CT灌注成像Tmax参数对急性前循环缺血性脑卒中患者预后的预测价值

Tmax parameter from CT perfusion predicts the prognosis of patients with acute ischemic stroke after thrombolysis

目的探讨急性前循环缺血性脑卒中患者CT灌注参数与临床预后的关系,以及Tmax参数是否可用于急性脑卒中患者良好预后的预测。方法回顾性纳入26例于2014-03—2020-09就诊于南京鼓楼医院的急性前循环缺血性脑卒中患者。所有患者均接受阿替普酶静脉溶栓治疗,患者发病后48 h内完成全脑CT灌注成像扫描,头颅磁共振弥散加权成像呈侧脑室旁单一病灶。通过病例系统筛选患者的临床资料,由2名影像科主治医师分别处理CT灌注原始图像。使用美国GE公司的CT Kinetics软件系统在原始灌注图像上勾画感兴趣区,软件在各个灌注参数的彩色函数图上叠加相应的区域并运算每个灌注参数。计算并记录Tmax与发病至灌注成像扫描间隔时间的比值(定义为Tmax-time index)。根据90 d mRS评分将患者分为m RS≤1组和mRS>1组,采用非参数检验对2组患者不同灌注参数和Tmax-time index参数进行分析,斯皮尔曼等级相关分析对Tmax参数及Tmax-time index与mRS≤1进行分析,通过COX回归分析对混杂因素进行校正,受试者工作特征曲线用于分析参数的预测价值并计算出最佳预测界值。结果2组患者入院时NIHSS评分有统计学差异(P=0.029),余基线资料差异均无统计学意义(P>0.05)。灌注参数仅Tmax存在组间差异,Tmax参数及Tmax-time index参数与m RS≤1(定义为良好预后)呈显著性相关(P值:0.016比0.006;ρ值:0.468比0.526)。校正入院NIHSS评分、扫描间隔时间、心房颤动等混杂因素后,Tmax参数是急性脑卒中患者预后良好的独立预测因子(P=0.042)。校正心房颤动、入院NIHSS评分因素后,Tmax-time index可独立预测急性脑卒中患者的良好预后(P=0.041)。受试者工作特征曲线提示Tmax<1.25 s是急性脑卒中患者良好预后的最佳预测界值(P=0.019,AUC=0.805,95%CI 0.618~0.991,特异性57.1%,敏感性94.7%),Tmax-time index<1.08是急性脑卒中患者良好预后的最佳预测界值(P=0.009,AUC=0.842,95%CI 0.661~1.000,特异性85.7%,敏感性89.7%)。上述两个参数在预测效力上无统计学差异(P>0.05)。结论CT灌注成像的Tmax参数及Tmax-time index参数可用于预测急性脑卒中患者的临床预后。

Objective To explore the relationship between time-to-maximum(Tmax)parameter from computed tomography(CT)perfusion and clinical prognosis of patients with acute anterior circulation ischemic stroke.Methods The clinical data of 26 acute anterior ischemic stroke patients who treated in the Nanjing Drum Tower Hospital from March 2014 to September 2020 were collected.All patients underwent CT perfusion imaging scan within 48 hours after thrombolysis with alteplase.All patients had a single focus around the lateral ventricle.Two imaging physicians used the CT Kinetics software provided by General Electric(GE)to outline the areas of interest(ROIs)on original images.CT Kinetics software provides the value of each perfusion parameter according to the defined ROIs.Researcher calculated the value of Tmax-time index(defined as the ratio of Tmax to the time from onset to CT perfusion scan).According to 90-day modified Rankin score,patients were divided into mRS≤1 group and mRS>1 group.Statistical methods,such as non-parametric test,Spearman rank correlation,receive operating characteristic(ROC)analysis and Cox regression model analysis,were applied in this study.Results Among general clinical data and perfusion parameters of the two groups of patients,NIHSS score and Tmax showed statistical differences.Correlation analysis suggested a significant correlation between Tmax and Tmax-time index with mRS≤1.Multivariate regression analysis showed that after adjusting for confounding factors,Tmax and Tmax-time index parameters were independently correlated with 90-day mRS≤1.ROC revealed that Tmax<1.25 s(AUC=0.805,P=0.019),Tmax/time<1.08(AUC=0.842,P=0.009)were the optimal thresholds.Tmax<1.25 s is a prediction of 90 days mRS≤1 with sensitivity of 94.7%,specificity of 57.1%.Tmax/time<1.08 is a prediction of 90-days mRS≤1 with sensitivity of 89.5%,specificity of 85.7%.There was no statistically significant difference in the predictive power of these two parameters(P>0.05).Conclusion Tmax and Tmax-time index from CT perfusion imaging can predict the clinical prognosis of patients with acute ischemic stroke after thrombolysis.