Development and validation of a nomogram for predicting malignant brain edema after successful recanalization in acute ischemic stroke patients undergoing mechanical thrombectomy based on dual-energy CT and clinical features

Wang Qibin; Wang Tianyu; Chen Qing; Ding Weili; Ding Zhongxiang; Xu Jiaona

doi:10.3969/j.issn.1672-5921.2025.12.004

Abstract：Objective To develop and validate a nomogram prediction model for malignant brain edema(MBE)in acute ischemic stroke(AIS)patients undergoing mechanical thrombectomy and achieved successful recanalization by combining dual-energy CT(DECT)imaging parameters with clinical features.Methods AIS patients who underwent mechanical thrombectomy and achieved successful recanalization(immediately post-procedural modified thrombolysis in cerebral ischemia[mTICI]grade≥2b)at the Department of Neurology,Hangzhou First People's Hospital between January 2019 and December 2024 were retrospectively and consecutively enrolled.All patients were divided into a training set and a validation set in a 7∶3 ratio based on stratified random sampling.Clinical data were collected for all patients,including baseline characteristics such as sex,age,hypertension,diabetes,atrial fibrillation,history of ischemic stroke,smoking history,oral anticoagulant use,responsible vessel(internal carotid artery,middle cerebral artery),hyperdense sign of middle cerebral artery on pre-operative CT,admission National Institutes of Health stroke scale(NIHSS)score,admission Alberta stroke program early CT score(ASPECTS),pre-operative blood pressure;and treatment details including pre-operative intravenous thrombolysis,intraoperative use of salvage measures(stent,balloon,stent+balloon),intraoperative use of tirofiban,number of thrombectomy attempts,and time from symptom onset to successful recanalization.All patients underwent non-contrast DECT within 1 hour after mechanical thrombectomy to identify post-interventional cerebral hyperdensities(PCHDs).The absolute iodine concentration and the average(mean),minimum(min),and maximum(max)CT values on virtual non-contrast(VNC)images,iodine overlay maps(IOM),and mixed images(MI)were recorded and calculated,represented as VNCmean,VNCmin,VNCmax,IOMmean,IOMmin,IOMmax,MImean,MImin,MImax.All parameters were measured three times and averaged.The relative iodine concentration was calculated as the ratio of the absolute iodine concentration to the iodine concentration in the superior sagittal sinus.A CT scan was performed within 72 hours postoperatively to assess whether the patient developed MBE.Using the occurrence of MBE after mechanical thrombectomy as the dependent variable,the least absolute shrinkage and selection operator(LASSO)regression was used in the training set to screen for the most predictive key variables,which were then incorporated into a multivariate Logistic regression model to identify independent risk factors for MBE after mechanical thrombectomy in AIS patients achieved successful recanalization,and a nomogram prediction model was constructed.The receiver operating characteristic(ROC)curve was used to evaluate the discriminatory performance of the nomogram model in the training and validation sets.Calibration curves and decision curve analysis(DCA)were plotted in both the training and validation sets to validate the consistency between the predicted and actual values and to assess the clinical applicability of the model.Results A total of 398 AIS patients who underwent mechanical thrombectomy and achieved successful recanalization were enrolled,among which 87 developed MBE post-operatively and 311 did not,yielding an MBE incidence of 21.9%.The training set comprised278patients(61with MBE,217without MBE),and the validation set comprised 120 patients(26 with MBE,94 without MBE).(1)No statistically significant differences were found in the clinical and imaging data between the training and validation sets(all P＞0.05).(2)Based on LASSO regression screening and subsequent multivariate Logistic regression analysis in the training set,the results showed that low admission ASPECTS(OR,0.194,95%CI 0.103-0.366,P＜0.01),high IOMmin(OR,1.680,95%CI 1.052-2.684,P=0.030),high relative iodine concentration(OR,1.816,95%CI 1.350-2.443,P＜0.01),and increased number of thrombectomy attempts(OR,2.324,95%CI 1.416-3.813,P＜0.01)were independent risk factors for MBE after mechanical thrombectomy in AIS patients achieved successful recanalization.A nomogram prediction model for MBE after mechanical thrombectomy in AIS patients achieved successful recanalization was constructed based on above four factors.Calibration curve analysis showed excellent calibration of the nomogram model in both the training set(calibration slope=0.95,intercept=0.02,mean absolute error=0.011)and the validation set(calibration slope=0.91,intercept=0.05,mean absolute error=0.022).ROC analysis results showed that the area under the curve of the nomogram model for predicting MBE after mechanical thrombectomy in AIS patients achieved successful recanalization was 0.909(95%CI 0.870-0.948)in the training set and 0.818(95%CI 0.733-0.903)in the validation set.DCA results showed that in the training set,the nomogram model provided significant net clinical benefit for predicting MBE after mechanical thrombectomy in AIS patients achieved successful recanalization when the threshold probability ranged from 0.00 to 0.91;in the validation set,this effective prediction range was from 0.04 to 0.78.Conclusion Low admission ASPECTS,high IOMmin,high relative iodine concentration,and increased number of thrombectomy attempts are independent risk factors for MBE after mechanical thrombectomy in AIS patients achieved successful recanalization,the nomogram model constructed based on these factors can predict the risk of MBE in a certain extent,and provide a reference for clinical decision-making.