Neurology. 2026 Apr 28;106(8):e214859. doi: 10.1212/WNL.0000000000214859. Epub 2026 Apr 7.
ABSTRACT
BACKGROUND AND OBJECTIVES: Despite a well-established amyloid-tau neurodegeneration cascade in Alzheimer disease (AD), the extent and mechanistic relevance of tau pathology in cerebral amyloid angiopathy (CAA), a disease of vascular amyloid deposition, remain unclear. Focusing on patients with CAA without cognitive impairment to minimize concurrent AD pathology, we hypothesized that tau extent would be greater in patients with CAA than in healthy controls (HCs), largely attributable to amyloid burden, but it would not be associated with CAA-related imaging markers or cognitive scores in CAA.
METHODS: This cross-sectional study included patients with CAA and age-matched and sex-matched HCs, none of whom had cognitive impairment or dementia. All participants underwent brain MRI, Pittsburgh compound B (PiB)-PET for amyloid, and 18F-flortaucipir (FTP)-PET for tau. Tau extent was estimated based on PET-Braak staging from cortical FTP uptake and categorized into earlier (Braak 0, I-II) vs later (Braak III-VI) PET-Braak stages. Cognitive scores obtained in CAA were standardized z-scores for memory, processing speed, and executive function. Multivariable regression models tested the association of tau extent with (1) CAA diagnosis and amyloid load in the whole cohort [CAA and HC] and (2) amyloid load, CAA-related imaging markers, and cognitive scores within the CAA cohort.
RESULTS: The mean age of patients with CAA (n = 50) was 70.3 ± 7.6 years and of HCs was 69.7 ± 7.5 years (p = 0.715) (n = 50); both groups were 56% male. Later PET-Braak stage was more frequent in patients with CAA than in HCs (46% vs 18%, p = 0.003). However, in multivariable models, later PET-Braak stage was independently associated with age (odds ratio [OR] 1.19, 95% CI 1.07-1.35, p < 0.001) and PiB uptake (OR 1.72, 95% CI 1.32-2.26, p < 0.001), but not with CAA diagnosis (p = 0.264). Within the CAA cohort, later PET-Braak stage was again independently associated with higher amyloid burden (OR 1.78, 95% CI 1.16-2.73, p = 0.008) but showed no relationship with CAA-related imaging markers or cognitive scores (all p > 0.2).
DISCUSSION: Our results suggest that tau extent is mainly driven by age and vascular amyloid among patients with CAA and that tau extent is not related to CAA-related imaging markers or cognitive scores. These findings further support the view that tau does not affect CAA-specific disease mechanisms among patients with CAA without cognitive impairment.
PMID:41945879 | DOI:10.1212/WNL.0000000000214859