Comparison of plasma ALZpath p-Tau217 with Lilly p-Tau217 and p-Tau181 in a neuropathological cohort

Abstract

There is an urgent need for accurate and validated methods to measure plasma phosphorylated tau (p-Tau) biomarkers in Alzheimer's disease (AD). In this study, we compared the performance of new commercially available plasma p-Tau217 assay developed by ALZpath to other established p-Tau assays in an autopsy-confirmed sample. We included 72 participants from the Arizona Study of Aging and Neurodegenerative Disorders cohort, with antemortem plasma assessments and postmortem neuropathological examination. Plasma p-Tau217 was measured using the single molecule arrays (SIMOA)-based ALZpath immunoassay and plasma p-Tau217 and p-Tau181 were analyzed with mesoscale discovery (MSD)-based immunoassays developed by Lilly Research Laboratories. Plasma biomarkers were compared with densities of post-mortem-assessed amyloid plaques and neurofibrillary changes. Levels of p-Tau217_ALZpath, p-Tau217_Lilly and p-Tau181_Lilly were significantly associated with plaque (partial Spearman ρ = 0.58, ρ = 0.78, ρ = 0.65; p < 0.001) and neurofibrillary (ρ = 0.26, ρ = 0.51, ρ = 0.37; p < 0.05) density scores, adjusting for age, sex, and time of interval between blood sampling and death. All three biomarkers were also significantly associated with plaque density scores (p-Tau217_ALZpath, ρ = 0.53; p-Tau217_Lilly, ρ = 0.73; p-Tau181_Lilly, ρ = 0.59; p < 0.001), when additionally adjusting for neurofibrillary density scores. However, only p-Tau217_Lilly was significantly associated with neurofibrillary density scores (ρ = 0.32, p = 0.022), when adjusting for plaque density scores. The correlations of p-Tau217_ALZpath and p-Tau181_Lilly with plaque density scores were comparable, whereas p-Tau217_Lilly exhibited significantly higher correlations with plaques (p_diff≤0.015) and neurofibrillary changes (p_diff≤0.004) than p-Tau217_ALZpath. While p-Tau217_ALZpath and p-Tau181_Lilly predicted the presence of Alzheimer's disease neuropathological change (ADNC), Braak staging (Braak 0-IV vs. Braak V-VI) and CERAD neuritic plaque classification (low/sparse vs. moderate/frequent) with similar accuracies (area under the curve [AUC]_range, 0.74-0.79), p-Tau217_Lilly AUCs were significantly higher (AUC_range,0.82-0.89, p_diff≤0.024) than the AUCs of p-Tau217_ALZpath. In conclusion, p-Tau217_ALZpath exhibited similar performance as p-Tau181_Lilly but its correlations with core measures of AD pathology were significantly lower in comparison with p-Tau217_Lilly. Future studies are warranted to replicate these findings in larger independent cohorts.

Keywords: Alzheimer’s disease; Amyloid plaques; Neurofibrillary changes; Phosphorylated-tau.

Fig. 1

A schematic overview of the p-Tau assays. Illustration of full-length tau, including N-terminal domain, proline rich domain and microtubule binding domain. The binding sites for antibodies are indicated under the respective epitope region for both assays (ALZpath and Lilly). The image was created using BioRender

Fig. 2

Correlation between different plasma biomarkers. Spearman’s correlation analysis was used to assess the association of p-Tau217_ALZpath with p-Tau217_Lilly (a) and p-Tau181_Lilly (b). Abbreviations: p-Tau, phosphorylated tau

Fig. 3

Association of p-Tau217_ALZpath, p-Tau217_Lilly and p-Tau181_Lilly with amyloid plaque density scores and neurofibrillary density scores. Association of p-Tau217_ALZpath (a-b), p-Tau217_Lilly (c-d) and p-Tau181_Lilly (e-f) with amyloid plaques (total density score) (a, c, e) and neurofibrillary changes (total density score) (b, d, f). Data are shown as partial Spearman correlation coefficients (ρ) and p-values. Plaque density scores and neurofibrillary density scores were measured on a semi-quantitative scale range from 0–3 using CERAD templates in five different brain regions and were added up to a total score range from 0–15. Data points are colored based on the ADNC classification. Abbreviations: ADNC, Alzheimer’s disease neuropathologic change; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; p-Tau, phosphorylated tau

Fig. 4

Independent associations between p-Tau217_ALZpath, p-Tau217_Lilly or p-Tau181_Lilly and both amyloid plaque density scores and neurofibrillary density scores. Correlations between plaque density scores and neurofibrillary density scores and p-Tau217_ALZpath, p-Tau217_Lilly or p-Tau181_Lilly after adjusting for other pathology load (i.e., when looking at neurofibrillary changes adjusting for plaques and conversely) as well as other covariates (age, sex and time between blood sampling and death). Partial Spearman’s rho (ρ) for each biomarker (p-Tau217_ALZpath, p-Tau217_Lilly or p-Tau181_Lilly) is indicated above the bars and percentual partial Spearman ρ over the sum of the partial Spearman ρ of the two pathologies (% partial ρ = 100*partial ρ/ (partial ρ_[plaque] + partial ρ_{[neurofibrillary changes]}) is indicated below the bar. Abbreviations: p-Tau, phosphorylated tau

Fig. 5

Levels of plasma biomarkers by ADNC classification. Plasma levels of p-Tau217_ALZpath, (a), p-Tau217_Lilly (b) and p-Tau181_Lilly (c) by ADNC classification. Group differences were assessed using Kruskal Wallis test and Mann-Whitney U tests (p-values were corrected for multiple comparisons using FDR). Boxes show interquartile range, the horizontal lines are the medians, and the whiskers are plotted using Tukey method. Abbreviations: ADNC, Alzheimer’s disease neuropathologic change; FDR, false discovery rate; p-Tau, phosphorylated tau

Fig. 6

Levels of plasma biomarkers by Braak Staging. Plasma levels of p-Tau217_ALZpath, (a), p-Tau217_Lilly (b) p-Tau181_Lilly (c) by Braak staging. Overall group differences were assessed using Mann-Whitney U tests (p-values were corrected for multiple comparisons using FDR). Boxes show interquartile range, the horizontal lines are the medians, and the whiskers are plotted using Tukey method. Abbreviations: p-Tau, phosphorylated tau

Fig. 7

Levels of plasma biomarkers by CERAD classification. Plasma levels of p-Tau217_ALZpath, (a), p-Tau217_Lilly (b) p-Tau181_Lilly (c) by CERAD classification. Group differences were assessed using Kruskal Wallis test and Mann-Whitney U tests (p-values were corrected for multiple comparisons using FDR). Boxes show interquartile range, the horizontal lines are the medians, and the whiskers are plotted using Tukey method. Abbreviations: CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; FDR, false discovery rate; p-Tau, phosphorylated tau

Fig. 8

Predicting neuropathological scales classification. ROC curves analysis for predicting ADNC (a), Braak (b), and CERAD (c) classification. ADNC was dichotomized as none/low or intermediate/high, CERAD was dichotomized as low/sparse, or moderate/frequent and Braak stages were also dichotomized 0-IV or V-VI. The DeLong test was used to determine whether the area under the curve (AUC) of two ROC curves were significantly different (reference biomarker marked as (**); p-value < 0.05). Abbreviations: ADNC, Alzheimer’s disease neuropathologic change; AUC, Area under the curve; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; p-Tau, phosphorylated tau; ROC, receiver-operating characteristic