Differential Distribution of Cerebral Microbleeds in Subtypes of Acute Ischemic Minor Stroke and TIA as well as its Association with Vascular Risk Factors
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.299147
Source of Support: None, Conflict of Interest: None
Keywords: Ischemic stroke, microbleed, risk factors, stroke subtypes
Cerebral microbleed (CMB) represents a novel neuroimaging marker of cerebral small vessel disease providing direct evidence of blood leakage from pathologically fragile small vessels primarily affected by hypertensive arteriopathy or cerebral amyloid angiopathy (CAA). CMB has been associated with white matter hyperintensities (WMHI) and is postulated to be a biomarker in the small vessel occlusion subtype of acute ischemic stroke and transient ischemic attack (TIA). On the contrary, nearly 30% of patients with nonvalvular atrial fibrillation was shown to have CMB in a recent cohort of patients with cardioembolic stroke.
Very few studies have primarily intrigued the differential distribution of CMB in various stroke subtypes till date., Likewise, there is scarcity of information on the correlation between the location of CMB and its association with stroke subtypes. A recent study noted that lobar distribution of CMB of non-CAA type was more frequent in cardioembolic subtype, whereas deep location was more common with small vessel disease and large artery atherosclerosis. There appears to be an underlying hitherto unexplored relationship between CMB location and stroke patho-mechanism. This raises pertinent questions into the various stroke risk factors and their correlation with CMB. Community-based studies have identified systemic hypertension and reduced triglyceride (TG) levels to be associated with CMB., But data are scarce on the impact of these vascular risk factors on CMB location in various stroke subtypes. Some studies have shed light on the predilection of CMB to affect non-lobar location in those with renal dysfunction, thereby pointing toward the differential prevalence of CMB in relation to stroke subtype and vascular risk factors. This study aimed to assess the de novo differential prevalence of CMB among the various TOAST stroke subtypes of acute ischemic minor stroke (AIS) and TIA in Indian population, and to identify the potential vascular risk factors associated with location and number of CMB in this study population.
Patients admitted within 48 h of an AIS (defined as those with acute onset of focal neurological deficits lasting more than 24 h and baseline National Institute of Health Stroke Scale [NIHSS] of less than or equal to 5) and TIA (defined as an episode of focal neurological dysfunction with abrupt onset and rapid resolution lasting for less than 24 h from symptom onset with or without a corresponding diffusion-weighted imaging lesion) evaluated in the Comprehensive Stroke Center, Department of Neurology, Sree Chitra Tirunal Institute for Medical Sciences and Technology (SCTIMST), Thiruvananthapuram, Kerala, India, between 2008 and 2015 were recruited into the study. Those with baseline modified Rankin scale (mRS) >2, history of intracerebral hemorrhage/traumatic brain injury in the past and history of dementia were excluded. Baseline demographic and clinical variables were noted. The neuroimaging studies were analyzed by independent blinded assessors, which included neuroradiologist and neurologist. Microbleed Anatomical Rating Scale (MARS) was used to analyze the CMB. The degree of periventricular and deep white matter changes were graded based on FAZEKAS system of grading. Stroke subtypes were delineated as per TOAST stroke subtype classification scheme into large artery atherosclerosis (LAA), cardio-embolism (CE), small vessel occlusion (Lacunar), stroke of other determined etiology (ODE), and stroke of undetermined etiology (UE). The prevalence of CMB in each subtype of AIS and TIA was analyzed to look for statistical significance. The study was approved by the Institutional Ethics Committee of SCTIMST, Thiruvananthapuram, Kerala, India.
For univariate analysis, Student's t-tests were used for continuous variables and Pearson χ2 tests and Fisher's exact tests for categorical variables. For ordinal variables, extended Mantel–Heanzel χ2 test for assessing a linear trend was used. P- value of less than 0.05 was considered statistically significant. A multivariable logistic regression analysis of all variables with statistical significance on univariate analysis was performed. All the statistical analysis was performed with SPSS statistical software version 21.
A total of 488 subjects who met the inclusion and exclusion criteria were included. The mean age of the study population was 57.5(standard deviation 14.4) years. In all, 385 (78.9%) had ischemic minor stroke, whereas 103 (21.1%) had TIA. The most common stroke subtype in our study population as per TOAST classification was undetermined etiology (31.6%) followed by large artery atherosclerosis (28.5%) and lacunar stroke subtype (25.6%). [Table 1] depicts the baseline demographic and clinical characteristics of the study population.
CMB was noted in 140 (28.7%) of the study population. About 42 of 140 (30%) had four or more CMB. [Table 2] depicts the differential distribution of demographic and clinical correlates among those with and without CMB. Around 80% of those with CMB had concomitant WMHI suggestive of microangiopathy, with 54% exhibiting moderate to severe grades on FAZEKAS scale (P = 0.000). In patients with four or more CMB, FAZEKAS grade 3 periventricular and deep WMHI were seen in 33.3%, whereas only 7.2% with one to three CMB had similar changes (P = 0.000).
The association of CMB with stroke subtype by TOAST classification was examined. About 35% of those with CMB had an underlying lacunar stroke subtype etiology, whereas 33.6% had large artery atherosclerosis and 10.7% were of cardioembolic subtype (P = 0.000). Approximately 40.5% of those with four or more CMB had an underlying lacunar stroke subtype, as against 28.6% with large artery atherosclerosis and 16.7% with cardioembolic etiology (P = 0.014).
The significance of the location of CMB and the number of CMB was analyzed [Table 3] and [Table 4]). Univariate analysis showed that CMB at all three locations (lobar, deep, and infratentorial) was associated with lacunar stroke, followed by large artery atherosclerosis. CMB was associated with TG levels less than 150mg/dL in all three locations, eventhough it was statistically significant only in lobar and infratentorial location. Serum creatinine levels more than 1.5mg/dL was more likely to be associated with lobar CMB (P = 0.033). Those with deep CMB were more likely to develop stroke than TIA (P = 0.008) [Table 4].
On multivariate logistic regression analysis, the presence of CMB was significantly associated with systemic hypertension, serum TG levels of less than 150 mg/dL, and presence of WMHI [Table 3]. The prior usage of antiplatelets and anticoagulants was not associated with the presence or number of CMB either on univariate or multivariate analysis.
The differential distribution of CMB in TOAST stroke subtypes of AIS/TIA and correlation of the presence, number, and location of CMB with known vascular risk factors were examined in this study. Our study demonstrated that CMB was more prevalent and numerous in lacunar subtype followed by large artery atherosclerosis etiology, irrespective of the location of CMB. Those with systemic hypertension, WMHI on magnetic resonance imaging brain, and serum TG levels of less than 150 mg/dL were more likely to harbor CMB.
Our study showed CMB to be present in 28.7% of the study population, which was comparable to previous studies.,, In unison with the well-established hypotheses, our study also highlighted that lacunar stroke subtype (small vessel occlusion) was not only likelier to harbor CMB but also to have multiple CMB (four or more in number), thereby indicating a plausible common underlying etiopathogenic link of small vessel disease contributing to occurrence of microangiopathy changes (WMHI), CMB, and small vessel occlusion (lacune), in turn exhibiting a common disease spectrum.,
Lacunar strokes were more likely to harbor CMB in all the three locations (lobar, deep, and infratentorial) followed by large artery atherosclerosis subtype in our study. Although the pathogenesis of large artery atherosclerosis and lacunar stroke is known to be different, it is possible that at the micro vascular level both share a similar patho-mechanism, as indicated by our study results. This hypothesis is strengthened by the possible predisposition to causation of small vessel disease by concomitant intracranial atherosclerosis in Asian population. Carotid atherosclerosis quantified at the level of internal carotid artery in the sub-analysis of Framingham cohort was associated with CMB primarily in deep regions. Even though cardioembolic strokes were noted to be associated with lobar location of CMB in a few recent studies, our study did not demonstrate the same. The clinical implication of CMB in terms of treatment decisions have been recently debated; however, our study did not reveal any association with antiplatelet or anticoagulation use to the presence, location, or number of CMB.
The traditional hypotheses of deep-seated CMB associated with ischemic strokes due to arteriosclerosis and lobar CMB having allegiance with CAA and resulting intracerebral hemorrhage have been questioned by recent evidence. Vascular amyloid deposition in CAA might also contribute to loss of contractile components of the vessel wall, impaired reactivity to physiologic stimulation in turn causing stenosis, and constriction of small vessels leading to ischemic events. Our study also provides evidence against the above-mentioned traditional hypotheses regarding CMB pathology, with lobar CMB also being prevalent in lacunar/small vessel occlusion subtype.
The enigmatic and paradoxical relationship of lipid levels with both ischemic and hemorrhagic strokes has been a focus of interest.,,,, In a recent meta-analysis, an inverse association was found for risk of hemorrhagic stroke with total cholesterol and low-density lipoprotein (LDL) fraction but not with high-density lipoprotein (HDL). More interestingly in the Rotterdam study, a lower TG fraction was strongly associated with both ICH and deep or infratentorial CMB. On the contrary, the Reykjavik study demonstrated lower TG levels to be associated with increased risk for lobar CMB. This observation is similar to our study which showed no relationship with total cholesterol levels, but with lower TG levels for lobar or infratentorial CMB. Although the exact mechanism remains obscure, it may be plausible that as TG is an integral component of cell membrane, levels below a threshold limit may cause endothelial weakness in small intracranial vessels contributing to microaneurysm and leak resulting in CMB.,,, A previous study based on post-mortem specimens have shown that pathology due to atherosclerosis progresses variably between cortical vessels, deep penetrating vessels, and basal arteries. The differential association between location of CMB with low TG levels seen in our study with higher predilection for lobar or infratentorial location of CMB associated with TG levels below 150 mg/dL may be a reflection of the above-mentioned phenomenon.
Renal dysfunction reflected by estimated glomerular filtration rate (eGFR) is independently associated with deep location of CMB and number of CMB, both in acute lacunar stroke and in neurologically healthy subjects., The pathophysiological similarity between deep WMHI and deep CMB due to hypertensive vasculopathy is postulated to be the connecting link as is being highlighted by the shared vulnerability of renal afferent arterioles and cerebral deep/superficial perforating arterioles to the pathological consequences of chronic systemic hypertension. Interestingly, such a correlation was noted only for lobar CMB in our study. This is in contradiction with previous studies,,, where renal dysfunction and chronic kidney disease (CKD) progression were associated with non-lobar CMB. The arteriopathy underlying strictly lobar CMB, which is CAA, appears to be less related to renal impairment. Traditionally, lobar CMB has been known to be more common with CAA and deep CMB with hypertensive arteriosclerosis, with recent studies providing evidence in contrary especially in the Asian population. The postulation is that not only the deep perforators are at risk due to deranged renal function but also factors independent of hypertension might be at play in turn causing vascular endothelial dysfunction leading to microhemorrhages affecting the lobar arterioles, eventually contributing to non-lobar CMB.
Our study is a cross-sectional retrospective observational study from a single center which focused on the select group of patients with AIS and TIA, thereby limiting the generalizability of our study results to the general population, on account of a potential selection bias. The use of absolute serum creatinine value as an indicator for renal dysfunction is also another potential limitation of our study, as eGFR or urine albumin–creatinine ratio would have been a better index. The uniformity of study design and meticulous scrutiny of neuroimaging by a dedicated stroke neuroradiologist (CKD) who is blinded to the clinical details are the advantages of our study.
Our study provides evidence on the differential distribution of CMB in TOAST subtypes of AIS and TIA in the Indian subcontinent, with demonstration of association of multiple vascular stroke risk factors with the presence, number and location of CMB. Even though no previous data are available from the Indian subcontinent, the results of our study were grossly comparable to those from Western population and other Asian cohorts.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2], [Table 3], [Table 4]