A Novel Bioengineered Model Of Osteosarcoma Recapitulates Critical Tumor Phenotypes Within A Native Bone Microenvironment Alan Chramiec1, Alessandro Marturano-Kruik1,2, Luke Hao1, Keith Yeager1, Max Summers1, Miranda Wang1, Roberta Lock1, and Gordana Vunjak-Novakovic1,3 1Department of Biomedical Engineering, Columbia University, USA; 2 Department of Chemistry, Materials and Chemical Engineering “G Natta”, Politecnico de Milano, Italy; 3 Department of Medicine, Columbia University, New York, NY, USA Traditional in-vitro models of solid bone tumors used in basic and preclinical research are unable to faithfully recapitulate human physiology. Specifically, monolayer cultures of osteosarcoma fail to recapitulate various features of the 3D tumor phenotype, and they lack the native bone milieu, where tumor growth, metastasis, and response to therapy are critically dependent on cancer cell interactions with the bone matrix, supporting cells, and secreted regulatory factors. Based on our previous success in developing a biomimetic tissue model of Ewing sarcoma, we extended our approach towards bioengineering a novel human 3D tumor model of osteosarcoma within a bone microenvironment that would overcome some of these limitations. Briefly, osteosarcoma cell lines were used to generate tumor-like aggregates which were subsequently introduced into and cultured within engineered bone tissue scaffolds. Our bioengineered osteosarcoma model was able to recapitulate several aspects of native tumors and their microenvironment missing in monolayer cultures. Significantly higher expression of cancer-related genes (as seen in patients), re-activation of the hypoxic and glycolytic tumor phenotype, re-activation of hypoxia-mediated expression of vascular endothelial growth factor VEGF-α (a critical component of tumor angiogenesis), induction of vasculogenic mimicry, more physiologically relevant growth rates, and the adjacent presence of functional osteoblasts within a demineralized native bone scaffold were all achieved. Additionally, we were able to capture a both a degree of intra-tumor heterogeneity, with the presence of a necrotic core and variability in proliferation across the tumor aggregates, as well as inter-tumor heterogeneity, with notable differences observed between our models generated from metastatic and non-metastatic osteosarcoma cell lines. Finally, the osteosarcoma cell lines were transduced with a GFP-luciferase reporter, allowing us to visualize the engineered tumors in situ within the bone scaffolds, and to evaluate tumor cell responses to various canonical and experimental drug treatments. Notably, unlike monolayers, our bioengineered osteosarcoma model could be cultured stably for several weeks. This allowed us to apply a clinical treatment regimen cycle of linsitinib, an experimental drug in an ongoing phase II clinical trial for Ewing sarcoma, in also being effective at preventing the growth of non-metastatic osteosarcoma bioengineered tumors. Overall, we were able to successfully generate a new, more native-like 3D bioengineered model of osteosarcoma with significant advantages over traditional in-vitro culture that has the potential to explore novel questions regarding tumor biology, the tumor microenvironment, and translational applications.