OBJECTIVES: A key challenge in early treatment for Alzheimer’s Disease (AD) is identifying the best population for treatment. Genetic tests and imaging technologies are now available to assess the presence of ApoE4 genes or β-amyloid plaques at preclinical stages of AD. Our objective was to assess the utility of these diagnostics for predicting conversion to AD.

METHODS: We simulated disease progression in cognitively normal patients using the AD Archimedes Condition-Event simulator (AD ACE). The AD ACE is constructed using predictive equations for change in disease biomarkers and clinical scales derived primarily from analyses of AD Neuroimaging Initiative (ADNI) data. We compared simulation results over 10 years for 400 patients in each of four patient subgroups (Aβ⁺/ApoE4⁺, Aβ⁺/ApoE4^-, Aβ^-/ApoE4⁺, Aβ^-/ApoE4^-) to evaluate the predictive power of ApoE4 genetic testing versus β-amyloid imaging. Patients with baseline florbetapir PET level ≥ 1.1 for the mcSUVR were considered as Aβ⁺