DETECT‐AD (Digital Evaluations and Technologies Enabling Clinical Translation for Alzheimer’s Disease): A simulated anti‐amyloid clinical trial using digital biomarkers as primary outcome measures

Abstract

AbstractBackgroundEarly detection and precise monitoring of cognitive and functional decline in persons with Alzheimer’s Disease (AD) in clinical trials is a high priority. Unfortunately, early detection and monitoring of the subtle and slow changes of disease with conventional cognitive testing and self‐report questionnaires has been fraught with poor sensitivity and lack ecological validity. Digital biomarkers using ambient sensing technologies may afford a more effective approach to conducting trials. Remaining to be established is the ability of this approach to fundamentally improve early AD treatment response readout in a formal clinical trial setting. A simulated clinical trial (DETECT‐AD: ClinicalTrials.gov, NCT05385913) is being conducted to determine how well continuously assessed, home‐based digital biomarkers may detect clinically meaningful changes in individuals with changing amyloid (Aß) burdens in pre‐symptomatic AD.MethodClinically prodromal AD patients with estimable rates of AD progression based on Aß Florbetapir‐PET status are enrolled divided n = 50:n = 50 into Aß “positive” (higher amyloid burden) patients who predictably progress (as if they were receiving placebo) and Aß “negative” patients progressing more slowly (representing the treatment group). A study‐provided multivitamin mimics RCT conditions and study drug adherence behavior. Baseline standard clinical and biomarker measures are obtained along with digital biomarkers (DB) captured with passive and active sensors (actigraphy, bed mats, IR sensing, e‐pillbox, computing/phone device use, weekly online reporting/cognitive assessment). Primary outcome is the change in a composite DB composed of measures in 4 domains: mobility, cognition, sleep, and socialization. Secondary measures are the relationship of DB to fluid biomarkers (plasma ptau181/217, amyloid, NfL) and MRI/PET measures.ResultTo date (01/16/2023), 46/100 have been enrolled with mean(SD): follow‐up = 27.0(7.9) weeks (no dropout); age = 78(6.6); 63% women; MoCA = 25.5(2.3); PET SUVR = 1.09(0.20), range: 0.850‐2.07. Exemplary baseline digital biomarker metrics are presented in the Table. Preliminary data shows an inverse Aß‐SUVR correlation with mean steps/day (r = ‐.41, p = .012) and a positive correlation of MoCA with computer use (r = .61; p = .036).ConclusionHome‐based continuous capture of multi‐domain DBs is feasible in a clinical trial setting and may yield early meaningful indicators of clinically relevant changes in older adults with pre‐AD dementia. Further evidence will evolve as sample sizes and longitudinal follow‐up increases.