Comparative and molecular approaches to improve identification, classification, and therapeutic approaches to cancer

Abstract

A major area of contemporary research in cancer is focused on improving tumor classification into clinically relevant subgroups of disease. To achieve this, it is important to understand the molecular events that driver tumor heterogeneity both at the cellular level and at the tissue level. I initially tested the hypothesis that canine lymphoma is composed of a group of molecularly distinct entities with prognostic significance. The results show that canine lymphoma can be stratified into molecular subgroups that have prognostic value and can assist to guide therapy. Next, I tested the hypothesis that canine hemangiosarcoma (HSA) is organized hierarchically with a cancer stem cell (CSC)-like population of cells at the apex. The data show that variable numbers of CSC-like cells are invariably present in HSA. These CSC-like cells retain the capacity to differentiate into vascular, inflammatory, or adipogenic tissue, suggesting that their multipotency is a contributing factor to the observed heterogeneity in this disease. Finally, I tested the hypothesis that CSCs, or CSC-like cells from three histologically distinct types of canine cancer (HSA, osteosarcoma, and glioblastoma) share molecular and functional properties. Using a system that allowed me to eliminate tumor-specific culture conditions, I showed that despite extensive heterogeneity in CSC-like cells from these tumors, they all showed reduced activity of pathways associated with proliferation and development. In summary, my results confirm that cellular heterogeneity exists both within and among tumors. A better understanding of the mechanisms that drive this will improve patient stratification and guide efforts to develop rational, more effective therapies.