Prognostic value of a novel plasma pTau217 assay for amyloid accumulation in healthy elderly

Abstract Background The performance of blood‐based phosphorylated tau (pTau) immunoassays to detect asymptomatic Alzheimer’s disease (AD) has important implications for therapeutic trials. pTau217 is often recommended as the preferred epitope due to its high fold changes in AD. The current study investigates the ability of a novel pTau217 assay to predict the dynamic phase of amyloid‐β (Aβ) accumulation in comparison to the best‐performing pTau181 assay. Methods Plasma pTau217 was quantified by the ALZPath Simoa assay at the University of Gothenburg in 109 cognitively unimpaired older adults (Flemish Prevent‐AD Cohort KU Leuven [F‐PACK] cohort). All subjects underwent baseline plasma sampling and longitudinal Aβ‐PET (median time interval = 6 years). For a subset of 72 subjects, pTau181 was quantified by the Homebrew ADx Simoa assay. Linear mixed‐effects models were used to calculate subject‐specific Aβ change and to assess the age‐, sex and APOE‐corrected predictive value of pTau species for Aβ change, which was assessed in a global and voxelwise manner. Accumulators were defined as having an Aβ rate of change z‐score > 1.5 (based on mean & standard deviation within the Aβ‐ subset). Performance to detect Aβ accumulators was assessed through receiver operating characteristic analyses. Plasma biomarkers were converted to z‐scores for effective comparison. Pearson correlations were calculated between pTau181 and pTau217. Results Plasma pTau217 levels were higher in Aβ accumulators than non‐accumulators with an area under the curve (AUC) of 0.69 (95%CI 0.58‐0.81). Moreover, higher pTau217 predicted steeper Aβ accumulation (β s =0.57, P<.001). Plasma levels of pTau217 strongly correlated with those of pTau181 (r=0.93, Figure 1). In the subgroup with matching pTau181 data, pTau181 was also higher in Aβ accumulators than non‐accumulators with an AUC of 0.68 (95%CI 0.53‐0.84), which was comparable to pTau217 (AUC=0.76; P DeLong =.12) or a multibiomarker model including both pTau species (AUC=0.80; P DeLong =.15) in the same subset (Figure 2A). pTau181 demonstrated comparable associations with Aβ accumulation in terms of both strength (β s =0.53, 95%CI 0.21–0.81 for pTau181 versus β s =0.48 for pTau217 in this subset) and spatial distribution (Figure 2B). Conclusions The novel ALZPath pTau217 assay predicted Aβ accumulation in asymptomatic elderly with comparable performance to the ADx pTau181 assay.