Performances of the plasma Aβ biomarker in longitudinal data

Background We recently developed a high‐performance plasma biomarker that reliably predicts individual brain status of amyloid‐β (Aβ) burden using immunoprecipitation‐mass spectrometry (IP‐MS). This study aimed to analyze performances of the plasma Aβ biomarker in a longitudinal dataset using J‐ADNI data. Method There were 165 subjects, including 38 cognitively normal (CN), 61 MCI and 46 AD individuals who participated in the J‐ADNI study, and 132 of them had two times of follow‐up data within 36 months. The plasma composite biomarker was generated by combining plasma ratios of APP669‐711/Aβ1‐42 and Aβ1‐40/Aβ1‐42, which were measured by the IP‐MS assay. Using PiB‐PET as the standard of truth (Aβ+ or Aβ‐), plasma biomarker performances were analyzed using the ROC analysis. Relationships of plasma composite biomarker values and longitudinal changes in Aβ deposition (assessed by PiB‐SUVR), degree of medial temporal atrophy (assessed by VSRAD® z‐scores of structural MRI), and cognitive decline (assessed by cognitive test scores) were analyzed. Result The ROC analyses showed that the composite biomarker performs good both in base line data and follow up data with AUC greater than 90%. There were no significant changes in plasma composite biomarker values across longitudinal observations. The baseline composite biomarker values were significantly correlated with annual rate of Aβ deposition, especially within subjects whose baseline PiB‐SUVR was relatively small (less than 2.0). The baseline composite biomarker values were also significantly correlated with the rate of medial temporal atrophy and the rate of cognitive decline. Conclusion Clinical utility of the plasma Aβ biomarker was validated in J‐ADNI longitudinal data, and it provided information for future disease progression.