PROJECT SUMMARY/ABSTRACT Immunotherapies hold immense promise to provide cures for many cancers and metastatic disease, but only benefit a fraction of patients. Tumors can still engage multiple mechanisms to avoid and escape anti-tumor immune responses, including suppression, inactivation, and exclusion of potential cytotoxic T cells, processes which collaborate with cells in the tumor microenvironment (TME). A better understanding of these barriers has led to a multitude of new immunomodulatory targets to be developed, some to be used in combination with e.g., checkpoint blockade or CAR T cells. On the other hand, dominant barriers to immunotherapy can be different among patients with the same cancer type, and thus there is a need for personalized approaches to immunotherapy, so that the appropriate targets are used. Here we develop a novel organotypic culture devices to maintain ex vivo cultures of primary tumors and an immune component (tumor-draining lymph nodes or circulating leukocytes), on a platform that enables precise control over spatial, molecular, cellular, and mechanical characteristics and that is relatively high-throughput to allow screening or large numbers of experimental variables. In preliminary data, we show that these devices mirror key features of in vivo responses to immunotherapy, such as improved tumor cell killing and increased markers of immunotoxicity (possible adverse events) in response to cytokine immunotherapy. We propose that these devices can be used both to screen for ideal immunotherapy combinations as well as to probe the basic mechanisms underlying the deficiencies in the anti-tumor immune response for tumors exhibiting varying levels of immune infiltration, neoantigen load, and baseline lymphatic densities. In this way, we can begin to build a stratification map that aligns key morphological features of individual tumors to treatment regimes that are most likely to lead to efficacy and tumor regression. Using a combination of both murine mouse models and primary patient-derived biospecimens, we will take advantage of the level of control afforded by our novel organotypic devices to mechanistically interrogate individual immune cell subsets and signaling axes, towards understanding their roles in influencing the course and outcomes of anti-tumor immune responses.