Confirmation Key Genes Associated with TNM Staging Classified as T3 and T4 in Lung Squamous Cell Carcinoma with Bioinformatics Analysis

Abstract

Background: Lung cancer is a high occurrence rate and mortality rate cancer. Non-small cell lung cancer (NSCLC) is confirmed in 80–85% of lung cancer cases. Lung squamous cell carcinoma (LUSC) is frequently diagnosed at the advanced stage with poor prognoses. The size of tumor was an important indicator of the prognosis. Methods: TCGA database and GEO database were performed to download transcriptome data and clinical information of LUSC. Firstly, we identified differentially expressed genes (DEGs) between TNM stage as T3-T4 and T1-T2 of LUSC patients in TCGA datasets. Furthermore, PPI was applied to identify proteins that interact actively during the process of tumorigenesis. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) were consulted to explore the enriched biological processes and pathways of the DEGs. After that, LASSO Cox regression algorithms were employed to confirm tumor size-related signature. In addition, survival analysis (including nomogram, Kaplan–Meier method, ROC curve, GSVA, and correlation matrix) was performed to achieve a accurate prognostic model. Finally, the GEO database was applied to check the tumor size-related prognostic features. Results: 1267 genes were identified as DEGs. And we can conclude that DEGs primarily concentrated in membranes, defence response to bacterium , transmembrane signaling receptor activity and olfactory transduction by the results from GO functions and KEGG pathways analysis. Five genes about tumour size-related risk signature including PCGF2, ULK3, MCRIP1,UCKL1, and CCDC18-AS1 were selected to forecast overall survival of LUSC patients. The credibility of prediction model was verified in GSE68825 and GSE68793. The LUSC patients were divided into low-risk score and high-risk score groups according to average value of risk score. Scatter plots show that patients in high-risk score group had shorter survival time. Conclusion: Our study identified five biomarkers that were related to tumor size in the LUSC. The prognostic model can efficiently predict the survival status of patients. In addition, the several biomarkers are conductive to further investigate therapies and forecast prognosis of LUSC.