Caliber of Intracranial Arteries as a Marker for Cerebral Small Vessel Disease

Background: The dilation of intracranial large arteries caliber, may transfer more hemodynamic burden to the downstream brain capillaries, which, in the long run, results in cerebral small vessel disease (CSVD). This study aimed to investigate the relationship between intracranial artery calibers and small vessel disease. Methods: Patients with first-ever ischemic stroke of lacunar infarction subtype were enrolled via Nanjing Stroke Registry Program. An intracranial arterial Z-score, named the brain arterial remodeling (BAR) score, was calculated by averaging the calibers of the seven main intracranial arteries. Among the enrolled patients, those with a BAR score < −1 SD were deemed to have small intracranial artery calibers; those with a BAR score >1 SD were deemed to have large intracranial artery calibers and those with a between BAR score were deemed to have normal intracranial artery calibers. Imaging markers of CSVD, including lacuna, white matter hyperintensity (WMH), enlarged perivascular spaces (EPVS) and cerebral microbleeds (CMBs) were rated and then summed to obtain a total CSVD score. Results: A total of 312 patients were involved in this study, patients with BAR score >1 SD were older (P = 0.039), and more prone to having a history of myocardial infarction (P = 0.033). The Spearman's rank correlation coefficient between the BAR score and total CSVD score is 0.320 (P < 0.001). Binary logistic regression found that BAR score >1 SD was correlated with lacuna (OR = 1.987; 95% CI, 1.037–3.807; P = 0.039); severe WMH (OR = 1.994; 95% CI, 1.003–3.964; P = 0.049); severe EPVS (OR = 2.544; 95% CI, 1.299–4.983; P = 0.006) and CSVD (OR = 2.997; 95% CI 1.182–7.599; P = 0.021). Ordinal logistic regression analysis found that age (OR = 1.028; 95% CI, 1.007–1.049; P = 0.009), hypertension (OR = 3.514; 95% CI, 2.114–5.769; P < 0.001) and BAR score >1 SD (OR = 2.418; 95% CI, 1.350–4.330; P = 0.003) were correlated with the total CSVD score. Conclusions: Patients with large intracranial arterial calibers may have heavier CSVD burden. The mechanisms of this association warrant further study.


INTRODUCTION
A previous study suggested extreme brain arterial diameters correlated with vascular death, myocardial infarction, and any vascular event (1). Our previous study also found that a dilated basilar artery correlated with stroke recurrence (2). As the intermediary arteries connect extracranial arteries and cerebral small vessels, intracranial large arteries dampen the systematic pressure and pulsatility that are transmitted to brain capillaries (3). The dilation of arterial caliber may reduce the capability of physiological cerebral autoregulation and cause end-organ damage, such as cerebral small vessel disease (CSVD).
A previous study found that a larger carotid lumen diameter (but not common carotid artery intima-media thickness) was associated with a higher prevalence of lacunar infarcts (LI) (4). In addition, carotid stiffness was associated with increased white matter hyperintensity (WMH) volume which was independent of carotid plaque (4). Middle cerebral artery (MCA) diameter, as a surrogate of stiffness, is associated with anterior enlarged perivascular spaces (EPVS), and the association is the strongest among individuals with dilated brain arteries (5). Additionally, extreme intracranial arterial enlargement, in some cases called intracranial arterial dolichoectasia, also correlated with lacuna, WMH, EPVS (6), and cerebral microbleeds (CMBs) (7). These findings suggested that intermediary arteries could modify the association of extracranial pressure and CSVD, as observed in LI, WMH, and EPVS.
However, the intracranial arterial caliber and its relationship with the total burden of CSVD remain unknown. We hypothesized that the brain capillaries of patients with large intracranial arterial calibers may be exposed to a more severe hemodynamic burden and may suffer a heavier burden of CSVD.

METHOD Patients
Consecutive patients with first-ever lacuna stroke proven by magnetic resonance imaging (MRI) were retrieved from Nanjing Stroke Registry (8) from September 1, 2015, through August 31, 2016. We excluded patients with symptomatic large artery stenosis (≥50%) and patients with possible cardioembolic sources such as atrial fibrillation, valvular heart disease or cardiac valve replacement. We also excluded patients if they had no brain MRI source images which were used for the measurement of intracranial artery diameters.

Vascular Risk Factors
Baseline characteristics, vascular risk factors, laboratory data, and medical documents were retrieved. Vascular risk factors, including hypertension (defined as a history of hypertension or diagnosed at discharge), diabetes (defined as a history of diabetes or diagnosed at discharge), hyperlipidemia (defined as a history of hyperlipidemia or received lipid-lowering treatments or diagnosed at discharge), history of myocardial infarction and smoking, were carefully identified according to our previous study (2).

Neuroimaging Examinations
A brain MRI examination was performed in either a 3.0-T (Magentom Trio, Siemens, Erlangen, Germany) or 1.5-T (GE Medical Systems, Milwaukee, WI) system to obtain axial T1weighted images, axial T2-weighted images, axial diffusionweighted imaging (DWI), fluid-attenuated inversion recovery (FLAIR), 3D time-of-flight MRA and gradient-echo T2 *weighted or susceptibility-weighted imaging (SWI) images. All MRI source images were evaluated by two neurologists (CZ and LH) who were blinded to the clinical information.

Intracranial Arterial Diameter Measurements and Brain Arterial Remodeling (BAR) Score
Diameters of the main seven intracranial arteries (Figures 1A-C), including the internal carotid arteries (ICA) (R and L) at the intra-cavernous segment, the MCA (R and L) at the M1 segment, the basilar artery (BA) at the mid-pons (Figures 1D-F) and the intracranial vertebral arteries (VA) (R and L) at the V4 segment, were measured according to our previous study (2). An average arterial Z-score, also named the brain arterial remodeling (BAR) score, for each individual was obtained by adding all measured arteries and dividing by the total number of identified arteries as described in a previous study (1). The BAR score was normalized and then used both continuously and categorically, using three categories: (1) "individuals with large diameters" for participants with a BAR score>1 SD, (2) "individuals with small diameters" for participants with a BAR score <-1 SD, and (3) "individuals with average diameters" for participants with a BAR score between −1 and 1 SD. Within each individual, the greater the number of arteries with negative scores, the more likely it was that the BAR score had a negative overall Z-score and vice versa ( Figure 1G).

Definition of the Total CSVD Score
The total CSVD score including 4 MRI markers of CSVD (lacuna, WMH, EPVS, and CMBs) was defined according to previous studies (9,10). Briefly, a lacuna was described as a round or ovoid hyperintense lesion on T2-weighted images, 3 and 15 mm in diameter, with a surrounding rim of hyperintensity on FLAIR but negative on DWI (11). WMH rated on FLAIR images was described using the modified Fazekas score (12), a periventricular WMH Fazekas score of 3 or a deep WMH Fazekas score of 2 or 3 was defined as severe WMH. EPVS in basal ganglia and centrum semiovale regions were rated. They were defined as small (<3 mm), punctate or linearshaped lesions with a cerebrospinal fluid-like signal on all MRI sequences but without a hyperintense rim on T2-FLAIR (11). The number of EPVS was rated as follows: 0 to 10 EPVS (mild); 11 to 25 EPVS (moderate); and >25 EPVS (severe) in both anatomic areas. CMBs were defined on gradient-echo T2 * or SWI as small (<10 mm), homogenous, round, lowsignal intensities (13). An ordinal total CSVD score ranging from 0 to 4 was calculated by counting the above 4 MRI features. One point was awarded for each of the following items: ≥1 asymptomatic lacuna (1 point if present); periventricular

Statistical Analyses
Continuous variables are presented as the mean ± SD. Categorical variables were recorded as proportions. Betweengroup comparisons of the distribution of continuous variables were performed using a one-way ANOVA or independent samples t-test. Comparisons of categorical variables were performed using the χ 2 -test or Fisher's exact test. We first investigated whether the demographic and vascular risk factors varied across the 3 remodeling groups. We then assessed the relationship between the BAR score and all CSVD signs with binary logistic regression. The correlation between the BAR score and total CSVD score was measured with the Spearman's rank correlation coefficient method. The relationship between the BAR score and the total CSVD score was evaluated with ordinal logistic regression analysis. Age, sex, and risk factors with a P-value of < 0.1 in the univariate analysis were included in multivariate analysis. All statistical testing was twotailed, and P < 0.05 was considered statistically significant. All analyses were performed with IBM SPSS Statistics 25.0 (IBM, Armonk, NY).

Sample Description
A total of 312 patients were involved in this study, with a mean age of 59.9 ± 11.1 years, and 73.4% were men. The mean body surface area was 1

Baseline Characteristics of the Participants According to BAR Category
Forty-five patients had small brain arterial diameters (BAR < −1 SD), 47 patients had large brain arterial diameters (BAR > 1 SD) and 220 patients had average diameters (−1 SD ≤ BAR ≤ 1 SD). Patients with large arterial diameters had older ages (P = 0.039) and more previous myocardial infarctions than other patients (P = 0.033). The hypertension rate was increased gradually according to the increase in diameter (51.1%, 63.6%, and 85.1% for BAR <-1 SD, −1 SD ≤ BAR ≤ 1 SD, and BAR >1 SD, respectively); however, it was not statistically significant (P = 0.647). We did not find a significant difference according to the gradual increase in intracranial arterial diameter concerning sex (P = 0.270), diabetes mellitus  Table 2). The percentage of total CSVD score = 0 declined gradually with increasing BAR score (Supplementary Figure 1) and the Spearman's rank correlation coefficient between BAR score and total burden of CSVD was 0.320 (P < 0.001).

DISCUSSION
A previous study has suggested that the incidence rates for death, vascular death, myocardial infarction, and any vascular event were higher in individuals with the largest arterial diameters (1). The hemodynamic burden may be transmitted through an intracranial large artery to the downstream small vessels (3) and then cause small vessel disease. In this study, we provided evidence that large brain artery caliber is associated with the total CSVD score, as well as lacuna, WMH, and EPVS. Our study found that large arterial diameter was correlated with age and myocardial infarction, but not with other atherosclerosis risk factors, such as diabetes, smoking, and hyperlipidemia. Previous studies also found that extreme intracranial arterial outward remodeling, called dolichoectasia, was also correlated with myocardial infarction (14), coronary arterial ectasia (15), and an enlarged descending thoracic aorta (16). This finding may indicate that brain arterial outward remodeling may be a biomarker of systemic arterial stiffness (17).
CSVD is a dynamic, whole-brain disorder and a common cause of dementia, stroke, and gait disturbances. Previous  studies suggested a strong relationship between large intracranial arterial diameter and lacuna infarction (4), WMH (18), EPVS (5), and CMBs (7). However, previous studies only studied one or two large brain arteries, and selected separated CSVD indicators. It has been widely accepted that the total CSVD score is a more complete overall gauge of the impact of CSVD on the brain than are the individual MRI features separately.
Our study found that large intracranial arterial diameter, as evaluated by the BAR score, was correlated with total CSVD score, which has seldom been reported to our knowledge. These findings suggested that the large intracranial arteries would be less able to dampen pressure and pulsatility, leading to more pulsatile energy dissipation in the brain and end-organ tissue damage (19). A previous study found a positive correlation between large intracranial artiries outward remodeling and the severity of MRI markers of small vessel disease (18). which implies that the long-term hemodynamic burden caused by vascular wall remodeling may play an important role in the development of both large arteriopathy and small vessel disease.
Our study has some limitations. First, we did not calculate cerebral vessel pulsatility, as previous studies have shown that high pulsatility in the ICA or MCA is associated with WMH (3). Cerebral veins and CSF are also thought to be important compartments for compensating arterial pulse pressure (20). Second, we did not evaluate the MMP level, as a previous study suggested that MMP dysfunction could bridge large intracranial arterial outward remodeling and CSVD (21). Third, the measurement of intracranial artery calibers is affected by cardiac cycle and artery pulse and it might depend on the imaging time on MRA. For example, the average distension of the MCA area from diastole to systole was 2.58%. However, the phasecontrast flow velocity profiles found no significant correlation between MCA distension and the pulsatility index (22). Four, we do not analyze the brain arterial wall by HRMRI, a fact that weakens our claim that the diameters phenotypes represent remodeling phenotypes. In summary, intracranial artery caliber is a biomarker for CSVD, and individuals with large arterial diameters have a greater risk of CSVD. The relationship between cerebral blood flow, cerebral vessel pulsatility, and CSVD needs further study.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the ethics committee of Jinling Hospital. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
ZC and GX: study design, interpretation of results and manuscript drafting. MW and CC: study design and interpretation of results. HL: data collection. XF: study design and statistical analysis. XL and GX: study design, statistical analysis and critical revision of manuscript. GX and ZC: have full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors: contributed to the article and approved the submitted version.