NIMG-56. PHENOTYPIC ANALYSIS OF RADIO-PATHOMIC MAPS IN DE-NOVO GLIOBLASTOMA IDENTIFIES DIFFERENCES IN BEVACIZUMAB TREATMENT RESPONSE

Abstract PURPOSE Bevacizumab (Bev) treatment for recurrent glioblastoma ceases tumor angiogenesis, which changes appearance on Gd enhanced T1-weighted imaging. However, little is known about how non-angiogenic portions of are affected by Bev. Therefore, this study tested the hypothesis that patients with specific phenotypes outside contrast enhancement, as identified using autopsy tissue-based radio-pathomic maps, would show differences in survival benefit from Bev treatment. METHODS Previously published methods were used to create machine learning models generate whole-brain maps cellularity (Cell), extracellular fluid (ECF) and cytoplasm (Cyt) glioma tissue aligned patient’s last MRI ground truth (Bobholz et al. 2022). These then generated an independent dataset (n=80) baseline imaging acquired prior Patients graded qualitative characteristics non-enhancing margin, including Well-Circumscribed (WC) no abnormal surrounding core, Hypercellular Front (HF) areas hypercellularity extending Necrotic (NF) necrosis Hybrid (HYF) both hypercellular necrotic presence beyond enhancing region. Kaplan Meier analysis was compare outcomes between who did not receive within each phenotype assess efficacy. RESULTS Only NF or HYF showed significant increase bevacizumab (HR=2.35, p=0.02; HR=2.45, p=0.03, respectively), trending difference observed WC HF patients. CONCLUSION Radio-pathomic identify greatest treatment, could be direct clinical decision-making.