Potential of Protein Risk Signatures in Lung Adenocarcinoma: A Comprehensive Bioinformatics Study

Abstract

Abstract Purpose Global leading cause of cancer deaths, lung adenocarcinoma (LUAD), is the focus of our study, which explores a protein-based risk signature for LUAD prognosis. Methods Utilizing the Cancer Genome Atlas, we used LASSO-COX to create a prognostic protein model. The model's effectiveness and clinical relevance were analyzed, and a predictive nomogram was built. Concurrently, potential pathways were identified through Gene Set Enrichment Analysis (GSEA) and the tumor immune microenvironment was examined alongside immunotherapy sensitivity. We corroborated model protein expression with immunohistochemistry images from the HPA database and immunofluorescence staining of clinical samples. Results Our six-protein model stratifies LUAD patients effectively into risk groups and shows strong pre-dictive power. The nomogram forecasts overall survival rates at one, three, and five-year intervals. GSEA highlighted enrichment of high-risk genes in metabolic pathways and low-risk genes in immune-related pathways—the latter indicating greater immunotherapy sensitivity. Validation experiments revealed high CD38 expression in lung cancer, presenting a novel paradox. Conclusion This study offers a valuable prognostic protein model for LUAD and uncovers a CD38 expression paradox, enhancing our understanding of protein involvement in lung cancer progression.