"The delay in Bristol Myers Squibb’s Cobenfy trials for Alzheimer’s disease psychosis pushes potential treatment further out, highlighting a significant unmet medical need and the challenging landscape of managing these symptoms in a vulnerable population."
Bristol Myers Squibb’s highly anticipated clinical trials for Cobenfy, a drug being investigated for psychosis associated with Alzheimer’s disease, have experienced a significant delay, with topline results now expected to begin arriving in early 2027 and spread throughout the year. This represents a shift from the company’s previous expectation of reporting all three ADEPT study results in 2026. The postponement underscores the persistent challenge of finding approved pharmacological treatments for a condition that affects a substantial portion of individuals with Alzheimer’s, leaving families to rely on off-label medications with significant safety concerns.
A Prolonged Wait for a Novel Treatment
The landscape of Alzheimer’s disease treatment is in a state of evolution, with emerging therapies aiming to address the underlying pathology of the disease. However, a critical and distressing symptom experienced by a significant number of Alzheimer’s patients – psychosis, which can manifest as hallucinations and delusions – remains largely unaddressed by approved medications in the United States. This unmet medical need is starkly illuminated by the latest developments from Bristol Myers Squibb (BMS), which announced a further delay in the reporting of topline results from its three Phase 3 ADEPT studies evaluating Cobenfy (a schizophrenia drug) for Alzheimer’s disease psychosis.
Originally anticipated to yield results throughout 2026, BMS has now informed investors that these crucial data readouts will commence in early 2027 and continue to be released over the course of that year. This marks the second public adjustment to the trial timelines. The initial delay was announced in December 2025, when BMS had to reopen enrollment for the ADEPT-2 study due to identified irregularities in trial execution at several sites. Data from these affected sites were subsequently excluded, and the Food and Drug Administration (FDA) was notified. Following a review by an independent data monitoring committee, the study continued, with the projected readout then shifting to late 2026.
The current delay is attributed by BMS Chief Executive Chris Boerner to slower-than-anticipated patient enrollment in the ADEPT-2 and ADEPT-4 studies, coupled with a slower accrual of relapse events in the ADEPT-1 study. While slower relapse accrual might, in some contexts, suggest a potential benefit of the investigational drug (as fewer relapses mean the trial needs more time to reach its primary endpoints), BMS Chief Medical Officer Cristian Massacesi acknowledged that the blinded nature of the study prevents them from definitively attributing this to Cobenfy’s efficacy versus a more robust performance from the control arm compared to historical data.
Despite the delay in definitive topline results, BMS plans to release safety and efficacy data from the open-label lead-in portion of ADEPT-1 and from a rollover study involving patients who have completed the trials. An interim analysis of ADEPT-1 may also be available later in 2026. It is important to note that open-label data, where participants and researchers are aware of the treatment being administered, cannot establish definitive efficacy due to the absence of a placebo control.
The Shadow of the "Boxed Warning"
The significance of the ADEPT trials and the potential for Cobenfy cannot be fully appreciated without understanding the current, precarious treatment paradigm for psychosis in Alzheimer’s disease. The FDA’s "boxed warning," the most serious type of warning issued by the agency, was applied in 2005 to all antipsychotic medications. This warning states that elderly patients with dementia-related psychosis who are treated with these drugs face an increased risk of death. This cautionary measure was based on pooled analyses of multiple placebo-controlled trials in this specific population, which revealed approximately 1.6 to 1.7 times higher mortality rates among treated patients.
Crucially, there is no antipsychotic medication specifically approved by the FDA for psychosis in Alzheimer’s disease. While two related approvals offer some context, they do not directly address the core unmet need. Pimavanserin, approved in 2016, is indicated for hallucinations and delusions associated with Parkinson’s disease psychosis, explicitly excluding dementia-related psychosis unrelated to Parkinson’s. Brexpiprazole, approved in May 2023, targets agitation associated with Alzheimer’s dementia, a distinct symptom, and notably, it still carries the class-wide boxed warning for increased mortality risk.
Consequently, the management of psychosis in individuals with dementia often relies on off-label prescribing of various drug classes, including second-generation antipsychotics, benzodiazepines, antidepressants, and mood stabilizers. These off-label uses are supported by modest evidence and are associated with a range of meaningful side effects such as sedation, increased risk of falls, and stroke. Current clinical guidelines strongly advocate for non-pharmacological interventions as the first line of treatment and recommend limiting antipsychotic courses to the shortest possible duration. The complexities surrounding this issue have even prompted an FDA public workshop, in collaboration with the Duke-Margolis Center for Health Policy, to examine whether the evidence base for the boxed warning should be re-analyzed. Laura Gault, BMS’s head of neuroscience development, has aptly described this condition as an area of "tremendous unmet medical need."
Interpreting the Delay: Caution Without Alarm
It is imperative to frame the delay in the ADEPT trials within its proper context. A pushed readout does not equate to a failed trial. What is currently known are specific operational challenges: slower enrollment in two studies, a more gradual accumulation of relapse events in a third, and prior issues with site conduct in one study that necessitated data exclusion. What remains unknown, and is the central question the trials aim to answer, is whether Cobenfy is effective for this specific indication. The blinded nature of the trials means that efficacy results have not yet been determined.
Cobenfy is already an approved medication for schizophrenia in adults. Its proposed mechanism of action – muscarinic receptor agonism – differs from the dopamine-blocking mechanisms of traditional antipsychotics that carry the class-wide boxed warning. This mechanistic distinction forms the basis for hope that Cobenfy might possess a more favorable safety profile in frail older patients with Alzheimer’s-related psychosis. However, this remains a hypothesis that requires robust clinical demonstration within this vulnerable population.
While BMS’s program is experiencing a delay, other investigational therapies for Alzheimer’s disease psychosis are progressing. Acadia Pharmaceuticals’ remlifanserin has received FDA fast-track designation for this indication, with Phase 2 RADIANT topline results expected in September or October 2026, and Phase 3 studies slated to commence thereafter. MapLight Therapeutics also anticipates releasing Phase 2 results for their candidate in the second half of 2027. Despite these ongoing efforts, none of these programs are yet close to seeking or receiving an approval decision.
Navigating Hallucinations and Delusions in the Interim
For families and caregivers currently managing hallucinations and delusions in individuals with Alzheimer’s disease, the delay in potential new treatments necessitates a focus on existing best practices. It is crucial to emphasize that any decisions regarding medication should be made in consultation with a healthcare professional who has a comprehensive understanding of the patient’s individual condition.
The initial and paramount step when psychotic symptoms emerge is a thorough medical evaluation to identify and address any potentially reversible causes. Conditions such as urinary tract infections, pneumonia, dehydration, uncontrolled pain, constipation, poor sleep hygiene, or the introduction of new medications can all precipitate or exacerbate psychotic symptoms in the context of dementia. Effectively treating these underlying issues can sometimes lead to the resolution of psychotic symptoms without the need for psychiatric medication.
Current clinical guidelines strongly recommend prioritizing non-drug approaches. These strategies, while not always simple, are concrete and actionable. They include identifying and mitigating potential triggers for agitation or confusion, maintaining consistent routines and familiar environments, optimizing lighting to reduce misperceptions, addressing sensory impairments such as hearing and vision problems, and avoiding confrontational interactions regarding the content of delusions. Caregivers often find that engaging in debates about the reality of a delusion can escalate the symptom.
When medication is deemed necessary, open communication with the prescribing physician is vital. Key questions to ask include the specific symptom being targeted by the medication, the plan for reassessment and eventual discontinuation, and the specific side effects that warrant immediate medical attention. Symptoms such as excessive sedation, new unsteadiness, or a fall are serious and require prompt contact with a healthcare provider.
Certain symptoms necessitate urgent medical evaluation, as opposed to a scheduled appointment. These include a sudden and significant decline in alertness, confusion developing over hours to days, fever, new onset of weakness, or any situation where the safety of the individual or their caregiver is compromised.
Support resources are available for families navigating these challenges. The Alzheimer’s Association offers a 24-hour helpline staffed by clinicians. Caregiver support programs are also often covered by Medicare. The path forward will involve the release of interim ADEPT-1 data and open-label findings later this year, followed by remlifanserin Phase 2 results in early fall, and the commencement of ADEPT topline readouts starting in early 2027.
Frequently Asked Questions
What was announced?
Bristol Myers Squibb announced that topline results from its three ADEPT trials, investigating Cobenfy for psychosis associated with Alzheimer’s disease, will now begin arriving in early 2027, a shift from the previously expected reporting throughout 2026.
Why was the readout delayed?
The delay is attributed to slower patient enrollment in the ADEPT-2 and ADEPT-4 studies, and a slower accrual of relapse events in the ADEPT-1 study. A previous delay was also linked to conduct irregularities at certain study sites.
Is there an approved treatment for psychosis in Alzheimer’s disease?
No, there is currently no medication FDA-approved specifically for psychosis in Alzheimer’s disease. Brexpiprazole is approved for agitation associated with Alzheimer’s dementia, which is a different symptom, and still carries a boxed warning.
What is the "boxed warning"?
Since 2005, all antipsychotic medications carry an FDA boxed warning indicating an increased risk of death in elderly patients with dementia-related psychosis treated with these drugs.
Does the delay mean the drug has failed?
No, a delay in reporting results does not indicate trial failure. The trials remain blinded, meaning efficacy data has not yet been analyzed. The company has suggested that slower relapse accrual could potentially be a positive indicator, but this cannot be confirmed while the study remains blinded.
What should families do when hallucinations or delusions begin?
It is recommended to seek an immediate medical evaluation to identify and treat any reversible causes, such as infections, dehydration, or medication side effects. Non-drug approaches should also be explored and implemented.
When is the next data expected?
Open-label ADEPT-1 data and a potential interim analysis are expected later in 2026. Phase 2 results for remlifanserin are anticipated in September or October 2026. Topline readouts for the ADEPT studies are scheduled to begin in early 2027.
