Pan-cancer analysis of necroptosis-related gene signature for the identification of prognosis and immune significance

Abstract

Background: Necroptosis is a novel programmed cell death mode independent on caspase. A number of studies have revealed that the induction of necroptosis could act as an alternative therapeutic strategy for drug-resistant tumors as well as affect tumor immune microenvironment. Methods: Gene expression profiles and clinical data were downloaded from XENA-UCSC (including The Cancer Genome Atlas and Genotype-Tissue Expression), Gene Expression Omnibus, International Cancer Genome Consortium and Chinese Glioma Genome Atlas. We used non-negative matrix factorization method to conduct tumor classification. The least absolute shrinkage and selection operator regression was applied to establish risk models, whose prognostic effectiveness was examined in both training and testing sets with Kaplan–Meier analysis, time-dependent receiver operating characteristic curves as well as uni- and multi-variate survival analysis. Principal Component Analysis, t-distributed Stochastic Neighbor Embedding and Uniform Manifold Approximation and Projection were conducted to check the risk group distribution. Gene Set Enrichment Analyses, immune infiltration analysis based on CIBERSORT, EPIC, MCPcounter, ssGSEA and ESTIMATE, gene mutation and drug sensitivity between the risk groups were also taken into consideration. Results: There were eight types of cancer with at least ten differentially expressed necroptosis-related genes which could influence patients’ prognosis, namely, adrenocortical carcinoma (ACC), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), acute myeloid leukemia (LAML), brain lower grade glioma (LGG), pancreatic adenocarcinoma (PAAD), liver hepatocellular carcinoma (LIHC), skin cutaneous melanoma (SKCM) and thymoma (THYM). Patients could be divided into different clusters with distinct overall survival in all cancers above except for LIHC. The risk models could efficiently predict prognosis of ACC, LAML, LGG, LIHC, SKCM and THYM patients. LGG patients from high-risk group had a higher infiltration level of M2 macrophages and cancer-associated fibroblasts. There were more CD8+ T cells, Th1 cells and M1 macrophages in low-risk SKCM patients’ tumor microenvironment. Gene mutation status and drug sensitivity are also different between low- and high-risk groups in the six cancers. Conclusions: Necroptosis-related genes can predict clinical outcomes of ACC, LAML, LGG, LIHC, SKCM and THYM patients and help to distinguish immune infiltration status for LGG and SKCM.