Diagnostic Accuracy of Plasma p-tau181 Measured Using Automated Chemiluminescent Immunoassays for Alzheimer's Disease: A Systematic Review and Meta-Analysis.

INTRODUCTION: To evaluate the diagnostic accuracy of plasma p-tau181 measured on fully automated immunoassay platforms for detecting Alzheimer's Disease (AD), and to determine how performance varies according to the reference standard used to define AD pathology. MATERIALS AND METHODS: A systematic search was conducted across major databases for studies reporting the diagnostic accuracy of plasma p-tau181. Eligible studies enrolled adults across the Alzheimer's disease continuum and applied either CSF Aβ42/40, combined CSF Aβ42/40 and ptau181, or clinical diagnostic criteria as the reference standard. Data were synthesized using random- effects models. HSROC modeling was applied to estimate pooled sensitivity, specificity, and AUC. RESULTS: A total of 19 studies involving 4266 participants were included. Most studies (18/19) used the Lumipulse platform, whereas only 1 study used Elecsys. Therefore, the pooled estimates mainly reflect Lumipulse-based evidence. Pooled sensitivity and specificity were 0.81 and 0.77 (AUC 0.84) when CSF Aβ42/40 defined amyloid pathology. When combined, CSF amyloid and tau positivity was required; sensitivity and specificity were 0.83 and 0.78 (AUC 0.86). Using clinical diagnostic criteria, pooled sensitivity reached 0.92, with a specificity of 0.78 (AUC 0.84). However, these subgroup estimates should be interpreted cautiously because substantial between-study heterogeneity remained, and some subgroups contained only a small number of datasets. DISCUSSION: P-tau181 measured by automated chemiluminescent immunoassays showed promising diagnostic performance across reference standards, but the summary estimates are influenced by substantial heterogeneity and should not be interpreted as universally transferable measures of performance. CONCLUSION: The results support plasma p-tau181 as a promising scalable biomarker for AD. However, because the available evidence is predominantly derived from Lumipulse-based studies, it remains insufficient to determine whether comparable performance extends to other automated CLIA platforms.