Abstract
BACKGROUND: With the swift advancements in immunotherapy for solid tumors, exploring immune characteristics of tumors has become increasingly important. The tumor microenvironment (TME) is closely related to the prognosis and treatment of tumor patients. This study aims to explore the expression characteristics and model construction of TME-related genes in lung adenocarcinoma (LUAD) patients, and provide help for clinical diagnosis and treatment. METHODS: Through the Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE) algorithm, we analyzed the transcriptomic data of 559 samples from The Cancer Genome Atlas (TCGA) data set to estimate the stromal cells and immune cells, and screened the immune-related differentially expressed genes (DEGs), namely, the TME-DEGs. Essential TME genes were then selected from the TME-DEGs by multivariate Cox and least absolute shrinkage and selection operator (LASSO) regression, and a prediction model of prognostic risk score (RS) was established. RESULTS: We identified 5 crucial TME genes: ABCC2, ECT2L, CD200R1, ACSM5, and CLEC17A. Analysis of the genes' associations with prognosis and clinical features showed that ABCC2 was significantly associated with poorer prognosis and decreased immune signatures, whereas the other 4 associated with improved prognosis and immune signatures. Further, a prognostic RS prediction model was constructed based on these 5 genes, and the results showed that patients with low RS had significantly higher overall survival (OS; P<0.001), relapse-free survival (RFS; P=0.009) and disease-free survival (DFS; P=0.005) than the high RS group, and it had a certain predictive accuracy [area under the curve (AUC)] of 5 years OS =0.70). Those were consistent in the GSE50081 cohort. CONCLUSIONS: Five crucial TME genes, ABCC2, ECT2L, CD200R1, ACSM5, and CLEC17A, are significantly correlated with the prognosis and tumor immune microenvironment (TIME) characteristic of LUAD patients, and the prognostic model has good prediction efficiency, which may improve clinical prognostic models and therapy selection.