NIMG-85. RADIOMIC FEATURES PREDICTIVE OF RESPONSE IN HGG-TARGETING CAR-T THERAPY

Abstract SIGNIFICANCE Radiomics may improve precision medicine in CAR-T (Chimeric Antigen Receptor T Cell) therapy patient selection. BACKGROUND High-Grade Glioma (HGG) is a heterogenous primary CNS neoplasm with high recurrence rate and poor outcomes. Many studies are exploring CAR cells to combat HGG. Radiomic models have shown value identifying biomarkers predictive of tumor genetics, response, prognosis. In this study, we explore radiomic features derived from four clinical sequences volumes edema enhancing predict treatment response the first three doses therapy. METHODS IRB-approved phase 1 trial (IRB 13384), patients underwent surgical resection received cell Of 82 accrued, 59 (20 females, median age = 49) completed cycles had 3T field strength MRI scans minimal imaging artifacts. T1 weighted pre-contrast, post-contrast, T2 weighted, FLuid Attenuated Inversion Recovery were used generate 3D 2D radiomics features. total 28,541 generated per using images prior administration. Each patient’s after was determined be either stable disease (29 patients) or progression. The feature set dimensionality reduced Maximum Relevance Minimum Redundancy. 10-fold cross-validation XGBoost determine randomized grid search for hyperparameter tuning. RESULTS Six (four shape-based), SHapley Additive exPlanations (SHAP)-based importance RANO an AUC >0.73. CONCLUSION Despite limited study size, such imaging-based can serve as potential basis optimizing design through more precise screening providing whether will respond HGGs.