Association between Enlarged Perivascular Spaces and Internal Carotid Artery Stenosis: A Study in Patients Diagnosed by Digital Subtraction Angiography

Abstract

Objective: An enlarged perivascular space (EPVS) is an imaging marker of cerebral small vessel disease, and its relationship with large artery disease is elusive. We investigated the EPVS in patients with internal carotid artery stenosis (ICAS) with the use of digital subtraction angiography (DSA) and tested the relationship between the degree of EPVS and the degree of ICAS; as well as the relationship between the degree of EPVS and white matter hyperintensity (WMH). Method: A total of 202 patients with or without ICAS diagnosed by DSA were enrolled. The maximal ICAS rates, the degrees of EPVS and WMH were measured. The patients’ clinical characteristics and laboratory parameters were recorded. Univariable analysis and multivariable regression were used to test their correlations. In a unilateral stenosis subgroup, the EPVSs in the ipsilateral hemisphere of stenosis and in the contralateral hemisphere were compared. Results: According to univariable analysis, there were significant differences in age (P = .000), Hg1bc (P = .035) and folic acid (P = .008) among the subgroups based on the degrees of EPVS in the basal ganglia (BG). Age (P = .000) and the level of fibrinogen (P = .018) differed statistically among the subgroups based on the degrees of EPVS in the white matter (WM). The correlation between the degrees of WM-EPVS and the ICAS levels was tested with a gamma test: G = .280, P = .001. The ordinal multivariable regression model showed that age was independently associated with both BG-EPVSs and WM-EPVSs. A current smoker status was also independently associated with WM-EPVSs. ICAS level was associated with the severity of WM-EPVSs after adjusting for other risk factors. The degree of BG-EPVS was not correlated with the degree of stenosis. (P = .101). In 59 patients with unilateral ICAS, as tested by the Wilcoxon signed ranks test, the WM-EPVS scores in the ipsilateral hemisphere of stenosis were higher than those in the contralateral hemisphere. (P = .004), but there was no difference in BG-EPVSs (P = .070). Both BG-EPVSs and WM-EPVSs were independently correlated with WMH. Conclusions: BG-EPVSs and WM-EPVSs have different risk factors. WM-EPVSs but not BG-EPVSs are correlated with ICAS.