Maridebart Cafraglutide (MariTide)

Amgen's investigational antibody-peptide conjugate that pairs a GIP receptor-blocking monoclonal antibody with two GLP-1 receptor agonist peptides, now in phase 3 obesity, diabetes and cardiovascular trials.

Also referenced as: AMG 133, AMG-133, MariTide

Also appears in: Metabolic

What is Maridebart Cafraglutide (MariTide)?

Maridebart cafraglutide (MariTide, formerly AMG 133) is not a conventional peptide. It is an antibody-peptide conjugate: a fully human monoclonal antibody that blocks the glucose-dependent insulinotropic polypeptide receptor (GIPR), with two GLP-1 receptor agonist peptide analogs attached to it through amino acid linkers at an engineered cysteine in the antibody’s constant region. The antibody scaffold gives the attached GLP-1 peptides a circulating half-life of roughly two weeks for the intact conjugate (about three weeks for total antibody), rather than the hours-to-days of unconjugated GLP-1 analogs, which is why the molecule is presented as a monthly-or-less-frequent injectable. Developed by Amgen, it is in late-stage clinical trials for obesity and related conditions and is not approved anywhere. It does not have an established presence on the research-peptide market, which is unsurprising for a roughly 153 kDa antibody conjugate that cannot be produced by solid-phase peptide synthesis.

How it works

The molecule combines two opposing incretin-receptor actions in one construct:

• GLP-1 receptor agonism — the two conjugated peptides activate GLP-1R, the same receptor targeted by semaglutide and liraglutide, reducing appetite and food intake. The peptide carries an α-aminoisobutyric acid substitution at position 2 for protease resistance, and conjugation to the IgG antibody markedly prolongs systemic exposure of the structurally intact GLP-1 peptide (Wu et al., Journal of Medicinal Chemistry, 2026;69(8):9348–9362).
• GIP receptor antagonism — the antibody backbone binds and blocks GIPR. In diet-induced obese mice and obese cynomolgus monkeys, GIPR-antagonist antibodies reduced body weight on their own and produced additive weight loss when combined with GLP-1R agonists, the preclinical rationale for building the conjugate (Killion et al., Science Translational Medicine, 2018;10(472):eaat3392).
• Bispecific pharmacology in one molecule — cell-based assays confirmed simultaneous GIPR antagonism and GLP-1R agonism, and the construct lowered body weight and improved metabolic markers in obese mice and monkeys before entering human testing (Véniant et al., Nature Metabolism, 2024;6(2):290–303).
• The GIP paradox — tirzepatide reaches strong weight loss by activating GIPR, while maridebart cafraglutide blocks it; both approaches appear to work, and the mechanism reconciling them (receptor desensitization, central versus peripheral effects, or β-arrestin signaling) remains an open research question (Novikoff and Müller, Trends in Endocrinology & Metabolism, 2024;35(7):566–568).

Research status

Maridebart cafraglutide is in active phase 3 development. It has no FDA, EMA or other regulatory approval, and no marketing application has been publicly announced; Amgen’s quarterly disclosures through August 4, 2026 describe only ongoing and newly started trials.

Phase 1 (NCT04478708). A randomized, double-blind, placebo-controlled single- and multiple-ascending-dose study in 110 participants with obesity found an acceptable safety profile with pronounced, dose-dependent weight loss, reaching −14.5% by day 85 in the highest multiple-dose group; weight remained reduced up to 150 days after the last dose in that group, consistent with the molecule’s long half-life (Véniant et al., Nature Metabolism, 2024;6(2):290–303).

Phase 2 (NCT05669599). Jastreboff et al. (New England Journal of Medicine, 2025;393(9):843–857; published June 23, 2025) randomized 592 adults across an obesity cohort (n=465) and an obesity-with-type-2-diabetes cohort (n=127) to placebo or one of several maridebart cafraglutide regimens for 52 weeks. On the treatment-policy (intention-to-treat) estimand, mean body-weight change at week 52 ranged from −12.3% to −16.2% in the obesity cohort versus −2.5% with placebo, and from −8.4% to −12.3% in the diabetes cohort versus −1.7% with placebo; glycated hemoglobin fell 1.2 to 1.6 percentage points in the diabetes cohort versus a 0.1-point rise with placebo. The roughly 20% figure often quoted for this trial (up to −19.9% in the obesity cohort and −17.0% in the diabetes cohort) is the efficacy estimand, which counts only participants who stayed on treatment, not the intention-to-treat result. Gastrointestinal adverse events were common and were less frequent in arms that used dose escalation and a lower starting dose; early discontinuation ranged from roughly 8% to 27% across arms depending on the escalation strategy, with the highest rates where treatment began at full strength. No unexpected safety signals were reported. A second, exploratory 52-week part of the trial re-randomized participants who had lost at least 15% of body weight to lower monthly or quarterly maintenance regimens; Amgen reported in January 2026 that the large majority maintained their weight loss with a low incidence of nausea and vomiting and no new safety signals, but those data had not appeared in a peer-reviewed journal by November 2026. A separate phase 2 study in type 2 diabetes (NCT06660173) reached its 24-week timepoint in the same period, with Amgen reporting reductions in both glycated hemoglobin and weight.

Phase 3 (MARITIME program). MARITIME-1 (NCT06858839; 3,853 adults with obesity or overweight without type 2 diabetes, started March 12, 2025) and MARITIME-2 (NCT06858878; 1,105 adults with type 2 diabetes, started March 17, 2025) completed enrollment in 2025 and list estimated primary completion dates of January 21, 2027. Companion phase 3 studies include MARITIME-CV (NCT07037433, cardiovascular outcomes in an estimated 12,800 participants with atherosclerotic disease), MARITIME-HF (NCT07037459, about 5,000 participants with heart failure with preserved or mildly reduced ejection fraction), MARITIME-OSA-1 and -2 (obstructive sleep apnea with and without positive airway pressure therapy), MARITIME-SWITCH (transition from weekly semaglutide or tirzepatide), extension studies of MARITIME-1 and -2, MARITIME-3-J in Japan, and MARITIME-PEN (autoinjector self-administration). Additional phase 3 trials in type 2 diabetes, including MARITIME-T2D-3 (NCT07866066), a head-to-head comparison with semaglutide, were registered in 2026. No phase 3 efficacy results had been reported by November 2026.

Common dosage forms

Maridebart cafraglutide exists only as clinical-trial material: a sterile solution for subcutaneous injection supplied to investigators. The phase 3 program is evaluating a prefilled autoinjector pen for at-home self-administration, and phase 1 studies have compared two subcutaneous presentations. No commercial presentation, strength or device has been approved by any regulator, and the compound is not legitimately available outside clinical trials.

Key considerations

• Not a peptide in the usual sense. The active molecule is a monoclonal antibody carrying two peptide payloads. It behaves pharmacokinetically like a biologic antibody, cannot be chemically synthesized like semaglutide or tirzepatide, and would be expected to be regulated as a biologic rather than a small-molecule or synthetic-peptide drug.
• Regulatory status. The Investigational status reflects an active phase 3 program with no US approval and no publicly announced marketing application as of November 2026. The pivotal MARITIME-1 and MARITIME-2 records list primary completion in January 2027.
• Gastrointestinal tolerability is the main safety theme. In phase 2, nausea and vomiting were frequent at treatment start, early discontinuation reached roughly 27% in the worst-tolerated arms, and gallbladder events occurred in 1% to 6% of treated participants in the obesity cohort versus none on placebo. Because the molecule persists for months after the last injection, adverse effects cannot be quickly reversed by stopping treatment. Whether the escalation-based phase 3 regimens reproduce phase 2 efficacy with better tolerability is the central open question.
• Mechanistic uncertainty. Blocking GIPR achieves weight loss comparable to agonizing it, and the biology behind that paradox is unresolved; the long-term consequences of chronic GIPR antagonism in humans (for example on bone or adipose tissue) have not been characterized.
• Naming. “MariTide” is Amgen’s trade-style name for the investigational product, “AMG 133” is the original development code, and “maridebart cafraglutide” is the international nonproprietary name, in which the “-bart” stem marks the engineered monoclonal antibody component and “-glutide” the GLP-1 analog component.