多参数弥散磁共振成像评估胶质瘤IDH1基因型及肿瘤增殖活性

Evaluation of IDH1 genotypes and tumor proliferation in gliomas by multiparametric diffusion magnetic resonance imaging

目的:探讨多参数弥散磁共振成像在评估胶质瘤异柠檬酸脱氢酶1(IDH1)基因型以及肿瘤细胞增殖活性中的应用价值。方法:回顾性分析106例胶质瘤患者的病例资料及术前高端弥散磁共振扫描数据。在评估IDH1基因型时,将其分成58例IDH1野生型和48例IDH1突变型两组。由于部分数据缺失,在评估肿瘤细胞增殖活性时,纳入94例患者,将其分为包括31例WHOⅡ级、20例Ⅲ级以及43例Ⅳ级三组。构建不同重建模型,包括弥散张量成像(DTI)、弥散峰度成像(DKI)、神经突方向离散与密度成像(NODDI)以及标准模型成像(SMI)。比较各模型衍生参数的组间差异及鉴别诊断效能。评价ki67标记指数(Ki67 LI)与各参数的相关性。结果:各模型各参数在IDH1野生型与突变型胶质瘤之间的差异均有统计学意义。其中SMI模型的轴突信号分数(f)具有最佳的诊断效能,AUC值为0.788,灵敏度为70.8%,特异度为75.9%。在较低级别胶质瘤(LrGG)亚组中,f具有最佳的诊断效能,AUC值为0.772,灵敏度为77.5%,特异度为70.6%。在胶质母细胞瘤(GBM)亚组中,各参数的组间差异均不具有统计学意义。随着病理级别升高,Ki67 LI显著升高。NODDI模型的细胞内体积分数(ICVF)具有最大的相关系数(0.579),与Ki67 LI呈显著正相关。结论:多参数弥散磁共振成像在评估胶质瘤IDH1基因型方面表现出良好的效能,其中SMI模型的f参数显著优于其他参数。ICVF在评估肿瘤细胞增殖活性方面表现出较大的潜力。

Objective:To explore the value of multiparametric diffusion magnetic resonance imaging(dMRI) in evaluating the isocitrate dehydrogenase 1(IDH1) genotypes and tumor cell proliferation in gliomas.Methods:The clinical and preoperative advanced dMRI data of 106 patients with gliomas were analyzed retrospectively.When evaluating IDH1 genotypes, patients were divided into two groups: 58 cases of IDH1 wild type and 48 cases of IDH1 mutant.Due to the lack of some data, 94 patients were included in the evaluation of tumor cell proliferative activity, and divided into three groups: 31 cases of WHO Ⅱ,20 cases of grade Ⅲ and 43 cases of grade Ⅳ.Different models of dMRI were reconstructed, including diffusion tensor imaging(DTI),diffusion kurtosis imaging(DKI),neurite orientation dispersion and density imaging(NODDI) and standard model imaging(SMI).The differences of the parameters derived from each model and the diagnosis efficiency were compared between groups.The correlations between Ki67 labeling index(Ki67 LI) and derived parameters were evaluated.Results:The parameters of each model showed significant differences between IDH1 wild type and mutant gliomas.Among them, fraction(f) derived from SMI had the best diagnostic efficiency, with AUC value of 0.788,sensitivity of 70.8% and specificity of 75.9%.In the lower grade glioma(LrGG),f had the best diagnostic efficacy, with AUC value of 0.772,sensitivity of 77.5%,and specificity of 70.6%.In the glioblastoma(GBM),the differences of parameters between groups were not statistically significant.With the increase of pathological grade, Ki67 LI increased significantly.Intracellular volume fraction(ICVF) derived from NODDI had the largest correlation coefficient of 0.579 and was significantly positively correlated with Ki67 LI.Conclusions:Multiparametric dMRI shows good efficacy in evaluating IDH1 genotypes of gliomas, and f derived from SMI is significantly better than those of others.ICVF shows great potential in evaluating the tumor cell proliferation.