Radiomic Analysis to Predict Outcome in Recurrent Glioblastoma Based on Multi-Center MR Imaging From the Prospective DIRECTOR Trial

摘要: Background Based on promising results from radiomic approaches to predict O6-methylguanine DNA methyltransferase promoter methylation status (MGMT status) and clinical outcome in patients with newly diagnosed glioblastoma, the current study aimed evaluate radiomics recurrent glioblastoma patients. Methods Pre-treatment MR-imaging data of 69 enrolled into DIRECTOR trial served as a training cohort, 49 independent formed an external validation cohort. Contrast-enhancing tumor peritumoral volumes were segmented MR images. 180 features extracted after application two intensity normalization techniques: fixed number bins linear rescaling. Radiomic feature selection was performed via principal component analysis, multivariable models trained MGMT status, progression-free survival first salvage therapy, referred herein PFS2, overall (OS). The prognostic power quantified concordance index (CI) for area under receiver operating characteristic curve (AUC) status. Results We established validated model using interpolation considering gadolinium-enhanced T1-weighted MRI (training AUC = 0.670, 0.673). Additionally, predicting PFS2 OS found cohort but not confirmed our Conclusions A prediction texture successfully established, providing non-invasive approach anticipate patient's response chemotherapy if biopsy cannot be performed. failed.