WATCH TIME: 4 minutes
“This tool doesn’t replace clinical evaluation or confirmatory biomarkers, but it adds meaningful value by helping clinicians better decide what the next step should be for their patients.”
Developed from data in the Bio-Hermes study, the CogniVue Amyloid Risk Measure (CARM) combines results from three cognitive subtests: adaptive motor control, visual salience, and shape discrimination to estimate the likelihood of amyloid pathology in patients with cog
nitive impairment. Using machine learning and PET-verified amyloid status, CARM was scaled from 0 to 100 and defined with 4 risk thresholds to guide interpretation. It was designed to help determine whether a patient’s cognitive impairment may be due to Alzheimer disease (AD), as well as help clinicians quickly identify patients who may benefit from confirmatory testing or who may be better suited for AD-directed therapies or trials.
The full CogniVue assessment takes just 10 minutes and provides not only cognitive screening but also insight into whether impairment is likely related to AD. In a recent interview with NeurologyLive®, dementia expert James Galvin, MD, MPH, emphasized its utility in primary care and memory care environments, especially where access to PET scans or lumbar punctures is limited. CARM can serve as an early filter, reducing unnecessary testing and guiding clinicians toward the most appropriate next steps.
Galvin, a professor of neurology at the University of Miami Miller School of Medicine, and chief scientific officer at CogniVue, sees promise for CARM not just in clinical use, but in research as well. By identifying patients more or less likely to have amyloid pathology, it can improve trial efficiency, reduce costs, and refine cohort selection. He also noted the broader implications of the Bio-Hermes data, including findings that nearly half of patients clinically diagnosed with early-stage AD had negative amyloid PET scans, reinforcing the need for more precise triage tools to avoid unnecessary testing and improve precision of early detection.