CNS-selective plasma p-tau217 accurately captures Alzheimer’s disease pathology and progression

Abstract Blood-based biomarkers have expanded access to biologically supported diagnosis of Alzheimer’s disease (AD), particularly through measurement of amyloid-beta (Aβ) and phosphorylated tau species 1–3 . Among these, plasma tau phosphorylated at threonine 217 (p-tau217) is currently the leading biomarker recommended by clinical guidelines 4–6 . However, circulating p-tau217 originates from both central nervous system (CNS) and peripheral tissues 7 , potentially limiting specificity, particularly in individuals with common age-related comorbidities 8 . Here we report a next-generation biomarker, brain-derived p-tau217%, which quantifies the proportion of circulating tau that is CNS-derived and phosphorylated at threonine 217. Across neuropathologically defined, Aβ- and tau-neuroimaging–characterized, and memory clinic cohorts, brain-derived p-tau217% consistently identified AD pathology and clinical AD with larger effect sizes, higher discriminative accuracy, and improved sensitivity and specificity, outperforming conventional non-CNS-selective plasma p-tau217, p-tau217/Aβ1-42 and p-tau217% alternatives as well as brain-derived-p-tau217 alone. Furthermore, the CNS-selective biomarker demonstrated more robust prediction of future clinical progression in individuals followed for up to two decades. Importantly, diagnostic performance remained high in older adults with diabetes and cardiovascular disease, populations in which standard p-tau217 showed reduced specificity. Moreover, superiority extended to comparisons against multiple CNS disease-related proteins in targeted proteomic analyses. These findings establish plasma brain-derived p-tau217% as a biologically grounded and clinically robust biomarker that advances molecular definition, detection, and prognosis of Alzheimer’s disease.