Can a once-monthly weight-loss injection do the work of a CPAP machine? That is the question Amgen has now formally posed to regulators with a Phase 3 trial called MARITIME-OSA-2 (ClinicalTrials.gov identifier NCT07226765), which began recruiting in late 2025 and was last updated on June 12, 2026. The study tests maridebart cafraglutide, the molecule Amgen markets in development circles as MariTide, against placebo in adults who have obstructive sleep apnea, are living with overweight or obesity, and are not currently using positive airway pressure (PAP) therapy. It is a small but pointed trial, enrolling 250 participants, and its design tells you a great deal about how the obesity-drug field is maturing.
Here is what the trial actually measures. MARITIME-OSA-2 is a randomized, double-blind, placebo-controlled study with two arms: one group receives maridebart cafraglutide subcutaneously, the other receives a matching placebo subcutaneously. There is a single primary endpoint, and it is refreshingly concrete: the change from baseline in the apnea-hypopnea index (AHI) at week 52. AHI counts how many times per hour a person stops breathing (apneas) or breathes too shallowly (hypopneas) during sleep. A higher number means more severe disease. By choosing AHI at one year as its sole primary measure, Amgen is betting that the drug's effect on the underlying mechanical problem will be large enough and durable enough to register cleanly over a full 52 weeks.
Why obstructive sleep apnea is the right second act for an obesity drug
Obstructive sleep apnea is, for a large share of patients, a disease of excess soft tissue. Fat deposited around the upper airway and neck narrows the passage; during sleep, the throat muscles relax and the airway collapses, cutting off airflow until the brain jolts the body awake enough to reopen it. This happens dozens or hundreds of times a night, fragmenting sleep and straining the cardiovascular system. The standard treatment, a PAP machine that splints the airway open with pressurized air, works well when patients tolerate it, but adherence is notoriously poor. That is precisely the population Amgen has targeted: people not on PAP therapy, the patients for whom the current standard of care has failed or been refused.
The logic is that if you remove the excess weight pressing on the airway, you remove the cause rather than managing the symptom. This is not a wild hypothesis. The incretin class to which MariTide belongs has already established that meaningful weight loss can translate into measurable AHI reductions, and obstructive sleep apnea has become a recognized proving ground for the category. By running a dedicated, indication-specific Phase 3 trial rather than relying on sleep-apnea data buried in a general obesity study, Amgen is positioning maridebart cafraglutide for a label that physicians and payers treat as a distinct medical claim, not a lifestyle benefit.
What maridebart cafraglutide is, and why the molecule matters
Maridebart cafraglutide is not a simple GLP-1 agonist. It is engineered to act on two metabolic pathways at once, combining agonism at the GLP-1 receptor with antagonism of the GIP receptor, an unusual dual-action design that distinguishes it from the single- and dual-agonist drugs that dominate the market. The practical consequence that matters most to patients and prescribers is dosing cadence: the drug is built for monthly subcutaneous administration rather than the weekly injections that define the incumbent franchises. A once-a-month shot is a materially different product experience, and if the efficacy holds, it is a competitive wedge.
In MARITIME-OSA-2, both the active and placebo arms are delivered subcutaneously to preserve the blind, and the trial spans a full year before the primary readout. The eligibility window is broad on age, enrolling adults from 18 to 99 and accepting all sexes, which keeps the study population representative of the real-world sleep-apnea patient pool. The 250-participant enrollment is modest by Phase 3 standards, which is consistent with a trial powered to detect a change in a continuous physiological measure (AHI) rather than a rare clinical event.
What a positive readout would mean, and what it would not
According to the registry, the study started on November 25, 2025, and its estimated primary completion date is September 29, 2027. That timeline is the first thing to internalize: this is a multi-year bet, and the data that decides it will not arrive soon. When it does, a clean win would look like a statistically significant and clinically meaningful drop in AHI in the maridebart arm relative to placebo at week 52, ideally large enough to move patients from a more severe to a milder disease category.
Strip away the enthusiasm and a few cautions remain. A single-endpoint design is efficient but unforgiving; if the AHI effect underwhelms, there is no secondary efficacy measure carrying the headline. The trial also excludes patients already on PAP, so a positive result speaks to the untreated population specifically and does not, on its own, establish the drug as a replacement for PAP in patients who tolerate their machines. And as with the entire incretin class, tolerability over a 52-week course, particularly gastrointestinal effects, will shape how the efficacy translates into a usable real-world therapy.
The bigger strategic story is what MARITIME-OSA-2 represents for the field. Obesity drugs are no longer being developed solely as weight-loss agents; they are being routed into the constellation of conditions that excess weight drives, with sleep apnea among the most mechanically obvious. A favorable readout would give Amgen a defined medical indication, a clearer reimbursement argument, and a monthly-dosing differentiator in a market hungry for one. Until September 2027, though, the honest answer is that this is a well-designed question, not yet an answer, and the apnea-hypopnea index at week 52 is the number that will settle it.