Dawson's Fingers in Cerebral Small Vessel Disease

To explore Dawson's fingers in cerebral small vessel disease (CSVD) and factors related to the development of Dawson's finger, we collected and analyzed clinical data of 65 patients with CSVD. We found a venous abnormality feature called Dawson's fingers around the ventricles in magnetic resonance images (MRIs) of 20 out of 65 patients with CSVD (30. 8%). A significant association between Dawson's fingers and diabetes mellitus (DM) was also detected (30 vs. 8.9%, P < 0.05). CSVD patients with Dawson's fingers had significantly increased cerebral microbleeds (CMB) (44.2 vs. 75.0%, p < 0.05), lacunae (66.7 vs. 95.0%, p < 0.05), and white matter hyperintensity (WMH) (p < 0.05) damage, and these patients exhibited significant cognitive domain impairment as assessed via Montreal Cognitive Assessment (MoCA) (18.9 ± 1.8 vs. 24.0 ± 0.8, p < 0.05) and Mini-Mental State Examination (MMSE) (24.5 ± 1.1 vs. 26.6 ± 0.6, p < 0.05). Our results show a distinctly high incidence of Dawson's fingers in CSVD patients and identify a significant association with DM, thus yielding insights about the appropriate use of Dawson's fingers, a venous imaging marker, to explore the basic pathophysiology of CSVD.

To explore Dawson's fingers in cerebral small vessel disease (CSVD) and factors related to the development of Dawson's finger, we collected and analyzed clinical data of 65 patients with CSVD. We found a venous abnormality feature called Dawson's fingers around the ventricles in magnetic resonance images (MRIs) of 20 out of 65 patients with CSVD (30. 8%). A significant association between Dawson's fingers and diabetes mellitus (DM) was also detected (30 vs. 8.9%, P < 0.05). CSVD patients with Dawson's fingers had significantly increased cerebral microbleeds (CMB) (44.2 vs. 75.0%, p < 0.05), lacunae (66.7 vs. 95.0%, p < 0.05), and white matter hyperintensity (WMH) (p < 0.05) damage, and these patients exhibited significant cognitive domain impairment as assessed via Montreal Cognitive Assessment (MoCA) (18.9 ± 1.8 vs. 24.0 ± 0.8, p < 0.05) and Mini-Mental State Examination (MMSE) (24.5 ± 1.1 vs. 26.6 ± 0.6, p < 0.05). Our results show a distinctly high incidence of Dawson's fingers in CSVD patients and identify a significant association with DM, thus yielding insights about the appropriate use of Dawson's fingers, a venous imaging marker, to explore the basic pathophysiology of CSVD.

INTRODUCTION
Cerebral small vessel disease (CSVD) refers to a broad class of cerebrovascular diseases that affect the small arterioles, venules, and capillaries with a variety of different causes (1,2). Age-related and hypertension-related CSVD are the most common forms (3,4). Despite its diverse pathogenesis modes, the consequences of CSVD on brain parenchyma elicit similar neuroimaging markers, including lacunes, white matter hyperintensity (WMH), enlarged perivascular space (EPVS), and cerebral microbleeds (CMB) (2). In general, small vessels cannot be visualized in computed tomography angiography (CTA) or magnetic resonance angiography (MRA), but the corresponding parenchymal lesions of small vessels can be captured on magnetic resonance imaging (MRI). The STandards for Reporting Vascular changes on nEuroimaging (STRIVE) is an international neuroimaging standard that classifies and defines CSVD markers (5). For the diagnosis of CSVD, MRI is more specific than clinical criteria. The characteristics of neuroimaging markers are indicative of etiological analysis (5). Notably, arteriolosclerosis is associated with the pathogenesis of sporadic CSVD, yet relatively little attention has been focused on potential roles of venules in CSVD. Thus, the role of venules in the pathogenesis of CSVD, as well as the potential for use of venule abnormalities as imaging markers for diagnosing CSVD, remain unclear.
Central vein sign (CVS), a vein positioned centrally in brain white matter lesions (WMLs), has been considered to be an imaging marker for multiple sclerosis (MS) (6-8), a chronic demyelination and neurodegenerative disease caused by central nervous system (CNS) inflammation. Some studies show that CVS can distinguish MS from mimic diseases (9,10) including CSVD (11)(12)(13)(14)(15)(16). Dawson's fingers are elongated, flameshaped lesions perpendicular to the lateral ventricle wall on fluid attenuated inversion recovery (FLAIR)/T2 weighted images. T2-weighted/FLAIR MRI allows visibility of venule within WMLs, thus representing inflammatory activity surrounding venule (17). It makes sense that Dawson's fingers are widely used as important imaging markers for the diagnosis and differential diagnosis of MS (18,19). Complicating matters, CSVD shares some imaging features with MS, including white matter demyelination and brain atrophy. Demyelination around the ventricle is a useful imaging marker for distinguishing MS from other diseases (18). Dawson's fingers, indicating perivascular inflammation around veins and venules, has close pathological association with CVS. Previous studies showed Dawson's fingers can distinguish MS from neuromyelitis optica spectrum disorder (18,20) and MOG antibody disease (20), but the condition of Dawson's fingers in CSVD is still unknown.
This study aimed to (1) determine the proportion of Dawson's fingers in CSVD; (2) explore the factors contributing to Dawson's fingers in patients with CSVD; (3) determine the imaging and clinical characteristic of CSVD patients with Dawson's fingers; and (4) explore if Dawson's fingers can be used as a specific sign to distinguish MS from CSVD.

Patients and Population
This was a prospective observational study. Patients were consecutively recruited from Tianjin Medical University General Hospital and Beijing Tiantan Hospital. Only patients with clinically manifested CSVD and at least one of the imaging features (lacuna, WMH, EPVS, and CMBs) were eligible to participate in this study. MS was diagnosed using 2017 McDonald criteria. Main exclusion criteria were patients with (1) unidentified vascular risk factors for CSVD such as a history of hypertension, hypercholesterolemia, diabetes mellitus (DM), smoking, and alcohol consumption; (2) evidence of other diseases that could potentially also cause WMLs; and (3) contraindications to MRI scanning.

Standard Protocol Approvals, Registrations, and Patient Consents
The Ethics Committees of Tianjin Medical University and Beijing Tiantan Hospital approved this study. We obtained informed consent from all the participants.

Image Acquisition Protocol
All brain MRI was performed on a 3.0-T scanner (Magnetom Trio Tim; Siemens). The routing sequences included T2, FLAIR,  Dawson's finger, an elongated, flame-shaped WMH lesion perpendicular to the lateral ventricle wall, was here defined based on T2/FLAIR images. All assessments were executed independently by two trained neurologists who were blinded to the clinical details. The interclass correlation coefficient was 0.86. If there was a discrepancy between the two raters, assessment from a third rater was used.

Clinical Assessment
Each patient underwent clinical assessment by two neurologists. The demographic factors, clinical factors, and vascular risk factors evaluated included the following: age, sex, hypertension

Statistical Analysis
Continuous variables with normal distributions, such as age and lesion volumes, are presented as the mean ± standard error of mean. All continuous variables were compared using the Student's t-tests for independent samples. Those with nonnormal distributions are presented as medians, and interquartile range (IQR) for non-parametric data are quoted. In the univariable analyses, we used χ 2 -tests for lacunes, CMBs, WMH, and EPVS (Fisher's exact test when the expected value was <5) between patients with or without Dawson's fingers. For variables that were non-normally distributed, Mann-Whitney or the Kruskal-Wallis tests were used, as appropriate, or the data were log transformed for parametric testing. ROC curves were used to evaluate the diagnostic value of Dawson's fingers. Statistical Product and Service Solutions (SPSS) for Windows version 17.0 software (Chicago, IL, USA) was used for the analyses. Values of p < 0.05 were considered statistically significant.

Data Availability
Data supporting the findings of this study are available from the corresponding author upon request.

RESULT Dawson's Fingers Appear in Patients With CSVD
Most often, CSVD is considered to be an arterial-related disease based only on radiological phenotypes; signs of CSVD on conventional MRI markers include CMB, lacunes, WMH, and moderate to severe EPVS (5). However, minimal attention has been paid to venous abnormalities in patients with CSVD. To verify the condition of venous vessels in patients with CSVD, we assessed the appearance of Dawson's fingers among 65 patients with CSVD. The baseline characteristics of the CSVD and MS groups are displayed in Table 1. We found a venous abnormality feature identified as Dawson's fingers around the ventricles in the MRI of CSVD patients (Figure 1). Dawson's fingers were present in 20 of 65 patients with CSVD (30.8%). There were, however, more patients with MS (32/46) than CSVD (20/65) that had Dawson's fingers (69.6 vs. 30.8%, P < 0.001); thus, Dawson's fingers can still be used as a neuroimaging marker to distinguish between CSVD and MS. The area under the ROC curve for Dawson's fingers was 0.694 (P = 0.001) (sensitivity 71% and specificity 69%).

Demographics and Vascular Risk Factors in Patients With and Without Dawson's Fingers
To explore the factors contributing to Dawson's fingers in CSVD, we performed demographic and vascular risk factor analysis of CSVD patients with and without Dawson's fingers. With the exception of patients with DM, these analyses did not identify differences between CSVD patients with Dawson's fingers compared to patients without Dawson's fingers in demographic factors, number of attacks, or disease duration, or with other vascular risk factors (e.g., hypertension, hyperlipidemia). There was a statistically significant difference between DM status and CSVD patients with (6/20) or without (4/45) Dawson's fingers (30 vs. 8.9%, P < 0.05, Table 2). The utility of Dawson's fingers to distinguish between CSVD from MS (32/46) decreased in CSVD patients with DM (6/10) (AUC = 0.454; P > 0.05).

Dawson's Fingers Affect Cognition and Imaging in Patients With CSVD
Dawson's fingers appear in patients with CSVD and are related to DM. To observe the impact of Dawson's fingers on clinical function and imaging of CSVD patients, we carefully analyzed imaging for all patients. WMH, lacunes, and CMB have a statistically significant difference between CSVD with and without Dawson's fingers, especially mixed lacunes (p < 0.001, Table 3). Then, we analyzed the function of CSVD with and without Dawson's fingers. No differences in mRS (1.3 ± 0.1 vs. 1.2 ± 0.1, p = 0.244) were observed, however, MoCA (18.9 ± 1.8 vs. 24.0 ± 0.8, p < 0.05) and MMSE (24.5 ± 1.1 vs. 26.6 ± 0.6, p < 0.05) were significantly different between CSVD with and without Dawson's fingers (Figure 2). Therefore, the cognitive level of CSVD patients with Dawson's fingers is significantly decreased compared to CSVD patients where Dawson's fingers were not detected.

DISCUSSION
CSVD mainly refers to a group of diseases in which arterioles are damaged; whether there is venous injury has been controversial. Dawson's fingers are an important imaging marker in the diagnosis of MS. To determine whether there is a venous abnormality in CSVD, we collected and analyzed the clinical data of small arteriosclerotic CSVD and found a venous abnormality feature (Dawson's fingers) around the ventricles in the MRI of 30.8% of patients. Statistical analysis revealed an association between Dawson's fingers and DM in the CSVD patients of our cohort. Moreover, we found that Dawson's fingers are associated with decreased cognitive ability. CSVD is typically considered to be a small artery systemrelated disease (24). The most prevalent etiological type for sporadic CSVD is arteriolosclerosis caused by aging or vascular risk factors (1,25). The pathological feature of arteriosclerosis is narrowing of the lumen and thickening of the vessel wall, and atherosclerosis has a strong relationship with the clinical symptoms of CSVD such as subsequent cognitive decline and neuroimaging features, e.g., CMB (26). Further, previous work has established that high blood pressure is as an independent risk factor for CSVD (27). It is notable that relatively few studies of CSVD have focused on veins. CVS, as a venous sign, is a common imaging feature identified with MS, which is only present in 6-7% proportion of CSVD patients. However, Dawson's fingers, another venous sign, were present in more than 30% of patients with CSVD in this study. These results clearly indicate that venous problems can occur amongst CSVD patients, thus supporting an important role for venous problems in the pathogenesis of CSVD. The pathogenesis of Dawson's fingers corresponds to the scope of perivenous inflammation, which is perpendicular to lateral ventricles along subependymal veins (28). As noted, Dawson's fingers can be used as an imaging marker in differential diagnosis of MS from its mimics. Additionally, it has been reported that Dawson's fingers can also help differentiate other autoimmune demyelinating diseases from MS (18,29,30). However, a common challenge in the clinic is difficulties in distinguishing MS from CSVD on the basis of MRI appearance. For example, in middle-aged or elderly patients, CSVD has a high prevalence of leukoencephalopathy (31), so a variety of imaging markers are used to distinguish MS from CSVD. In our study, 20 of 65 CSVD patients (30.8%) had Dawson's fingers, whereas 32 of 46 MS patients (69.6%) had Dawson's fingers, and the difference in proportions between MS and CSVD groups was significant. Thus, despite its apparently abnormally high prevalence in CSVD patients in our cohort, Dawson's fingers can still be used to distinguish between MS and CSVD. However, our data also shows that the utility of this marker decreased in patients with DM. Therefore, careful consideration of this and other imaging markers for the differential diagnosis of MS and CSVD, and especially in patients with DM, merits attention.
DM can contribute to the development of CSVD by initiating and accelerating the development of arteriolosclerosis (26). It is known that CSVD patients with DM exhibit CMB, WMH, and lacunes more commonly than do CSVD patients without DM (32). It is also long been known that DM can cause venous damage (33). Our study showed significantly higher prevalence of Dawson's fingers in CSVD patients with DM. This result suggested that DM may be associated with the occurrence of venous problems in CSVD, but this needs to be confirmed in a large longitudinal study.
Our study has some limitations, including its nature as an explorative study with a relatively small patient cohort. Therefore, we encourage researchers to use caution in directly extrapolating our results to other research; future studies should include additional participants. In addition, this study did not conduct any follow-up scientific explorations of the potential pathogenic mechanism of Dawson's fingers in CSVD patients. We did rigorously evaluate the clinical data of all CSVD patients. Our work also therefore suggests that future examinations of the pathophysiology of CSVD should pay attention to venous abnormalities and should certainly consider potential impacts from diabetes.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committees of Tianjin Medical University and Beijing Tiantan Hospital. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

AUTHOR CONTRIBUTIONS
YF conceptualized the study, acquired funding for this study, and designed the study. YF, ZZ, and LY collected the data. LY, YY, WZ, and AL analyzed the data. YF, WZ, and AL wrote the manuscript. All authors contributed to the article and approved the submitted version.