Abstract 118: Cerebral Venous Reflux And Dilated Basal Ganglia Perivascular Space In Hypertensive Small Vessel Disease

Abstract

Background: Cerebral venous flow alterations have been considered as a contributor in white matter ischemia and cognitive impairment, but their role in small vessel disease has never been investigated. Method: 297 patients with spontaneous hypertensive intracerebral hemorrhage who underwent brain MRI were included. Presence of cerebral venous reflux (CVR) was defined as abnormal signal intensity in the dural venous sinus (cavernous, inferior petrosal, sigmoid or transverse sinus) or internal jugular vein using MR angiography (Figure). Dilated basal ganglia (BG) perivascular spaces (PVS) was defined as the number of PVS >20 or any PVS with size > 3mm in short axis (L-PVS). The association between CVR and dilated BG-PVS were evaluated in univariate and multivariable models. Stroke recurrence was also analyzed in cox regression model. Results: Compared to patients without CVR, patients with CVR (n=37, 12,5%) had similar demographics, but showed a trend toward more severe SVD markers including microbleeds, white matter hyperintensities and lacunes. In addition, presence of CVR is significantly associated with dilated BG-PVS (BG-PVS > 20, 62.2% vs. 34.6%, P = 0.002; L-PVS, 48.6% vs. 18.8%, P < 0.001). In multivariable model, the presence of CVR remained independently associated with dilated BG-PVS (odds ratio 4.03 [1.68-9.68], p=0.002) after adjustment for age and other SVD markers. In cox model, CVR predicts higher risk of stroke recurrence independent of conventional SVD markers (odds ratio 2.32 [1.00-5.34], P = 0.049). Conclusion: We demonstrated a close relationship between CVR and dilated BG PVS, suggesting the potential role of venous drainage dysfunction in the formation of enlargement of PVS in hypertensive SVD.