Polygenic risk score predicts pathologically confirmed cerebral amyloid angiopathy.

INTRODUCTION: Cerebral amyloid angiopathy (CAA) and Alzheimer's disease (AD) both involve amyloid beta (Aβ) accumulation in vessels and brain parenchyma, respectively. As Aβ-targeting therapies emerge, CAA draws attention due to its link with amyloid-related imaging abnormalities (ARIA), underscoring the need for biomarkers beyond magnetic resonance imaging (MRI). METHODS: CAA polygenic risk scores (CAA-PRS) were generated in 105 ADNI participants, and their predictive ability for pathological CAA was evaluated, including in subgroups with high amyloid burden or without MRI-visible CAA markers. RESULTS: CAA-PRS was significantly associated with pathological CAA (OR = 1.766, p < 0.001), with an AUC of 0.783 overall and 0.766 (OR = 1.780, p = 0.002) in those with high amyloid burden. A marginal association was observed in MRI-negative individuals. DISCUSSION: CAA-PRS may serve as a complementary biomarker to imaging for identifying high-risk individuals with CAA, particularly in the context of Aβ-targeting therapies and ARIA risk. HIGHLIGHTS: CAA-PRS is generated. CAA-PRS is associated with pathologically confirmed CAA. CAA-PRS is associated with CAA in individuals with high amyloid burden.