MariTide

Overview

MariTide is the development name for maridebart cafraglutide (AMG 133), an investigational, long-acting antibody-peptide conjugate from Amgen. It combines a human antibody that antagonizes the glucose-dependent insulinotropic polypeptide receptor (GIPR) with two modified peptides that agonize the glucagon-like peptide-1 receptor (GLP-1R).

Randomized phase 1 and phase 2 studies provide human evidence, and a broad phase 3 programme is under way. The product nevertheless remains investigational. As of 20 August 2026, bounded FDA and EU exact-name searches identified no approved indication, regulator-reviewed commercial product or approved dosing schedule.

Verified identity: an antibody-peptide conjugate

FDA GSRS identifies maridebart cafraglutide, UNII Z24U3U73HN and CAS 2760218-55-9, as a human IgG1 monoclonal-antibody conjugate. Its record describes an anti-GIPR antibody dimer connected through engineered thioether and amino-acid linkers to two modified GLP-1 analogue peptides.

This is not a simple free peptide with one short linear sequence. FDA's record contains two 450-residue heavy chains, two 214-residue light chains, disulfide links, protein modifications and the two attached peptide components. It does not publish one whole-product molecular formula or molecular weight. Those fields are therefore withheld rather than reconstructed or guessed. The GLP-1 component sequence reported in the discovery paper is not a sequence for the entire conjugate.

FDA also states that assignment of a UNII is a substance-identity function and does not imply regulatory review or approval.

Mechanism and the unresolved contribution

Cell and preclinical experiments (PMID 38316982) support two simultaneous activities: the antibody blocks GIPR, while the attached peptide components activate GLP-1R. This differs fundamentally from medicines that agonize both GIPR and GLP-1R. Calling MariTide a “dual incretin agonist” reverses half of its reported pharmacology.

The phase 1 human design did not include component-only arms and therefore could not determine how much of the observed weight change came from GIPR antagonism versus GLP-1R agonism. Preclinical synergy is a rationale for the construct, not human proof that each mechanism contributes a known amount or that either pathway guarantees a clinical outcome.

Pharmacokinetics and structural boundaries

The phase 1 paper reported mean half-life estimates of roughly 21 days, with values spanning 14–24 days across intact and total AMG 133 assays. Intact and total assays did not measure exactly the same molecular species, so the range should not be reduced to an absolute product constant.

The long exposure supported testing at intervals of four weeks, and later phase 2 protocols also examined an eight-week arm. This is investigational conjugate pharmacokinetics under defined study conditions—not evidence of an approved monthly or bimonthly regimen, bioequivalence outside the trial product or safe self-administration.

Current regulatory status

An exact-name search of FDA's public Drugs@FDA dataset returned no match for maridebart cafraglutide on 20 August 2026. A case-insensitive search of the current EU Union Register dataset returned 0 matches for maridebart, cafraglutide or MariTide. These are dated, bounded public-data checks, not a claim about every national database or a future regulator decision.

EMA decision EMA/PE/0000239753, dated 24 April 2025, grants a product- specific paediatric waiver for prevention of cardiovascular events. A waiver is a development decision, not a marketing authorisation. No Russian marketing authorisation was located in the accessible indexed regulator and medicine-register surfaces reviewed.

Phase 1 human evidence

NCT 04478708 was a randomized, double-blind, placebo-controlled single- and multiple-ascending-dose study; its current registry record lists 110 participants. In the published multiple-dose cohorts, three subcutaneous doses were tested at four-week intervals. Mean weight change was -7.4% at day 78 in the 140 mg cohort and -14.5% at day 85 in the 420 mg cohort, versus +1.5% with placebo at day 85.

In the 420 mg cohort, mean weight remained -11.2% from baseline 150 days after the last dose. These small early cohorts had no active comparator, and the paper notes that the human design could not separate the two mechanistic components. The results support further study; they do not establish a recommended amount, superiority to another medicine or durable long-term benefit.

Phase 2 design and research integrity

NCT 05669599 was a 52-week, double-blind, randomized, placebo-controlled, dose-ranging phase 2 trial with 592 participants. The obesity cohort included 465 participants without type 2 diabetes, while the obesity- diabetes cohort included 127. Protocol arms tested 140, 280 or 420 mg every four weeks, an eight-week arm and different escalation strategies.

The primary endpoint was percentage change in body weight at week 52. The primary publication foregrounds the treatment-policy estimand, an intention-to-treat approach that includes outcomes regardless of treatment discontinuation or rescue. Sponsor headlines based on an efficacy/trial-product estimand answer a different question and must not replace the primary treatment-policy result. Amgen funded the trial, several authors were Amgen employees, and the registry currently has no structured results for this study; the peer-reviewed paper remains the principal result source.

Phase 2 outcomes

Participants without type 2 diabetes

Under the treatment-policy estimand, mean weight change at week 52 ranged from -12.3% to -16.2% across maridebart cafraglutide groups, versus -2.5% with placebo. Confidence intervals and differences between protocol arms matter; the range is not a guaranteed individual response and does not validate every dosing interval equally.

Participants with type 2 diabetes

In the 127-participant obesity-diabetes cohort, treatment-policy mean weight change ranged from -8.4% to -12.3%, versus -1.7% with placebo. Mean HbA1c change ranged from -1.2 to -1.6 percentage points, versus +0.1 percentage points with placebo.

The trial did not directly compare MariTide with semaglutide, tirzepatide or another approved obesity medicine. Cross-trial ranking would mix populations, estimands, follow-up and handling of discontinuations and is therefore not a valid head-to-head conclusion.

Safety and tolerability boundaries

Gastrointestinal adverse events were common in phase 2 and were less frequent with a lower starting dose and dose escalation. The authors reported no unexpected safety signal in this 52-week trial, but that phrase does not mean that long-term or rare-event safety has been established.

Phase 1 also mainly identified gastrointestinal treatment-emergent events, and the programme measures anti-drug antibodies as well as intact and total conjugate exposure. Sample size, selected eligibility criteria and trial-product quality limit generalisation. Phase 3 cardiovascular, heart-failure and other outcome studies are still needed to determine longer-term benefit-risk.

Current phase 3 and registry programme

An exact ClinicalTrials.gov intervention query for maridebart cafraglutide, checked on 20 August 2026, returned 25 records: 10 completed, 7 active, not recruiting, 7 recruiting and 1 not yet recruiting. The same snapshot contained 10 phase 3 records; registry existence is not a positive result.

The main active programme includes MARITIME-1 (NCT 06858839, planned enrolment 3,853) without type 2 diabetes and MARITIME-2 (NCT 06858878, 1,105) with type 2 diabetes. Larger recruiting outcome studies include a cardiovascular trial (NCT 07037433, 12,800) and a heart-failure trial (NCT 07037459, 5,056). The register also contains obstructive-sleep- apnoea, switching and extension studies. No phase 3 outcome was posted at review; planned or active studies cannot support efficacy or approval claims.

Russian literature and research

A 2026 Russian Journal of Cardiology review by Mamedov, Druk and Akhundova correctly describes AMG 133 as an antibody-peptide conjugate combining an anti-GIPR antagonist antibody with two GLP-1 agonist peptides. It is useful Russian-origin terminology and regional context, but it summarizes the same international phase 1 evidence and is not an independent Russian clinical trial.

The 2025 Terapevticheskii Arkhiv review by Shestakova and Bashlykova contains an identity error, labelling “Maridabart/Cafraglutide” as cagrilintide, and places it among dual incretin agonists. Those labels conflict with FDA GSRS and the primary papers, so they were rejected for identity and mechanism. This is why adding Russian literature requires source criticism rather than automatic agreement.

No Russian-origin primary human MariTide trial or Russian marketing authorisation was located in the accessible indexed sources reviewed. Russian-language summaries or registry mirrors of international studies do not become Russian-origin research merely through translation.

WADA boundary and conclusion

An exact-name review of WADA's official 2026 Prohibited List found no named match for maridebart, cafraglutide or MariTide. The separate 2026 Monitoring Program names markers of semaglutide and tirzepatide, but does not name MariTide. Monitoring is also not the same as inclusion on the Prohibited List.

No S0 or other WADA class is inferred, and absence of a named match is not permission. Athletes must check the current list and their anti-doping body. The evidence-based conclusion is narrower: MariTide has substantive phase 2 human evidence and a major phase 3 programme, while approval, phase 3 outcomes, long-term benefit-risk and an authorised regimen remain unestablished.