Prospective evaluation of plasma phopshorylated tau in real‐life memory clinic in Thailand

Abstract

AbstractBackgroundDespite the substantial accuracy of plasma phosphorylated tau (p‐tau) in diagnosing Alzheimer’s disease (AD) in research cohorts, data on real‐life memory clinic patients are limited. Additionally, previous biomarker studies have often been geographically restricted, raising concerns about generalizability to underrepresented populations. We prospectively assessed the diagnostic accuracy of plasma p‐tau181 in a memory clinic in Thailand.MethodPatients at their early symptomatic stages (i.e. no more than mild dementia) who presented with memory impairment were consecutively enrolled between October 2021 and August 2022. They prospectively underwent routine clinical assessment and plasma p‐tau181 quantification (Simoa). Within 10 weeks, amyloid and tau‐positron emission tomography (PET) were performed using [18F]‐Florbetaben and [18F] PI‐2620, respectively. The diagnostic performance of plasma p‐tau181 (using a predefined cut‐off derived from an independent cohort), neurocognitive specialists and regional tau‐PET (Braak stage III‐IV) were compared head‐to‐head using amyloid‐PET as the reference standard.ResultAmong 51 participants enrolled, 36 (70.6%) were female. The median age was 72 years (interquartile range: 64.5‐76.0). Amyloid‐PET was positive in 43 (70.6%) of the participants. Plasma p‐tau181 had an area under the curve (AUC) of 0.84 (95% confidence interval (CI) 0.73‐0.94), sensitivity of 0.80 (95%CI 0.64‐0.90), specificity of 0.75 (95%CI 0.51‐0.90), and accuracy of 0.78 (95%CI 0.65‐0.88) for detecting amyloid‐PET positivity in early symptomatic patients. In comparison, clinical diagnosis by neurocognitive specialists had an AUC of 0.70 (95%CI 0.56‐0.85), sensitivity of 0.65 (95%CI 0.49‐0.79), specificity of 0.75 (95%CI 0.51‐0.90), and accuracy of 0.69 (95%CI 0.55‐0.80), whereas tau‐PET had an AUC of 0.88 (95%CI 0.79‐0.97). The AUC of p‐tau181 was not significantly different from clinical diagnosis (p = 0.15) or tau‐PET (p = 0.48), however, tau‐PET significantly outperformed clinical diagnosis (p = 0.03).ConclusionThe diagnostic performance of plasma p‐tau181 is substantial in the diverse population of a specialized memory clinic. Its discriminative accuracy was comparable to a second‐generation tau‐PET and showed tendency towards superiority to clinical assessment. The findings of our study may serve as a guide for the cost‐effective implementation of plasma p‐tau into real‐life clinical practice.