Abstract 3995: pVAC-Seq: A genome-guided in silico approach to identify tumor neoantigens for personalized immunotherapy

Abstract

Abstract Despite the recent surge in high-throughput sequencing of cancer genomes, the challenge of translating these data into clinically actionable information remains. One promising approach involves the identification of tumor-specific mutant antigens (‘TSMAs’ or neoantigens) via massively parallel sequencing and analysis of matched cancer and normal samples that can be used to create personalized vaccines. In the past, this effort has primarily focused on targeting selection of ‘shared’ tumor antigens, found across many patients. Here, we advocate a more ‘personalized’ approach. These unique antigenic markers or TSMAs arise from numerous genetic changes, acquired somatically that are present exclusively in tumor (mutant) and not in normal (wild- type) cells. Vaccines incorporate these short, antigen-derived peptides (called epitopes) that aim to enhance the immune system's anti-tumor activity by selectively increasing the frequency of anti-tumor specific CD8+ T-cell antigens, and hence expand the ability of the immune system to recognize and destroy cancerous cells. Selecting the best/most immunogenic epitopes from a large number of mutations is an important challenge, in particular in cases of high mutational load such as melanoma and lung cancer. To address this need, we have developed an in silico based sequence analysis method for identification and subsequent refinement of patient-specific antigens for use in personalized vaccines. This flexible and streamlined computational workflow for identification of personalized variant antigens by cancer sequencing (pVAC-Seq) integrates tumor mutation and expression data (DNA- and RNA-Seq) to shortlist candidate neoantigen peptides for a personalized vaccine. Harnessing existing class I prediction algorithms, high-affinity neoantigens over varying peptide lengths are evaluated. To demonstrate the workings of pVAC-Seq, we applied it to four metastatic melanoma patients, the clinical results for three of whom were described previously. These patients were enrolled in a phase 1 vaccine clinical trial employing autologous, functionally mature, interleukin (IL)-12p70-producing dendritic cells (DC). Since melanoma patients harbor hundreds of mutations, it can be challenging to filter down and target the best set of potentially immunogenic neoantigens for vaccine design. By implementing the methods developed in pVAC-Seq, we were able to rapidly streamline the screening and identification of a smaller number of potentially immunogenic neoepitopes within the landscape of all neoepitopes. Citation Format: Jasreet Hundal, Beatriz M. Carreno, Allegra A. Petti, Gerald P. Linette, Obi L. Griffith, Malachi Griffith, Elaine R. Mardis. pVAC-Seq: A genome-guided in silico approach to identify tumor neoantigens for personalized immunotherapy. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3995.