AUTHOR=Vils Alex , Bogowicz Marta , Tanadini-Lang Stephanie , Vuong Diem , Saltybaeva Natalia , Kraft Johannes , Wirsching Hans-Georg , Gramatzki Dorothee , Wick Wolfgang , Rushing Elisabeth , Reifenberger Guido , Guckenberger Matthias , Weller Michael , Andratschke Nicolaus 

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

JOURNAL=Frontiers in Oncology

VOLUME=11

YEAR=2021

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.636672

DOI=10.3389/fonc.2021.636672

ISSN=2234-943X

ABSTRACT=<sec><title>Background</title><p>Based on promising results from radiomic approaches to predict <italic>O<sup>6</sup>-methylguanine DNA methyltransferase promoter methylation status</italic> (<italic>MGMT</italic> status) and clinical outcome in patients with newly diagnosed glioblastoma, the current study aimed to evaluate radiomics in recurrent glioblastoma patients.</p></sec><sec><title>Methods</title><p>Pre-treatment MR-imaging data of 69 patients enrolled into the DIRECTOR trial in recurrent glioblastoma served as a training cohort, and 49 independent patients formed an external validation cohort. Contrast-enhancing tumor and peritumoral volumes were segmented on MR images. 180 radiomic features were extracted after application of two MR intensity normalization techniques: fixed number of bins and linear rescaling. Radiomic feature selection was performed <italic>via</italic> principal component analysis, and multivariable models were trained to predict <italic>MGMT</italic> status, progression-free survival from first salvage therapy, referred to herein as PFS<sub>2</sub>, and overall survival (OS). The prognostic power of models was quantified with concordance index (CI) for survival data and area under receiver operating characteristic curve (AUC) for the <italic>MGMT</italic> status.</p></sec><sec><title>Results</title><p>We established and validated a radiomic model to predict <italic>MGMT</italic> status using linear intensity interpolation and considering features extracted from gadolinium-enhanced T1-weighted MRI (training AUC = 0.670, validation AUC = 0.673). Additionally, models predicting PFS<sub>2</sub> and OS were found for the training cohort but were not confirmed in our validation cohort.</p></sec><sec><title>Conclusions</title><p>A radiomic model for prediction of <italic>MGMT</italic> promoter methylation status from tumor texture features in patients with recurrent glioblastoma was successfully established, providing a non-invasive approach to anticipate patient’s response to chemotherapy if biopsy cannot be performed. The radiomic approach to predict PFS<sub>2</sub> and OS failed.</p></sec>