Identification of replication factor C subunit 4 as a potential therapeutic target in esophageal squamous cell carcinoma based on bioinformatic analysis and machine learning

BACKGROUND: Esophageal squamous cell carcinoma (ESCC) is one of the highly lethal and aggressive malignant tumors worldwide. To effectively prevent and treat this disease, the search for novel molecular targets is of great significance for promoting the molecular diagnosis and targeted therapy of ESCC. METHOD: Gene expression profiles from gene expression omnibus (GEO) datasets were normalized and analyzed to identify differentially expressed genes. Functional enrichment, protein-protein interaction network, and machine learning algorithms were applied for biomarker screening. Immune infiltration analysis and immunohistochemistry were performed to assess clinical relevance. RESULTS: Analysis identified 752 differentially expressed genes in ESCC, with enrichment in upregulated pathways including DNA replication and mismatch repair, and downregulated pathways such as autophagy. Gene Ontology/Kyoto Encyclopedia of Genes and Genomes analyses revealed complex molecular networks driving ESCC. Key hub genes and diagnostic biomarkers aurora kinase A (AURKA), kinesin family member 4 A (KIF4A), and replication factor C subunit 4 (RFC4) were identified, with high diagnostic area under the receiver operating characteristic curve values from 0.976 to 0.983. RFC4 expression correlated with mast cell infiltration patterns, showed elevated expression in ESCC tissues via immunohistochemistry, and was associated with poor prognosis. CONCLUSION: This study identifies AURKA, KIF4A, and RFC4 as potential in silico biomarkers for ESCC. This study further highlights RFC4 as a promising candidate for diagnostic and prognostic applications, offering new insights into prevention strategies for ESCC.