Recent evidence indicates that the presence of conventional dendritic cells type 1 (cDC1) in the tumor microenvironment (TME) is required for response to immune checkpoint blockade (ICB) therapy. In addition to cross- presenting cancer cell-derived antigens to CD8+ and CD4+ T cells, cDC1 promote tumor infiltration by effector T cells, and support their survival and function. Thus, interventions that improve cDC1 recruitment to the TME could enhance patient responses to ICB. Focal radiation therapy (RT) increases responses to ICB therapy, at least in part by inducing type I interferon (IFN-I) and driving cDC1 into the irradiated tumor. We have previously shown that cDC1 are essential for immune-mediated regression of irradiated and synchronous non-irradiated tumors (abscopal effect) in mice treated with RT and ICB. Abscopal responses have also been achieved in metastatic cancer patients treated with RT and ICB, but less reliably than expected, and the determinants of such responses remain unclear. We hypothesize that a previously unexplored barrier to abscopal responses is the limited infiltration of poorly immunogenic tumors by cDC1, which precludes effector T cells generated at the irradiated tumor site from rejecting non-irradiated tumors. Moreover, we hypothesize that activation of a strong IFN-I response in the irradiated tumor is essential for achieving systemic activation of natural killer (NK) cells, which can home to non-irradiated tumors and foster the recruitment of cDC1. This hypothesis is supported by a strong scientific premise which is based on the recent literature and on our extensive published and unpublished data, including the fact that increased serum IFNb post-RT was the top predictor for abscopal responses in metastatic lung cancer patients treated with RT+anti-CTLA4 (Nat Med 2018). To test this hypothesis three independent but related aims that address different mechanistic questions are planned. Aim 1 will investigate the role of RNA:DNA hybrids, which accumulate in the cytosol of irradiated cancer cells and in the cargo of small extracellular vesicles (sEV) they produce, in activating the IFN-I pathway via cGAS/STING in cancer cells and locoregional DCs. The role of RT-induced IFNb in systemic NK cell activation will be confirmed by using IFNAR1-deficient NK cells. Aim 2 will determine the contribution of sEV to RT-induced IFN-I activation in vivo by using Rab27a-deficient cancer cells. Aim 3 will directly address the role of NK cells in driving cDC1 infiltration in abscopal tumors and abscopal responses to RT+ICB. In addition, NK cell functional subsets present in the blood of lung cancer patients with abscopal response to RT+anti- CTLA4 will be investigated by single cell analysis. Results of proposed studies will identify a novel mechanism whereby local IFN-I induction by RT activates a systemic cross-talk between NK cells and cDC1, required for T-cell mediated rejection of abscopal tumors.