Pyroptosis, apoptosis, and necroptosis molecular subtype derived prognostic signature universal applicable for gastric cancer-A large sample and multicenter retrospective analysis

Abstract

Background: Whether pyroptosis, apoptosis, and necroptosis (PAN) molecular subtypes exist in gastric cancer (GC) remains unclear.

Methods: Seven independent cohorts including a total of 1901 GC patients were enrolled in our research. TCGA (n = 371) and GSE84437 (n = 433) were combined into one cohort (n = 804) to screen for prognosis-related PAN genes using a univariate Cox regression analysis. The R package "ConsensusClusterPlus" was applied to conduct a clustering analysis of the combination set based on prognosis-related PAN genes. The R package "limma" was used for the identification of differentially expressed genes (DEGs) between different PAN clusters (FDR <0.05 and |logFC|>1). The combined cohort was randomly divided into a training group (n = 484) and a test group (n = 320) at a ratio of 6:4 to establish and verify the prognostic model. A univariate Cox regression analysis, least absolute shrinkage and selection operator method (LASSO) regression analysis, and multivariate Cox regression analysis were used for the identification of prognostic genes and the construction of risk scores. Another five independent cohorts (GSE62254, n = 300; GSE15459, n = 191; GSE26901, n = 109; GSE26253, n = 432; and GSE13861, n = 65) were used for external validation to verify the accuracy and stability of the prognostic signature.

Results: The internal and external validation demonstrated that the 5-gene risk score (LOXL4, SLCO2A1, CST2, PDK4, and MMP11) was an effective instrument for the prognostic risk classification of GC patients. The overall survival (OS) and relapse-free survival (RFS) in the high-risk group were significantly lower than those in the low-risk group and were accompanied by a larger proportion of macrophage and regulatory T cell infiltration. The low-risk group had a good prognosis, with a high tumor mutation burden (TMB), strong cytolytic activity, and a higher proportion of activated CD4 T cell infiltration. In addition, compared with the low-risk group, the cancer-related pathways in the high-risk group were overactivated, and the function of DNA damage repair (DDR) was significantly weakened. Regarding drug sensitivity, the high-risk group was more suitable for targeted drugs, such as axitinib, lapatinib, and nilotinib. The low-risk group was more sensitive to chemotherapy, such as cisplatin, gemcitabine, and vinorelbine.

Conclusion: A universally applicable prognostic signature of GC is proposed in this research based on pyroptosis, apoptosis, and necroptosis (PAN) molecular subtypes.

Keywords: Apoptosis; Gastric cancer; Necroptosis; Prognostic; Pyroptosis.