Association of decreased visibility on deep medullary vein gray-matter volume mediated by increased extracellular fluid in the white matter of patients with cerebral small vessel disease.

Abstract

Background: The visibility and signal continuity of deep medullary veins (DMVs) play an important role in cerebral small vessel disease (CSVD). However, the relationship between DMV and gray-matter atrophy remains unclear. This study sought to investigate the link between DMV scores, extracellular fluid, and gray-matter atrophy in patients with CSVD.

Methods: We reviewed the clinical and multimodal magnetic resonance imaging data from 123 patients diagnosed with CSVD between January and December 2022. The DMV score was assessed using a scoring system (0 to 3 points) based on DMV visibility on susceptibility-weighted images across six anatomical regions, yielding a final score from 0 to 18. Extracellular fluid was assessed through the metric of free water (FW) in normal-appearing white matter (NAWM). Normalized gray-matter volume (GM_N) was used to quantify the gray-matter volume, defined as the ratio of gray-matter volume to intracranial volume. Spearman correlation, general linear model, and mediation analyses were employed to evaluate the relationships among variables.

Results: Spearman correlation analysis revealed a positive correlation between DMV score and FW in NAWM (r=0.603; P<0.001). General linear model analysis confirmed this association as independent [β=0.656, 95% confidence interval (CI) 0.521-0.790; P<0.001]. Conversely, FW in NAWM showed a negative correlation with GM_N (r=-0.485; P<0.001), with an independent association confirmed by general linear model analysis (β=-0.630, 95% CI: -0.769 to -0.491; P<0.001). Additionally, the DMV score was negatively correlated with GM_N (r=-0.390; P<0.001), as supported by a significant association in general linear model analysis (β=-0.502, 95% CI: -0.657 to 0.348; P<0.001). Mediation analysis indicated a significant indirect effect of FW in NAWM on the relationship between DMV score and GM_N (β=-0.346, 95% CI: -0.534 to -0.187; P<0.001). All associations were remained significant after adjustments were made for age, gender, vascular risk factors, normalized white-matter hyperintensity volume, and CSVD burden.

Conclusions: The observed link between DMV disruption and FW in NAWM-GM_N suggests that DMV dysfunction may contribute to gray-matter atrophy in CSVD by increasing extracellular fluid. This identifies DMV changes as a key factor in CSVD pathology and supports the potential of targeting extracellular fluid as a therapeutic strategy to mitigate gray-matter loss.