Gadolinium- and water-based blood-brain barrier dysfunction measures in patients with sporadic small vessel disease

Subtle blood-brain barrier (BBB) leakage has been detected in small vessel disease (SVD). While established methods rely on gadolinium-based contrast agents (GBCA), diffusion-weighted arterial spin labelling (DW-ASL) is a promising alternative which assesses water exchange rate (k_w ) without injected contrast. However, DW-ASL has not been widely applied in sporadic SVD. We aimed to determine how k_w varied with GBCA BBB leakage measures, baseline and 1-year change in SVD burden.

We recruited patients with mild ischaemic stroke (lacunar or cortical) all characterised for SVD features. We assessed k_w using DW-ASL and GBCA measures of BBB leakage (permeability-surface area product (PS), blood plasma volume and exchange rate of GBCA) using dynamic-contrast enhanced MRI. We used separate linear regression models to assess how k_w varied with GBCA-derived metrics, baseline and 1-year change in WMH volume in co-variate-adjusted analyses.

We included 24 with complete MRI (61±10 years; 71% male). Patients with higher k_w tended to have more severe baseline SVD (e.g. subcortical grey matter (SGM): B=14.59 min^-1/%WMH volume, 95% confidence interval (95%CI)=-1.00,28.18, p=0.04) and greater 1-year increase (B=0.0013 %WMH volume increase/min^-1, 95%CI=-0.0001, 0.0026, p=0.06). We generally found k_w and GBCA BBB leakage measures were not meaningfully associated (e.g. SGM k_w ∼PS: B=-0.23 min^-1/10^-4 min^-1, 95%CI=-7.47, 7.01, p=0.95).

BBB water exchange estimated using DW-ASL tended to be greater with higher WMH burden and progression, suggesting k_w may be a sensitive measure of BBB dysfunction in SVD. However, non-concordance between k_w and GBCA metrics suggests the two methods probe different aspects of BBB function.