From the Department of Radiology and Biomedical Imaging (J.L., T.V., L.J., M.v.R., N.T., S.M., G.C.P., R.B., A.G., M.A.H., H.S., W.B., B.V.M.-N., A.A., M.L., M.A.), Yale School of Medicine, New Haven, Connecticut.
Department of Neurosurgery (J.L., M.S.), Heinrich-Heine-University, Duesseldorf, Germany.
AJNR Am J Neuroradiol. 2023 Oct;44(10):1126-1134. doi: 10.3174/ajnr.A8000.
The molecular profile of gliomas is a prognostic indicator for survival, driving clinical decision-making for treatment. Pathology-based molecular diagnosis is challenging because of the invasiveness of the procedure, exclusion from neoadjuvant therapy options, and the heterogeneous nature of the tumor.
We performed a systematic review of algorithms that predict molecular subtypes of gliomas from MR Imaging.
Data sources were Ovid Embase, Ovid MEDLINE, Cochrane Central Register of Controlled Trials, Web of Science.
Per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, 12,318 abstracts were screened and 1323 underwent full-text review, with 85 articles meeting the inclusion criteria.
We compared prediction results from different machine learning approaches for predicting molecular subtypes of gliomas. Bias analysis was conducted for each study, following the Prediction model Risk Of Bias Assessment Tool (PROBAST) guidelines.
Isocitrate dehydrogenase mutation status was reported with an area under the curve and accuracy of 0.88 and 85% in internal validation and 0.86 and 87% in limited external validation data sets, respectively. For the prediction of promoter methylation, the area under the curve and accuracy in internal validation data sets were 0.79 and 77%, and in limited external validation, 0.89 and 83%, respectively. PROBAST scoring demonstrated high bias in all articles.
The low number of external validation and studies with incomplete data resulted in unequal data analysis. Comparing the best prediction pipelines of each study may introduce bias.
While the high area under the curve and accuracy for the prediction of molecular subtypes of gliomas are reported in internal and external validation data sets, limited use of external validation and the increased risk of bias in all articles may present obstacles for clinical translation of these techniques.
脑肿瘤的分子特征是预测生存的指标，指导着治疗的临床决策。由于该检测程序具有侵袭性、不能用于新辅助治疗选择，以及肿瘤存在异质性，因此基于病理的分子诊断具有挑战性。
我们对从磁共振成像预测脑胶质瘤分子亚型的算法进行了系统评价。
数据来源为 Ovid Embase、Ovid MEDLINE、Cochrane 中央对照试验注册库、Web of Science。
根据系统评价和荟萃分析的首选报告项目（PRISMA）指南，筛选了 12318 篇摘要，对 1323 篇进行了全文审查，其中 85 篇文章符合纳入标准。
我们比较了不同机器学习方法预测脑胶质瘤分子亚型的预测结果。按照预测模型风险偏倚评估工具（PROBAST）指南，对每项研究进行了偏差分析。
内部验证中，异柠檬酸脱氢酶突变状态的曲线下面积和准确率分别为 0.88 和 85%，有限外部验证数据集中分别为 0.86 和 87%。对于启动子甲基化的预测，内部验证数据集中的曲线下面积和准确率分别为 0.79 和 77%，在有限的外部验证中，分别为 0.89 和 83%。PROBAST 评分显示所有文章都存在高度偏倚。
外部验证数量少，且数据不完整，导致数据分析不均衡。比较每项研究中最佳预测方案可能会引入偏差。
尽管内部和外部验证数据集中报告了较高的曲线下面积和预测脑胶质瘤分子亚型的准确率，但所有文章中有限的外部验证和增加的偏倚风险可能会对这些技术的临床转化构成障碍。
