Tau-PET is increasingly used in Alzheimer disease (AD) clinical trials, both for participant selection based on disease stage and as an outcome measure. However, the longitudinal patterns of tau accumulation according to AD stages remain unclear. We aimed to examine the patterns of tau-PET accumulation in different AD stages defined by the Alzheimer's Association Workgroup to inform trial design. This was a retrospective analysis of data collected from participants in 2 longitudinal observational cohort studies: the Translational Biomarkers in Aging and Dementia (TRIAD) study from Montreal, Canada, and Alzheimer's Disease Neuroimaging Initiative (ADNI), a multisite study in North America. Amyloid-PET-positive participants were biologically staged as one of the following based on tau-PET uptake: A+T2- (initial), A+T2MTL+ (early), A+T2MOD+ (intermediate), and A+T2HIGH+ (advanced). Amyloid-PET-negative participants were included as non-AD controls. All participants underwent amyloid-PET and tau-PET imaging, with longitudinal follow-up (mean: 2.96 ± 1.35 years). Linear mixed-effects models evaluated stage-specific regional longitudinal tau-PET changes; statistical power analyses estimated the sample sizes needed to detect differences in tau-PET accumulation for future AD trials. The study included 542 participants (mean age: 67.9 ± 15.3 years; 56.3% female), comprising 321 non-AD controls and 221 individuals with AD. The baseline AD stage determined both the regional distribution and magnitude of tau accumulation. No significant accumulation occurred in A+T2- individuals over 4-6 years. In those with early-stage AD, tau accumulation was localized to early affected regions (TRIAD: β = 0.15, 95% CI 0.09-0.21, p < 0.001; ADNI: β = 0.21, 95% CI 0.03-0.40, p = 0.03). Individuals with intermediate-stage AD showed accumulation in intermediate regions (TRIAD: β = 0.16, 95% CI 0.10-0.22, p < 0.001; ADNI: β = 0.37, 95% CI 0.15-0.59, p = 0.001) while those with advanced-stage AD exhibited accumulation in later affected regions (TRIAD: β = 0.45, 95% CI 0.39-0.50, p < 0.001; ADNI: β = 0.31, 95% CI 0.14-0.49, p < 0.001). Stage-specific tau-PET region-of-interest selection reduced required sample sizes by 30%-93% for detecting hypothetical disease-modifying drug effects. Stage-specific patterns of tau-PET change highlighted the importance of baseline biological AD staging for selecting outcome regions of interest. Statistical power analyses indicated that aligning outcome regions with disease stage reduced sample size estimates, suggesting potential gains in trial efficiency. While findings were consistent across 2 independent cohorts, differences in tracers and demographic composition represent key limitations.
Trudel et al. (Tue,) studied this question.