ABSTRACT The complexity of cell-cell interactions in the tumor represents one of the biggest barriers to understanding cancer and to developing effective therapeutics. Cancer cells constantly interact with fibroblast cells, endothelial cells, immune cells, signaling molecules and the extracellular matrix in the tumor microenvironment (TME). The interactions between host and cancer cells are complex, with effects that may be tumor-suppressive or tumor-promoting. For example, macrophages are first recruited to fight cancer; however, interactions with cancer cells can render them tumor-supportive. Thus, enabling a greater understanding of the complexities of the interaction between cancer cells and their microenvironment can lead to a better understanding of mechanisms of drug resistance, identification of new molecular targets and help address the large unmet needs in treating cancer. Quantitative technologies, such as ours, that integrate ex vivo drug treatments and assess pharmacological responses with cellular and molecular phenotypes in native tissues, should accelerate the discovery and development of novel therapeutics. Yet present tools to study drug responses and the TME have not kept up with drug testing needs. Given the nearly infinite number of potential combinations and limited resources to actually test them, there is a great need for testing platforms that use human, intact tumor tissue ex vivo to predict in vivo responses to combination immunotherapies in miniaturized, multiplexed formats that retain as much of the TME as possible. In our first R01 cycle, we developed a microfluidic platform (called Oncoslice) that allows for selective spatiotemporal exposure of organotypic cultures to dozens of drug conditions, and we demonstrated its utility with cell death assays on xenograft and patient GBM slices. In this second R01 cycle, our goal is to apply our Oncoslice platform to evaluate combination immunotherapies and their interaction with the TME. The platform will be applied to two difficult-to-treat solid tumor types – pancreatic cancer (PCa, mouse, and human) and triple-negative breast cancer (TNBC, mouse). Our recent finding that intratumoral migration of CD8+ T cells is a necessary precursor to anti-tumor activity in response to immune checkpoint inhibitor therapy in pancreatic cancer lends credence to our assertion that developing a micro-scale understanding of cell-cell interactions in the TME is critical. Both PCa and TNBC are extremely heterogeneous and respond poorly to current immune checkpoint inhibitors. New developments to our platform will include live imaging of dynamic changes in the immune TME, tissue and cytokine sampling, selective transcriptomics, and protein pathway profiling. We will integrate phenotypic responses in slices with molecular data to build predictive statistical models. Together, these approaches will establish an innovative platform for immuno-modulatory drug discovery designed to provide insights i...