Premium
Phospho‐tau217 and phospho‐tau181 in plasma and CSF as biomarkers for Alzheimer’s disease
Author(s) -
Janelidze Shorena,
Palmqvist Sebastian,
Quiroz Yakeel T.,
Lopera Francisco,
Stomrud Erik,
Su Yi,
Chen Yinghua,
Serrano Geidy E,
Leuzy Antoine,
Mattsson Niklas,
Strandberg Olof,
Smith Ruben,
Villegas Andres,
Sepulveda Diego,
Chai Xiyun,
Proctor Nicholas,
Zetterberg Henrik,
Beach Thomas G.,
Blennow Kaj,
Reiman Eric M.,
Dage Jeffrey L.,
Hansson Oskar
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.037520
Subject(s) - neuropathology , cohort , biomarker , cerebrospinal fluid , medicine , dementia , psen1 , oncology , amyloid (mycology) , alzheimer's disease neuroimaging initiative , pathology , disease , alzheimer's disease , amyloid precursor protein , biology , biochemistry
Background Cerebrospinal fluid (CSF) p‐tau181 (tau phosphorylated at threonine 181) is an established biomarker of Alzheimer’s disease (AD) reflecting abnormal tau metabolism in the brain. We have recently shown that also plasma p‐tau181 is a promising biomarker for AD (Janelidze et al, Nature Medicine , 2020). We now evaluate whether CSF p‐tau217 or plasma p‐tau217 are even better biomarkers. Methods We compared CSF p‐tau217 to CSF p‐tau181 in the Swedish BioFINDER‐1 cohort (n=194). We next evaluated plasma p‐tau217 and plasma p‐tau181 in three cohorts with 1,438 participants: an Arizona‐based neuropathology cohort, including 34 AD and 47 non‐AD participants; the Swedish BioFINDER‐2 cohort, including cognitively unimpaired (n=312) participants, and clinically diagnosed patients with mild cognitive impairment (MCI, n=188), AD dementia (n=126), and other non‐AD neurodegenerative diseases (n=109); a Colombian autosomal‐dominant AD kindred, including 365 PSEN1 E280A‐carriers and 257 non‐mutation carriers. Results CSF p‐tau217 had stronger correlations with the tau‐PET tracer, and more accurately identified individuals with abnormal tau‐PET scans (AUC=0.93) than CSF P‐tau181. CSF P‐tau217 correlated better than p‐tau181 with CSF and PET measures of neocortical amyloid‐β burden and more accurately distinguished AD dementia from non‐AD neurodegenerative disorders. Antemortem plasma P‐tau217 differentiated individuals with intermediate‐to‐high likelihood of AD according to neuropathology from those without AD (AUC=0.89) and performed significantly better than plasma P‐tau181. Plasma P‐tau217 also differentiated clinical AD dementia from non‐AD neurodegenerative diseases (AUC=0.96) significantly better than plasma P‐tau181, plasma neurofilament light, and established MRI measures, and similar to CSF P‐tau217, CSF P‐tau181, CSF Aβ42/40, and tau‐PET. Increased plasma P‐tau217 was observed already in the pre‐symptomatic stages of AD. In PSEN1 mutation carriers the increase started at age 25, about 20 years prior to estimated onset of MCI. Plasma P‐tau217 correlated with cerebral tau tangle densities in subjects with neuritic plaques (r=0.64). It predicted abnormal tau‐PET scans (AUC=0.95) significantly better than plasma P‐tau181, plasma neurofilament light, CSF P‐tau181 and CSF Aβ42/Aβ40, and similar to CSF P‐tau217. Conclusions Plasma and CSF P‐tau217 reflect brain tau burden, increases early in AD, and differentiates AD from other neurodegenerative diseases.