323 Pre-Operative MRI Radiomics Model Non-Invasively Predicts Key Genomic Markers and Survival in Glioblastoma Patients. (April 2023)

• Record Type: Journal Article
• Authors: Pease, Matthew; Gersey, Zachary Charles; Colen, Rivka R.; Zinn, Pascal O.
• Is Part Of: Neurosurgery. Volume 69(2023)Supplement 1
• Page Start: 47
• Page End: 47
• Publication Date: 2023-04
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1227/neu.0000000000002375_323
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract : INTRODUCTION: Although Glioblastoma (GBM) is the most common type of primary brain malignancy, few tools exist to pre-operatively risk stratify patients by overall survival (OS) or common genetic alterations. METHODS: We retrospectively identified 236 patients with pathologically confirmed GBMs from the Cancer Genome Atlas (89; TCGA) and MD Anderson Cancer Center (147; MDACC). After two neuro-radiologists segmented MRI tumor volumes, we extracted first order and second order radiomic features (gray-level co-occurrence matrixes). We used the Maximum Relevance Minimum Redundancy technique to identify the 100 most relevant features and validated models using leave one out cross validation (LOOCV) and validation on external datasets (i.e., TCGA). Our results were reported as area under the curve (AUC). RESULTS: MDACC has significantly higher OS (22 months) compared to the TCGA dataset (14 months). On both LOOCV and external validation, our radiomics models were able to identify EGFR amplification (all AUCs >0.83), MGMT methylation (all AUCs >0.85), GBM subtype (classical; proneural; mesenchymal; all AUCs >0.92), and OS (AUC >0.91 on LOOCV and 0.71 for TCGA validation). We developed a personalized radiomics "report card" that pre-operatively risk stratifies patients by likely genetic alteration. CONCLUSIONS: Our robust radiomics pipeline has the potential to pre-operatively discriminate common genetic alterations and patients with favorable survival.