Fully automated measurement of plasma A42/40 and p-tau181: Analytical robustness and concordance with cerebrospinal fluid profile along the Alzheimer's disease continuum in two independent cohorts

INTRODUCTION: For routine clinical implementation of Alzheimer's disease (AD) plasma biomarkers, fully automated random-access platforms are crucial to ensure reproducible measurements. We aimed to perform an analytical validation and to establish cutoffs for AD plasma biomarkers measured with Lumipulse. METHODS: Two cohorts were included. UNIPG: n = 450 paired cerebrospinal fluid (CSF)/plasma samples from subjects along the AD-continuum, subjects affected by other neurodegenerative diseases, and controls with known CSF profile; AMS: n = 40 plasma samples from AD and n = 40 controls. Plasma amyloid beta (A beta)42, A beta 40, and p-tau181 were measured with Lumipulse. We evaluated analytical and diagnostic performance. RESULTS: Lumipulse assays showed high analytical performance. Plasma p-tau181 levels accurately reflected CSF A+/T+ profile in AD-dementia and mild cognitive impairment (MCI)-AD, but not in asymptomatic-AD. Plasma and CSF A beta 42/40 values were concordant across clinical AD stages. Cutoffs and probability-based models performed satisfactorily in both cohorts. DISCUSSION: The identified cutoffs and probability-based models represent a significant step toward plasma AD molecular diagnosis.