Establishment and validation of an immune-associated signature in lung adenocarcinoma.

Recent studies demonstrated that immune associated genes (IAGs) played an important role in the treatment of lung adenocarcinoma (LUAD). In the research, we established an IAGs signature and validated its prognostic value in LUAD by using bioinformatic methods and public databases. Based on the RNA-Seq samples from The Cancer Genome Atlas (TCGA), 576 differentially expressed IAGs were firstly identified. The R package coxph was used to select significant prognostic IAGs using both univariate and multivariate analyses. As a result, four IAGs (SCG2, CCL20, CAT, S100P) were finally screened in an IAGs signature. Based on these four IAGs, LASSO (least absolute shrinkage and selection operator) Cox regression analysis was used to construct a Risk score prognostic model and survival analysis revealed that high risk score was significantly associated with poor survival outcomes, which was validated in the external datasets GSE68465 and GSE31210. In addition, Risk score was found to be significantly associated with stage, lymphatic involvement, tumor metastasis and immune cells (B cells and dendritic cells) infiltration. Moreover, it was found that TP53 and EGFR had a higher mutation frequency in high risk group. Then a nomogram with clinical characteristics was established to superiorly predict prognosis of LUAD patients, and calibration plots and ROC analysis proved its accuracy. We believe that our findings can be conveniently used for individualized prediction of the clinical prognosis for LUAD patients, but further clinical trials and experimental exploration are needed to validate our observations.