Alberta stroke programme early CT score on diffusion-weighted imaging and clot burden scoring on MR angiography in the prediction of hemorrhagic transformation after thrombolysis in acute cerebral infarction
Xinying Wang,Xinfeng Yu,Jianzhong Sun,Fang Cao,Minming Zhang
DOI: https://doi.org/10.3760/cma.j.issn.1005-1201.2014.06.004
2014-01-01
Abstract:Objective To evaluate Alberta stroke programme early CT score on diffusion-weighted imaging (DWI-ASPECTS)and clot burden score on MR angiography (MRA-CBS)in predicting hemorrhagic transformation(HT) in acute anterior circulation cerebral infarction after thrombolysis in diffusion-weighted imaging Alberta stroke program.Methods A total of 37 consecutive patients with acute anterior circulation cerebral infarction were treated with thrombolysis.The clinical information , score of DWI-ASPECTS before thrombolysis , score of MRA-CBS before thrombolysis and images of enhanced gradient echo T 2*-weighted angiographywithin ( ESWAN) 24 hours before and after thrombolysis were all collected.The interval between onset and the two MRI scans were recorded respectively.We identified HT according to the images of ESWAN scanned after thrombolysis , and divided patients into 2 groups:with HT(14 cases) and without HT (23 cases).Differences of clinical data and imaging indicators between the two groups were compared by using Fisher′s exact test and Wilcoxon rank sum test.Logistic regression analysis was performed by taking HT as the dependent variable , and the scores of NIHSS , DWI-ASPECTS and MRA-CBS at admission were taken as independent variables.The variables which were statistically significant in logistic regression analysis were enrolled in receiver operating characteristic analysis.Results In HT group, the scores of NIHSS, DWI-ASPECTS and MRA-CBS were 15.00 ±5.30, 6.00(4.75,7.00) and 7.00(0.75,8.50) respectively.In the other group without HT, these scores were 7.00 ±4.80, 9.00(8.00,10.00)and 10.00(6.00,10.00) respectively.Compared with patients without HT , patients with HT had a higher baseline NIHSS score ( Z=-3.72,P<0.01), a lower DWI-ASPECTS (Z=-4.13,P<0.01) and a lower MRA-CBS (Z=-2.00, P<0.05).Logistic regression analysis showed that the scores of DWI-ASPECTS ( OR 0.42,95%CI 0.21-0.87,P <0.05 ) and NIHSS ( OR 1.22, 95%CI 1.00-1.48, P <0.05 ) at baseline predicted HT development independently.Receiver operating characteristic analysis showed that the optimal cut -off point of DWI-ASPECTS to predict the development of HT was≤7.Its sensitivity, specificity and area under ROC curve were 92.9%, 78.3% and 0.902 respectively ( P<0.01 ).Conclusions ASPECTS on DWI is of great value in predicting HT after thrombolysis in acute cerebral infarction.CBS on MRA can provide additional information for predicting HT.