磁共振灌注参数对静脉溶栓后出血转化的预测研究

目的 明确磁共振成像（magnetic resonance imaging，MRI）灌注参数脑血流达峰时间（Tmax）是否能预测急性缺血性卒中静脉溶栓后脑实质出血（parenchymal hemorrhage，PH），并确定其最佳阈值。方法 回顾性分析接受静脉溶栓治疗的急性缺血性卒中患者，依据欧洲协作急性卒中研究Ⅱ(European Cooperative Acute Stroke Study Ⅱ，ECASS Ⅱ）标准在溶栓后24 h复查的MRI或计算机断层扫描（computed tomography，CT）影像上评估出血转化。受试者工作特征曲线（receiver operatingcharacteristic curve，ROC）分析Tmax预测PH型出血的最佳阈值以及该阈值下的最佳预测体积，并用Logistic回归分析PH型出血的独立预测因素。结果 纳入分析112例患者，其中11例（9.8%）发生PH型出血转化。Tmax＞8 s为预测PH型出血的最佳阈值（曲线下面积=0.706，P =0.025），低灌注（Tmax＞8 s区）体积大于45 ml为PH型出血的独立预测因素，优势比（odds ratio，OR）=13.95，P =0.014；PH型出血组与无PH型出血组之间Tmax＞14 s体积、Tmax＞14 s/Tmax＞8 s的比值均无统计学差异（P =0.064；P =0.74）。结论 磁共振灌注参数Tmax＞8 s为预测急性缺血性卒中溶栓后PH型出血转化的最佳阈值；低灌注体积，而非低灌注严重程度，与PH型出血转化相关。

Magnetic Resonance Perfusion Parameter to Predict Hemorrhagic Transformation after Intravenous Thrombolysis

Objective To test whether Tmax of magnetic resonance imaging (MRI) could predict cerebral
parechymal hemorrhage (PH) after intravenous thrombolysis and then investigate its optimal
threshold.
Methods Acute ischemic stroke patients undergoing intravenous thrombolysis were retrospectively
analyzed. Follow-up imaging of MRI or computed tomography (CT) within 24 hours was assessed
for hemorrhagic transformation according to European Cooperative Acute Stroke Study (ECASS Ⅱ)
system. The optimal threshold and the optimal volume of Tmax to predict PH were analyzed with
receiver operating characteristic curve (ROC). The association of Tmax with PH was examined
using Logistic regression.
Results One hundred and twelve patients were analyzed, of whom 11 (9.8%) deveolped PH after
thrombolysis. Tmax>8 s was the optimal threshold to predict PH (area under curve [AUC]=0.706,
P =0.025). The volume >45 ml of hypoperfusion region (Tmax>8 s) was independently predictive
of PH (P =0.014, odds ratio [OR]=13.95). There were no significant differences of the Tmax>14 s
volumes or the ratio of Tmax>14 s/Tmax>8 s between PH group and non-PH group (P =0.064;
P =0.74).
Conclusion The MR perfusion parameter of Tmax>8 s was the optimal threshold to predict PH.
Hypoperfusion volume, rather than hypoperfusion intensity, was associated with PH.