Recent clinical studies indicate that immunotherapy utilizing blockade of checkpoint receptors on T cells may represent a promising new approach for cancer treatment.However, the response rates are still relatively low (e.g.~20% with PD-1 blockade in lung cancer).While these findings underscore the need for efficacious combinatory strategies, the best rationale for identification of optimal agents is not clear.We hypothesized that an unbiased in vitro screen to identify FDA-approved oncology agents with immuno-stimulatory properties may help select agents capable of promoting the anti-tumor immune response and the response to immunotherapy.By screening for ability to induce T cell chemokine expression in cancer cells, we found that only a single agent, the HDAC inhibitor romidepsin, induced expression of multiple T cell chemokines in a wide array of mouse and human lung cancer cells.Importantly, romidepsin ability to induce T cell chemokines was dependent on NF-B and STATI pathways.Consistent with modulation of these key immune surveillance pathways, genome-wide studies showed that romidepsin enhanced expression of a broad spectrum of immuno-stimulatory genes.Importantly, romidepsin induced a strong anti-tumor response in mice which was entirely dependent on T cell presence.Finally, romidepsin co-treatment significantly enhanced the response to PD-1 blockade immunotherapy.These findings provide proof-of-concept evidence for drug screening approaches based immuno-stimulatory ability of oncology agents.More specifically, these results indicate that combination of romidepsin with PD-1 blockade may represent a promising approach for lung cancer treatment.Using human lung adenocarcinoma samples, we are also investigating the impact of common oncogene and tumor-suppressor mutations on NF-B and STAT1 pathway activity.Our findings suggest that different mutations are critically associated with activity of these immune surveillance pathways and may therefore also impact the response to immunotherapy.