Abstract: Objective To identify clinical and imaging predictors of outcome and to develop a clinical/apparent diffusion coefficient score that will enable better patient selection for thrombolytic therapyin acute ischemic stroke.Methods Baseline clinical and radiologic data variables (including predicted volumes obtainedfrom apparent diffusion coefficient (ADC) based imaging analysis system) that consideredpossibly related to outcome were selected in 40 patients with acute ischemic stroke of the anteriorcirculation. A univariate analysis was conducted to explore the association between these factorsand bad outcome, defined as a modified Rankin scale score≥2. Logistic regression was thenperformed to select the most important variables independently affecting prognosis. Receiveroperator characteristic curve (ROC) was then used to obtain cut-off points for each independentvariable. A risk score (clinical/apparent diffusion coefficient score) was then developed based onthese variables. The predictive value of the clinical/multimodal magnetic resonance imaging (MRI)score was compared with single clinical data with respect to the clinical outcome 3 months afterstroke onset by use of modified Rankin Scale (mRS).Results In the univariate analysis, variables associated with poor outcome were: age, baselineNational Institutes of Health Stroke Scale (NIHSS), predicted infarct core volume, predictedsalvageable ischemic tissue volume, predicted final infarct volume, final infarct volume andadmission diffusion-weighted imaging (DWI) lesion volume. Four variables independentlyassociated with poor outcome were identified by Logistic regression: age, predicted infarct corevolume, predicted final infarct volume and NIHSS. We then developed the clinical/ apparentdiffusion coefficient score: 1 point for age>58 years, 1 for predicted infarct core volume>5.84 ml, 1for predicted final infarct volume>10.6 ml and 1 for>12 for NIHSS. The factor that most accuratelypredicted good clinical outcome 3 months after stroke was clinical/apparent diffusion coefficientscore (area under the curve=0.878, P <0.001), followed by final infarct volume (area under thecurve=0.802, P =0.001), predicted final infarct volume (area under the curve=0.797, P =0.001),predicted infarct core volume (area under the curve=0.739, P =0.01), baseline NIHSS (areaunder the curve=0.759, P =0.005), predicted salvageable ischemic tissue volume (area under thecurve=0.719, P =0.018) and baseline DWI lesion volume (area under the curve=0.693, P =0.037).Conclusion Clinical/apparent diffusion coefficient score based on clinical and radiologic data issuperior to single clinical and ADC in estimating the chances of poor outcome.

Key words: Stroke; ischemic; Thrombolytic therapy; Magnetic resonance imaging; Diffusionmagnetic resonance imaging

摘要： 目的 探讨影响急性缺血性卒中预后的因素，建立一种基于临床和多模式磁共振成像（magneticresonance imaging，MRI）的急性前循环缺血性卒中预后评估系统。方法 选择发病9小时内完成多模式MRI的前循环急性缺血性卒中患者40例。按照改良的Ranking量表（modified Ranking Scale，mRS）分为预后良好组（0～1分）和预后不良组（2～6分）。评价两组年龄、基线美国国立卫生研究院卒中量表评分（national institutes of health stroke scale，NIHSS）、基线弥散加权像（diffusion-weighted imaging，DWI）体积、基线灌注加权像（perfusion-weighted imaging，PWI）体积以及由基于表观弥散系数（apparent diffusion coefficient，ADC）的图像分析方法获得的预测梗死核心体积、预测可挽救脑组织体积等临床/影像信息对预后的影响；采用多因素分析筛查出单因素分析中具有统计学意义的变量作为预后评估系统的组成部分，应用受试者工作特征曲线（receiver operatorcharacteristic curve，ROC）分析获得各变量的阈值评分，整合后获得临床/ADC评分，应用ROC曲线下面积（area under curve，AUC）分析各评分模式判断预后的效能。结果 预后良好组与预后不良组在年龄、基线NIHSS、预测梗死核心体积、预测可挽救脑组织体积、预测最终梗死体积、实际最终梗死体积和基线DWI异常区域体积的差异均具有统计学意义。多因素分析显示年龄、预测梗死核心体积、预测最终梗死体积和基线NI HSS能作为判断预后的风险因素，构成临床/ADC预后评分系统的四个因素。应用ROC分析获得以上四个变量判断预后不良的阈值分别为＞58岁、＞5.84 ml 、＞10.6 ml 和＞12分。该评分系统的AUC最大（AUC=0.878，P＜0.01），其判断急性缺血性卒中患者90 d预后的效能最高，其次是实际最终梗死体积（AUC=0.802，P =0.001）、预测最终梗死体积（AUC=0.797，P =0.001）、预测梗死核心体积（AUC=0.739，P =0.01）、基线NIHSS（AUC=0.759，P =0.005）、预测可挽救脑组织体积（AUC=0.719，P =0.018）和基线DWI异常区域体积（AUC=0.693，P =0.037）。其中，临床/ADC预后评分系统与预测梗死核心体积、预测可挽救脑组织体积、基线DWI异常区域体积AUC之间的差异具有统计学意义（P分别为0.043，0.035和0.01）。结论 临床/ADC预后评分系统比基线NIHSS评分和各影像参数判断90 d急性缺血性卒中患者预后的效能高；制定急性缺血性卒中患者治疗方案时，应结合患者临床和影像信息综合考虑。

关键词: 卒中; 缺血性; 血栓溶解疗法; 磁共振成像; 磁共振成像; 弥散