Identification of Molecular Features Correlating with Tumor Immunity in Gastric Cancer by Multi-Omics Data Analysis
Yin He,Xiaosheng Wang
DOI: https://doi.org/10.21037/atm-20-922
IF: 3.616
2020-01-01
Annals of Translational Medicine
Abstract:BACKGROUND:Although immunotherapy has achieved success in treating various refractory malignancies including gastric cancers (GCs) with DNA mismatch repair deficiency, only a subset of cancer patients are responsive to immunotherapy. Therefore, the identification of useful biomarkers or interventional targets for improving cancer immunotherapy response is urgently needed.METHODS:We investigated the associations between various molecular features and immune signatures using three multi-omics GC datasets. These molecular features included genes, microRNAs (miRNAs), long non-coding RNAs (lncRNAs), proteins, and pathways, and the immune signatures included CD8+ T cell infiltration, immune cytolytic activity (ICA), and PD-L1 expression. Moreover, we investigated the association between gene mutations and survival prognosis in a gastrointestinal (GI) cancer cohort receiving immunotherapy and two GC cohorts not receiving such a therapy.RESULTS:The mutations of some important oncogenes and tumor suppressor genes were appreciably associated with immune signatures in GC, including PIK3CA, MTOR, RNF213, TP53, ARID1A, PTEN, ATM, and CDH1. Moreover, a number of genes exhibited a significant expression correlation with immune signatures in GC, including CXCL9, CXCL13, CXCR6, CCL5, GUCY2C, MAP3K9, NEK3, PAK6, STK35, and WNK2. We identified several proteins whose expression had a significant positive correlation with immune signatures in GC. These proteins included caspase-7, PI3K-p85, PREX1, Lck, Bcl-2, and transglutaminase. In contrast, acetyl-CoA carboxylase (ACC) had a significant inverse expression correlation with immune signatures in GC, suggesting that inhibiting ACC could enhance antitumor immunity in GC. Furthermore, we identified numerous miRNAs and lncRNAs with a significant expression correlation with GC immunity, including hsa-miR-150, 155, 142, 342, 146, 101, 511, 29, AC022706.1, LINC01871, and AC006033.2. We also identified numerous cancer-associated pathways whose activity was associated with GC immunity, including mTOR, PI3K-AKT, MAPK, HIF-1, and VEGF signaling pathways. Interestingly, we found seven genes (ARID1A, BCOR, MTOR, CREBBP, SPEN, NOTCH4, and TET1) whose mutations were associated with better OS in GI cancer patients receiving anti-PD-1/PD-L1 immunotherapy but were not associated with OS in GC patients without immunotherapy.CONCLUSIONS:The molecular features significantly associated with GC immunity could be useful biomarkers for stratifying GC patients responsive to immunotherapy or intervention targets for promoting antitumor immunity and immunotherapy response in GC.