Figure 1

Personalized medicine approaches in which therapies are matched to individual patient characteristics hold tremendous promise for the treatment of solid tumors. These approaches most often use genomic profiling of patient tumors to match individual tumors to treatment; however, these personalized strategies are limited by substantial intratumoral heterogeneity, both genomically and phenotypically. Genomically, tumors can be made up of many genetically-distinct subclones, each harboring tens to hundreds of mutations and with distinct gene expression pathways and signaling. Phenotypically, these subclones possess diverse responses to tumor microenvironment signaling, immune infiltration, and treatment. This genomic and phenotypic heterogeneity poses a substantial barrier to the broad implementation of personalized medicine strategies for many solid tumors.

This presentation will summarize translational work using patient-derived models of cancer, genomics, and comparative oncology to develop a personalized medicine platform for identifying new therapies to treat osteosarcoma, an aggressive bone malignancy of canines and humans. This platform provides a framework to reimagine the limitation of intratumoral heterogeneity not as a burden, but as an advantage.

Specifically, by using patient-derived models to simultaneously profile, both genomically and phenotypically, multiple unique locations within a single tumor, the possibility exists to develop a rapid, predictive biomarker of therapy response that can also identify and validate new therapies in a personalized way. The potential advantages and current challenges of these strategies will be discussed, focusing on paths to overcome remaining barriers to successfully implement these platforms.