Abstract Immunotherapy is a promising therapeutic strategy for many cancers. Anti-tumor immune responses are elicited against neoantigens that arise from genetic alterations within tumor cells that give rise to a repertoire of peptides that the immune system recognizes as non-self and attack, killing the tumor cells harboring them. Cancers with a high mutational load generally benefit from immune checkpoint therapy, but some solid malignancies, like pancreatic cancer, have a low mutational burden that hampers immunotherapy. One strategy to overcome this limitation is to deliver exogenous neoantigens to cancer cells. Here we propose an approach that uses the tumor penetrating peptide iRGD, to deliver neoantigens to pancreatic tumors. Unlike conventional RGD peptides, iRGD, is not only able to target and the tumor vasculature through αv integrin, but also to extravasate and penetrate tumor tissue via neuropilin-1 delivering conjugated or co-administered drugs or peptides. Pancreatic cancer long term survivors often have neoantigens that mimic common viral epitopes, suggesting the possibility that these individuals benefited from an immune response against neoantigens that mimic viral components. We posit that iRGD-mediated delivery of these antigens to pancreatic tumors will redirect pre-existing antiviral immunity against them. For this purpose, we will use peptides activating cytomegalovirus (CMV)-specific T cells. CMV is a β-herpesvirus infects >60% of the population and elicits a strong immune response accounting for >10% of all circulating CD4 and CD8 T cells. Our preliminary data show that mice latently infected with CMV containing orthotopic pancreatic tumors respond to treatment with iRGD plus CMV peptides with tumor regression associated to increased necrosis, and marked T cell infiltration. Here we propose experiments to determine in an orthotopic pancreatic tumor mouse model, what are the best neoantigens and the optimal treatment conditions for achieving long lasting tumor regression (Aim 1). We will also assess the presence of CMV specific T cells in human pancreatic tumors, a necessary step toward the eventual translatability of this approach (Aim 2).