Bioinformatics-based analysis of programmed cell death pathway and key prognostic genes in gastric cancer: Implications for the development of therapeutics

Abstract

Background: Gastric cancer (GC) represents a major global health burden as a result of its high incidence and poor prognosis. The present study examined the role of the programmed cell death (PCD) pathway and identified key genes influencing the prognosis of patients with GC.

Methods: Bioinformatics analysis, machine learning techniques and survival analysis were systematically integrated to identify core prognostic genes from the The Cancer Genome Atlas Stomach Adenocarcinoma (TCGA-STAD) dataset. A prognostic model was then developed to stratify patients into high-risk and low-risk groups, and further validated in the GSE84437 dataset. The model also demonstrated clinical relevance with tumor staging and histopathology. Immune infiltration analysis and the potential benefits of immunotherapy for each risk group were assessed. Finally, subgroup analysis was performed based on the expression of three key prognostic genes.

Results: Three core prognostic genes (CAV1, MMP9 and MAGEA3) were identified. The prognostic model could effectively differentiate patients into high-risk and low-risk groups, leading to significantly distinct survival outcomes. Increased immune cell infiltration was observed in the high-risk group, and better potential for immunotherapy outcomes was observed in the low-risk group. Pathways related to cancer progression, such as epithelial-mesenchymal transition and tumor necrosis factor-α signaling via nuclear factor-kappa B, were enriched in the high-risk group. By contrast, the low-risk group showed a number of pathways associated with maintenance of cell functionality and immune responses. The two groups differed in gene mutation patterns and drug sensitivities. Subgroup analysis based on the expression of the three key genes revealed two distinct clusters with distinct survival outcomes, tumor immune microenvironment characteristics and pathway enrichment.

Conclusions: The present study offers novel insights into the significance of PCD pathways and identifies key genes associated with the prognosis of patients with GC. This robust prognostic model, along with the delineation of distinct risk groups and molecular subtypes, provides valuable tools for risk stratification, treatment selection and personalized therapeutic interventions for GC.

Keywords: bioinformatics; gastric cancer; machine learning; prognostic model; programmed cell death.