Cerebrospinal fluid AΒ42, total tau, and phosphorylated tau in the evaluation of Alzheimer’s dementia: Experience from a commercial reference laboratory

Abstract

AbstractBackgroundAβ amyloid plaques and tau neurofibrillary tangles play an important role in Alzheimer’s disease (AD) pathophysiology, and the CSF biomarkers Aβ42, total tau (T‐tau), and phosphorylated tau181 (P‐tau) have been extensively used in diagnosing AD. Here we describe our experience with use of the Aβ42/T‐tau index (ATI) and P‐tau for classification of patients with suspected AD.MethodA convenience sample of results from CSF specimens submitted to a commercial reference laboratory for measurement of Aβ42, T‐tau, and P‐tau (presumably for diagnosis in patients with suspected AD) was selected for analysis. Each marker was measured using sandwich ELISA−based tests. The ATI was calculated as Aβ42/(240 + (1.18 × T‐tau)). AD classification was based a combination of ATI and P‐tau thresholds: not consistent with AD (not‐AD; P‐tau <54 pg/mL and ATI >1.2); borderline (P‐tau 54‐68 pg/mL and/or ATI 0.8‐1.2); AD (P‐tau >68 pg/mL or ATI <0.8); indeterminant—possible non‐AD dementia (NADD; ATI <0.8, P‐tau <54); and “other” (highly elevated ATI). Differences in age at testing among the classification groups were assessed using ANOVA with Tukey pairwise analysis.ResultResults from 42,030 specimens (49.1% women) were included. The most frequent classifications based on ATI and P‐tau values were AD (30.0%) and borderline (30.2%), followed by indeterminate‐NADD (21.7%), not‐AD (16.3%), and other (1.9%). Individuals with an AD classification were significantly older than those classified as not‐AD (mean [SD] age 66.8 [10.1] vs 62.5 [12.3] years; P<0.001), and a higher proportion were female (55.6% vs 42.8%). The ANOVA model indicated significant differences in age at testing among the classification groups (P<0.001). In post‐hoc pairwise comparisons, the not‐AD group was significantly younger than the AD group (Tukey multiple comparison test: difference= ‐4.3 years, Padj <0.001). This pattern was consistent in analyses stratified by sex (women: difference=‐4.8 years, Padj <0.001; men: difference= ‐4.4 years, Padj <0.001).ConclusionNearly 30% of specimens submitted for AD testing yielded results consistent with AD, underscoring the value of the ATI and P‐tau measurements in CSF in patients with suspected AD. The greater age among those classified with AD vs non‐AD likely reflects the higher pretest probability of AD in older individuals.