Following the advent of immunotherapy as a possible cure for cancer, remarkable insight has been gained on the role of mutational load and neoantigens as key ingredients in T cell recognition of malignancies. However, not all highly mutational tumors react to immune therapies, and even initial success is often followed by eventual relapse. Recent research points out that high heterogeneity in the neoantigen landscape of a tumor might be key in understanding the failure of immune surveillance. In this work we present a mathematical framework able to describe how neoantigen distributions shape the immune response. Modeling T cell reactivity as a function of antigen dominancy and distribution across a tumor indicates the existence of a diversity threshold beyond which T cells fail at controling heterogeneous cancer growth. Furthemore, an analytical estimate for the evolution of average antigen clonality indicates rapid increases in epitope heterogeneity in early malignancy growth. In this scenario, we propose that therapies targeting the tumor prior to immunotherapy can reduce neoantigen heterogeneity, and postulate the existence of a time window, before tumor relapse due to de novo resistance, rendering immunotherapy more effective.
Major FindingsGenetic heterogeneity affects the immune response to an evolving tumor by shaping the neoantigen landscape of the cancer cells, and highly heterogeneous tumors seem to escape T cell recognition. Mathematical modeling predicts the existence of a well defined neoantigen diversity threshold, beyond which lymphocites are not able to counteract the growth of a population of highly heterogeneous subclones. Furthermore, evolutionary dynamics predict a fast decay of neoantigen clonality, rendering advanced tumors hard to attack at the time of immunotherapy. Within this mathematical framework we propose that targeted therapy forcing a selective pressure for resistance might as well increase neoantigen homogeneity, providing a novel possibility for combination therapy.The authors declare no potential conflicts of interest.