Maridebart Cafraglutide (MariTide): What It Is, Phase 2 Results and Trial Status

As of October 5, 2026, maridebart cafraglutide (MariTide, code name AMG 133) is an investigational obesity drug from Amgen: an antibody that blocks the receptor for the hormone GIP, linked to two peptides that activate the GLP-1 receptor, given as an injection once a month or less often^[2,10]. It is not FDA approved or approved anywhere else: it does not appear among the FDA’s novel drug approvals for 2026 or in Drugs@FDA, and as of September 4 it was not on the European Medicines Agency’s list of applications under review^[26,27,28]. Its strongest published data come from phase 2: in people with obesity and without diabetes, body weight fell by an average of 12.3% to 16.2% over 52 weeks, depending on dose and schedule, versus 2.5% on placebo^[1].

This is general information, not treatment guidance. If you are considering this or any other medication, talk to your doctor.

What maridebart cafraglutide is

Most drugs in this family are peptides, short chains of amino acids. Maridebart cafraglutide is a fully human monoclonal antibody against the GIP receptor, with two copies of a GLP-1-like peptide attached to it^[2]. It does two things at once: it activates the GLP-1 receptor, like semaglutide (Ozempic, Wegovy), and it blocks the GIP receptor, the receptor for another gut hormone in the same family, the incretins^[1,2].

Attaching the peptides to an antibody keeps them in the bloodstream much longer, according to the Amgen team that designed it, and that is what could allow monthly dosing^[3]. Amgen says the design supports starting with a monthly injection and staying on treatment with as few as 4 or 6 doses a year^[11]. That second part is what its extension trials are testing; it has not been shown yet^[22,23]. How each molecule fits into the family is covered in what GLP-1 medications are, and the names in GLP-1 generic and brand names.

If it blocks GIP, why does it resemble tirzepatide?

Tirzepatide activates the GIP receptor; maridebart cafraglutide blocks it. Yet both, combined with GLP-1 action, are linked to more weight loss than GLP-1 alone. Scientific reviews openly call this a paradox^[4,5,6]. A 2025 review sums up three hypotheses^[4]:

1. GIP may matter little, and tirzepatide’s edge could come from the way it activates the GLP-1 receptor. The authors think this does not explain everything.
2. Constant activation ends up as blocking. A receptor that is stimulated nonstop becomes desensitized and stops responding. This has been seen in fat cells at high doses, but the authors doubt it explains what happens in the body.
3. They work through different pathways in the brain. In rodents, activating the GIP receptor reduces food intake through one type of neuron, while blocking it seems to release a brake on GLP-1 action in others. That is why the blocker alone causes little weight loss in these models.

A 2026 review adds that blocking GIP may boost GLP-1 signaling and counter the body’s own GIP, which promotes fat storage^[6]. A third review points out that in humans there is no evidence that activating or blocking GIP on its own reduces appetite^[5]. Almost everything known about the mechanism comes from animal studies.

Phase 2 results

The phase 2 trial, funded by Amgen, enrolled 592 adults for 52 weeks in two cohorts^[1,12]:

• Obesity without diabetes (465 people), mean BMI 37.9. Six schedules were tested against placebo: 140, 280 or 420 mg every 4 weeks, 420 mg every 8 weeks, and 420 mg every 4 weeks after a 4- or 12-week dose escalation^[1].
• Obesity with type 2 diabetes (127 people), mean BMI 36.5: 140, 280 or 420 mg every 4 weeks, or placebo^[1].

One detail changes the numbers a lot. The paper’s primary analysis uses the treatment-policy estimand, which counts every participant whether or not they stayed on the drug. Amgen’s releases highlight the efficacy estimand: what would have happened if everyone had stayed on treatment^[1,8]. I give both.

Mean change in body weight at 52 weeks. Ranges run from the least to the most effective schedule^[1,8].

The published abstract gives the range across groups, not the figure for each schedule, so I cannot say here which was which. In the diabetes cohort, A1C (average blood sugar over the past few months) fell by 1.2 to 1.6 percentage points, versus a rise of 0.1 on placebo^[1]. According to Amgen, weight loss had not leveled off by week 52^[7]. As with every drug in this class, people with diabetes lose less weight.

The 2026 data exist only as press releases. In February 2026, Amgen said the second part of this trial had finished: an exploratory extra year of treatment, only for people who had lost at least 15% in part 1. According to the company, “the large majority” kept the weight off on a lower monthly dose or a quarterly dose^[9]. In the same release, it said a separate phase 2 trial in 409 people with type 2 diabetes had lowered A1C and weight at 24 weeks^[9,24]. The release gives no figures for either, and as of October 5, 2026, I have not found these results in PubMed. Until they are published, they cannot be assessed.

Before that, phase 1, with 110 participants split into small groups^[13]: on 420 mg every 4 weeks, weight fell 14.5% by day 85, after three doses, and 150 days after the last dose it was still 11.2% below baseline^[2].

Side effects in the trials

With no FDA label, there is no official list of side effects, warnings or contraindications. This is what the paper and the press releases report:

• Digestive. Common, mainly nausea, vomiting and constipation, and less frequent when the dose was stepped up gradually or started lower. No unexpected safety signals emerged^[1,7]. According to Amgen, they clustered around the first dose, and in the dose-escalation groups nausea lasted a median of six days and vomiting one to two days^[7].
• Stopping treatment. According to Amgen, in the dose-escalation groups about 11% stopped because of any side effect, and up to 7.8% because of digestive ones, fewer than in the groups without escalation^[7,8]. The paper’s abstract does not give these figures.
• Bones. Amgen says there was no association between treatment and changes in bone mineral density^[7], another figure that is not in the abstract.

For digestive symptoms on approved GLP-1s: what to eat with nausea and constipation on GLP-1s.

Phase 3: the MARITIME program

Amgen calls its phase 3 program MARITIME^[7]. None of these trials has reported results. The phase 3 doses are not in the registry records, which only refer to low, medium and high doses^[14].

In addition, the MARITIME-1 and -2 extensions test keeping weight off with injections every 4 or 8 weeks (and every 12 in the MARITIME-1 extension), versus switching to placebo, with primary completion estimated for December 2027^[22,23]. An open-label trial in the United States and Puerto Rico moves about 300 people who already lost at least 10% on a weekly GLP-1 over to MariTide^[21], and a phase 2b trial measures liver fat^[25]. Amgen announced three phase 3 trials in type 2 diabetes for 2026^[11]; as of October 5, I have not found them on ClinicalTrials.gov.

Every trial in the table except MARITIME-3-J has U.S. sites^{[14,15,16,17,18,19]}. If you are curious about taking part, ClinicalTrials.gov lists the open sites and contacts, and your doctor can help you judge whether a trial fits you. For sleep apnea with an approved drug, see Zepbound for sleep apnea.

Is MariTide FDA approved? Regulatory status

No. It is not among the FDA’s novel drug approvals for 2026 (list current as of September 28) and a Drugs@FDA search returns nothing for it^[26,27]. It is not approved in the European Union either: as of September 4, it was not on the European Medicines Agency’s list of applications under review^[28].

The FDA does not publish the applications it is reviewing. But the two key phase 3 trials do not reach primary completion until 2027, according to their records, and Amgen described them as “ongoing” in August 2026^[11,14,15].

Can you buy it? Not legally: outside its clinical trials, there is no authorized way to get it. The FDA warns that companies have illegally sold unapproved GLP-1 drugs falsely labeled “for research purposes” or “not for human consumption,” with dosing instructions, and that these products are of unknown quality and may be harmful^[29]. That FDA page names semaglutide, tirzepatide, retatrutide, survodutide and mazdutide, not maridebart cafraglutide, and I have not found an FDA warning letter that names it. A product sold online under the MariTide name is not an approved drug.

MariTide vs. semaglutide, tirzepatide and other drugs in development

Within this table, it is the only one that blocks GIP, the only antibody and the only one tested with monthly or less frequent doses^[1,2]. Whether any of that is an advantage for a given person has yet to be shown.

There is no head-to-head comparison. I have not found any trial on ClinicalTrials.gov that pits it against another drug; the switching trial gives MariTide to everyone^[21]. Putting its phase 2 numbers next to the others’ phase 3 results does not work either: it lasted 52 weeks, versus 68 to 80, with much smaller groups^[1,30,31,32]. More trial figures, charted, are in the trial results chart.

When could it be approved?

There is no date. MARITIME-1 and MARITIME-2 have January 2027 as their estimated primary completion^[14,15]. After that, Amgen would need to analyze and publish the results and submit an application, and the FDA would need to review it. Approval would not mean immediate availability or insurance coverage either.

What is still unknown

• Whether the phase 2 results hold up in phase 3, with thousands of participants over 72 weeks^[14,15].
• Whether maintenance with 4 or 6 doses a year works in controlled trials. So far there is only a press release about part 2 of phase 2, without figures^[9,22].
• Long-term and cardiovascular safety. MARITIME-CV and MARITIME-HF are not expected to reach primary completion before mid-2028^[16,17].
• How it compares with approved drugs. Without head-to-head trials, nobody knows.

Frequently asked questions

What is MariTide?

MariTide is the name Amgen uses for maridebart cafraglutide, an investigational injection for obesity. It combines an antibody that blocks the GIP receptor with two peptides that activate the GLP-1 receptor, and in the published part of phase 2 it was given every 4 or 8 weeks^[1,2]. It is not approved.

When will MariTide be FDA approved?

There is no date. Its two main phase 3 trials have January 2027 as their estimated primary completion, and Amgen has not said when it will file^[11,14,15]. As of October 5, 2026, it is not approved anywhere^[26,28].

Is MariTide better than Zepbound or Wegovy?

There is no way to know: no trial has compared them. MariTide’s numbers come from a 52-week phase 2 trial with small groups, while tirzepatide’s and semaglutide’s come from longer phase 3 trials^[1,30,31]. Zepbound and Wegovy are FDA approved; MariTide is not.

Can you buy MariTide online?

Not legally. It is not approved, and outside its clinical trials there is no authorized way to get it^[26,27]. The FDA warns against unapproved GLP-1 products sold online as “research” chemicals, which are of unknown quality^[29].