Wei Guo1, Sijin Sun1, Lei Guo2, Peng Song1, Xuemin Xue2, Hao Zhang1, Guochao Zhang1, Zhen Wang1, Bin Qiu1, Fengwei Tan1, Qi Xue1, Yibo Gao3, Shugeng Gao4, Jie He5. 1. Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Panjiayuannanli No. 17, Chaoyang District, Beijing, 100021, The People's Republic of China. 2. Department of Pathology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, The People's Republic of China. 3. Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Panjiayuannanli No. 17, Chaoyang District, Beijing, 100021, The People's Republic of China. gaoyibo@cicams.ac.cn. 4. Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Panjiayuannanli No. 17, Chaoyang District, Beijing, 100021, The People's Republic of China. gaoshugeng@vip.sina.com. 5. Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Panjiayuannanli No. 17, Chaoyang District, Beijing, 100021, The People's Republic of China. prof.jiehe@gmail.com.
Abstract
PURPOSE: Lung cancer has the highest morbidity and mortality among all cancer types. Reliable prognostic biomarkers are needed to identify high-risk patients apart from TNM system for precision medicine. The present study is designed to identify robust prognostic biomarkers in lung adenocarcinoma (LUAD) based on integration of multiple GEO datasets, The Cancer Genome Atlas (TCGA) database and Clinical Proteomic Tumor Analysis Consortium (CPTAC) database. METHODS: Four LUAD GEO datasets (GSE10072, GSE2514, GSE43458, and GSE32863) and TCGA database were implemented to analyze the differently expressed genes (DEGs). Gene ontology, KEGG pathway, and protein-protein interaction network (PPI) were conducted based on the above DEGs. Hub genes were selected based on connectivity degree in the PPI network. Expression analysis and Kaplan-Meier survival analysis were conducted in CPTAC lung adenocarcinomas cohort. Kaplan-Meier survival analysis and Cox proportional hazards regression were performed on these hub genes using TCGA and our own cohort. RESULTS: A total of 430 shared genes in all five datasets were identified as DEGs. Based on their PPI network, nine hub genes were selected and all of them were significantly associated with overall survival using GEPIA analysis. Two hub genes, TOP2A and UBE2C, were further combined and showed poorer prognosis in both TCGA dataset and our validated cohort. Analysis in CPTAC revealed that TOP2A and UBE2C were significantly highly expressed in tumor sample. Multivariable analysis suggested TOP2A and UBE2C as independent prognostic factors in LUAD. CONCLUSION: Using data mining approach, we identified TOP2A and UBE2C as two robust prognostic factors in LUAD. We also demonstrated the TOP2A/UBE2C co-expression status in LUAD, and TOP2A/UBE2C co-expression correlated with poorer prognosis. More in-depth research is needed for transforming this result into clinical setting.
PURPOSE: Lung cancer has the highest morbidity and mortality among all cancer types. Reliable prognostic biomarkers are needed to identify high-risk patients apart from TNM system for precision medicine. The present study is designed to identify robust prognostic biomarkers in lung adenocarcinoma (LUAD) based on integration of multiple GEO datasets, The Cancer Genome Atlas (TCGA) database and Clinical Proteomic Tumor Analysis Consortium (CPTAC) database. METHODS: Four LUAD GEO datasets (GSE10072, GSE2514, GSE43458, and GSE32863) and TCGA database were implemented to analyze the differently expressed genes (DEGs). Gene ontology, KEGG pathway, and protein-protein interaction network (PPI) were conducted based on the above DEGs. Hub genes were selected based on connectivity degree in the PPI network. Expression analysis and Kaplan-Meier survival analysis were conducted in CPTAC lung adenocarcinomas cohort. Kaplan-Meier survival analysis and Cox proportional hazards regression were performed on these hub genes using TCGA and our own cohort. RESULTS: A total of 430 shared genes in all five datasets were identified as DEGs. Based on their PPI network, nine hub genes were selected and all of them were significantly associated with overall survival using GEPIA analysis. Two hub genes, TOP2A and UBE2C, were further combined and showed poorer prognosis in both TCGA dataset and our validated cohort. Analysis in CPTAC revealed that TOP2A and UBE2C were significantly highly expressed in tumor sample. Multivariable analysis suggested TOP2A and UBE2C as independent prognostic factors in LUAD. CONCLUSION: Using data mining approach, we identified TOP2A and UBE2C as two robust prognostic factors in LUAD. We also demonstrated the TOP2A/UBE2C co-expression status in LUAD, and TOP2A/UBE2C co-expression correlated with poorer prognosis. More in-depth research is needed for transforming this result into clinical setting.