BMS Pipeline Hits Snags: Cobenfy Alzheimer's Trial Delayed to 2027, Raising Execution and Translation Questions

BMS Q2 Financial Beat Overshadowed by Cobenfy and Milvexian Delays

Bristol Myers Squibb reported strong second-quarter financial results, exceeding revenue and earnings per share estimates, and subsequently raised its full-year guidance. However, this positive news was overshadowed by significant delays in key Phase 3 data readouts for two of its investigational drugs. The ADEPT program evaluating Cobenfy for Alzheimer’s disease psychosis has been pushed to early 2027 due to enrollment challenges and slower-than-expected relapse event accrual. Additionally, data from the Johnson & Johnson-partnered milvexian's LIBREXIA-AF trial for atrial fibrillation has also been delayed, following an earlier discontinuation of its acute coronary syndrome study.

• Bristol Myers Squibb delivered a robust financial performance in the second quarter, with revenues reaching $12.97 billion, a 5% increase year-over-year and surpassing consensus estimates. EPS also significantly beat expectations at $2.04, a 28% increase. This strong showing prompted the company to raise its annual revenue guidance to a midpoint of $49.5 billion and boost EPS expectations to $6.88, driven by strong sales of drugs like Orencia, Reblozyl, Camzyos, and Opdivo.
• The Phase 3 ADEPT program for Cobenfy, an oral neuropsychology drug acquired from Karuna Therapeutics, has faced another delay, with topline data readouts for Alzheimer’s disease psychosis now expected to begin in early 2027. This postponement is attributed to ongoing enrollment challenges in ADEPT-2 and ADEPT-4 studies, as well as slower accrual of relapse events in ADEPT-1, marking a repeated delay for the program.
• Further clouding the financial beat, the Phase 3 LIBREXIA-AF trial for milvexian, an investigational oral Factor XIa inhibitor partnered with Johnson & Johnson, has also seen its data readout pushed back. This delay follows the early discontinuation of a separate Phase 3 study for milvexian in acute coronary syndrome late last year, which an independent data committee determined was unlikely to meet its efficacy goal, raising questions about broader trial conduct.

Addressing the Unmet Needs in Alzheimer's Disease Psychosis Treatment

Current management of Alzheimer's disease psychosis (AD+P) remains fundamentally constrained by the absence of an FDA-approved therapy specifically indicated for this condition, forcing reliance on off-label atypical antipsychotics with a challenging risk-benefit profile. Mounting evidence suggests AD+P may represent a mechanistically distinct entity from other psychotic disorders, which helps explain the modest and inconsistent efficacy observed with existing pharmacotherapy. These limitations are compounded by significant safety concerns and a paucity of viable non-pharmacological alternatives with robust evidence.

• No approved indication and mechanistic uncertainty: No medication carries FDA approval for AD+P specifically; atypical antipsychotics are used off-label despite evidence that AD+P may arise from mechanisms distinct from schizophrenia — only 115 of over 2,000 genes implicated in each condition overlap, and network analyses show antipsychotics perform significantly worse in AD+P than in schizophrenia (P < 0.001).

• Limited and narrow efficacy: Antipsychotics demonstrate only modest, short-term benefit (6–12 weeks), are more effective for aggression than agitation, and provide no control over pain — a symptom tightly linked to behavioral and psychological symptoms of dementia (BPSD). Recent studies have also found some patients show no measurable benefit over placebo, and the co-occurrence of overlapping neuropsychiatric symptoms makes targeting a single symptom domain difficult.

• Serious safety and mortality signals: The FDA issued black box warnings in 2005 (atypical antipsychotics) and extended these to typical antipsychotics in 2008, citing increased all-cause mortality and cerebrovascular adverse events (including stroke) in elderly dementia patients. Pneumonia is among the most frequently reported causes of death, and stroke risk is dose-dependent, with an adjusted odds ratio of 1.60 (95% CI, 1.51–1.69) and particularly elevated risk in older patients and those with high-affinity M1 muscarinic/α2 adrenergic receptor-binding agents.

• Compounded cardiovascular risk with polypharmacy and dosing: Concurrent use of antipsychotics with cognitive enhancers raises the odds of a major adverse cardiovascular/cerebrovascular event (MACCE) by 96.3%, while high-dose antipsychotic regimens increase this risk by 238% — underscoring that dosing strategy and drug combinations substantially influence outcomes.

• Inadequate safety reporting in trials: Adverse events are inconsistently and incompletely documented across randomized controlled trials, with no RCT meeting full CONSORT reporting standards for harms — limiting the ability to make well-informed risk-benefit assessments.

• Broader systemic and adherence challenges: Antipsychotic use is associated with increased likelihood of long-term care institutionalization and may not be cost-effective for healthcare systems. Additionally, Alzheimer's patients experience significantly more psychotropic-related adverse events than non-AD older adults (OR = 1.66; 95% CI, 1.20–1.82), often involving benzodiazepines, antipsychotics, and autonomic nervous system-active agents, and frequently requiring inpatient admission.

• Lack of robust non-pharmacological alternatives: While guidelines recommend non-pharmacological strategies as first-line treatment, evidence supporting alternatives such as aromatherapy (e.g., Melissa officinalis, Lavandula officinalis) remains weak, particularly for symptoms like pain that frequently drive BPSD.

Cobenfy's Expanding Potential Beyond Schizophrenia and ADP

Cobenfy (xanomeline-trospium chloride, known as KarXT) has established schizophrenia as its primary approved indication, but the literature also points to emerging investigation in bipolar disorder. Notably, the available data do not specify the intervention models (e.g., parallel assignment, crossover, single-group) used across these trials.

• Schizophrenia (approved indication): Cobenfy received FDA approval in September 2024, based on three double-blind, randomized, placebo-controlled trials (EMERGENT-2, EMERGENT-3, EMERGENT-4) in adults with acute psychosis. It is a first-in-class dual muscarinic M1/M4 receptor agonist (xanomeline) combined with trospium, a peripherally restricted muscarinic antagonist that mitigates peripheral side effects. Trials demonstrated significant improvements in positive, negative, and cognitive symptoms, with an 8.4-point greater reduction in PANSS total score versus placebo. Long-term open-label data (52 weeks) showed >75% of participants achieving a >30% PANSS improvement, with a mean decrease of 33.3 points. The drug's safety profile favors reduced weight gain and extrapyramidal symptoms compared to conventional antipsychotics, with mild GI-related side effects (nausea, dyspepsia, constipation) that rarely led to discontinuation.

• Bipolar disorder (emerging indication): Cobenfy is being investigated for manic episodes and cognitive impairment associated with bipolar disorder, supported by its pharmacological rationale and potential for transdiagnostic efficacy across mania and cognitive domains. Existing schizophrenia data suggest the drug does not exhibit depressogenic effects or increase suicidality risk, which may support further exploration in mood disorder populations.

• Mechanistic distinction driving expanded interest: Unlike traditional antipsychotics, Cobenfy operates via cholinergic (muscarinic M1/M4) receptor activation rather than dopamine receptor blockade, a novel mechanism that underpins its exploration beyond schizophrenia and may explain its differentiated tolerability profile in adjacent psychiatric indications.

• Data limitations: A systematic review identified only four studies with available results out of fourteen KarXT-focused studies reviewed, indicating that while short-term efficacy and tolerability data are promising, broader clinical recommendations—particularly for real-world use and indications beyond schizophrenia—require additional evidence before being finalized.

The Competitive Landscape for Novel ADP Therapies

Following the recent approval of Cobenfy (xanomeline-trospium), a first-in-class M1/M4 muscarinic receptor agonist, the competitive landscape for schizophrenia therapies with this mechanism is expanding. Several other agents targeting muscarinic receptors are also advancing through clinical development. While specific trial intervention models are not detailed in the literature, key information on these emerging competitors is summarized below.