アルツハイマー病の診断における血漿中pTau₂₁₇/Aβ₄₂比および3つの閾値モデルの診断性能

Early and accurate diagnosis of Alzheimer's disease (AD) typically relies on invasive or expensive methods like cerebrospinal fluid (CSF) biomarkers and amyloid PET imaging. Blood-based biomarkers, particularly plasma phosphorylated tau (pTau₁₈₁, pTau₂₁₇) and amyloid-beta ratios (Aβ₄₂/₄₀), offer a more accessible diagnostic alternative. This study assessed the diagnostic accuracy of plasma biomarkers and developed a three-zone classification model to reduce reliance on invasive confirmatory tests. We retrospectively evaluated 109 participants referred to a tertiary memory clinic. Participants underwent cognitive assessments, brain MRI, CSF biomarker analyses (pTau₁₈₁, Aβ₄₂/₄₀), and plasma biomarker measurements (pTau₁₈₁, pTau₂₁₇, Aβ₄₂/₄₀, pTau₂₁₇/Aβ₄₂ ratio). Diagnostic performance was evaluated using ROC analyses, and thresholds achieving ≥ 95 % sensitivity and specificity were used to define low, intermediate and high-risk zones. Plasma biomarkers correlated significantly with CSF biomarkers. For identifying AD pathology (A+/T + vs. others), plasma pTau₂₁₇ and the pTau₂₁₇/Aβ₄₂ ratio demonstrated the highest accuracy (both AUC=0.95), outperforming plasma pTau₁₈₁ (AUC=0.88) and Aβ₄₂/₄₀ ratio (AUC=0.73). At optimal thresholds, plasma pTau₂₁₇ showed 87.5 % sensitivity and 93.4 % specificity, whereas the pTau₂₁₇/Aβ₄₂ ratio showed higher sensitivity (95.8 %) but lower specificity (85.2 %). Using the three-zone model, plasma pTau₂₁₇ enabled definitive classification in 80.7 % of patients, increasing to 84.4 % with the pTau₂₁₇/Aβ₄₂ ratio. Among patients with mild cognitive impairment, plasma pTau₂₁₇ achieved excellent accuracy (AUC=0.98). Plasma pTau₂₁₇, alone or combined with Aβ₄₂, provides highly accurate and scalable identification of AD pathology, substantially reducing the need for invasive diagnostic procedures.