Cognivue Clarity® in the Detection of Biomarker Confirmed Mild Cognitive Impairment and Alzheimer’s Disease
James Galvin1, Michael Kleiman1, Heather Harris2, Paul Estes2
1University of Miami, 2Cognivue Medical Affairs
Objective:
Demonstrate the utility of Cognivue Clarity as a digital biomarker of mild cognitive impairment (MCI) and Alzheimer’s disease (AD).
Background:
AD and MCI treatment and prevention trials are underway; however, recognition of potentially eligible individuals can be challenging which leads to delayed recruitment and high screen failure. Digital biomarkers that can accurately identify MCI and AD and correspond to AD biomarkers could greatly facilitate enrollment and track disease progression and/or response to treatment. Cognivue Clarity is an FDA-cleared 10-minute, digital cognitive testing platform using adaptive psychophysics to capture global and domain-specific performance.
Design/Methods:
The Bio-Hermes study, funded by the Global Alzheimer’s Platform, enrolled Healthy controls, MCI and AD. Assessments included sociodemographic, Cognivue Clarity, plasma AD biomarkers, and amyloid PET scans. Cognivue Clarity psychometric properties, diagnosis discrimination, correlations with AD biomarkers, and diagnostic power were analyzed.
Results:
Cognivue Clarity was analyzed in 555 participants (297 Controls, 113 MCI, 145 AD) with a mean age of 71.9+6.6y, 15.6+2.8y of education, 56.4% female, 21.7% underrepresented minorities, a mean MMSE of 26.6+3.0, and a mean Cognivue Clarity score of 64.0+17.4. Cronbach’s alpha was 0.882 (95%CI: 0.867-0.897) indicating high reliability. Cognivue Clarity average and subtest scores robustly differentiated (p<0.001) between healthy controls (71.9+12.5), MCI (60.1+16.2) and AD (49.4+17.4). Cognivue Clarity® correlated with Aβ42/40 (r=.253, p<.001), Amyloid Probability Score (r=-.339, p<.001), Amyloid PET (r=-.486, p<.001), ptau217 (r=-.426, p<.001), and ptau181 (r=.-287, p<.001). Cognivue Clarity had a positive likelihood ratio of 2.17, negative likelihood ratio of 0.29, and diagnostic odds ratio of 7.48.
Conclusions:
Cognivue Clarity discriminated healthy controls from MCI and AD, strongly correlated with AD biomarkers, and had strong psychometric properties. Individuals with Cognivue Clarity average scores <70 were 7.5-fold more likely to be impaired. Cognivue Clarity can be used to screen individuals for cognitive impairment, enrich clinical trial inclusion, and track disease progression.