Evaluation of Vascular Permeability in Alzheimer's Disease using Dynamic Susceptibility Contrast Magnetic Resonance Imaging (DSC-MRI)

Abstract

Background Although amyloid plaques and neurofibrillary tangles have been considered main pathology of Alzheimer’s disease (AD), therapeutic trials targeting these molecules have not been successful. Recently, vascular contributions to Alzheimer’s disease (AD) are increasingly reported, suggesting that the integrity of blood-brain barrier (BBB) may play a critical role in disease mechanism. Detecting subtle BBB leakage in AD have been challenging, but recently several studies showed evaluation of BBB disruption with perfusion MRI. Dynamic susceptibility contrast (DSC) perfusion imaging is applied to evaluate cerebral blood volume in tumors and strokes in the brain, but has not applied to detect the disintegrity of BBB, especially in AD. Therefore, this study aimed to investigate the preclinical and clinical applicabilities of DSC-MRI to detect BBB disruption in AD-model mouse and AD. Methods For the preclinical study, DSC-MRI was obtained in two strains of Alzheimer transgenic mice (4 to 20 months old), which were 13 5xFAD mice (8 with transgenic mice, 5 with wild-type mice) and 8 Tg2576 mice (4 with transgenic, 5 with wild-type) and their age-matched wild-type mice using Gadolinium contrast injection. Cerebral blood volume (CBV), cerebral blood flow (CBF), extraction fraction and leakage parameters were calculated and analyzed. Group comparison with region-of-interest (ROI) and voxel-based analysis were performed to compare the difference of permeability index in two groups. After MR imaging, mice were sacrificed and immunohistology of their brains were evaluated to compare the extent of amyloid plaques and endothelium structure. For the clinical study, DSC-MRI was also obtained in 10 AD patients and 13 age-matched controls for the pilot study. Perfusion indices were calculated to compare between the two groups with voxel-based and ROI-based analyses. Results For the animal study, the ROI-based quantitative measurements of DSC map showed higher permeability indices in the hippocampus of 5xFAD mice compared with those of WT mice. VBM analysis of leakage maps revealed significant differences in left frontal cortex and dorsal hippocampus between two groups (p=0.01). Temporal increase of leakage was more prominent in AD group than in control group (p=0.042). For the clinical study, in AD group, leakage value increased at right putamen (p= 0.012), thalamus (p= 0.035) and extraction fraction also increased at left corpus callosum (p< 0.001), right cingulate gyrus (p= 0.004) and parahippocampal gyrus (p= 0.016) compared with the control group. Conclusions DSC-MRI revealed BBB disruption in AD mouse models, and most frequently affected locations were left frontal cortex and hippocampus. Human pilot study also showed higher permeability index in several gray matter and white matter regions of AD patients, indicating that DSC-MRI also detects BBB disruptions in a human study. Our findings suggest that abnormalities in DSC-MRI may be a feasible biomarker to detect impaired BBB function in AD patients, which should be explored in further clinical studies.