NIMG-62. RADIOMIC-BASED PROGRESSION-FREE SURVIVAL STRATIFICATION OF PEDIATRIC LOW-GRADE GLIOMA IS ASSOCIATED WITH MOLECULAR ALTERATIONS

Abstract Pediatric low-grade glioma (pLGG) encompasses a variety of tumor subtypes with heterogeneous treatment response and relatively long progression-free survival (PFS). Radiomics may serve as non-invasive in-vivo tool for early prediction PFS surrogate marker to objectively gauge the efficacy novel strategies. Here, we present multivariate model based on radiomic features clinical variables risk stratification pLGGs in terms seek associations predicted groups mutations key molecular markers using data from PedCBioportal. Pre-operative multi-parametric MRI scans (T1-pre, T1-post, T2, T2-FLAIR) 129 patients newly diagnosed pLGG (median age, 7.76, range, 0.35-19.58 years; median PFS, 28.5, 1.1-124.8 months) were collected quantitative (n = 881) extracted. A Cox proportional hazard’s (Cox-PH) regression was fitted (age, sex, extent resection) 4-fold cross-validation. subset 27) that most predictive selected by applying Elastic Net regularization penalty during Cox-PH fitting. High-, medium- low-risk determined predictions. modeling showed excellent performance suggested concordance index 0.78. Radiogenomic assessment (data available 94/129 patients) more enrichment NF1 RB1 genes high-risk group, compared low- medium-risk groups. We potential value radiomics providing upfront which further be used an added arm enrolled trials. In next step this work, will expand cohort cross-validate these results external cohort.