After delaying the phase 3 readout of its Alzheimer’s psychosis program for Cobenfy earlier this year, Bristol Myers Squibb is offering an early peek that appears to bode well for the high-stakes data drops ahead, even as the fine details remain hazy.
BMS on Wednesday rolled out preliminary findings from its late-stage Adept program, which is assessing its Cobenfy cocktail of xanomeline and trospium chloride—approved for schizophrenia back in 2024—as a treatment for patients with psychosis associated with Alzheimer’s disease.
The first look at the results comes courtesy of the 12-week open-label period of Bristol’s Adept-1 study, plus a yearlong open-label extension trial coded Adept-3, with Cobenfy tied to improvements in hallucinations and delusions, as well as overall disease severity, in both. BMS shared the data as part of the Clinical Trials on Alzheimer’s Disease conference this week in Boston.
Notably, with the full slate of Adept trials still ongoing, data from double-blind treatment periods across the program remain under wraps. BMS clarified that its analysis tapped patients who completed or prematurely discontinued the Adept-1 and -3 open-label periods. Participants still involved in those open-label portions weren’t included in Wednesday’s data tease, which reported out findings based on descriptive analyses.
Looking at Adept-1, 273 patients of 320 enrolled had completed the three-month open-label lead-in period by Bristol’s data cutoff, the company said in a Nov. 18 release. Over that span, treatment with Cobenfy, also known as KarXT, yielded a mean reduction of 12 points from baseline on the Neuropsychiatric Inventory-Clinician Hallucinations and Delusions (NPI-C H+D) scale, plus a 1.9-point mean reduction on participants’ Clinical Global Impression-Severity (CGI-S) score.
On the Adept-1 safety front, treatment-emergent adverse events were logged in 56.6% of study participants, with serious treatment-related side effects in 5% and adverse events leading to discontinuation in 4.4% of patients, per BMS.
On the rollover Adept-3 study, which absorbed patients completing the double-blinded Adept studies -1, -2 and -4, 273 people of 401 enrolled had completed the 52-week treatment period in time for the analysis, BMS said.
Here, Cobenfy was tied to 5.1-point and 0.8-point mean reductions from baseline on the NPI-C H+D scale and CGI-S score, respectively.
“ADEPT-1 and ADEPT-3 are designed to answer two different, important questions,” Cristian Massacesi, M.D., Bristol’s head of development and chief medical officer, said in a statement. “The open-label period of ADEPT-1 provides an initial look at how patients respond to KarXT over 12 weeks, while ADEPT-3 is evaluating the long-term safety and tolerability of KarXT over 52 weeks.”
The findings so far are “encouraging,” Massacesi said, and that sentiment was echoed by several analysts this week, even as those shops caveated that much remains unknown about Cobenfy’s overall performance.
Investors are indeed closely watching how the drug fares in Alzheimer’s psychosis, hoping for a market expansion opportunity following Cobenfy’s rather subdued launch in schizophrenia beginning in the fall of 2024.
While admitting they were “still slightly perplexed by the ability to disclose data from these ongoing trials,” the team at William Blair agreed with Bristol’s optimism around the open-label results, adding that improvements across hallucinations, delusions and disease severity, coupled with an apparently manageable tolerability profile, support future use of Cobenfy in Alzheimer’s psychosis patients, even as the lack of a placebo comparator in today’s announcement hampers read-through to Bristol’s other trials.
Analysts at Citi, meanwhile, referred to Cobenfy’s impact on hallucinations and delusions “substantial” in the open-label readout. Moreover, the Citi team noted that the long-term safety data garnered in Adept-3 is “particularly noteworthy” given the advanced age of the patient population Cobenfy would be appealing to with an Alzheimer’s psychosis nod.
Treatment-emergent adverse events cropped up in 65.1% of participants in Adept-3, with serious treatment-emergent events recorded in 7.7% and event-related discontinuations in 4.7% of patients.
Over the summer, BMS kept investors waiting when it said it expected phase 3 data from Adept to now come through in early 2027 and be spread across the year, versus prior estimates that readouts would land in 2026.
At the time, Bristol CEO Chris Boerner attributed the delay to slower enrollment in Adept-2 and Adept-4, plus slower relapse event accrual in Adept-1. Suggestion has also been made that Adept-1 could undergo an interim analysis as soon as later this year.
An expansion for Cobenfy could help boost investor sentiment as Bristol grapples with a disconnect between respectable sales growth and market expectations.
Despite growing sales 30% in the third quarter to hit $57 million, for instance, Cobenfy’s most recent earnings performance came in some $11 million below analyst expectations. BMS’ chief commercialization officer Adam Lenkowsky noted on a conference call last month that the company may work on speeding up new patient starts and bolstering repeat prescribing.
BMS picked up Cobenfy as the crown jewel in its $14 billion buyout of Karuna Therapeutics.