Abstract Body: Background: Cerebral small vessel disease is characterized on MRI in part by the accumulation of white matter hyperintensities (WMH). Progression of WMH is associated with the development of vascular cognitive impairment and dementia. Disruption of the blood-brain barrier (BBB) has been implicated in the pathogenesis of WMH. We conducted a prospective study to test if BBB disruption (BBBD), detected using dynamic susceptibility contrast (DSC) imaging, would predict progression of WMH over one year.
Methods: These are the primary findings for study NCT03366129 registered on clinicaltrials.gov. Participants with a remote history of stroke (>3 months) were eligible if they had evidence of early confluence of WMH on a FLAIR MRI scan. The primary outcome was based on research MRI scans performed at baseline and one year later. FLAIR scans were segmented to create WMH masks to calculate the WMH percent of total brain volume at each time point. BBBD was measured using DSC imaging on the baseline MRI. WMH masks were co-registered to the BBB imaging and dilated by 3mm to create a second mask of the adjacent normal appearing white matter (penumbra). BBBD was averaged in the WMH and in the penumbra. BBBD and change in WMH volume were compared with linear regression; Shapiro–Wilk test of normality was used to determine normality of residuals, and Spearman’s rank correlation was used when they were not.
Results: Of 81 consecutive participants enrolled in the study, 50 completed the primary endpoint and were included in the analysis (median age 69; 46% women). The mean WMH fraction was 1.25% at baseline and 1.36% at one year. The mean baseline BBBD was 0.20% in the WMH and 0.22% in the penumbra. With linear regression, more severe BBBD was associated with greater WMH progression when measured in the WMH (ß=0.95, 95% CI 0.39 to1.51, r²=0.19, p=0.001) and in the penumbra (ß=0.81, 95% CI 0.10 to 1.53, r²=0.10, p=0.027) as shown in the figure; however, the residuals where not normally distributed for any of the variables. Spearman’s rank correlation found strong correlations between and worse BBBD and greater change in WMH fraction when measured in the WMH (Spearman's rho=0.47, p=0.0005) and the penumbra (Spearman's rho=0.45, p=0.0011).
Conclusions: Greater BBBD measured with DSC MRI was predictive of more WMH progression over the subsequent year in this population of participants with chronic cerebrovascular disease and may be a biomarker of small vessel disease progression.