Metabolic · 9 min read · Published September 8, 2026

MariTide (Maridebart Cafraglutide): The Once-Monthly GLP-1 Agonist and GIP Blocker, Explained

How Amgen's antibody-peptide conjugate works, why it blocks GIP when tirzepatide activates it, what the 592-person Phase 2 in NEJM showed at 52 weeks, what the second-year maintenance data mean, the MARITIME Phase 3 program, the gut-tolerability problem and the dose-escalation fix, and where Pfizer's berobenatide fits as the other monthly.

What is MariTide and is it approved?

MariTide (maridebart cafraglutide, formerly AMG 133) is Amgen's investigational obesity drug: a monoclonal antibody that blocks the GIP receptor, conjugated to two GLP-1 receptor agonist peptides. It is dosed monthly by injection. As of September 2026 it is in Phase 3 (MARITIME) and is not approved by the FDA or any other regulator.

Every other injectable in this class is a peptide dosed weekly. MariTide is a different kind of molecule: an antibody-peptide conjugate, where the antibody supplies both the long half-life and one of the two pharmacological actions. Amgen's Phase 1 (Nature Metabolism, 2024) reported dose-dependent weight loss with weight still below baseline up to 150 days after the last dose, which is what made monthly dosing plausible (Nature Metabolism, 2024). Profile: MariTide. Side-by-side data with the approved dual agonist: MariTide versus tirzepatide.

How can blocking GIP and activating GIP both cause weight loss?

This is the GIP paradox. Tirzepatide activates the GIP receptor alongside GLP-1 and reached 20.9 percent weight loss at 72 weeks in SURMOUNT-1 (n=2,539). MariTide antagonizes the GIP receptor alongside GLP-1 and reached up to 19.9 percent at 52 weeks in Phase 2. Both directions work in humans; one explanation is that chronic GIP agonism desensitizes the receptor, mimicking blockade.

GIP is the incretin that fat cells respond to most, and animal data pointed both ways for years: blocking GIP signaling reduced weight in Amgen's preclinical models, yet GIP agonists also reduce weight when combined with GLP-1. Amgen bet on antagonism, building a fully human antibody against the GIP receptor and attaching GLP-1 analogues to it so both signals arrive at once. Tirzepatide bet on agonism (NEJM, SURMOUNT-1, 2022). Both bets paid out. That means the GIP receptor's role in humans is still not understood at the level of mechanism, and it is one reason to be cautious about anyone selling a confident story about what 'GIP' does. Evidence tier for both: human RCT.

What did the Phase 2 trial show?

In the 52-week Phase 2 (NEJM, June 2025; 465 adults with obesity, 127 with obesity and diabetes), monthly MariTide produced 12.3 to 16.2 percent weight loss without diabetes and 8.4 to 12.3 percent with diabetes on the treatment-policy estimand, versus 2.5 and 1.7 percent for placebo. On-treatment figures reached 19.9 and 17.0 percent, with no plateau.

The two numbers matter differently. The treatment-policy figure counts everyone randomized, including people who stopped; the on-treatment figure is what those who stayed achieved. Amgen led with the ~20 percent figure in November 2024 and the journal led with the lower one in June 2025; both are correct readings of the same trial (NEJM, 2025). The gap between them is mostly the discontinuation story below. On the comparison table, MariTide's 52-week number sits close to tirzepatide's 72-week number, at half the injections.

What did the second-year maintenance data show?

Participants who had lost at least 15 percent at week 52 (more than 90 percent of those eligible joined) were re-randomized to placebo, 70, 140 or 420 mg monthly, or 420 mg every 12 weeks. On January 14, 2026 Amgen said most kept their weight off for a further 52 weeks on lower monthly or quarterly dosing.

Part 2 is the part of the program built for the question payers ask: can a monthly drug become a quarterly drug once the weight is off? Amgen's answer at the J.P. Morgan conference was yes in principle, with efficacy 'on a monthly dosing schedule' and 'potential when administered quarterly' (Clinical Trials Arena, January 2026). The full Part 2 numbers, including how much weight the placebo re-randomized group regained, had not been published in a journal as of September 2026, so the claim rests on a company presentation. Evidence tier: human RCT, company-reported.

What is the MARITIME Phase 3 program?

As of Amgen's first-quarter 2026 report, MARITIME-1 (obesity without diabetes) and MARITIME-2 (with diabetes) are enrolling, alongside MARITIME-CV (cardiovascular outcomes), MARITIME-HF (heart failure), MARITIME-OSA-1 and OSA-2 (sleep apnea) and MARITIME-SWITCH, which moves people from weekly tirzepatide or semaglutide to every-eight-week or quarterly MariTide. Extension studies test monthly, eight-week and quarterly maintenance.

• MARITIME-1 and MARITIME-2: the pivotal chronic weight management trials, with EXTENSION arms that re-randomize completers to monthly, every-eight-week or quarterly dosing to measure maintenance.
• MARITIME-CV and MARITIME-HF: outcomes trials in atherosclerotic disease and heart failure with obesity, the studies that would support cardiovascular labeling.
• MARITIME-SWITCH: the first Phase 3 built around switching from an approved weekly drug to a less frequent one, a direct test of the convenience thesis.
• Three additional type 2 diabetes Phase 3 trials planned to start in 2026, plus a Phase 2b in liver fat.

No Phase 3 results had been reported as of September 2026 (Amgen first quarter 2026 results). The program's size, six global Phase 3 studies running at once, is itself a signal of how much Amgen expects from the monthly schedule.

Why did so many people stop, and what changed for Phase 3?

In Phase 2, about 11 percent of participants in dose-escalation arms stopped for adverse events, under 8 percent for gastrointestinal ones; arms starting at full dose did worse. Nausea resolved in a median of 6 days, vomiting in 1 to 2 days. Phase 3 escalates from 21 mg to 35 mg to 70 mg over eight weeks.

The Phase 2 design deliberately included arms that went straight to 280 or 420 mg monthly with no ramp, and those arms had the worst tolerability. The lesson was the same one every GLP-1 program has learned: the first exposure drives the nausea, and a slower climb helps. Amgen's June 2025 update set the Phase 3 escalation at 21, 35 and 70 mg 'over a further optimized eight-week dose escalation period' (Amgen, June 2025). Note that 70 mg is far below the 140 to 420 mg used in Phase 2; whether the lower Phase 3 target dose gives up efficacy is one of the questions MARITIME-1 will answer. A monthly drug also cannot be 'paused' the way a weekly one can: once injected, the antibody stays for weeks, which is a tolerability consideration unique to this class.

Where does berobenatide, the other monthly, fit?

Berobenatide (formerly MET-097i) is Pfizer's ultra-long-acting GLP-1 peptide from its Metsera acquisition. It is not an antibody and has no GIP activity. In VESPER-3 (64-week Phase 2b, about 54 per arm), monthly maintenance after weekly titration produced 10 and 12.3 percent placebo-adjusted weight loss at 28 weeks (February 3, 2026), and Pfizer plans ten Phase 3 trials in 2026.

Berobenatide's data are younger and shorter: 28 weeks of monthly maintenance versus MariTide's 52 weeks of monthly treatment and a second year of maintenance (Pfizer, February 2026; Pfizer, June 2026). Pfizer's pitch is tolerability, with 'low gastrointestinal adverse events and discontinuations despite rapid dose escalation'. Whether a pure GLP-1 can reach MariTide's or tirzepatide's ceiling is the open question.

The bottom line

MariTide is the first drug to show that blocking GIP, rather than activating it, works for weight loss in a large trial, and the first to show a year of monthly dosing with no plateau. The published evidence is a single Phase 2 in 592 people; the maintenance data and the gentler Phase 3 dosing are company-reported so far. MARITIME-1 had not reported as of September 2026, and nothing sold as MariTide or 'AMG 133' today is the Amgen molecule. Related: the 2026 GLP-1 field guide and oral GLP-1 pills.