Lung cancer is the most lethal malignancy in Canada, accounting for more deaths than colon, prostate and breast cancer combined. Unfortunately, the vast majority of patients are initially diagnosed with advanced disease, which make them not amenable to a curative surgical approach, and only eligible for systemic treatments. In the past few years, treatments targeting the immune system have drastically modified the therapeutic landscape of advanced non-small lung cancer (NSCLC), notably with the use of immune checkpoint inhibitors (ICIs). These drugs facilitate the recognition of tumor cells by the immune system, enabling immune cells to eliminate them more effectively. However, only 20 to 30 % of patients will benefit from this type of treatment: it is therefore critical to accurately identify those who will be susceptible to respond to these expensive treatments. This project aims to develop and validate a clinical tool relying on artificial intelligence (AI) technology to generate a model integrating clinical, radiological and molecular data of a patient and his tumor in order to predict the likelihood of treatment response. We will first fully characterize the tumor and the clinical profile of patients treated with ICI after a diagnosis of advanced NSCLC. Parameters from radiologic features, tumor immune microenvironment as well as molecular profile in relation to response to treatment will be analysed in order to generate a predictive biomarker algorithm based on artificial  intelligence. In the era of precision medicine, this research will allow us not only to improve patient’s management but also to optimize the costs related to these expensive treatments. In order to successfully carry out this project, we will bring together the expertise in radiology, pathology, oncology, immunology and big data management from three Quebec
academic hospitals as well as from a world-leader biotech in the fields of artificial intelligence and oncology.