Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Olympic-ro 33, Seoul 05505, Republic of Korea (C.H.S., H.S.K., J.E.P., S.C.J., C.G.C., H.B.L., S.J.K.), and Bioimaging Center, Biomedical Research Center, Asan Institute for Life Sciences, Asan Medical Center, Seoul, Republic of Korea (D.C.W.).
Radiol Imaging Cancer. 2020 Jul 31;2(4):e190083. doi: 10.1148/rycan.2020190083. eCollection 2020 Jul.
To compare the ability of 2-hydroxyglutarate (2HG)-to-lipid and lactate (2HG/[lipid + lactate]) ratio with the ability of 2HG concentration alone to predict the isocitrate dehydrogenase mutation status in patients with glioma.
In this retrospective study, consecutive patients with histopathologically proven glioma were enrolled between July 2016 and February 2019. A total of 79 patients were enrolled (mean age, 44 years; 49 men). The 2HG concentration and other MR spectroscopic parameters were measured by single-voxel point-resolved spectroscopy before surgery. The diagnostic performance of the 2HG concentration and 2HG/(lipid + lactate) ratio were calculated. Internal validation was assessed by the bootstrap approach with 1000 bootstrap resamples. Differences in the predictive accuracy of 2HG/(lipid + lactate) ratio and 2HG concentration were determined by calculating the integrated discrimination improvement. The diagnostic accuracy (sensitivity, specificity, and area under the receiver operating characteristic curve [AUC]) of these measures was also compared separately in patients with glioblastomas and patients with lower-grade gliomas.
Of the 79 enrolled patients, 28 had mutations and 51 had wild-type . The sensitivity, specificity, and AUC of 2HG concentration for predicting -mutant gliomas were 89% (25 of 28), 67% (34 of 51), and 0.80 (95% confidence interval [CI]: 0.70, 0.88; C statistic, 0.80), respectively. The sensitivity, specificity, and AUC of the 2HG/(lipid + lactate) ratio for predicting -mutant gliomas were 79% (22 of 28), 92% (47 of 51), and 0.90 (95% CI: 0.81, 0.96; C statistics, 0.90), respectively. The optimal cutoff value for the 2HG/(lipid + lactate) ratio was 0.63. The 2HG/(lipid + lactate) ratio was significantly better for predicting mutation status than the 2HG concentration alone ( < .01). In glioblastoma, the 2HG/(lipid + lactate) ratio was also better for predicting mutations than the 2HG concentration alone, with borderline significance ( = .052). In lower-grade glioma, the 2HG/(lipid + lactate) ratio and the 2HG concentration showed comparable diagnostic performance ( = .72).
The 2HG/(lipid + lactate) ratio is more accurate for predicting mutation status in patients with glioma than the 2HG concentration alone. Brain/Brain Stem, CNS, MR-Imaging, MR-Spectroscopy, Neoplasms-Primary, Neuro-Oncology© RSNA, 2020.
比较 2-羟戊二酸（2HG）与脂质的比值（2HG/[脂质+乳酸盐]）与 2HG 浓度单独预测脑肿瘤患者异柠檬酸脱氢酶突变状态的能力。
本回顾性研究纳入了 2016 年 7 月至 2019 年 2 月间经组织病理学证实的连续脑胶质瘤患者。共纳入 79 例患者（平均年龄 44 岁，49 例男性）。手术前通过单体素点分辨波谱法测量 2HG 浓度和其他磁共振波谱参数。计算 2HG 浓度和 2HG/（脂质+乳酸盐）比值的诊断性能。通过 1000 次 bootstrap 重采样的 bootstrap 方法评估内部验证。通过计算综合判别改善来确定 2HG/（脂质+乳酸盐）比值和 2HG 浓度预测准确性的差异。还分别比较了这些措施在胶质母细胞瘤患者和低级别胶质瘤患者中的诊断准确性（敏感性、特异性和受试者工作特征曲线下面积[AUC]）。
在纳入的 79 例患者中，28 例有 突变，51 例有野生型。2HG 浓度预测 -突变型脑肿瘤的敏感性、特异性和 AUC 分别为 89%（25/28）、67%（34/51）和 0.80（95%置信区间[CI]：0.70，0.88；C 统计量，0.80）。2HG/（脂质+乳酸盐）比值预测 -突变型脑肿瘤的敏感性、特异性和 AUC 分别为 79%（22/28）、92%（47/51）和 0.90（95%CI：0.81，0.96；C 统计量，0.90）。2HG/（脂质+乳酸盐）比值的最佳截断值为 0.63。2HG/（脂质+乳酸盐）比值预测 突变状态明显优于单独的 2HG 浓度（<0.01）。在胶质母细胞瘤中，2HG/（脂质+乳酸盐）比值也优于单独的 2HG 浓度预测 突变，具有边缘显著性（=0.052）。在低级别胶质瘤中，2HG/（脂质+乳酸盐）比值和 2HG 浓度的诊断性能相当（=0.72）。
与单独的 2HG 浓度相比，2HG/（脂质+乳酸盐）比值更能准确预测脑肿瘤患者的 突变状态。脑/脑干、中枢神经系统、磁共振成像、磁共振波谱、肿瘤-原发性、神经肿瘤学©RSNA，2020。
