Establishment and validation of a gene signature for predicting metastasis and prognosis of lung adenocarcinoma based on anoikis-related genes.

Objective: Anoikis is a programmed cell death process that plays a crucial role in tumor metastasis. Lung adenocarcinoma frequently results in multi-organ metastasis, which significantly impacts patient prognosis. This study aims to identify new anoikis-related gene signature to predict metastasis and prognosis of lung adenocarcinoma. Methods: The study obtained gene expression profiles and clinical data of patients with metastatic and non-metastatic lung adenocarcinoma from the TCGA and GEO databases. Additionally, 293 genes related to anoikis were downloaded from the GeneCard database. Unsupervised cluster analysis was used to divide patients with metastatic lung adenocarcinoma into two tumor subtypes. The study assessed immunoinfiltration and immune cell function in two groups using the TIMER database and single sample gene set enrichment analysis (ssGSEA). A prognostic model of genes related to anoikis was constructed using the minimum absolute contraction and selection algorithm (LASSO) and Cox regression model, which was then validated using external data sets. The predictive power of the model was further evaluated using ROC curves and a nomogram. The study evaluated the differences in immunotherapy and drug therapy between high-risk and low-risk groups. Selective gene expression was verified using real-time quantitative fluorescent PCR (qRT-PCR), and immune cell infiltration was verified using multiple immunofluorescence histochemistry. Results: The two molecular subtypes exhibited significant differences in clinicopathological features, prognosis, and immune cell infiltration. Lasso and multivariate Cox regression identified three prognostic genes related to anoikis (TLE1, EIF2AK3 and BIRC3), and a risk model was constructed using these genes. The patients were categorized into high- and low-risk groups based on their median risk scores. The low-risk group exhibited higher overall survival (OS) time, immune activity, tumor mutation burden (TMB), and PD1/ PD-L1 expression, which is consistent with a better response to immune checkpoint inhibitors. The nomogram demonstrates the model's superior predictive value. The expression of three prognostic genes was higher in lung adenocarcinoma cell lines and tissue, as determined by qRT-PCR. Furthermore, this expression was positively correlated with the degree of immune cell infiltration. Conclusion: This study has established a prognostic risk model characterized by anoikis-related genes. The model has good prognostic value in patients with lung adenocarcinoma and can be used as a potential diagnostic marker and therapeutic target for evaluating patient prognosis. [ABSTRACT FROM AUTHOR]