ADC三维直方图分析对胶质瘤MGMT启动子甲基化程度的预测探究  

ADC-Based Three-Dimensional Histogram Analysis in Prediction of Methylation Degree of MGMT Promoter in Glioma

作　　者：邓凯吉 傅雅兰 王凌寒 林怡珊 蒋日烽[1] DENG Kaiji;FU Yalan;WANG Linghan(Department of Radiology,Fujian Medical University Union Hospital,Fuzhou,Fujian Province 350001,P.R.China)

机构地区：[1]福建医科大学附属协和医院放射科,福州350001 [2]福建医科大学临床医学部,福州350101

出　　处：《临床放射学杂志》2021年第7期1266-1270,共5页Journal of Clinical Radiology

基　　金：福建省自然科学基金项目(编号:2018J05135);福建省卫生计生科研人才培养项目(编号:2018-1-37)。

摘　　要：目的探讨基于表观扩散系数(ADC)的三维直方图分析在预测胶质瘤MGMT启动子甲基化程度中的价值。方法回顾性分析经手术病理证实的97例胶质瘤患者的MRI和临床病理资料。男51例,女46例,平均年龄(51.14±14.48)岁。所有患者术前均行平扫和增强结构MRI及DWI检查。术后对病灶MGMT分子表达进行免疫组织化学检测,将患者分为MGMT启动子高甲基化组(MGMT≤10%,n=54)和MGMT启动子低甲基化组(MGMT>10%,n=43)。采用ImageJ软件,逐层勾画肿瘤实质部分边界,将各层ROI组成三维的VOI,测量VOI的直方图参数,包括最大值、最小值、平均值、ADC_10th、ADC_20th、ADC_30th、ADC_40th、ADC_50th、ADC_60th、ADC_70th、ADC_80th、ADC_90th、ADC_峰度和ADC_变异系数。采用Mann-Whitney U检验比较MGMT启动子高甲基化组和MGMT启动子低甲基化组间各项直方图参数值的差异,进一步绘制受试者工作特征曲线(ROC),确定最佳诊断阈值,并计算相应的敏感度和特异度。结果MGMT启动子高甲基化组中肿瘤病灶的ADC_60th、ADC_70th、ADC_80th、ADC_90th和平均值均显著低于MGMT启动子低甲基化组(P=0.029、0.018、0.011、0.018、0.044),其中ADC_80th具有最大的曲线下面积(AUC=0.650),最佳的诊断灵敏度和特异度分别为62.79%和75.93%。结论基于ADC的三维直方图分析可用于预测胶质瘤MGMT启动子甲基化的程度,其中以ADC_80th的预测效能最好。Objective To explore the value of ADC-based three-dimensional histogram analysis in predicting the methylation degree of MGMT promoter in glioma.Methods MR imaging and clinicopathological data from 97 glioma patients confirmed by pathology were retrospectively analyzed,including 51 males and 46 females,with an average age of 51.14±14.48 years.All patients underwent non-enhanced and enhanced structural MRI and DWI scans before surgery.Postoperatively,the expression of MGMT molecules in the tumors was examined by immunohistochemistry.Patients were divided into MGMT promoter hypermethylation group(MGMT≤10%,n=54)and MGMT promoter hypomethylation group(MGMT>10%,n=43).Using ImageJ,we delineated the solid part of the tumor slice by slice,and the ROIs of each slice were composed into a three-dimensional VOI.The histogram parameters of VOI were measured,including maximum ADC,minimum ADC,average ADC,ADC_10th,ADC_20th,ADC_30th,ADC_40th,ADC_50th,ADC_60th,ADC_70th,ADC_80th,ADC_90th,ADC_kurtosis and ADC_variation.The Mann-Whitney U test was used to compare the differences in the histogram parameter values between the MGMT promoter hypermethylation group and the MGMT promoter hypomethylation group,and the receiver operating characteristic(ROC)curve was further used to determine the best diagnostic threshold,optimal sensitivity and specificity.Results The ADC_60th,ADC_70th,ADC_80th,ADC_90th and average ADC values of the tumor in the MGMT promoter hypermethylation group were significantly lower than those in the MGMT promoter hypomethylation group(P=0.029,0.018,0.011,0.018,0.044).ADC_80th had the highest area under the curve(AUC=0.650),with a optimal sensitivity and specificity of 62.79%and 75.93%,respectively.Conclusion ADC-based three-dimensional histogram parameters can be used to predict the degree of methylation of the MGMT promoter in glioma,and ADC_80th has a best predictive performance.

关 键 词：胶质瘤 MGMT 表观扩散系数 直方图分析 

分 类 号：R739.41[医药卫生—肿瘤]