Cancer Immunotherapy Is More Effective in Irradiated Tumors in Preclinical Mouse Model of Melanoma

Abstract

Post-radiotherapy local recurrence is a common challenge in cancer management. Since radiotherapy induces DNA damage, tumor cells that survived radiotherapy have higher mutation burden. It is recognized that high mutation burden and high clonal neoantigen burden are associated with checkpoint inhibitor sensitivity. Therefore, we hypothesized that tumors that survived radiotherapy are more responsive to immune checkpoint treatment and cancer immunotherapy maybe an effective treatment for local recurrences after radiotherapy. To simulate recurrent tumor post radiotherapy, we irradiated B16F10 melanoma cells with 5 Gy and selected the surviving colonies for further passage. This process was repeated twice to obtain the final irradiated B16F10 cells. These irradiated B16F10 cells and parent B16F10 cells were used to generate flank xenograft tumors (5*104) in B57BL/6 mice. The mice were then treated with saline, αPD-L1 (10mg/kg), αCTLA4(5mg/kg) or αPD-L1+αCTLA4 on days 3, 6 and 9. Tumor volumes were obtained for each mouse. The tumor volumes were assessed every 2–3 days starting day 3. In the irradiated B16F10 melanoma model, our results demonstrate that both αCTLA4 alone and combination groups achieved tumor growth control (p<0.005), and there is also significant survival difference (p<0.005) with median survival 22.5, 26, 35 and 38.5 days in control, αPD-L1, αCTLA4, and combination subgroups. In the parent B16F10 melanoma model, immunotherapy treatments improved survival, though the increases were not statistically significant. Immunotherapy treatment arms were significantly more effective in irradiated B16F10 model (p value<0.005). We demonstrated that irradiated B16F10 tumors are more responsive to checkpoint inhibitor treatment than non-irradiated tumors. This finding suggest that immunotherapy may be effective for recurrent tumors after radiation. Our results are preliminary and further preclinical and clinical validation are needed.