Anterior Watershed Angle And Collateral Circulation In Atherosclerotic Middle Cerebral Artery Occlusive Disease

Part I Anterior Watershed Angle and evaluation of cerebral blood flow in Patients with Symptomatic Middle Cerebral Artery Occlusive DiseasesObjectives:To establish a novel concept of anterior watershed angle (AWSA), by which the normal location of borderzone between anterior cerebral artery (ACA) and middle cerebral artery (MCA) can be quantitatively described, as well as the shift extent of ACA-MCA borderzone in the setting of MCA stenosis or occlusion (MCAS), and subsequently uncovered the potential clinical implications of AWSA.Methods:Patients were retrieved from Nanjing Stroke Registry Program in Jinling hospital between June 2006 and December 2012) 100 patients with a normal cerebrovascular digital subtraction angiography (DSA); 2) 275 patients were selected from 750 patients who had one or both side of MCAS (50-100%) with the flowing excluded criteria:A non-atherosclerotic MCA-M1 occlusive diseases, such as arteritis, moyamoya disease, or radiation-induced arteriopathy; asymptomatic MCA-M1 disease; with a tandem lesion of common carotid artery or internal carotid artery; lesions located beyond bifurcation of MCA-M1; with a history of atrial fibrillation or cardiogenic stroke. The 100 normal subjects and 275 patients with symptomatic MCAS were included in the first part of our study. Firstly, the concept of AWSA and its measuring method were presented in detail. In the DSA imaging with a Towne projection (a projection of posterior-anterior and appropriate head inclination), AWSA is defined as an angle with an apex that is the point of vertical projection of intracranial internal carotid artery bifurcation on the midline of sagittal sinus, and with a beginning line of the midline of sagittal sinus, and with a final line that links the apex with the central point of ACA-MCA cortical borderzone. Secondly, the 95% and 99% normal reference ranges were figured out according to 200 sides of AWSA among the 100 normal subjects. Thirdly, the correlations of AWSA with extent of MCA stenosis, antegrade flow (by thrombolysis in cerebral infarction, TICI), and collateral circulation with the American Society of Interventional and Therapeutic Neuroradiology and the Society of Interventional Radiology (ASITN/SIR) grading and patterns of cerebral computed tomography perfusion (CTP) were analyzed, respectively, in order to explore the potential clinical implications of AWSA.Results:The 95% and 99% normal reference range of AWSA originated from the 100 normal subjects was 0-22.5 and 0-24.7 degrees, respectively. Analysis of 750 AWSAs (100 normal subjects and 275 patients with MCAS) showed AWSA increased significantly along with the extent of MCA stenosis in setting of moderate, severe stenosis and total occlusion (P<0.001). The specificity of predicting an ipsilateral MCA stenosis of≥50% or≥70% with the condition of AWSA≥22.5 degrees (the 95% normal upper limit) was 93.0% and 85.6%, respectivley, which indicated the shift of ACA-MCA borderzone might imply a hemodynamic MCA stenosis. Multiple linear regression equation was modeled for AWSA with the two independent variables of TICI and ASITN-grading, and showed AWSA was closely associated with antegrade flow and retrograde flow (P<0.001). Spearmann correlation analysis showed that the transformation of AWSA (AWSA-trans), defined as AWSA divided by 22.5, was closely and linearly related to ASITN-grading (r=0.921, P<0.001), and TICI (r=-0.752, P<0.001). Of the 275 patients with MCAS,26 with CTP were categorized into three groups:normal or mild defect group, moderate defect group and severe defect group, according to the profiles of cerebral perfusion parameters. Patients who had a severe defect of cerebral perfusion had a larger AWSA when adjusted the confounder factor of the extent of MCA stenosis by ANCOVA (F=3.578, P=0.045).Conclusions:AWSA can be used to quantitatively describe the shift of ACA-MCA cortical borderzone. The 95% and 99% normal reference range of AWSA is 0-22.5 and 0-24.6 degrees, respectively. MCAS can be specifically indicative of hemodynamic influence in the setting of AWSA being more than the normal upper limit. AWSA was closely correlated with antegrade flow grading (TICI) and collateral circulation grading (ASITN-grading), especially the AWSAtrans is closely and linearly correlated with ASITN-grading and TICI, and may have some potential values in evaluation of cerebral blood flow of MCAS.Part II Predictors of Internal Borderzone Infarcts in Patients with Symptomatic MCASObjectives:To investigate the independent predictors of internal borderzone infarcts (IBZ) in patients with MCAS, and explore the potential clinical roles of AWSA in evaluating the cerebral blood flow beyond MCA occlusive lesion.Methods:193 subjects were selected from the above 275 patients with MCAS with the following criteria:1) the initial onset time to admission was less than 3 months; 2) the contralateral MCA-M1 was normal or only mild stenosis (< 50%); 3) the quality of MRI met with the requirement of analysis (performed within 2 weeks). Clinical data were collected including demographics, traditional vascular risk factors, patterns of infarcts on MRI, lesion characteristics of MCAS and blood flow related variables including TICI, ASITN-grading and composite circulation score (CCS, including CCS1=TICI+ASITN-grading and CCS2=TICI+AWSAtrans). Patients were divided into "with IBZ group" and "without IBZ group". These above-mentioned variables were compared between the two groups by univariate analysis. A full model (Model I) was initially established by including variables with a P value lower than 0.1 into multivariate Logistic regression model, using a method of Enter; Second, Model II was fitted with all variables in Model I using a method of backward Likelihood Ration (LR), to select the independent predictors of IBZ; Subsequently, Model III-VI were one by one established after manually removing the most significant (with lowest P value) blood flow related variable in the former model. Receiver operator curve (ROC) was employed to evaluate the quality of individual model by comparing their area under curve (AUC), thus weighted values of these above-mentioned blood flow related variables in precition of IBZ.Results:The median age of 193 patients was 54 (45-64) years, and 146 (75.6%) patients were male. Sixty-six patients (34.2%) presented with IBZ, of which 55 (83.3%) were concurrent with other pattern of infarction. Multivariate Logistic analysis showed that these following variables were predictors for IBZ:CCS2 (poor vs good, OR,3.58; 95%CI:1.42-9.02; P=0.007), TICI (involved vs un-involved, OR,3.65; 95%CI:1.62-8.23; P=0.002), the extent of MCA stenosis (P=0.042; severe vs moderate stenosis:OR,3.10; 95%CI:1.25-7.69; P=0.014; total occlusion vs moderate:OR,2.79; 95%CI:1.09-7.17; P=0.033) and CCS1 (poor vs good:OR,2.72; 95%CI:1.12-4.59; P=0.022); Of the four varibles, CCS2 might be the best predictors of IBZ for it was ininitially selected out with a similar AUC as that of full model. Without a history of hypertension was also identified with about a double higher risk of IBZ. Unexpectedly, the ASITN-grading was not identified as an independent predictors of IBZ (P=0.065) even further stratified analysis according to MCA stenosis or total occlusion.Conclusion:IBZ is one of the most common infarct patterns of patients with MCAS, frequently accompanied with other infarct patterns. History of hypertension, CCS2, TICI, extent of MCA stenosis, and CCS1 were the independent predictors of IBZ, but collateral grading was not identified as an independent predictor of IBZ, which indicated that CCS might be more reasonable than collateral grading in evaluating cerebral blood flow beyond MCA stenosis, and AWSA-trans played a role in blood flow evaluation of MCAS.Part Ⅲ Composite Circulation Score Predicting Risk of stroke Recurrence in Territory of Symptomatic MCASObjectives:To investigate independent predictors of ischemic stroke recurrence in the territory of symptomatic MCAS, and explore the potential clinical values of AWSA in predicting the recurrence of stroke.Methods:206 subjects were selected from the above-mentioned 275 patients with a similar inclusion criteria of part 2 except the item3. Clinical data were collected including demographics, vascular risk factors, MCA lesion profiles, blood flow related variables as described in part 2. In addition, follow-up data were also reviewed to determine whether a patient suffered from a primary outcome, which was defined as stroke recurrence in the territory of diseased MCA. The included patients were divided into "with primary outcome group" and "without primary outcome group". The baseline clinical data, MCA lesion profiles, blood flow related variables were compared between the two groups. Similar to part 2, those variables with a P value lower than 0.1 were included with each blood flow related variable in a multivariate Cox hazard proportional regression model by a way of Enter to select the independent predictors of stroke recurrence. The potential clinical implications of AWSA-trans were also evaluated.Results:Of 206 patients included in part 3,13(6.3%) were lost to follow up. The remaining 193 patients with follow-up records were included into analysis, with a mean age of 54.4±11.5 years, and 139 patients were male (72.0%). There were 48 patients with moderate stenosis (24.9%),78 patients (40.4%) with severe stenosis, and 67 patients (34.7%) with total occlusive MCAS. The median follow-up time was 15 (6-27) months and 25 (13.0%) primary outcome were identified, with an 8% of 12-month and 15% of 24-month of accumulated incidence of primary outcome. After adjusting other confounders with multivariate Cox hazard proportional regression model, SBP on admission, history of hyperlipidemia, ASITN-grading (P=0.032), CCS1(HR,3.28; 95%CI:1.42-7.58; P=0.005) and CCS2 (HR,3.23; 95%CI: 1.35-7.68;P=0.008) remained to be the independent predictors of stroke recurrence in the territory of MCAS, but the subtype comparison within ASITN-grading (HR, poor vs good,2.95; 95%CI:0.99-8.80; P=0.052; none vs good,0.67; 95%CI: 0.12-3.75; P=0.645) did not reach statistical significance. Further Kaplan-Meier analyses stratified to the extent of MCA stenosis showed that for patients with MCA stenosis (50-99%), poor CCSi (HR,3.28; 95%CI:1.42-7.58; P=0.005) and poor CCS2 (HR,3.23; 95%CI:1.35-7.68; P=0.008) were associated with a higher risk of stroke recurrence, but the accumulated rsik of stroke recurrence was found no statistical difference among the three ASITN-grading (none, poor and good); For patients with total occlusion, CCS1, CCS2, and ASITN-grading were not associated with the accumulated rsik of stroke recurrence (P=0.469,0.206 and 0.401, respectively, by Log-rank). These results indicated that CCS was superior to collateral circulation in prediction of stroke recurrence for patients with MCAS, and there might be a different prognostic profile between MCA stenosis and toatal occlusion.Conclusions:In this study, the accumulated risk of stroke recurrence in the MCAS territory was 8% and 15% in 12-month and 24-month of initial onset. The independent predictors of stroke recurrence in the MCAS territory included SBP on admission, history of hyperlipidemia, CCS1, CCS2 and collateral circulation grading (ASITN-grading). Composite circulation score might be more reasonable than ASITN-grading in evaluating the blood supply beyond MCA stenosis, and the CCS2 might have a similar clinical effect to CCS1 in prediction of stroke recurrence.