Correlations between tumor mutation burden and immune infiltrates and their prognostic value in pancreatic cancer by bioinformatic analysis

Abstract

PurposeWe aimed to investigate the patterns and prognostic roles of tumor mutation burden and immune microenvironment in pancreatic cancer. MethodsThe somatic mutation data, transcriptome profiles and clinical information were downloaded from the Cancer Genome Atlas database. Gene expression difference, Gene ontology, KEGG, gene set enrichment analyses and “CIBERSORT” algorithm were performed to screen differentially expressed genes, enriched functions or pathways and immune infiltrates differences between high and low TMB groups. Single sample gene set enrichment and unsupervised consensus clustering analyses were used for immunity grouping. Immune cell infiltration and expressions of HLA and checkpoint genes were investigated. Finally, a nomogram model integrating TMB and immune infiltration was established. ResultsA total of 608 differentially expressed genes were identified between high and low TMB groups, KEGG base excision repair and DNA replication pathways were enriched in high TMB group. Infiltration levels of M0 macrophages were higher and dendritic resting cells were lower in high TMB group. The risk model based on TMB-related immune genes, FAM19A2 and SLC22A17 was established and high risk scores indicated poorer prognosis. The expressions of HLA genes and immune checkpoint genes were higher in high immunity group. The nomogram showed remarkable ability for individualized survival estimation with good AUC values (0.794 and 0.800, respectively) for 3- and 5-year survival rates prediction. ConclusionsThe characteristics of tumor mutation burden and immune infiltration in pancreatic cancer provide new insights into the tumor microenvironment, immunotherapies and a novel prognostic nomogram model for pancreatic cancer patients.