Basic Science and Pathogenesis

Quantitative susceptibility mapping(QSM) is a MRI processing technique that relies on gradient-echo sequences to quantify the magnetic susceptibility of the brain tissue. QSM has been associated to leakage of contrast-agent in dynamic-contrast-enhanced(DCE) MRI and has thus been proposed as a potential marker of blood-brain-barrier(BBB)-leakage¹. BBB-leakage is likely involved in the early-stages of remodeling of blood vessels with cerebral amyloid angiopathy (CAA)², but its prevalence in the MTL of CAA cases remains unexplored. Furthermore, a pathological validation of QSM as a marker of BBB-leakage is lacking.

Two autopsy-cases from a cohort of donors with a clinical diagnosis of CAA were selected. A block of formalin-fixed brain tissue (∼4x2.5x1cm) from the MTL was scanned at ultra-high resolution 7 Tesla MRI. The protocol included a T2-weighted turbo spin echo(TSE) sequence for anatomical reference and a multi-echo fast-low-angle-shot(FLASH) sequence (voxel-size 100 μm³ isotropic). QSM was calculated from the latter using the open-source toolbox SEPIA³. The tissue-blocks were embedded in paraffin and the posterior part was additionally imaged using micro-computer tomography (CT) (voxel-size 11.86 μm³). Subsequently, 6-μm-thick serial sections were obtained and each seventh section successively stained for H&E, Perls Prussian blue for iron deposits, and van Kossa for calcium. Immunohistochemistry was performed on adjacent sections for amyloid-β and fibrin. Whole-slide sections were imaged with a digital microscope. MRI, micro-CT and histological sections were aligned to each other using anatomical landmarks of the hippocampus adopting the software Imaris⁴.

Regional (hippocampus and rhinal cortex) measures of QSM and deep-learning based measures of iron, calcium, and fibrin deposits obtained through Imaris will be compared to each other. Furthermore, three-dimensional histology will be used to perform a single-vessel analysis and validate QSM as a measure of BBB-leakage by correlating perivascular QSM values with corresponding perivascular fibrin-density. The ex vivo study will be complemented by the analysis of in vivo MRI data acquired in a cohort of CAA cases (n =  14) and control cases (n =  7). Regional MTL QSM values will be compared to k-trans values calculated based on DCE-MRI.

This study will contribute to validate QSM as a potential, non-invasive measure of BBB-leakage.