Subdominant neo-epitopes in combination with checkpoint blockade modulators are protective in a model of pancreatic adenocarcinoma

Pancreatic adenocarcinoma (PDA) remains one of the most lethal cancer types, with patients presenting with late-stage disease at time of diagnosis. Surgical resection, the only curative treatment, is limited to patients without metastases, a minority of those diagnosed. PDA is refractive to radiation and chemotherapy due to the dense stroma that is a hallmark of the tumor microenvironment. Despite the abundance of novel therapies available for other tumors, little progress has been made in the development of treatments for PDA. Immunotherapy in the form of induced adaptive immunity has the potential to infiltrate the tumor microenvironment and overcome the characteristic resistance mechanisms of PDA. Mutations that arise during the development of PDA are specific to cancer cells and have therefore escaped the regulatory mechanisms that prevent cytolytic T cells from recognizing antigens that are also present on normal cells. Recent sequencing technology has allowed the genomic analysis of many cancers, including brain, breast, colorectal, ovarian, and pancreatic cancers. By targeting immunotherapy to the specific mutome of an individual tumor, a personalized treatment can be developed that is not limited by the presence of specific molecules involved with tumor cell proliferation, such as EGFR overexpression. Therapeutic vaccines for tumor immunotherapy rely on activating immune effector cells, especially CD4+ helper T cells and CD8+ cytolytic T cells, capable of specifically recognizing epitopes expressed by tumor cell antigens. An obstacle has been identifying epitopes that are specifically recognized by tumor cells and not by normal tissues. Although genome-wide sequencing technologies have shown that human pancreatic cancer cells are not considered a hypermutated cancer and the tumor is not well infiltrated with effector T cells, this thesis shows that subdominant epitopes are capable of inducing an anti-tumor response when coupled with appropriate immune-activating agents, including checkpoint blockade antibodies. Furthermore, we show that both unmutated and mutated tumor antigens can induce tumor-specific immunity, and that the addition of an agonist OX40 antibody fundamentally reprograms these effector cells to overcome tumor-induced exhaustion that synergizes with an anti-PD-1 antibody to provide near-complete, durable tumor rejection