Clinical performance and decision thresholds of an automated plasma p-tau217 assay for amyloid pathology in a Taiwanese memory clinic

Fully automated plasma p-tau217 assays may facilitate clinical implementation of blood-based testing for Alzheimer’s disease (AD) amyloid pathology. However, data on assay-specific decision thresholds and clinical utility in East Asian memory-clinic populations remain limited. We evaluated plasma Roche p-tau217 in 139 participants from a Taiwanese memory clinic using amyloid PET or CSF biomarkers as reference standards. Diagnostic performance and single- and dual-cutoff decision thresholds were assessed. Head-to-head comparisons with Simoa p-tau217, Roche p-tau181, and neurofilament light chain (NfL) were performed in participants with complete biomarker profiles. Associations with body mass index and renal function were also examined. Roche p-tau217 showed high diagnostic accuracy for amyloid β (Aβ) positivity (AUC 0.973). In the primary threshold-analysis cohort, a dual-cutoff approach using thresholds of 0.225 and 0.336 pg/mL classified 88.1% of participants into definitive negative or positive zones, with an overall percent agreement of 95.0% within these zones. Sensitivity analyses showed similar performance. In participants with complete biomarker profiles, no statistically significant difference in AUC was detected between Roche p-tau217 and Simoa p-tau217, whereas Roche p-tau217 showed significantly higher AUC than p-tau181 and NfL. Lower estimated glomerular filtration rate (eGFR) was independently associated with higher p-tau217 concentrations. Fully automated Roche p-tau217 showed high diagnostic performance for identifying Aβ pathology. Cohort-derived dual decision thresholds classified most participants into definitive zones and may support plasma p-tau217-based triage for confirmatory amyloid testing in memory-clinic practice. External validation is required before clinical implementation.