动态磁敏感对比增强参数联合动脉脑血容量多参数预测脑胶质瘤术后复发的效能研究OA

Objective:To investigate application value of the multi-parameter of dynamically magnetic sensitivity contrast parameters of relative cerebral blood volume(rCBV)and relative cerebral blood flow(rCBF)combined with artery CBV(CBVa)in predicting postoperative recurrence of high-grade gliomas.Methods:The clinical data of 112 patients with high-grade glioma,who were confirmed by postoperative pathology at Guangde Traditional Chinese Medicine Hospital during January 2022 and December 2023,were retrospectively analyzed.All patients received regular follow-up of magnetic resonance imaging(MRI)after surgery,and they were divided into recurrence group(72 cases)and pseudoprogression group(40 cases)according to the results of follow-up.The post-processing software was adopted to sketch the new abnormal enhancement area in the surgical area.The rCBV,rCBF,mean transit time(MTT),time to peak(TTP)and CBVa histogram feature related parameters[including Mean,Median,Majority,maximum value(Max)and minimum value(Min),skewness,kurtosis and the 10-90th percentile]were obtained to compare the difference of each parameter between the two groups.Multivariate Logistic regression was used to analyze independent risk factors for postoperative recurrence of high-grade glioma,and regression equations were constructed.The receiver operating characteristic(ROC)curve analysis was used to assess predictive efficacy of a single indicator or combined model.Then,the optimal truncation value was calculated and obtained.Results:The rCBV and rCBF levels of recurrence group were respectively(2.62±1.29)ml/100g and(2.66±1.35)ml(100g·min),which were higher than(1.23±0.45)ml/100g and(1.12±0.31)ml(100g·min)of pseudoprogression group,and the differences were significant(t=4.78,2.65,P<0.05).The CBVa Mean,CBVa Median,CBVa Max value of recurrence group were higher than those of pseudoprogression group,and the differences were significant(t=5.54,4.91,2.40,P<0.05).The CBVa(20%,25%,30%,40%,60%,70%,75%,80%,90%)values of recurrence group were significantly higher than those of pseudoprogression group,and the differences were statistically significant(t=4.15,4.50,4.90,5.00,5.20,5.30,4.90,4.75,4.10,P<0.05).The MTT and TTP levels of recurrence group were significantly lower than those of pseudoprogression group,and the differences were statistically significant(t=-2.71,-3.27,P<0.05).There were not significant differences in CBVa Min,Majority,skewness,kurtosis and CBVa 10%value between the two groups(P>0.05).ROC curve analysis showed that rCBV,rCBF,MTT,TTP,CBVa mean,median,maximum and CBVa(20%,25%,30%,40%,60%,70%,75%,80%,90%)values all had favorably predictive efficacy.Among of them,rCBV had the best predictive efficacy in identifying recurrence group and pseudoprogression group,and area under curve(AUC)value of ROC curve was 0.865.When the truncation value was 1.25,the sensitivity and specificity of rCBV were respectively 93.0%and 86.7%in predicting recurrence.The parameters with statistically significance of the above differences were included into the multivariate logistic regression model,and the results showed that rCBV,rCBF level and CBVa Median were independent risk factors of predicting postoperative recurrence of high-grade glioma(OR=3.12,2.87,2.54,P<0.05).Logistic regression equation was constructed:Logit(P)=-2.35+0.85×rCBV+0.72×rCBF+0.63×CBVa Median,whose AUC value was 0.902,which indicated the combination of rCBV,rCBF and CBVa Median had higher predictive value for postoperative recurrence of high-grade glioma,and the predictive efficacy of that was better than single parameter.Conclusion:In MRI perfusion imaging,rCBV,rCBF combined with CBVa Median can effectively identify postoperative recurrence and pseudoprogression of high-grade glioma,which can provide a high-precisely noninvasive prediction tool for clinical works.