"Neoantigen-specific T cell responses can drive tumor control, yet the ways in which immunity shapes tumor evolution and neoantigen quality remain poorly understood. To dissect immune driven selection, we studied three Kras/p53 driven lung cancer lines (KPA, KPC, HKP1) implanted into immunocompetent or immunodeficient mice. These models displayed a spectrum of immunogenicity: KPA was highly immunogenic, with low mutational burden, robust T cell infiltration, and complete responsiveness to immune checkpoint blockade (ICB); KPC showed intermediate sensitivity; and HKP1 was immune cold, enriched for regulatory T cells, and resistant to ICB. Vaccination with irradiated KPA conferred protection against KPA and partially against KPC, but not HKP1, indicating limited sharing of functionally relevant neoantigens. To quantify immune selection at clonal resolution, we reconstructed evolutionary trees for each tumor, predicted MHC restricted candidate neoantigens using netMHCpan, and integrated node frequency with cancer cell fraction across immune competent versus immune deficient hosts. This approach identified nodes under strong positive immune pressure and highlighted 86 high priority neoantigens across the three models. Together, these data link clonal fitness costs to functional immunity and provide a generalizable strategy to prioritize neoantigens most likely to mediate tumor rejection.

"