Studies in osteosarcoma over the past 40 years have led to a steady improvement in the overall outcome of patients with osteosarcoma. In the year 2008, we can expect greater than 60% overall survival for newly diagnosed non-metastatic appendicular osteosarcoma. However, to achieve this current outcome, many patients are treated with aggressive cytotoxic chemotherapy and ultimately are not cured, and some patients who would be curable even without this aggressive approach are likely treated and cured. And finally, patients presenting with metastatic disease and those whose tumors recur after standard approaches continue to do very poorly. We believe that in order to continue to make progress in the treatment of this disease, we must achieve two main objectives. Firstly, we must find biomarkers that prospectively and accurately identify newly diagnosed non-metastatic patients who will not be cured with current modalities. We hope that the achievement of this goal will allow for innovative clinical studies in this high-risk population while not jeopardizing those patients who currently are cured using the available treatment approaches, and ultimately accelerate progress toward curing more patients. Secondly, we must develop entirely new approaches to the treatment of metastatic and recurrent osteosarcoma. Our approach has been to develop models of highly aggressive and less aggressive osteosarcoma, and to use these models to identify genetic alterations and signaling pathways that distinguish the two phenotypic behaviors. We have identified plasma membrane-cytoskeletal linker protein, ezrin, as one pathway that identifies aggressive biological behavior in mouse and dog osteosarcoma. Using ezrin as the initial discriminator, we have high ezrin expression to activation of mTOR signaling, suggesting a possible novel target for therapy of aggressive osteosarcoma. We have also linked β4 integrin signaling to metastatic behavior that also appears to be linked to mTOR signaling. Most recently, we have identified a critical relationship between mTOR signaling and the IGF I signaling pathway that may help point the way to combination targeting therapy aimed at blocking both mTOR and IGF signaling in these tumors. Finally, we have proposed a novel clinical trial design to begin to test agents targeted at recurrent, metastatic disease, and this also will be discussed.