Project 460772

Cancers that arise from cells of the immune system are broadly divided into leukemias and lymphomas and the latter forms solid tumours at various places in the body. The most common variety of lymphoma that originates from B cells is known as diffuse large B-cell lymphoma (DLBCL). The genetic changes (mutations) that convert a healthy B cell to DLBCL can affect any of more than 100 genes. Recent studies have led to the model that DLBCL is a collection of diseases that appear similar under a microscope but have unpredictable responses to therapy. Some cases of DLBCL evolve from a slow-growing (indolent) cancer years after a patient is diagnosed. Additional DLBCLs share many genetic features with indolent lymphomas but can appear without warning. The genetic similarities and differences between cancers can be inferred by comparing the patterns of mutations across the entire genome of tumours from many patients. We will use data from over 3000 lymphoma patients and will generate data from additional lymphomas for which such information is currently lacking. We will use a combination of artificial intelligence (AI) techniques to identify genetic features and patterns that are shared among some DLBCL patients and other lymphomas. Our goal is to refine existing categorization methods to more accurately describe the nature and complexity of this cancer. By better understanding how DLBCL can evolve from indolent lymphomas, we may be situated to predict which patients are in need of an alternative treatment strategy and aid in identifying more suitable therapies when patients need them.