Characterization of the Neoantigen Profile in a Tumor Mutation Burden-high Melanoma Patient With Multiple Metastases.

Background/aim: Recently, neoantigen (NA) profiling has been intensively performed for the development of novel immunotherapy. We previously reported a melanoma case with a high tumor mutation burden that achieved complete remission after anti-programmed death-1 therapy. We herein revisited the same case, characterized the NA profiles of other metastatic lesions using in silico algorithms and in vitro CTL assays, and investigated the immunological status, including tumor-infiltrating lymphocytes and the T cell receptor (TCR) repertoire profile, in metastatic sites.

Materials and methods: NA candidates obtained from whole-exome sequencing were applied to the HLA-binding prediction algorithm, NetMHCpan4.1. HLA-A*2402-restricted sequence candidates with a strong binding capacity (<50 nM) and elution affinity (<1%) were selected and evaluated for synthetic peptide candidates. The immunological status in metastatic sites was characterized using gene expression profiling, immunohistochemistry, and a TCR repertoire analysis.

Results: The genomic analysis revealed that all metastatic sites, such as costal, intra-muscular, and brain lesions, had >1,500 SNVs, and 12 driver mutations were common to all sites. New driver mutations were identified in intra-muscular (KMT2C: p.P3292S) and brain (JAK1: p.S404P) metastases and a functional analysis of these mutations revealed that JAK1 mutation exhibited a promoting effect on invasion activity. CTL assays using synthetic NA peptides identified more NA epitopes in brain metastasis.

Conclusion: These results might suggest that the heterogeneity of driver gene mutations is unremarkable, while immunological response is variable in metastatic sites. As a result, the genomic and immunological investigation has provided a very valuable and informative suggestion regarding better cancer therapy decisions.