Radiomics to understand pre-treatment tumor biology for resectable non-small cell lung cancer

Lung cancer is the leading cause of cancer related mortality in the United States. The standard of care for early-stage non-small cell lung cancer (NSCLC) is surgical resection. However, with the increasing usage of sublobar resections for small stage IA tumors and neoadjuvant chemoimmunotherapy for resectable stage IB-IIIA tumors, the selection of appropriate patients and the prediction of how they might respond to such therapies is vital. Radiomics is the usage of extracted imaging features as analyzable data. Radiomic modeling with machine learning and artificial intelligence can be used to provide non-invasive and predictive information about tumor biology and aggressiveness before treatment is initiated. Radiomic modeling has been utilized to identify NSCLC histopathological features, proteomic mutational burden, responsiveness to surgical and immunotherapy interventions, and occult lymph node metastasis. This review article provides an overview of studies using radiomic features to model risk and radiomic predictive tools such as the Computer-Aided Nodule Assessment and Risk Yield (CANARY) that have been developed to provide insight and pre-operative risk stratification for resectable NSCLC.