Tumor Immunologic Heterogeneity Influences Response To Radiation And Anti-Pd-1 Immunotherapy

Clinical trials of CTLA-4 and PD-1/PD-L1 antibodies (Ab) have shown responses limited to 10-30% of patients. Increasing evidence suggests radiation (RT) can enhance responses to checkpoint therapies; however, understanding of tumor-derived factors that influence response is limited, and many preclinical models are manipulated to expressing specific antigens. We sought to develop a model of immunologic heterogeneity to study factors that influence responses to RT and immunotherapy. Tumor clones PyA1 and PyB2 exhibiting efficient orthotopic engraftment were derived from the PyMT mammary mouse carcinoma model. Cells were assayed by flow cytometry for MHCI and PD-L1 expression to evaluate antigen presentation and responses after interferon-γ (IFNγ) and 10 or 20 Gy RT treatment. Tumors were then grafted into naïve mice and radiated to 12 or 20 Gy. Immune infiltrates were evaluated by flow cytometry after tumor dissociation. With evidence of immunologic responses, mice were treated with 12 Gy and PD-1 Ab in addition to a single pretreatment dose of CTLA-4 Ab to evaluate combination therapy. To elucidate factors influencing responsiveness of combination treatment in PyA1 but not PyB2 tumors, we inoculated mixed tumor cell populations at 80/20, 50/50, and 20/80 ratios. PD-L1 and MHCI expression increased significantly after INF-γ and RT to a greater extent in PyA1 than PyB2 cells. Upon orthotopic implantation and RT, PyA1 but not PyB2 tumors regressed significantly. Dissociated tumors 10 days after 12 Gy treatment revealed a significant increase in CD45 leukocytes, CD8 effector T cells, and decreased CD4 T cells, myeloid derived suppressor cells, and macrophages in PyA1 tumors but minimal changes in PyB2 tumors. Secondly, there was elevated MHCI and induced PD-L1 expression in the responsive PyA1 compared to PyB2 tumors suggesting a greater antitumor response may be obtained with immunotherapy. Mice were treated with 12 Gy, 12 Gy + PD-1 Ab, and a single dose of CTLA-4 Ab 3 days prior to RT and PD-1 therapy. There was prolonged tumor suppression in the combination RT + PD-1 and RT + PD-1 + CTLA-4 groups with repeated measures P values of .0003 and .0039, respectively. Notably 2/8 tumors treated with RT + PD-1 and 6/8 mice treated with RT + PD-1 +CTLA-4 were undetectable at day 81. Mixing tumors with different ratios of PyA1 and PyB2 revealed that 50% inoculation of PyA1 is sufficient for an immune response against both tumor clones. This suggests a mechanism where responsive tumors can induce antitumor activity against unresponsive tumors. These results support that tumor immunologic heterogeneity can influence immune responses after radiation. This is an excellent model to study tumor-derived factors that enhance or suppress the immune response after RT and could inform clinical approaches to radiation and immunotherapy combinations.