Weichang Yang, Zhijian Wu, Shanshan Cai, Jiajia Xiang and Xiaoqun Ye* Pages 1 - 20 ( 20 )
Background: Metastasis is the leading cause of death in lung cancer patients. Pre-metastatic niche (PMN) plays an important role in pre-metastatic tumors. However, the development of clinical applications of PMN is still limited.
Methods: Expression data for lung adenocarcinoma (LUAD) patients and PMN-related genes were downloaded from the UCSC Xena website and GeneCards database, respectively. Multiple combinations based on machine learning algorithms were used to screen signature genes and construct a PMN-associated index. Spearman analysis explored the correlation between the PMN-associated index and immune cell infiltration. In addition, we analyzed the clinical value of the PMN-associated index based on drug sensitivity analysis and TIDE scores.
Results: The enrichment analyses suggested that PMN-related genes were mainly enriched in the PI3K-Akt and HIF-1 signaling pathways. We chose random survival forest, Lasso, and multivariate Cox regression analyses to construct the PMN-associated index based on the results of multiple machine learning algorithms. Six signature genes (SNAI2, CXCR4, TNFSF11, ENG, TIMP1, and PDGFB) were screened to construct the PMN-associated index. KM analysis suggested that the survival probability was greater in the low PMN-associated index group than in the high PMN-associated index group. In addition, we confirmed that LUAD patients with a low PMN-associated index were more likely to benefit from immunotherapy.
Conclusion: We confirmed that the PMN-associated index is a valid predictor of prognosis, immune characteristics, and antitumor therapy efficacy in LUAD patients, which provides additional evidence for the potential clinical value of PMN development.
Pre-metastatic niche, lung adenocarcinoma, immune cell infiltration, prognosis, CXCR4, tumor microenvironment.