BACKGROUND: Lung cancer remains the leading cause of cancer-related mortality, with poor outcomes driven by late presentation and therapy resistance. Although genes encoding secreted proteins may reflect tumor biology and have biomarker potential, systematic multi-cohort studies identifying and validating prognostically relevant secreted-protein candidates in non-small cell lung cancer (NSCLC) are limited. METHODS: We conducted a multi-cohort transcriptomic screen of 1,896 genes encoding secreted proteins to nominate prognostic candidates in NSCLC. Candidates were prioritized based on differential expression between tumor and normal tissues and survival associations across public NSCLC cohorts, and SERPINA3 was selected for further validation in the TCGA NSCLC dataset. Tumors were stratified by SERPINA3 expression for Kaplan-Meier survival analysis and gene set enrichment analysis using the Hallmark and Kyoto Encyclopedia of Genes and Genomes (KEGG) collections. Translational relevance was assessed in a retrospective bronchoalveolar lavage fluid (BALF) registry using multivariable Cox models. RESULTS: In the TCGA NSCLC dataset, SERPINA3-high tumors had shorter overall and progression-free survival (OS: log-rank p=0.016; PFS: log-rank p=0.018). GSEA showed enrichment of TNF-α/NF-κB, inflammatory response, IL6-JAK-STAT3, and IL-17 signaling, and the SERPINA3-high state displayed a T-cell-infiltrated yet checkpoint-high/exhausted phenotype with increased TGF-β, IL-10, and IDO. In the BALF registry, higher BALF-but not plasma-SERPINA3 independently predicted shorter PFS, with a nonsignificant trend for OS after adjustment. CONCLUSION: Across datasets, SERPINA3 marks an immune-inflammatory tumor state associated with worse survival. BALF-based measurement captured tumor-proximal signals that may be less apparent in plasma, supporting its potential clinical utility for risk stratification and warranting prospective validation.
Kim et al. (Mon,) studied this question.