Imaging And Pathological Study Of Deep Spinal Cord Veins

Part 1.Risk factors of Deep medullary veins abnormality on 3-T Magnetic Resonance Imaging in a population-based StudyBackground and purpose:Pathological and imaging studies indicate an association of pathology of deep medullary veins with aging and neurodegenerative diseases.However,the characteristic of deep medullary venous changes in normal aging population is still unknown.Whether deep medullary venous changes are associated with age,vascular and dementia risk factors,or stroke has not been examined in large sample population.This study is aimed to characterize deep medullary veins on 3T susceptibility weighed imaging in an aging community-based population,and to verify its association with cerebrovascular and dementia risk factors and risk of incident stroke.Methods:Participants completed baseline Magnetic Resonance Imaging examination from the Shunyi cohort were included.A block with a thickness of 6 mm including the body of the corpus callosum was selected.Minimum intensity projection was then computed perpendicular to the block on the susceptibility weighted imaging.Two rectangles(6cm×1cm)parallel to the long axis of lateral ventricle were positioned in each hemisphere,and the deep medullary crossing the rectangle were visually counted,the average of left and right veins was taken as the index of deep medullary venous number.Potential risk factors included age,sex,body mass index,current smoker,alcohol user,hypertension,diabetes,hyperlipidemia,serum homocysteine level,folic acid concentration,vitamin B12 level and C-reactive protein concentration,and APOE4 gene.Univariate and multivariate linear regression models were used to analyze the association between deep medullary venous number and these potential risk factors,and univariate and multivariate Cox regression models utilized to analyze the correlation of deep medullary venous number with risk of incident stroke.Results:A total of 1056 participants were included,with an average age of 55.7 years(standard deviation,9.1).Among those,the median number of deep medullary veins was 19 with a range of 12.5 to 24.The deep medullary venous number decreased with age(P=0.005).The deep medullary venous number was significantly negative associated with C-reactive protein level(P=0.002 after adjusting for age and sex),and positively correlated with folate level(P=0.026 after adjusting for age and sex)Other included risk factors were not associated with the deep medullary venous number(P>0.05 after adjusting for age and sex).As for the risk of incident stroke,during the 3403 person-year of follow-up,15 cases of stroke events occurred,including 13 cases of cerebral infarction and 2 cases of cerebral hemorrhage;there was no significant correlation between the deep medullary venous number and the incident cerebrovascular events,cerebral infarction,or cerebral hemorrhage(uncorrected P values>0.05).Conclusions:The deep medullary venous number decreased with age,but not associated with other classic vascular risk factors or risk of incident stroke.The significant association of deep medullary venous density with serum C-reactive protein level and serum folic acid concentration suggesting cerebral venules might be closely linked to systemic inflammation and nutrient and metabolism disorder,which deserves further investigation.Part 2.The associations between abnormality of deep medullary venous density and cerebral small vessel disease imaging markers.Background and purpose:One challenge in the study of cerebral small vessel disease is the lack of direct marker of cerebral small vessel.A recent imaging study reported a decrease of deep medullary venous density in patients with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy(CADASIL).However,studies are lacking that systemically examined the relationship of deep medullary venous density with small vessel neuroimaging markers in general populations.This study was designed to verify the relationship of periventricular deep medullary venous changes with brain structural imaging markers in a large sample community dwelling population.Methods:Participants were enrolled from an ongoing community-based cohort in a rural area of Beijing.Deep medullary veins were visually counted in the paraventricular region on 3T susceptibility weighted imaging.On brain MRI,lacunes,cerebral microbleeds and dilated perivascular spaces(PVS)were visually evaluated,white matter hyperintensities(WMH)volume were computed from automatically segmentation,and brain atrophy was indicated by brain parenchymal fraction(BPF).Linear regression models were used to analyze the association between the number of deep medullary veins venous and imaging markers.In addition,the Voxel Based Morphology was used to identify the specific regions of cortical atrophy associated with decreased venous number.Results:We included a total of 1056 participants who completed brain MRI with an average age of 55.7 years(standard deviation 9.1).The average of deep medullary venous number was 19(range 12.5 to 24).Of those,14.3%had at least one lacune,11.2%,had at least one cerebral microbleed,14.0%and 14.6%had severe dilated perivascular space in basal ganglia and white matter respectively;the median volume of WMH was 0.8 ml(inter-quantile range 0.24-2.60),and the average of brain parenchymal fraction was 76.5%(standard deviation 3.1).Deep medullary venous number showed a correlation trend with lacunes but not statistically significant(P=0.05),while deep medullary venous number was not associated with WMH volume,cerebral microbleeds or dilated PVS(P>0.05 after adjusting for age and vascular risk factors).The deep medullary venous number was significantly positively correlated with brain parenchymal fraction,whole brain gray matter volume fraction,whole brain white matter volume fraction and hippocampus volume fraction(adjusted vascular risk factors,lacunar and white matter high signal volume P<0.05).Voxel-Based Morphometry analysis showed that the gray matter atrophy significantly associated with decreased deep medullary venous density was mainly in the regions of bilateral hippocampus and thalamus.Conclusions:Deep medullary venous number was significantly associated with brain atrophy,but not independently associated with lacune,WMH volume,cerebral microbleed or enlarged perivascular space,suggesting the potential role played by venules in age-related neurodegenerative process.Longitudinal cohort studies are needed to further verify the relationship between brain deep medullary venous loss and brain atrophy.Part 3.Exploration of brain small vein specific staining methodBackground and purpose:There have been few studies that have investigated venule pathology,this may in part be due to difficulties in differentiating between venules with thickened walls and arteriole hyalinization using routine stains such as hematoxylin and eosin.Recently,animal and cell experiments reported that monocarboxylate transporter-1(MCT-1)is a specific molecular marker of venous endothelium,but no studies have examined this in autopsy brain tissue.Through using MCT-1 immunohistochemistry staining,this study was designed to verify whether brain venules can be specifically stained by MCT-1 in a setting of small vessel pathology,thus establishing a robust method for specific identification of cerebral venules in the future.Methods:The autopsy brain specimens were from Human Brain Tissue Bank of Chinese Academy of Medical Sciences&Peking Union Medical College.We included four autopsy brain specimens with severe AD neuropathological burden from body donors with age of 80-90 years old,and 4 without AD pathological changes from age and sex matched controls.Three regions(anterior,middle and posterior)of periventricular white matter were sampled from the archived formalin-fixed autopsy brain,and 24 tissue blocks were embedded in paraffin,cut into 5-10mm-thick sections,and stained with hematoxylin and eosin staining,Masson's trichrome stains staining,?-actin immunohistochemistry and monocarboxylate transporter-1 immunohistochemistry staining.Results:We observed that the cerebral venules and capillary endothelium were stained by immunohistochemical staining with monocarboxylate transporter-1,while the arterioles were not stained.Also,we found that sclerotic arterioles lack a robust media with collagen replacing the smooth muscle.Conclusion:Monocarboxylate transporter-1 immunohistochemical staining could be a useful staining method to distinguish cerebral venules from arterioles.