AbstractBackgroundAccumulation of β‐amyloid (Aβ) plaques and neurofibrillary tangles are core components of Alzheimer's disease (AD) pathogenesis. Longitudinal in vivo observations of AD pathologic changes using amyloid and tau positron emission tomography (PET) enable us to obtain AD biomarkers' hypothetical trajectory curves. However, the difference in disease progression by striatal Aβ deposition, sex, and apolipoprotein E (APOE) genotype has not been explained with trajectory curves. The aim of this study is to develop the trajectory in accumulation of cortical and striatal Aβ and tau deposits. We also tried to investigate the differences in sex and APOE genotype on trajectory curves.MethodWe obtained 817 subjects from the Alzheimer’s Disease Neuroimaging Initiative database, who showed either 18F‐florbetapir (AV45) PET positivity at least once or AV45 PET negativity with normal cognition. We Z‐transformed each value based on Aβ‐negative, cognitively normal subjects. Instead of obtaining a single rate value (∆Z/∆t) per subject, we calculated the slope for every interval of PET scans. For each biomarker, we fitted a penalized spline to model the slope of Z value as a nonlinear function of baseline Z value. Using the fitted splines, we obtained the predicted Z curves as a function of time and developed trajectory curves.ResultAmong the subjects, 534 subjects were available for AV45 PET analysis and 163 subjects underwent 18F‐flortaucipir PET. While cortical Aβ accumulation rate increased until Z value reached 4.33 and decreased subsequently, ∆Z/∆t for the striatal Aβ and tau stayed more constant. Time from Z=0 to Z=2 was the longest for tau PET (22.6 years) and shortest for cortical Aβ (14.9 years). For cortical Aβ, Z values of males reached 1.7 years earlier than females. However, for striatal Aβ, males took 2.8 more years than females to reach Z=2. Also, APOE ɛ4 carriers showed faster progression for all biomarkers.ConclusionThis study shows that Aβ accumulates in the striatum after cortex and then tau protein, which is consistent with the amyloid cascade hypothesis. The trajectory curves differ according to sex and APOE genotype. Therefore, prognosis and treatment should be approached differently in clinical practice with respect to striatal involvement, sex, and APOE genotype.