KarXT for Alzheimer’s psychosis (ADEPT-4)

Resolution Criteria Trial: NCT06585787 Selected endpoint key: 481b6986df1af1738a9bde9fe21249eed02d76332694a3de58d8199740038c53 Selected endpoint: Change from baseline in Neuropsychiatric Inventory-Clinician: Hallucinations and Delusions (NPI-C: H+D) score Endpoint type: functional_scale Protocol outcome group: primary Endpoint description: Not provided by ClinicalTrials.gov. Endpoint timeframe: Up to Week 14 ClinicalTrials.gov population: Participants with psychosis associated with Alzheimer's disease Market population: Participants with psychosis associated with Alzheimer's disease Analysis set: Prespecified primary efficacy analysis set; the enumerated registry outcome does not further specify the analysis set. Comparison: KarXT versus placebo Success direction: lower Success rule: In participants with psychosis associated with Alzheimer's disease, KarXT must produce a greater reduction from baseline than placebo in Neuropsychiatric Inventory-Clinician: Hallucinations and Delusions (NPI-C: H+D) score at up to Week 14, with the prespecified statistical success criterion met in the prespecified primary efficacy analysis set. ClinicalTrials.gov source: https://clinicaltrials.gov/study/NCT06585787?tab=researcher#outcome-measures:~:text=Change%20from%20baseline%20in%20Neuropsychiatric%20Inventory-Clinician%3A%20Hallucinations%20and%20Delusions%20(NPI-C%3A%20H%2BD)%20score YES 1. The exact selected endpoint (Change from baseline in Neuropsychiatric Inventory-Clinician: Hallucinations and Delusions (NPI-C: H+D) score) satisfies the approved success rule in the specified population, analysis set, comparison, and timepoint. 2. The result is supported by a qualifying source in the approved evidence hierarchy. NO 1. The exact selected endpoint (Change from baseline in Neuropsychiatric Inventory-Clinician: Hallucinations and Delusions (NPI-C: H+D) score) misses its approved success rule. 2. No qualifying disclosure of the selected endpoint is available by 2026-12-31 at 23:59:59 UTC. NO DECISION / ADMIN REVIEW 1. Available evidence is contradictory, incomplete, pooled, interim, subgroup-only, or does not match the approved population, analysis set, comparison, and timepoint. Resolution deadline: 2026-12-31T23:59:59Z Deadline evidence: AI bounded approximation (confidence 94/100): official source supports 2026-01-01 through 2026-12-31; market deadline 2026-12-31 is the conservative end of that window. Bristol Myers Squibb's official December 3, 2025 press release states that additional ADEPT-program trial results, explicitly including ADEPT-4, are expected to read out by the end of 2026. Because the source provides a year-end window rather than an exact disclosure day, the conservative market deadline is December 31, 2026. Deadline evidence source: https://investors.bms.com/iframes/press-releases/press-release-details/2025/Bristol-Myers-Squibb-Announces-Continuation-of-ADEPT-2-Phase-3-Study-in-Psychosis-Associated-with-Alzheimers-Disease/default.aspx Evidence precedence 1. ClinicalTrials.gov posted results 2. Sponsor regulatory filing or official release 3. Peer-reviewed publication or official conference abstract Only the selected endpoint in the specified population, analysis set, comparison, and timeframe can resolve YES. Overall positive messaging, another endpoint, interim results, pooled analyses, or an unmatched subgroup cannot resolve YES.