Computational protocol: Novel technologies and emerging biomarkers for personalized cancer immunotherapy

Similar protocols

Protocol publication

[…] Cancer is a genetic disease. The accumulation of genetic mutations in a tumor leads to a change in its proteome. A “cancer anti-genome” generated during this process can be recognized by T cells [, ]. Somatic mutations in cancer may give rise to mutated proteins that are degraded into peptides (neoepitopes) presented in the complex with major histocompatibility complex (MHC) molecules on the cell surface as neoantigens. There is a long standing interest in mutated antigens as discussed in a landmark review by Dr. Gilboa in 1999 []. Heroic efforts were made by multiple groups to assess reactivity against such antigens using DNA library screens. Although these studies illustrated proof of principle, it was not practically feasible to assess this class of antigen in a systemic manner. Only a minority of mutations are shared between patients; thus, the vast majority of mutated antigens are patient-specific. Therefore, the assessment of neoantigens needs to be based on the genome of individual tumors. The revolution in next-generation sequencing technology at affordable costs along with the progress in bioinformatics has now made it feasible to describe the full mutation load (i.e., the ‘genetic landscape’) of human tumors [–]. Specifically, a comparison of the genomic sequence of cancer tissue to that of non-transformed cells from the same patient can be used to reveal the full range of genomic alterations within a tumor, including nucleotide substitutions, structural rearrangements and copy number alterations [].Several preclinical and clinical reports underscore the importance of understanding the immunogenicity of neoantigens and their potential application in cancer immunotherapies. Two studies in mouse models provided the first evidence that cancer exome based approaches can be utilized to identify neoantigens recognized by CD8+ T cells [, ]. Moreover, a recent study showed that tumor specific mutant antigens are important targets of immune checkpoint blockade therapy []. Subsequently, it has likewise been demonstrated that similar approaches can be utilized in the clinical setting to identify immunogenic neoantigen specific CD8+ T cells in patients treated with tumor infiltrating lymphocyte (TIL) therapy and checkpoint targeting therapies [, ]. Two human studies reported that neoantigens were recognized by intratumoral CD4+ T cells in patients with epithelial cancer and melanoma [, ]. This accumulating evidence suggests that the immune response to mutant neoepitopes plays a dominant role in tumor rejection. Due to the uniqueness of neoantigens, research into tumor immunogenicity has shifted the interest from TAAs (differentiation antigen, cancer/testis antigen and overexpressed self-antigen) to patient-specific mutation antigens.The studies in both mouse models and human material used exome sequencing, computer algorithm-guided epitope prediction and the tandem minigene library approach to identify MHC Class I- or II-binding neoepitopes that were processed and presented by APCs and recognized by neoantigen specific CD8+ and CD4+ T cells. A tumor harbors hundreds of putative neoepitopes per the analysis of the current TCGA database. It is imperative to differentiate and identify actual tumor protective neoepitopes from the putative neoepitopes defined in silico. There are two major factors that can be subject to variability when identifying tumor specific mutated antigens using these novel approaches. First, multiple computational tools to identify tumor specific mutations have been developed simultaneously. Different mutation calling tools such as EBcall, JointSNVMix, MuTect, SomaticSniper, Strelka and VarScan 2 have been developed to compare tumor samples with normal tissue samples at each variant locus in order to increase the accuracy of somatic single nucleotide variant (sSNV) calling. These tools used to identify mutations have a high degree of overlap [, ]. As a next step to identify neoepitopes, algorithms to predict binding affinity to patient specific HLA alleles can be used together with predictions on proteasomal processing. The accuracy of the prediction algorithms mostly depends upon calculating the score of binding to the MHC complex. Recent studies showed that combined use of multiple tools gave a better prediction [–]; however, more work is needed to accurately assess the immunoprotective properties of mutation-derived neoepitopes. Second, it has been demonstrated by unbiased screens that not all mutations result in neoantigens that are recognized by autologous T cells. Therefore, it would be valuable to have robust pipelines to filter whole exome data, especially for tumors with high mutation loads. Multiple groups have made significant efforts to establish such pipelines. The filtering steps that have been applied are based on the expression level of the mutations, e.g., RNA sequencing data, and the likelihood that a given mutated epitope will be processed by the proteasome and presented by patient specific MHC molecules [, , , ]. The two latter filtering steps can be assessed using algorithms that are already established to identify pathogen-derived epitopes. Currently, the data is still too sparse to know which of these filters is most relevant and how to accurately apply thresholds these filters to include immunogenic and exclude non-immunogenic neoepitopes. However, the most significant improvement in these predictions might be on the T cell side; the establishment of algorithms that can identify the subset of epitopes that are most likely to be recognized by TCR repertoire.The development of robust in vitro T cell culture protocols, high-throughput combinatorial encoding of MHC multimer flow staining and high-throughput TCR gene capture allows us to assess the frequency, phenotype and polyfunctionality of the particular neoantigen specific T cell response [–]. These high-throughput technologies further reduce the large number of potential neoepitopes to a small number of real immunogenic neoepitopes. Therefore, these technologies will help us reevaluate the accuracy of computational tools as well as select candidate neoepitopes for vaccines and subsequently monitor the neoepitope specific T cell response during therapy. We will discuss the potential application of these high-throughput assays in the corresponding section. Overall, this approach, despite being in its early stages, has shown that the level of mutation load as a potential biomarker can correlate with clinical outcome to checkpoint blockade therapy in patients with advanced melanoma, colorectal cancer and NSCLC [, , , , ]. Patients with highly mutagenized tumors are most likely to respond to ipilimumab treatment. However, some melanoma patients with low mutation load have also experienced long-term clinical benefit. In addition, similar observations were reported in patients with NSCLC treated with anti-PD-1 antibody []. […]

Pipeline specifications

Software tools EBCall, JointSNVMix, MuTect, SomaticSniper, Strelka, VarScan
Application WES analysis
Organisms Homo sapiens
Diseases Neoplasms, Drug-Related Side Effects and Adverse Reactions