Analytical and clinical performance of eight Simoa® and Lumipulse® assays for automated measurement of plasma p-tau181 and p-tau217

Among the Alzheimer's disease (AD) biomarkers measured in blood, phosphorylated forms of tau (p-tau) have been shown to exhibit a particularly high diagnostic potential. Here, we performed a comprehensive method comparison study, followed by evaluation of the diagnostic performance of eight recent plasma p-tau immunoassays targeting different tau phosphorylation sites, different tau fragments, and that are measured by two distinct platforms. We enrolled a cohort of 40 patients with AD at the stage of dementia (AD-dem) characterized by positive CSF A + T + profile, and a control group of 40 cognitively healthy participants (Control), to conduct a comprehensive method comparison for three plasma p-tau181 and five plasma p-tau217 assays run on the Simoa® HD-X™ or Lumipulse® G600II/G1200 platforms. Design of the compared assays differed in regard to: (1) tau phosphorylation site targeted by the capture antibody (T181 or T217), and (2) epitope of the pan-tau detector antibody (N-terminal or mid-region). For each of the assays we determined precision and analytical sensitivity parameters and used Passing-Bablok regression and Bland-Altman plots for pairwise comparison of p-tau181 or p-tau217 assays. Subsequently, we evaluated the diagnostic accuracy of all the assays for discrimination between AD-dem and Control groups. We found a strong, positive correlation between all the measurements. Fixed and/or proportional bias was observed for each of compared p-tau181 assay pairs or p-tau217 assay pairs. While both plasma p-tau181 and p-tau217 levels were significantly increased in AD-dem vs. Control groups as measured by all assays, higher median concentration AD-dem/Control fold change and AUC values were observed for p-tau217 (assays range: fold change 3.72–6.74, AUC 0.916–0.956) compared with p-tau181 (assays range 1.81–2.94, AUC 0.829–0.909), independently of the platform used. No significant differences were observed between diagnostic performance of p-tau181 assays or p-tau217 assays targeting tau N-terminus or mid-region. Although all plasma p-tau measurements enabled discrimination between clinical groups, p-tau217 assays showed the highest robustness, independently of the pan-tau detector antibody targeting N-terminal or mid-region, and independently of the platform used. Considering the observed method disagreement in measured absolute concentrations, we stress the need for development of certified reference material, harmonizing measurements across different platforms.