Background: Cerebral amyloid angiopathy-related inflammation (CAA-ri) is associated with a cerebrospinal fluid (CSF) biomarker profile similar to that observed in CAA. Few CAA-ri patients have been studied by fibrillar amyloid-β (Aβ) imaging (using 11C-Pittsburgh compound B and 18F-florbetapir, but not 18F-florbetaben).
Objective: To describe CSF biomarkers, magnetic resonance imaging (MRI), and 18F-florbetaben (FBB)-positron emission tomography (PET) changes in CAA-ri patients.
Methods: CSF levels of total tau, phosphorylated tau, Aβ1-42, and Aβ1-40, MRI (cerebral microbleeds count on susceptibility-weighted imaging and semi-quantitative analysis of fluid-attenuation inversion recovery white matter hyperintensities), and FBB-PET (using both cerebellar cortex and pons to calculate standardized uptake value ratios) were analyzed in nine consecutive CAA-ri patients.
Results: A median number of 769 cerebral microbleeds/patient were counted on MRI. When using the pons as reference region, amyloid load on FBB-PET was very strongly correlated to CSF Aβ1-40 levels (rho = -0.83, p = 0.008) and moderately correlated to cerebral microbleed numbers in the occipital lobes (rho = 0.59, p = 0.001), while comparisons with other CSF biomarkers were not statistically significant (total tau, rho = -0.63, p = 0.076; phosphorylated tau, rho = -0.68, p = 0.05; Aβ1-42, rho = -0.59, p = 0.09). All correlations were weaker, and not statistically significant, when using the cerebellum as reference region. A non-significant correlation (rho = -0.50, p = 0.18) was observed between CSF Aβ1-40 levels and cerebral microbleed numbers.
Conclusion: In CAA-ri, CSF Aβ1-40 levels correlated well with amyloid load assessed by FBB-PET when the pons was used as reference, and to a lesser degree with cerebral microbleeds count on MRI. This confirms earlier data on CSF Aβ1-40 as an in vivo marker for CAA and CAA-ri.
Keywords: Amyloid; florbetaben; imaging.