Unbiased testing of several hundred tumor-specific single nucleotide variants of a tumor for protective immunogenicity and CD8(+) response reveals surprises.

Abstract Somatic mutations in cancer cells can give rise to new MHC epitopes referred to as neoepitopes. Neoepitopes have been clearly demonstrated to elicit an antigen specific T cell response, which are capable of mediating targeted killing of cancer cells. Despite recent advancements, exactly what criteria make a good neoepitope for personalized therapy is currently unknown. Here, using a syngeneic tumor of C57BL/6 mice, named FABF, we present results of an exhaustive and exhausting study where we have tested hundreds of long peptides each containing a single nucleotide variant (SNV) (with the mutation in the center of the peptide) for their ability to elicit tumor rejection and independently, CD8+ T cell response. We observe that (i) about 2% of all SNVs lead to generation of peptides that can mediate tumor rejection to any significant degree; (ii) each peptide alone elicits a modest protection, and a combination elicits stronger protective immunity, (iii) all the neoepitopes that elicit tumor rejection have poor binding affinity for MHC I, and have positive values for Differential Agretopic Index (1). (iv) Even though the protective responses are CD8-mediated, there is no correlation between a neoepitope’s ability to elicit a CD8 response and tumor rejection. Some of these observations are inconsistent with some aspects of the current consensus about the defining characteristics of good neoepitopes.