 |
|
||||
| Home | Login | |||||
| About | Editorial board | Articles
|
NSI Publications
|
Search | Instructions | Online Submission | Subscribe | Videos | Etcetera | Contact |
|
||||||||||||||||||||
|
|
|
Pattern and Severity of Leukoaraiosis and Microvascular Resistance- Inputs from a TCD Study from South Asia
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.344637
Keywords: Leukoaraiosis, microvascular resistance, pulsatility Index, TCD
Leukoaraiosis is a pathological condition, predominantly of the brain white matter, thought to be related to long standing hypoxic ischemic insult due to microcirculatory disturbances. Leukoaraiosis is prevalent among the elderly population and can lead to strokes, cognitive and gait disturbances, depression and pseudobulbar manifestations. It can predispose to recurrent strokes and poor stroke outcomes.[1] In addition to leukoaraiosis, the spectrum of cerebral small vessel disease (cSVD) includes subcortical infracts, lacunes, prominent perivascular spaces, cerebral microbleeds, and brain atrophy. These changes are postulated to be related to microangiopathic changes.[1] Impaired perfusion, endothelial dysfunction, blood brain barrier (BBB) disruption, etc., have been implicated as the cause of leukoaraiosis. Leakage of fluid and plasma proteins could lead to progressive lipohyalinosis and induce inflammation in the vicinity of cerebral small vessels leading to progression of the white matter lesions. The underlying pathology of leukoaraiosis is a combination of demyelination, loss of oligodendrocytes, and axonal damage. In imaging modalities like MRI, leukoaraiosis is usually symmetrically and bilaterally distributed in the hemispheric white matter (periventricular or deep). It can also occur in the deep gray matter and brain stem including the pons. The severity of leukoaraiosis can be assessed using both CT and MRI.[2] However, MRI is preferred since it has higher sensitivity and specificity for detecting parenchymal changes related to small vessel disease.[3] In MRI, leukoaraiosis is seen as bilateral, patchy, or diffuse areas of hyperintensities on T2-weighted or FLAIR sequences involving the periventricular and centrum semiovale (deep) white matter. The severity of leukoaraiosis can be rated on axial FLAIR MRI using the Fazeka's scale.[4] Unlike structural imaging, there is no optimal investigative modality to understand the hemodynamic changes coexisting with cSVD. Recently, transcranial ultrasound is increasingly being used to evaluate the hemodynamic changes of the cerebral vessels including microcirculatory disturbances.[5] Transcranial Doppler ultrasonography (TCD) is a promising investigative tool which is non-invasive, inexpensive, and mobile without having hazard of radiation exposure. It can indirectly evaluate the functional status of small vessels with a good inter-observer reliability.[6] It provides physiologic data such as mean velocity, peak systolic velocity, and end diastolic velocity from intracranial vessels. Among the various TCD parameters, the Gosling's Pulsatility Index (PI) described by Gosling and King[7],[8] was found to be the most useful parameter for evaluation of downstream resistance in cerebral vascular bed. It reflects the microvascular resistance and compliance of the small vessels.[9],[10] Cerebral small vessel disease can potentially increase downstream resistance in the cerebral circulation resulting from lipohyalinosis and micro-atherosclerosis. This can be reliably evaluated with TCD.[11] Previous studies have indicated that TCD derived PI can be a useful tool in detection of leukoaraiosis with sensitivity of 70–89% and specificity of 73–86%[11] with high negative predictive value.[12] The correlation between severity of leukoaraiosis and changes in the PI was also noted to have good correlation in previous studies.[13],[14],[15] However, no such study has been carried out from South Asia. There are significant racial and ethnic variability in cerebrovascular disease patterns especially between Asians and Caucasians.[16],[17] Hence, the first objective of this study was to investigate the correlation between the presence and severity of leukoaraiosis as assessed by Fazeka's grading with changes in mean MCA PI in South Asian patient population. Current evidence indicates that leukoaraiosis is heterogeneous, reflecting different etiopathogenic factors, and varying disease stages. Some of the lesions could even be reversible if contributed by disturbances in interstitial fluid (ISF) dynamics before chronic changes set in. In addition to the classical mechanisms, a non-inflammatory periventricular venulopathy was also recently implicated in the pathogenesis of periventricular leukoaraiosis, suggesting that venous collagenosis and resultant vasogenic oedema and defective ISF circulation as potential additional mechanisms.[18] Therefore, the pathogenic mechanisms as well as the resultant microvascular changes including hemodynamic alterations could be different in periventricular and deep white matter leukoaraiosis. Therefore, it would be interesting to examine whether the Pulsatility Index (TCD parameter reflecting microvascular resistance) is different for leukoaraiosis in periventricular and deep locations, which formed the second objective of our study. To summarize, we examined the correlation of pulsatility index in TCD with the severity and location of leukoaraiosis in South Asian patients.
Patient and settings This was a prospective and cross-sectional study, conducted in the Comprehensive Stroke Care unit of a tertiary referral center for Neurological disorders in India over a period of 12 months from January 2017. The study included consecutive adult patients (age >18 years) with acute or recent strokes within 2 weeks of evaluation, who underwent MR imaging of the brain in addition to angiographic evaluation (CT or MR Angiography). Patients with modified Rankin Scale (mRS) score of >3, those with co existent large vessel atherosclerosis with ≥50% stenosis of the extra or intracranial vessels in any of the imaging modalities (CT or MR Angiography), large artery territory infarcts or major intracranial hemorrhage at any point of time and any concomitant central nervous system diseases (except vascular cognitive impairment) were excluded. Voluntary informed written consent was obtained from all the subjects prior to enrolment in accordance with the guidelines provided in the Declaration of Helsinki and the ICH E6 Guideline for Good Clinical Practice. The Institutional Review Board and the Institute Ethical Standards Committee had approved the study. Methods Baseline clinical and demographic details were collected. Brain MRI was acquired with MAGNETOM® Avanto 1.5 Tesla scanner in T1-weighted, T2-weighted, FLAIR, and SWI sequences. Vessel imaging of all patients were done either with CT Angiography or MR Angiography to rule out any significant major vessel stenosis as per exclusion criteria. The severity of small vessel disease was graded as per Fazeka's grading by a radiologist (KS) who was blinded to the clinical characteristics and transcranial doppler findings. Two neurologists (DHK and SS), blinded to the MRI findings, performed TCD (Nicolet® Sonara® digital TCD system) for all participants using 2 Hz probe. TCD was done on both sides through the temporal window on the middle cerebral artery (MCA). Optimal depth and angle of insonation, yielding the best spectral wave form and highest mean flow velocity were selected. PI in the MCA (MCA PI) was derived from the difference in the systolic and diastolic velocity divided by the mean flow velocity; PI = (PSV – EDV)/MFV, provided by automated in-built calculation by the machine. Mean MCA PI was calculated by averaging bilateral MCA PI. If the subject had good temporal window only on one side, unilateral MCA PI was considered as mean MCA PI. An average over at least 10 heart beats was applied in order to have a representative value of the TCD measures. Statistical analyses Data analysis was done using IBM SPSS Statistics for Windows version 21. Differences between patient groups in continuous variables were assessed by t tests or analysis of variance with tests for linear trend for severity of leukoaraiosis, whereas differences in nominal variables were compared by Pearson's Chi square tests. Univariate analysis was done to look for association between mean MCA-PI with risk factors and presence of white matter ischemic changes and microbleeds. A stepwise multiple linear regression analysis was performed to identify independent predictors of higher TCD PIs. Spearman's correlation coefficients were calculated to characterize the associations between TCD-PI scores and small vessel ischemic changes, separately for periventricular, and deep white matter locations. For determining the specificity and sensitivity of PI in different cut-off points we used ROC (receiver operator curve) analysis. A P value of <0.05 was considered significant for all analyses.
Out of 128 patients screened, 98 patients who met the inclusion criteria were recruited. Eight patients were subsequently excluded due to poor temporal window and the entire evaluation could be completed in 90 patients [Figure 1] of which 29% were females. Demographic characteristics of patients are presented in [Table 1] and [Table 2].
Mean age was 61 years (±10.9). 20 patients did not have leukoaraiosis, either periventricular or deep white matter, and were scored as grade 0 on Fazeka's scale. Higher age, presence of hypertension, diabetes mellitus, CAD and leukoaraiosis were strongly associated with elevated mean MCA PIs in univariate analysis [Table 3]. In multivariate analysis, presence of leukoaraiosis retained significant association with higher mean MCA PI after adjusting for age, diabetes, hypertension, and CAD [Table 4].
Correlation analysis showed both periventricular (Spearman's correlation coefficient 0.56, P = 0.01) and deep white matter (Spearman's correlation coefficient 0.63, P = 0.01) leukoaraiosis strongly correlating with higher mean MCA PI. The results of the receiver operator curve analyses are presented in [Table 5]. These analyses identified near identical cut-off points for PI values that allowed discrimination of periventricular leukoaraiosis with 69% sensitivity and 72% specificity, discrimination of deep leukoaraiosis with 74% sensitivity and 75% specificity, and discrimination of combined hemispheric leukoaraiosis with 69% sensitivity and 74% specificity.
The spectrum of cerebral small vessel disease includes leukoaraiosis (of varying severity and location), acute lacunar infarcts, lacunes, microbleeds, prominent perivascular spaces, and brain atrophy. Leukoaraiosis is usually symmetrically and bilaterally distributed in the hemispheric white matter (periventricular or deep). It can also occur in the deep gray matter and brain stem including pons. These changes are postulated to be related to microangiopathic changes which could either be causal or consequential (to some extent). Histopathology of these vessels have shown fibrosis in the wall which narrows the lumen, making the vessel rigid, with consequent loss of autoregulation.[19] This leads to increase in downstream vascular resistance and elevation of Pulsatility index in TCD. Kidwell et al. retrospectively compared TCD PIs and MRI manifestations of small vessel disease in 55 consecutive patients.[11] There were strong correlations between MCA PI and periventricular hyperintensities, deep white matter hyperintensities, lacunar disease, and combined hyperintensities. Similar correlation also found between pontine ischemia and vertebro-basilar PIs. In their multivariate analysis after adjusting for age, PI remained an independent predictor of white matter disease. Similar findings have been reported by Webb, et al, from UK, with MCA PI being the strongest physiological correlate of leukoaraiosis, independent of age in a dose-dependent manner.[13] Xiong, et al. from Hong Kong have found mean MCA PI and mean VB PI (Vertebro basilar Pulsatility Index) were independent predictors for total WMC (White matter changes) volume.[15] In our study, we looked into the correlation of location and severity of leukoaraiosis on MRI with Pulsatility index in TCD in the regional (South Asian) patient population. Similar to previous studies from other parts of the world, in our study, presence of leukoaraiosis was found to be associated with higher mean MCA PI, even after adjusting for age, diabetes, hypertension, and CAD. Previous studies have reported that age,[20],[21] diabetes,[22] hypertension,[23] and vascular dementia[24] were associated with increase in PI. In our study, univariate analysis showed age, diabetes, hypertension, CAD, and presence of white matter ischemic changes having significant association with elevated mean PI. Stepwise multiple regression analysis showed that age and diabetes along with presence of leukoaraiosis were independently associated with elevated PI. We found significant association between presence of microbleeds and mean MCA PI in univariate analysis; however, losing significance in multivariate analysis. This observation was similar to Mok, et al. where they did not find any correlation between presence of microbleeds and PI, even though their study was confounded by small number of patients.[12] ROC analysis in our study demonstrated PI cut points that allowed discrimination of leukoaraiosis at different locations—periventricular Leukoaraiosis with 69% sensitivity and 72% specificity, deep leukoaraiosis with 74% sensitivity and 75% specificity, and discrimination of combined hemispheric leukoaraiosis with 69% sensitivity and 74% specificity. In the study by Mok et al., MCA PI correlated significantly with WMC volume, independent of age, sex, and vascular risk factors. However, MCA PI was found to have low specificity for this association.[14] In a similar study by Xiong et al., the PI correlated with the volume of white matter changes.[15] However, we did not evaluate volume of white matter changes in our study. Our study demonstrates strong correlation between presence of leukoaraiosis (both periventricular and deep) with elevated MCA pulsatility index. This is the first study evaluating TCD assessment of small vessel disease from South Asia. The information from the receiver operator curve (ROC) analyses demonstrates strikingly similar PI cut-off points that allowed discrimination of leukoaraiosis at various locations—periventricular, deep locations as well as a combined hemispheric locations. This is a strong, albeit indirect, evidence for comparable changes in microvascular resistance for a given severity of leukoaraiosis irrespective of its location, viz. periventricular or deep. Since we have ruled out significant proximal vessel stenosis with vessel imaging, confounding effect of co-existing large vessel disease on the TCD parameters could be avoided. This is an advantage compared to other similar studies. Limitations in our study are mainly related to the inherent limitations of TCD which is operator dependent and the subjective visual quantification of leukoaraiosis in brain MRI, both of which could differ between observers.
Our study confirms the association of elevated microvascular resistance in cerebral circulation (as evidenced by higher pulsatility index) with presence of leukoaraiosis, irrespective of its location (periventricular or deep white matter) in South Asian patients. However, further studies are needed to ascertain whether these hemodynamic changes are causal or consequential to the hemispheric leukoaraiosis. Acknowledgements We acknowledge all our patients for their kind co-operation, and all consultants, technologists and all other supporting staff who have been part of our Comprehensive Stroke Care Centre at various periods of time for their dedicated patient care and contribution to the program in various forms. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.
[Figure 1]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||
