High tau PET binding reveals a constellation of beta‐amyloid positivity and fast cognitive decline

Abstract

AbstractBackgroundBeta‐amyloid Positron Emission Tomography (PET) efficiently detects beta‐amyloid plaques supporting the presence of a pathophysiological landmark of Alzheimer’s disease (AD) continuum, but lacks specificity regarding the prediction of future cognitive status. The increasing validation of tau PET tracers raises questions about whether a tau PET biomarker could predict clinically relevant cognitive decline and its potentially beneficial role in the assessment of individuals who belong to the AD continuum.MethodWe identified 335 subjects [cognitive normal (CN; n=217), mild cognitive impairment (MCI; n=99), dementia (n=19)] with baseline beta‐amyloid and tau PET scans that have completed a follow‐up of at least two years, from the Alzheimer’s Disease Neuroimaging Initiative dataset. First, we assessed the accuracy of tau PET to discriminate the MCI patients that will convert to AD. Based on Alzheimer’s Disease Assessment Scale–Cognitive Subscale, we used linear mixed‐effects models to define the annual rate of cognitive change of all individuals and Gaussian mixture modelling to cluster them as either fast or slow decliners. Next, group comparisons and the accuracy of tau PET to discriminate between them were tested. Finally, the association between baseline tau PET SUVR and cognitive rate of change was assessed with linear modelsResultThe individuals with beta‐amyloid positivity and fast cognitive decline showed the highest tau PET binding values independently of baseline cognitive status, compared to the other groups (beta‐amyloid negative with fast or slow decline and beta‐amyloid positive with slow decline). Baseline tau PET binding could determine fast decliners with a beta‐amyloid positive scan with an accuracy of 85% (composite temporal region of interest). Similarly, tau PET binding accuracy was high in determining MCI patients that would be both beta‐amyloid positive and show progress in their diagnosis. Linear models showed that baseline tau PET binding is related to the annual rate of cognitive decline in beta‐amyloid positive individuals.ConclusionHigh baseline tau load is coupled with both beta‐amyloid pathology and an accelerated profile of cognitive decline. Tau PET imaging could offer substantial advantages over beta‐amyloid PET imaging since it provides important information about both the underlying pathophysiology and the prognosis of the disease.