Direction-specific biomarker profiles and clinical significance of plasma p-tau217/amyloid PET discordance in cognitively normal adults.

Abstract

Introduction: Plasma phosphorylated tau (p-tau)217 is increasingly used to triage amyloid testing, but the biological and clinical meaning of discordant plasma p-tau217 and amyloid positron emission tomography (PET) results in cognitively normal adults remains uncertain.

Methods: We analyzed cognitively normal participants from the Chinese Preclinical Alzheimer's Disease Study (CPAS; n = 594) and the Alzheimer's Disease Neuroimaging Initiative (ADNI; n = 475). Amyloid PET positivity was defined as Centiloid ≥ 20. Cohort-specific p-tau217 positivity was defined using PET-anchored cutoffs with ≥ 0.90 specificity, yielding four p-tau/PET groups. Linear models assessed cross-sectional biomarker and cognitive differences; mixed-effects and Cox models assessed ADNI longitudinal Centiloid accumulation, cognitive trajectories, and clinical progression.

Results: In cross-sectional analyses, compared to p-tau-/PET-, p-tau+/PET- had higher glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) levels across cohorts; p-tau-/PET+ showed a lower plasma amyloid beta 42/40 ratio in ADNI; and p-tau+/PET+ showed the broadest multimodal abnormalities. In ADNI, both p-tau+ groups had higher clinical progression risk and faster decline in selected cognitive domains; both PET+ groups showed faster Centiloid accumulation. Faster Centiloid accumulation in p-tau+/PET- was observed only at lower specificity thresholds.

Discussion: Discordant plasma p-tau217 and amyloid PET results should be interpreted according to both the positivity pattern and the p-tau217 cutoff applied. Distinguishing these patterns may aid risk stratification and longitudinal monitoring. Different p-tau217 operating points may support risk enrichment and prioritization for confirmatory amyloid testing, but require validation before clinical use.