Cognition Therapeutics said on October 22, 2026 that the FDA's Divisions of Neurology and Psychiatry, within the agency's Office of Neuroscience, have aligned with the company on key design elements for a planned registrational Phase 3 trial of zervimesine (CT1812) in people with mild to moderate dementia with Lewy bodies (DLB), a disease for which no drug is approved in the United States. The agreement, disclosed in a filing with the Securities and Exchange Commission, covers patient selection, dosing, trial duration and the endpoints the agency would accept as evidence in a future application.
As described in that filing, the proposed study would randomize participants to a single 100 milligram daily dose of oral zervimesine or to placebo for nine months. Its primary endpoint would be a two-item version of the Neuropsychiatric Inventory, called NPI-2, which scores the severity and frequency of hallucinations and delusions, both core features of DLB; a key secondary endpoint would be the ADCS-ADL scale, which measures everyday function. Lisa Ricciardi, the company's president and chief executive, said the agreement followed "collaborative discussions with FDA's Divisions of Neurology and Psychiatry within the Office of Neuroscience."
The company's basis for moving forward is a 130-person Phase 2 trial called SHIMMER, which randomized adults with mild to moderate DLB, screened to MMSE scores of 18 to 27, to 100 milligram or 300 milligram doses of zervimesine or to placebo for 26 weeks. According to results published in the journal Alzheimer's & Dementia in December 2025, the 88 participants who received the drug, combined across both doses, worsened more slowly than the 42 on placebo: 86 percent slower on the Neuropsychiatric Inventory, 52 percent slower on the ADCS-ADL scale and 62 percent slower on a Parkinson's disease movement rating scale. These were topline, exploratory findings from a single mid-stage trial, and the new FDA agreement means only that a confirmatory study can proceed, not that the drug has been shown to work.
Wednesday's announcement is the latest in a string of FDA exchanges this year: the company reported reaching alignment in June on using the same symptom scale as a primary measure in DLB patients with psychosis, and said in September it had scheduled a further meeting to settle remaining endpoint questions. It also follows earlier UK regulatory movement on the same medicine that this newspaper has previously reported. None of these steps is a finding that zervimesine is safe or effective.