Plasma p-tau217 as a Potential Independent Biomarker for Predicting the Progression from MCI to AD Dementia

 

Qingjie Lian^{1, 2}, Qingyuan Sun², Xiaodong Han², Aidi Shan², Cuibai Wei^{2, 3, 4}

¹School of Biological Science and Medical Engineering, Beihang University, 100191, Beijing, China. ²Innovation Center for Neurological Disorders and Department of Neurology, Xuanwu Hospital, Capital Medical University, National Clinical Research Center for Geriatric Diseases, 100053, Beijing, China. ³Jinan Hospital of Xuanwu Hospital, Shandong First Medical University, 250102, Jinan, China. ⁴Clinical Research Center, Xuanwu Hospital, Capital Medical University, National Clinical Research Center for Geriatric Diseases, 100053, Beijing, China.

INTRODUCTION: Alzheimer's disease (AD) is the leading cause of dementia. Mild cognitive impairment (MCI) represents a critical stage for the early prevention and treatment of AD. Multiple plasma biomarkers, including phosphorylated tau 217 (p-tau217), have shown predictive potential for AD at the MCI stage. However, the comparative predictive performance of p-tau217 and Aβ-PET for AD remains uncertain. Furthermore, whether multi-biomarker plasma panels can improve clinical utility for AD prediction is unclear. Accordingly, this study aimed to assess whether plasma p-tau217 could serve as an alternative to Aβ-PET and multi-biomarker plasma panels for predicting AD in MCI individuals.

METHODS: Using follow-up diagnostic information, we included 47 MCI-Converters (patients who progressed to AD) and 66 MCI-Nonconverters (patients who did not progress) from the Alzheimer’s Disease Neuroimaging Initiative. Wilcoxon rank-sum tests, five-fold cross-validated logistic regression models, DeLong tests, Akaike information criterion, Bayesian information criterion, Cox proportional hazards models, Spearman correlations, and linear mixed-effects models (LMM) were applied to evaluate inter-group differences, discriminative performance, conversion risk, and associations of plasma biomarkers with Aβ-PET measures (centiloid, summary SUVR) and longitudinal cognitive decline. Risk stratification analyses were further conducted using LMM after categorizing participants into low-, moderate-, and high-risk strata according to a p-tau217-based two-cut-point approach.

RESULTS: p-tau217 demonstrated the most significant inter-group difference (p < 0.001) and the highest Cox hazard ratio (7.88 [95% CI: 4.38-14.19]), and best distinguished MCI-Converters from MCI-Nonconverters (AUC = 0.886 [95% CI: 0.790-0.983]), with its discriminative performance comparable to that of Aβ-PET measures and multi-biomarker plasma models (DeLong p > 0.05). p-tau217 exhibited the strongest associations with cognitive decline and Aβ-PET (p < 0.001). High-risk participants showed the steepest cognitive decline (p < 0.001), with a positive predictive value (PPV) of 100% for Aβ-PET amyloid positivity and a PPV of 82.2% for MCI-Converters. Low-risk participants showed a negative predictive value (NPV) of 78.3% for Aβ-PET amyloid negativity and an NPV of 84.9% for MCI-Nonconverters.

CONCLUSIONS: p-tau217 is the optimal plasma biomarker with the potential to serve as an alternative to Aβ-PET and multi-biomarker panels for predicting conversion from MCI to AD dementia, thereby facilitating plasma biomarker–based prediction of MCI-to-dementia conversion.