Screening for cerebral amyloid angiopathy based on serological biomarkers analysis using a dielectrophoretic force-driven biosensor platform

Abstract

We aimed to analyze plasma amyloid-β (Aβ)_1–40 and Aβ_1–42 using a highly sensitive dielectrophoretic-driven biosensor platform to demonstrate the possibility of precise cerebral amyloid angiopathy (CAA) diagnosis in participants classified according to Aβ-positron emission tomography (PET) positivity and the neuroimaging criteria for CAA. We prospectively recruited 25 people with non-Alzheimer's disease (non-AD) and 19 patients with Alzheimer's disease (AD), which were further classified into the CAA− and CAA+ (possible and probable CAA) groups according to the modified Boston criteria. Patients underwent plasma Aβ analysis using a highly sensitive nano-biosensor platform, Aβ-PET scanning, and detailed neuropsychological testing. As a result, the average signal levels of Aβ_{1–42/1–40} differed significantly between the non-AD and AD groups, and the CAA+ group exhibited significantly higher Aβ_1–40 signal levels than the CAA− group in both non-AD and AD groups. The concordance between the Aβ_1–40 signal level and the neuroimaging criteria for CAA was nearly perfect, with areas under the curve of 0.954 (95% confidence interval (CI) 0.856–1.000), 0.969 (0.894–1.000), 0.867 (0.648–1.000), and 1.000 (1.000–1.000) in the non-AD/CAA− vs. non-AD/possible CAA, non-AD/CAA− vs. non-AD/probable CAA, AD/CAA− vs. AD/possible CAA, and AD/CAA− vs. AD/probable CAA groups, respectively. Higher Aβ_1–40 signal levels were significantly associated with the presence of CAA according to regression analyses, and the neuroimaging pattern analysis partly supported this result. Our findings suggest that measuring plasma Aβ_1–40 signal levels using a highly sensitive biosensor platform could be a useful non-invasive CAA diagnostic method.