Epitope placement tunes CSF tau biomarkers: Comparison of P-tau217 and BD-tau designs

Abstract

Measurement of different tau species using immunoassays has proven effective for detecting early Alzheimer’s disease (AD) pathology. However, assay performance depends on factors such as the tau form targeted (phosphorylated or non-phosphorylated) and the partner antibody used. Here, we evaluated five different antibody combinations for their ability to differentiate amyloid-positive (Aβ +) from amyloid-negative (Aβ −) individuals in cerebrospinal fluid (CSF). A total of 290 patients were included in the study grouped by amyloid status. We compared four CSF tau assays designed to separate two axes of assay biology, phosphorylation state and fragment targeting (N-terminal vs mid-domain partner antibodies), to determine how these design choices influence capture of AD pathology. All CSF tau biomarkers were markedly elevated in the Aβ + group. Phosphorylated tau assays showed larger fold increases (approximately 3–sixfold) compared with non-phosphorylated assays (~ twofold). In linear regression models, all biomarkers were associated with log-transformed pTau217. ROC analyses demonstrated high accuracy for detection of amyloid positivity across assays (AUC range 0.86–0.95). BD-ptau217 achieved the highest AUC (0.95, 95% CI 0.92–0.97) and fold increase in the Aβ + compared to Aβ − . Our results show that, irrespective of the partner antibody, phosphorylated tau assays are more accurate than assays that also measure non-phosphorylated forms of the protein in differentiating Aβ + from Aβ − individuals. Moreover, among the immunoassays evaluated, BD-pTau217 exhibited a greater fold increase in the Aβ + group than N-terminal pTau217 and Mid-pTau217. Together, these findings underscore the importance of assay configuration for biomarker performance and indicate that BD-pTau217 may offer superior sensitivity for the early identification of AD pathology. Future studies evaluating these assays in plasma are warranted.