Diagnostic performance of plasma Aβ42/40 ratio, p‐tau181, GFAP, and NfL along the continuum of Alzheimer's disease and non‐AD dementias: An international multi‐center study
James D. Doecke, Giovanni Bellomo, Lisa Vermunt, Daniel Alcolea, Steffen Halbgebauer, Sjors G. J. G. In ‘t Veld, Niklas Mattsson, Kateřina Veverová, Christopher Fowler, Lynn Boonkamp, Isabel Houtkamp, Marleen JA Koel‐Simmerlink, Inge M.W. Verberk, Lorenzo Gaetani, Andrea Toja, Anna Lidia Wojdała, Juan Fortea, Yolande Pijnenburg, Afina W. Lemstra, Wiesje M. van der Flier, Jakub Hort, Markus Otto, Oskar Hansson, Lucilla Parnetti, Colin L. Masters, Alberto Lleó, Armand González‐Escalante, José Contador, Marc Suárez‐Calvet, Aida Fernández‐Lebrero, Albert Puig‐Pijoan, Paula Ortiz‐Romero, Esther Jiménez‐Moyano, Carolina Minguillón, Marta del Campo, Charlotte E. Teunissen
Abstract
INTRODUCTION: Plasma phosphorylated tau (p-tau)181, glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), and amyloid beta ratio (Aβ42/40) may have diagnostic and prognostic value in Alzheimer's disease (AD). Here we assess which markers can best identify AD from controls and other non-AD dementias in a large international multi-center study. METHODS: Plasma samples (n = 1298) were collected from six international centers. Aβ40, Aβ42, GFAP, NfL, and p-tau181 were measured using single molecule array. In each group, AD diagnosis/co-pathology was defined according to cerebrospinal fluid biomarkers or amyloid positron emission tomography. Validations were performed in three separate cohorts via single and dual cut-off models. RESULTS: p-tau181 showed the best area under the curve value to separate AD from frontotemporal dementia, controls, and Aβ- dementia with Lewy bodies. However, this discriminative power could not be reproduced by applying pre-defined cut-offs. DISCUSSION: p-tau181 was the best single plasma marker for detecting AD at any stage. Specific cut-offs are needed to maximize diagnostic performances. HIGHLIGHTS: Phosphorylated tau (p-tau)181 provided a clear differentiation between controls and Alzheimer's disease (AD) participants, with evidence of increased levels in the preclinical stage of AD. Plasma biomarkers demonstrated that when amyloid co-pathology is removed from dementia with Lewy bodies (DLB), only glial fibrillary acidic protein and neurofilament light chain remain to predict DLB. Given the low prevalence of amyloid co-pathology in frontotemporal dementia (FTD), p-tau181 and its ratio with amyloid beta 42 are strong biomarkers to differentiate FTD from AD.