Prognostic utility of p‐tau217 and tau PET in Alzheimer's disease

TL;DR

This study examines how combining p-tau217 blood tests with tau PET scans can predict cognitive decline in early Alzheimer's disease.

Contribution

The study evaluates the added value of plasma p-tau217 to tau PET in predicting clinical progression in preclinical Alzheimer's.

Findings

p-tau217 improved the predictive power of medial temporal lobe tau PET but not neocortical tau PET.

Tau PET remains the primary tool for predicting cognitive decline in early Alzheimer's stages.

Higher tau PET positivity correlated with increased risk of clinical progression.

Abstract

Recently proposed biomarker‐based biological staging schemes may improve risk prediction of cognitive impairment in Alzheimer's disease (AD). Evidence demonstrates that tau positron emission tomography (PET) reliably identifies individuals at high risk for clinical progression. While plasma phosphorylated tau at threonine 217 (p‐tau217) has been proposed as a cost‐effective biomarker, its added prognostic value to tau PET remains under‐explored. Here, we tested the utility of combining plasma p‐tau217 and tau PET for predicting risk of clinical progression in cognitively unimpaired (CU) individuals.

We evaluated 156 CU individuals from the A4 Study placebo group with positron emission tomography (PET) for amyloid‐β (Aβ) plaques ([18F]Florbetapir) and tau tangles ([18F]Flortaucipir), plasma p‐tau217 and longitudinal neuropsychological testing. Aβ‐positive (A+) individuals were classified as p‐tau217‐positive (A+Tp‐tau217+), tau PET‐positive in the medial temporal lobe (A+TMTL+) and in the neocortex (A+TNEO+). Cutpoints were determined as the mean + 2.0 standard deviations (SD) of the corresponding tau biomarker in Aβ‐negative controls from the LEARN substudy. Time‐to‐event analyses considered clinical progression as a 1 point increase in the Clinical Dementia Rating ‐ Sum of Boxes (CDR‐SB) score, and dichotomized A/T biomarkers were used as predictors.

Demographics of the study population are reported in Table 1. Cox proportional‐hazard models showed a gradual increase in the risk of clinical progression in the A+TMTL+ (HR=2.25, p = 0.0034) and A+TNEO+ (HR=3.14, p < 0.0001) groups versus the A+ (reference) group. In analyses incorporating p‐tau217 to the models, we found that the A+Tp‐tau217+ group showed a significantly increased risk for clinical progression compared to A+ group when added to the TMTL model but not to the TNEO model. (Figure 1). Model fit indexes indicated that adding p‐tau217 to the models improved the predictive performance of the TMTL model, but not of the TNEO model (Table 2).

We found that p‐tau217 positivity does not lead to a clear added prognostic value to tau PET in assessing risk of clinical progression in individuals with preclinical AD, particularly when measuring neocortical tau PET signal. These findings support imaging biomarkers as the primary prognostic tools for predicting cognitive impairment in early AD stages.

Figures