Pyroptosis-related genes expression and nomogram predict overall survival of gastric cancer

Abstract Background The prognosis of gastric cancer remains poor. Pyroptosis-related genes (PRGs) have been investigated as a potential biomarker in several types of cancer, including gastric cancer. This study aimed to investigate the expression, mutation and diagnostic and prognostic value of PRGs, analyzing data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Methods RNA-sequencing data (RNA-seq), somatic datasets, and copy number variation (CNV) data for gastric cancer were also collected from the TCGA. Gene expression matrix and clinical information of GSE84437 were obtained from GEO data. Bioinformatics analysis was performed to investigate expression profiles of PRGs and their infiltration of immune cells, as well as prognostic significance in gastric cancer. Results A total of 22 out of 33 PRGs were up-regulated, only one PRGs was down-regulated in GC compared to normal tissues, while 10 of them showed no difference between the two groups. A total of 117 out of 433 (27.02%) gastric cancer samples demonstrated genetic mutations, missense mutation was the most common variant classification. More than half of the 33 PRGs had copy number amplification. We performed unsupervised consensus clustering based on the expression of PRGs. Two clusters associated with PRGs named cluster A and cluster B were identified in gastric cancer. Compared with cluster B, cluster A not only had worse overall survival, more patients younger than 65 years, and more deaths, but also had a lower infiltration level of T cell and greater activation B cells and mast cells. According to Gene set variation analysis, cluster A showed greater enrichment of vascular smooth muscle contraction, ECM receptor interaction and KEGG pathways of dilated cardiomyopathy. PRGs cluster B was markedly enriched in cytosolic DNA sensing, non-homologous end joining, and basal transcription KEGG pathways. Multivariate cox analyses revealed that CASP5 was the independent factor affecting the prognosis of patients with gastric cancer. The discriminative ability of the final model for overall survival was assessed using the C statistics, 0.651 for overall survival. A predictive nomogram suggested that 3-year and 5-year overall survival rates could be predicted relatively well compared to an ideal model across the entire cohort. Conclusions PRGs was relatively up-regulated in gastric cancer, it was associated with worse overall survival. The overall survival risk for an individual patient can be estimated using PRGs-based nomograms, which can lead to individualized therapeutic choices.