AbstractBackgroundPrior studies of tau PET tracers in participants with normal cognition and/or a broad range of Alzheimer’s disease (AD) severity suggest baseline tau PET signal is associated with subsequent rates of cognitive decline. However, the prognostic utility of tau PET imaging is less well‐established within more homogenous cohorts limited to specific AD stages and for cognitive versus functional decline. We addressed these questions by analyzing baseline [18F]GTP1 tau PET signal and longitudinal changes in cognitive and functional measures in the Phase 2 Tauriel study of semorinemab in prodromal‐to‐mild AD.MethodsWe analyzed data from a subset of Tauriel participants with prodromal AD (pAD; n=127) or mild AD (mAD; n=233) who underwent baseline [18F]GTP1 tau PET and longitudinal assessments with the Clinical Dementia Rating‐Sum of Boxes (CDR‐SB), Alzheimer’s Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog13), Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), and/or Alzheimer’s Disease Cooperative Study‐Activities of Daily Living (ADCS‐ADL). As no treatment effects were observed with semorinemab in Tauriel, data were pooled across semorinemab and placebo arms. Associations between baseline [18F]GTP1 SUVRs and annualized longitudinal change on each measure were explored using Pearson correlations and multiple linear mixed effects regression models adjusting for relevant baseline covariates.ResultsBaseline cortical [18F]GTP1 SUVRs and longitudinal decline on clinical measures were significantly greater in mAD relative to pAD cohort. Prognostic associations with baseline cortical [18F]GTP1 were similar in pAD and mAD and more robust for predominantly cognitive measures (ADAS‐Cog13: pAD rs=0.37, mAD rs=0.35; RBANS Total Index: pAD rs=‐0.40, mAD rs=‐0.41) than measures incorporating functional status (CDR‐SB: pAD rs=0.27, mAD rs=0.29; ADCS‐ADL: pAD rs=‐0.12, mAD rs=‐0.26). Analogous results were seen using temporal or Braak stage regions of interest. [18F]GTP1 associations with cognitive decline generally remained significant after adjusting for baseline cognitive score, age, hippocampal volume, and education.ConclusionsHigher baseline [18F]GTP1 tau PET signal was associated with faster subsequent clinical decline in both prodromal and mild AD. More robust associations were seen with cognitive relative to functional measures. These results bolster prior work that suggests a role for tau PET in patient selection/stratification in AD clinical trials.