Proteomic profiling of tumor microenvironment and prognosis risk prediction in stage I lung adenocarcinoma

Objectives: With the increasing popularity of CT screening, more cases of early-stage lung cancer are being diagnosed. However, 24.5% of stage I non-small-cell lung cancer (NSCLC) patients still experience treatment failure post-surgery. Biomarkers to predict lung cancer patients at high risk of recurrence are needed. Materials and methods: We collected protein mass spectrometry data from the Taiwan Lung Cancer Moonshot Project and performed bioinformatics analysis on proteins with differential expressions between tumor and adjacent normal tissues in 74 stage I lung adenocarcinoma (LUAD) cases, aiming to explore the tumor microenvironment related prognostic biomarkers. Findings were further validated in 6 external cohorts. Results: The analysis of differentially expressed proteins revealed that the most enriched categories of diseases and biological functions were cellular movement, immune cell trafficking, and cancer. Utilizing proteomic profiling of the tumor microenvironment, we identified five prognostic biomarkers (ADAM10, MIF, TEK, THBS2, MAOA). We then developed a risk score model, which independently predicted recurrence-free survival and overall survival in stage I LUAD. Patients with high risk scores experienced worse recurrence-free survival (adjusted hazard ratio = 8.28, p < 0.001) and overall survival (adjusted hazard ratio = 6.88, p = 0.013). Findings had been also validated in the external cohorts. Conclusion: The risk score model derived from proteomic profiling of tumor microenvironment can be used to predict recurrence risk and prognosis of stage I LUAD.