Precision medicine for Alzheimer's disease in Down syndrome

Abstract INTRODUCTION Down syndrome (DS) exhibits a genetic form of Alzheimer's disease (AD). We used a blood‐based proteomic algorithm to predict cognitive status, treatment responders, and change to vitamin E in DS adults from a completed clinical trial, “Vitamin E in Aged Persons with Down Syndrome,” which originally showed no significant cognitive benefit using the primary endpoint cognition (Brief Praxis Test BPT). METHODS Plasma and extracellular vesicle (EV; astrocytic and neuronal) biomarkers were assayed at baseline and 36 months ( n = 138 each). Cognitive response was measured using combined scores from the BPT, vocabulary, and behavior and function DS tests. Support vector machine (SVM) analyses predicted diagnostic and treatment responders and change accuracy. RESULTS SVM classified demented versus non‐demented with up to 99% accuracy and predicted treatment response and changes with up to 100% accuracy in plasma and EV. DISCUSSION Our study supports blood‐based screening and precision diagnostics for AD therapy in DS.