M. Teke1, A. L. Sarvestani1, A. Saif1, K. Remmert1, M. Garmendia-Cedillos1, R. Pursley1, E. Verbus1, S. Gregory1, C. Ryan1, M. Yaffe1, A. Blakely1, J. Davis1, J. Hernandez1 1National Cancer Institute, Surgical Oncology Program, Bethesda, MD, USA
Introduction:
To guide the application of immuno-oncology agents, a greater understanding of the tumor microenvironment (TME), including dynamic cell interactions, is required. Immunotherapies have had enormous impact on treatment of solid tumors, but some cancers are resistant to current treatments, and many individuals fail to respond to therapies that are highly beneficial to others. To address, we have engineered a translational platform that utilizes thin sections of human peritoneum containing macroscopically apparent tumor affixed to small platforms and placed into a chamber integrated into a perfusion circuit. This setup, termed SMART (Sustained Microenvironment for Analysis of Resected Tissue) System, utilizes perfusate made from the patient’s own plasma to simulate in vivo physiology, but it currently lacks the ability to model lymphocyte recruitment and infiltration from the periphery.
Methods:
Tumor-bearing peritoneum was affixed to a specialized 3D printed platform. Simultaneously, autologous peripheral blood mononuclear cells were isolated and labeled with a fluorescent cytoplasmic dye (f-PBMCs). To enable reproducible introduction of f-PMBCs into tissue matrices, we developed the SMART cell injector (SCI). The SCI integrates a commercially sourced microscope, fluorescent adapters, motorized micromanipulator and a microsyringe pump. We wrote code to automate and integrate the components for injection of f-PBMCs via pulled glass capillaries (tip diameter 50 um). Advanced white light laser confocal microscopy was used to elucidate the dynamic interaction between injected f-PBMCs and the TME.
Results:
Flow cytometry of the f-PBMCs identified labeling of greater than 90% of CD45+ cells, including T cells, B cells, natural killer cells, and macrophages. Depot injection of 5 million ± 0.5 million cells were introduced strategically adjacent to the centralized tumor nodule in the ostensibly normal peritoneal matrix. Using advanced confocal microscopy, we identified morphologic changes in the f-PBMCs at 4h ± 1h that consisted of elongation and alignment with matrix fibers. At 6h ± 1h, we observed motility of f-PBMCs along matrix fibers. After 48hrs within the SMART system, imaging revealed f-PBMCs infiltrating the centralized tumor distant from the injection site, demonstrating lymphocyte recruitment into the TME. Furthermore, a portion of the infiltrated f-PBMC were noted to interact and periodically “dock” onto tumor cells (Fig).
Conclusion:
We developed a technique to model peripheral lymphocyte recruitment into intact human tumors. We envision this technology to be highly informative for the evaluation of immune-oncology agents and cell therapies such as CAR-T products.
