Plasma and CSF biomarkers in a memory clinic: Head-to-head comparison of phosphorylated tau immunoassays

Abstract

Introduction: Direct comparisons of the main blood phosphorylated tau immunoassays in memory clinic populations are needed to understand possible differences.

Methods: In the BIODEGMAR study, 197 participants presenting with cognitive complaints were classified into an Alzheimer's disease (AD) or a non-AD cerebrospinal fluid (CSF) profile group, according to their amyloid beta 42/ phosphorylated tau (Aβ42/p-tau) ratio. We performed a head-to-head comparison of nine plasma and nine CSF tau immunoassays and determined their accuracy to discriminate abnormal CSF Aβ42/p-tau ratio.

Results: All studied plasma tau biomarkers were significantly higher in the AD CSF profile group compared to the non-AD CSF profile group and significantly discriminated abnormal CSF Aβ42/p-tau ratio. For plasma p-tau biomarkers, the higher discrimination accuracy was shown by Janssen p-tau217 (r = 0.76; area under the curve [AUC] = 0.96), ADx p-tau181 (r = 0.73; AUC = 0.94), and Lilly p-tau217 (r = 0.73; AUC = 0.94).

Discussion: Several plasma p-tau biomarkers can be used in a specialized memory clinic as a stand-alone biomarker to detect biologically-defined AD.

Highlights: Patients with an Alzheimer's disease cerebrospinal fluid (AD CSF) profile have higher plasma phosphorylated tau (p-tau) levels than the non-AD CSF profile group. All plasma p-tau biomarkers significantly discriminate patients with an AD CSF profile from the non-AD CSF profile group. Janssen p-tau217, ADx p-tau181, and Lilly p-tau217 in plasma show the highest accuracy to detect biologically defined AD. Janssen p-tau217, ADx p-tau181, Lilly p-tau217, Lilly p-tau181, and UGot p-tau231 in plasma show performances that are comparable to their CSF counterparts.

Keywords: biomarker; dementia; disease; phosphorylated tau; plasma; tau.

FIGURE 1

Diagram of tau depicting the antibodies and the epitopes recognized in each plasma immunoassay. Schematic diagram of the tau isoform (2N4R) of 441 amino acids, which comprises two N-terminal domains (N) and the four microtubule-binding domains (R). The phosphorylation sites T181, T217, and T231 are also shown. (A) The tau protein fragments recognized by the combination of antibodies of the immunoassays tested in this study. (B) The combination of antibodies employed in Lumipulse cerebrospinal fluid (CSF) phosphorylated tau (p-tau) assay, which was used to determine the Alzheimer’s disease (AD) CSF profile (amyloid beta (Aβ)42/p-tau ratio). The epitopes recognized by each antibody are shown. ^aCSF Janssen p-tau217 used PT82 as detector antibody, which targets aa119–126 (Janssen R&D, Springhouse, PA). ^bPlasma Lilly t-tau assay, the capture antibody is Tau12, and detector antibody is 4G10-E2.

FIGURE 2

Levels of the plasma tau biomarkers in the non-AD versus the AD cerebrospinal fluid (CSF) profile groups, visualized as raincloud plots. Each point indicates a tau measurement of an individual. The box plot indicates the median (horizontal line), interquartile range (box), and 1.5× interquartile range (whiskers). The width of the shaded area (violin plot) represents the proportion of the data located there. AD CSF profile was defined by a CSF amyloid beta (Aβ)42/p-tau181 ratio < 10.25 (as measured by Lumipulse, Fujirebio). P-values were computed with a Mann-Whitney U test for all plasma biomarkers. The percentage increase of the tau biomarkers in the AD CSF profile from the non-AD profile is shown in green.

FIGURE 3

Discrimination accuracy (receiver-operating characteristic [ROC] analyses) of plasma and cerebrospinal fluid (CSF) tau immunoassays. Forest plot depicting the area under the curve (AUC) and the 95% confidence interval (CI) of each tau biomarker to discriminate between the Alzheimer’s disease (AD) CSF profile group form the non-AD CSF profile groups. AD CSF profile was defined by a CSF amyloid beta (Aβ)42/p-tau181 ratio < 10.25 (as measured by Lumipulse G600II, Fujirebio). CI, confidence interval.