BMS to continue Phase III ADEPT-2 for Cobenfy in Alzheimer’s-related psychosis after site irregularities

Bristol Myers Squibb will enrol additional patients into its Phase III ADEPT-2 trial of Cobenfy/KarXT (xanomeline/trospium) for psychosis associated with Alzheimer’s disease (AD), after identifying “clinical trial execution irregularities” at a small number of sites. The decision — made jointly with the US FDA and the independent Data Monitoring Committee (DMC) — reflects regulators’ view that the study remains scientifically and clinically viable.

During a blinded internal review, BMS excluded data from several sites prior to database lock. While the company did not disclose the nature of the irregularities, they were significant enough to warrant removal from the primary analysis. An independent interim efficacy and safety review, evaluated by the DMC, recommended continuing the trial with additional enrollment to restore the planned sample size. BMS remains fully blinded to outcomes.

About ADEPT-2 and Cobenfy

ADEPT-2 is a multicenter Phase III randomized, double-blind, placebo-controlled trial (NCT06126224) evaluating Cobenfy for hallucinations and delusions in Alzheimer’s dementia. The primary endpoint is change in NPI-C (Hallucinations + Delusions), with CGI-S severity score as a key secondary endpoint. Safety and tolerability are also being assessed.

Because of added recruitment, top-line ADEPT-2 results are now expected end-2026, a year later than originally planned. Readouts from companion studies ADEPT-1 and ADEPT-4 are also anticipated in that timeframe.

Cobenfy — a combination of the muscarinic agonist xanomeline and the peripheral blocker trospium — is already approved for schizophrenia in adults and is viewed by BMS as a potential first-in-class therapy for Alzheimer’s-related psychosis.

The wider context

Psychosis affects a large share of people living with Alzheimer’s, yet treatment options remain limited and often rely on antipsychotics with significant safety concerns. The continuation of ADEPT-2 preserves a critical neuroscience asset for BMS at a time of mounting patent pressures across its portfolio.

Other molecules in development for Alzheimer’s-related psychosis include:

• ACP-204 (remlifanserin) — Acadia Pharmaceuticals’ selective 5-HT₂A inverse agonist in Phase III development.
• LY03020 — Luye Pharma’s dual TAAR1 and 5-HT2C receptor agonist, currently in Phase I.

These programs underscore growing interest in non-dopaminergic approaches to treating psychosis in dementia populations.