA approval and no FDA label, and Amgen describes it as investigational.^[7] The only way to receive it is inside one of Amgen’s clinical trials, and ClinicalTrials.gov lists which are still recruiting.^[10] Any product sold online as MariTide or maridebart cafraglutide cannot be verified as the real drug.

FDA’s warnings so far have named other unapproved drugs, such as retatrutide and survodutide, sold with “for research purposes” or “not for human consumption” labels. It urges people not to buy such products, which it calls “of unknown quality.”^[13] The same reasoning applies to anything sold as MariTide. A research-only label does not make a sale legal, as our guide on whether peptides are legal explains.

If you want treatment now, the approved drug that also acts on both GLP-1 and GIP is tirzepatide (Zepbound), a weekly shot that activates GIP rather than blocking it.^[11]