DKI术前预测脑胶质瘤IDH突变状态的研究

Study of Diffusion Kurtosis Imaging Preoperatively Predicting IDH Mutation Status in Glioma

目的探讨术前磁共振扩散峰度成像(diffusion kurtosis imaging,DKI)对评估脑胶质瘤异柠檬酸脱氢酶(isocitrate dehydrogenase,IDH)突变状态的临床应用价值。方法纳入行DKI及常规磁共振成像(magnetic resonance imaging,MRI)平扫及增强扫描的71例(其中IDH突变型41例,IDH野生型30例)经手术及病理检测的脑胶质瘤患者。运用后处理工作站分别测量肿瘤实质区、瘤周水肿区及对侧正常脑白质区的平均峰度(mean kurtosis,MK)、轴向峰度(axial kurtosis,Ka)、径向峰度(radial kurtosis,Kr)、峰度各向异性(kurtosis anisotropy,FAK)、平均扩散系数(mean diffusivity,MD)、各向异性分数(fractional anisotropy,FA)、轴向扩散系数(axial diffusivity,Da)和径向扩散系数(radial diffusivity,Dr),并规范化肿瘤实质区(肿瘤实质区/对侧正常脑白质区)的各参数值。运用统计学软件对所得参数进行各组间比较,并和脑胶质瘤IDH突变状态进行相关性以及各参数的诊断效能分析。结果在肿瘤实质区,IDH突变型规范化FA、MK、Ka和Kr值低于IDH野生型,而规范化Dr值则高于IDH野生型(P<0.05)。规范化FA、MK、Ka和Kr值与IDH突变状态呈负相关(P<0.01),而规范化Dr值与其呈正相关(P<0.05)。对IDH突变状态的评估,各参数中均以MK值诊断效能最好,其曲线下面积、诊断阈值、敏感度、特异性分别为0.794、0.518、81.1%、75.8%。DKI预测IDH突变状态的准确性为77.1%。结论DKI是术前无创评估脑胶质瘤IDH突变状态的有效影像学方法,通过肿瘤弥散峰度等信息的检测,判断其IDH突变状态,为患者精准诊治提供有效依据。

Objective To investigate the clinical application value of preoperative magnetic resonance diffusion kurtosis imaging(DKI)in evaluating isocitrate dehydrogenase(IDH)mutation status.Methods Seventy-one patients with surgical and pathologically confirmed gliomas(41 cases of IDH mutation,30 cases of IDH wild type)were received conventional magnetic resonance imaging(MRI)plain scan,enhanced scan and DKI.Mean kurtosis(MK),axial kurtosis(Ka),radial kurtosis(Kr),kurtosis anisotropy(FAK),mean diffusivity(MD),fractional anisotropy(FA),axial diffusivity(Da)and radial diffusivity(Dr)parameter images were used to measure the parameter values of the tumor lesion area,peripheral edema,and contralateral normal-appearing white matter(CNAWM),and to normalize the parameter values of tumor lesion area(tumor lesion area/CNAWM).Statistical software was used to compare the parameters among each group,and the correlation between the parameters and IDH mutation status of glioma was analyzed as well as the diagnostic efficacy of each parameter.Results In the tumor lesion area,the normalized FA,MK,Ka and Kr values of IDH mutation were lower than IDH wild type,the normalized value of Dr was higher than those of IDH wild type(P<0.05).The normalized FA,MK,Ka and Kr values were negatively correlated with IDH mutation status(P<0.01),while normalized Dr value was positively correlated with IDH mutation status(P<0.05).In the evaluation of IDH mutation status,MK value was the best among all parameters,and its area under curve,diagnostic threshold,sensitivity and specificity were 0.794,0.518,81.1%,75.8%.The accuracy of DKI in predicting IDH mutation status was 77.1%.Conclusion DKI is an effective imaging method for preoperative noninvasive evaluation of IDH mutation status of glioma.Through the detection of tumor diffusion kurtosis and other information,the IDH mutation status can be determined,providing an effective basis for accurate diagnosis and treatment of patients.