ABSTRACT T CELL IMMUNOSURVEILLANCE IN PANCREATIC DUCTAL ADENOCARCINOMA Rebecca A. Evans Robert H. Vonderheide The prevailing theory of cancer immune surveillance, as understood from carcinogen-driven mouse models of highly mutated tumors, states that T cell-mediated immune pressure drives a continuum of tumor elimination, equilibrium, and escape. This “Triple E Hypothesis” reflects host-tumor interactions in a subset of human cancers that have responded well to cancer immunotherapy, including melanoma, small cell lung carcinoma, and bladder carcinoma; yet, many tumors remain refractory to such interventions. These clinical failures suggest that immunosurveillance in other cancers manifests with a fundamentally different biology than previously described. Here, using a genetic mouse model of spontaneous pancreatic carcinoma that features an immunosuppressive microenvironment and few non-synonymous mutations, we report that the natural history of such cancers is T cell-independent. Furthermore, tumor escape from T cell surveillance is not required for cancer progression, as tumor cells arising in T cell-depleted genetic mice grow unchecked in immune-competent hosts upon implantation. Checkpoint blockade with CTLA-4 and PD-1 antibodies does not expose mutant epitopes strong enough to elicit therapeutic responses; combined with whole exome sequencing of PDA-derived murine cell lines, these findings confirm that that PDA does not harbor a mutational burden commensurate with the current hypothesis of immunosurveillance. However, ectopic expression of a neo-antigen in PDA tumor cells is sufficient to restore immunosurveillance, override its immunosuppressive microenvironment, and establish T cell memory against “quiescent” endogenous antigens. Thus, cardinal features of tumor immunosurveillance are elicited by an antigen of sufficient strength irrespective of microenvironmental immunosuppression. The finding that antigen strength itself is a critical determinant of cancer immunosurveillance informs future clinical approaches for the majority of human tumors that are “immunologically cold” and refractory to current immunotherapies.