Analysis of Magnetic Resonance Imaging Features of Virchow-robin Spaces Associated With Cerebral Small Vessel Disease

Abstract

Background: A critical component of the radiological diagnosis of cerebral small vascular disease (CSVD) is magnetic resonance imaging (MRI). Subcortical infarcts, microbleeds, lacunes, and white matter hyperintensities (WMH), which are all late signs of the disease, are imaging characteristics that are considered diagnostic. Virchow-Robin spaces have been identified as potential early biomarkers of CSVD. However, there is still a lack of clarity in their association yet this is important in gaining insight into the mechanistic process of cerebral small vascular disorders and the design of potential targeted therapies. Thus, the goal of this study was to examine the MRI characteristics of CSVD-related Virchow-Robin space features. Materials and methods: Magnetic resonance images of study subjects with clinical and supporting radiologic features of CSVD were obtained. Age and sex-matched images, from study subjects with no clinical or radiologic features of CSVD, were used for comparison. Data collected included age, sex, history of alcohol consumption, smoking, medical comorbidities (hypertension and diabetes), absolute count of Virchow-Robin spaces, their anatomic location, length, diameter, area and white matter hyperintensities– using Fazekas scale. Data management: Independent student’s T-test was used to compare age, substance use, morphometric characteristics and the total white matter hyperintensity burden score. The association between sex, comorbidities, substance use and anatomic location were measured using crude and adjusted odds ratio in univariate and multivariate analysis respectively. Receiver operating characteristic analysis was used to obtain the specificity and sensitivity for each variable. Results: In the study, 118 study subjects were included, with 43 (36%) diagnosed with CSVD. The CSVD group was significantly older (55.1±19.0 vs 44.9±16.4 years, p=0.003), and had higher prevalence of hypertension (OR=9.3, 95% CI 9.1-9.3, p<0.001) and diabetes (OR=9.8, 95% CI 7.3-13.3, p<0.001) than the non-CSVD group. CSVD study subjects had a greater number of Virchow-Robin spaces (28.5±11.2 vs 12.2±8.5, p<0.001), higher Potter's rating scale scores (p<0.001), and more frequent type 3 location (p=0.007), periventricular hyperintensities (p=0.005), deep white matter hyperintensities (p=0.001), and total white matter hyperintensity burden score (p=0.001). They also had wider (2.4±0.9 vs 1.7±0.7 mm, p<0.001), longer (2.4±0.8 vs 1.9±0.7 mm, p<0.001), and larger area (18.4±10.9 vs 11.5±6.5 mm2, p<0.001). Conclusion: MRI features associated with CSVD include: a high Potter’s scale, a high frequency of type III Virchow-Robin spaces, larger Virchow-Robin spaces and a high absolute count of Virchow-Robin spaces. The absolute number of Virchow-Robin spaces provided the best prediction model for CSVD, with the highest sensitivity and specificity, in comparison to the routinely used total white matter hyperintensity burden score