Utility Of ADC Map In Predicting Different Brain Tissue Outcomes And Hemorrhagic Transformation After Endovascular Therapy In Patients Of Acute Ischemic Stroke

Background and PurposeEndovascular therapy（EVT）is a standard and effective therapy for patients of acute large vessel occlusive ischemic stroke in anterior circulation.The key to achieving a favorable prognosis through EVT is to rescue as much ischemic brain tissue as possible while avoiding the occurrence of hemorrhagic transformation（HT）and other complications.Preoperative imaging evaluation is an important part for the systematic management of acute ischemic stroke（AIS）.For patients with a preoperative large infarct area and severe ischemia,EVT will fail to achieve a good prognosis,and it can even lead to the occurrence of HT which is also associated with early neurological deterioration and unfavorable clinical outcomes.Diffusion weighted imaging（DWI）and apparent diffusion coefficient（ADC）sequence are routinely used for preoperative diagnosis of AIS and evaluation of infarct core area.However,part of the brain tissue in the DWI hyperintensity lesion area can be reversed after successful recanalization of occluded vessel.If there is no recanalization,the salvageable brain tissue in penumbra can progress to infarction to some extent over time.And there is no widely accepted predictive method for HT being used prior to EVT.This study aims to evaluate the predictive ability of ADC map for brain tissue outcome after EVT,investigate the predictive value of preoperative ADC value and corresponding volume on HT after EVT and analyze the factors affecting HT in patients with AIS treated with EVT.MethodsWe performed a retrospective analysis based on the prospective database of the People’s Hospital of Zhengzhou University from January 2018 and June 2020.Patients with anterior circulation large vessel occlusion who received EVT within 24hours after symptom onset and had both baseline MRI and follow-up CT or MRI data within a week were included.Olea Sphere23.0 image post-processing software was used to process the MRI before and within a week after the operation,and the ADC value,pixel,and volume of preoperative DWI lesion reversal area,final infarct area of preoperative DWI lesion and increased infarction area as well as volumes below different ADC thresholds(0.3-0.6*10^-3mm²/s,interval 0.1*10^-3mm²/s)were respectively measured.The baseline data,onset time,imaging time and successful recanalization time were recorded.Then we analyzed the difference of ADC value among three areas using Kruskal-Wallis H test,and explored the optimal ADC threshold for predicting DWI lesion reversal using receiver operating characteristic curve（ROC）analysis.We performed univariate analysis to find the difference of clinical and image data between HT group and non-HT group.And we performed ROC analysis to explore the optimal ADC threshold and corresponding optimal volume for predicting HT.And then multivariate analysis was performed to assess independent factors associated with HT.ResultA total of 119 patients were included in the study.45 patients had complete preoperative and postoperative follow-up MRI imaging and they all had different degree of DWI lesion reversal and infarct growth after successful recanalization.Of the 119 patients,41（34.45%,41/119）patients had HT after EVT,and 65（54.62%,65/119）patients achieved favorable clinical outcome at 90 days.The preoperative median ADC values of the final infarct core of preoperative DWI lesion,the area of DWI lesion reversal and the area of infarct growth are0.555*10^-3mm²/s（0.462,0.648）,0.637*10^-3mm²/s（0.509,0.765）and0.948*10^-3mm²/s（0.905,0.991）respectively.The ADC values among three regions are significantly different.The optimal cut-off point to predict DWI lesion reversal after successful recanalization is ADC<0.57*10^-3mm²/s with an accuracy of 87.1%,a sensitivity of 97.2%and a specificity of 68.3%.The volume below each ADC threshold had a certain predictive value for HT.ADC achieved a predictive accuracy of 82.2%for HT,and the sensitivity was 61%and specificity was 91%at an optimal operating point threshold volume of 6.46ml below optimized ADC threshold of 0.4*10^-3mm²/s.The predictive ability of other ADC thresholds was relatively lower.The accuracy is 75.7%for volume of ADC<0.6*10^-3mm²/s,79.6%for volume of ADC<0.5*10^-3mm²/s and 70%for volume of ADC<0.3*10^-3mm²/s.Multivariate analysis showed lager volume of ADC<0.4*10^-3mm²/s,intravenous thrombolysis,tandem occlusion and atrial fibrillation were independent predictors of HT.Conclusions1.All patients had different degree of DWI lesion reversal and infarct growth after successful recanalization with EVT.The preoperative ADC values are significantly different in brain tissues with different outcomes after successful recanalization of occluded vessel.2.The degree and the extent of ischemia before EVT are important factors of HT after recanalization.The volume below each ADC threshold had a certain predictive value for HT.The optimal threshold to predict HT is volume of ADC<0.4*10^-3mm²/s greater than 6.46ml with a accuracy of 82.2%.