Search

PurposeMalignant head and neck squamous cell carcinoma (HNSCC) is characterized by a poor prognosis and resistance to conventional radiotherapy. Infiltrating myeloid-derived suppressive cells (MDSCs) is prominent in HNSCC and is linked to immune suppression and tumor aggressiveness. This study aimed to investigate the impact of boron neutron capture therapy (BNCT) on the MDSCs in the tumor microenvironment and peripheral blood and to explore the potential for MDSCs depletion combined with BNCT to reactivate antitumor immunity.Methods and materialsCarcinogen, 4-NQO, -induced oral tumors were irradiated with a total physical dose of 2 Gy BNCT in Tsing Hua Open Reactor (THOR). Flow cytometry and immunohistochemistry accessed the dynamics of peripheral MDSCs and infiltrated MDSCs within the tumor microenvironment. Mice were injected with an inhibitor of CSF-1 receptor (CSF-1R), PLX3397, to determine whether modulating M-MDSCs could affect mice survival after BNCT.ResultsPeripheral CD11b+Ly6ChighLy6G- monocytic-MDSCs (M-MDSCs), but not CD11b+Ly6CloLy6Ghigh polymorphonuclear-MDSCs (PMN-MDSCs), increased as tumor progression. After BNCT treatment, there were temporarily decreased and persistent increases of M-MDSCs thereafter, either in peripheral blood or in tumors. The administration of PLX-3397 hindered BNCT-caused M-MDSCs infiltration, prolonged mice survival, and activated tumor immunity by decreasing tumor-associated macrophages (TAMs) and increasing CD8+ T cells.ConclusionM-MDSCs were recruited into 4-NQO-induced tumors after BNCT, and their number was also increased in peripheral blood. Assessment of M-MDSCs levels in peripheral blood could be an index to determine the optimal intervention window. Their temporal alteration suggests an association with tumor recurrence after BNCT, making M-MDSCs a potential intervention target. Our preliminary results showed that PLX-3397 had strong M-MDSCs, TAMs, and TIL (tumor-infiltrating lymphocyte) modulating effects that could synergize tumor control when combined with BNCT.

Tumor vasculatures and hypoxia are critical tumor micro-environmental factors associated with tumor response to the therapy and heterogeneous in both time- and location-dependent manner. Using a murine orthotopic anaplastic astrocytoma model, ALTS1C1, this study showed that brain tumor edge had a very unique microenvironment, having higher microvascular density (MVD) and better vessel function than the tumor core, but on the other hand was also positive for hypoxia markers, such as pimonidazole (PIMO), hypoxia inducible factor-1α (HIF-1α), and carbonic anhydrase IV (CAIX). The hypoxia at tumor edge was transient, named as peripheral hypoxia, and caused by different mechanisms from the chronic hypoxia in tumor core. The correlation of CAIX staining with astrocyte activation marker, glial fibrillary acid protein (GFAP), at the tumor edge indicated the involvement of astrocyte activation on the development of peripheral hypoxia. Peripheral hypoxia was a specific trait of orthotopic brain tumors at tumor edge, regardless of tumor origin. The hypoxic cells were resistant to the therapy, regardless of their location. Surviving cells, particularly those at the hypoxic region of tumor edge, are likely the cause of tumor recurrence after the therapy. New therapeutic platform that targets cells in tumor edge is likely to achieve better treatment outcomes.

Macrophages display different phenotypes with distinct functions and can rapidly respond to environmental changes. Previous studies on TRAMP-C1 tumor model have shown that irradiation has a strong impact on tumor microenvironments. The major changes include the decrease of microvascular density, the increase of avascular hypoxia, and the aggregation of tumor-associated macrophages in avascular hypoxic regions. Similar changes were observed no matter the irradiation was given to tissue bed before tumor implantation (pre-IR tumors), or to established tumors (IR tumors). Recent results on three murine tumors, TRAMP-C1 prostate adenocarcinoma, ALTS1C1 astrocytoma, and GL261 glioma, further demonstrate that different phenotypes of inflammatory cells are spatially distributed into different microenvironments in both IR and pre-IR tumors. Regions with avascular hypoxia and central necrosis have CD11bhigh/Gr-1+ neutrophils in the center of the necrotic area. Next to them are CD11blow/F4/80+ macrophages that sit at the junctions between central necrotic and surrounding hypoxic regions. The majority of cells in the hypoxic regions are CD11blow/CD68+ macrophages. These inflammatory cell populations express different levels of Arg I. This distribution pattern, except for neutrophils, is not observed in tumors receiving chemotherapy or an anti-angiogenesis agent which also lead to avascular hypoxia. This unique distribution pattern of inflammatory cells in IR tumor sites is interfered with by targeting the expression of a chemokine protein, SDF-1α, by tumor cells, and this also increases radiation-induced tumor growth delay. This indicates that irradiated-hypoxia tissues have distinct tumor microenvironments that favor the development of M2 macrophages and that is affected by the levels of tumor-secreted SDF-1α. [ABSTRACT FROM AUTHOR]