Construction of a novel prognostic model for evaluation of immune infiltration and immunotherapy in lung adenocarcinoma

Background: Immunotherapy is an indispensable treatment in advanced lung adenocarcinoma (LUAD). However, only a small percentage of patients have experienced the specificity and effectiveness of immune checkpoint inhibitors. To maximize the therapeutic benefits for LUAD patients, a bioinformatics analysis was used to validate a prognostic signature for LUAD based on immune-related genes. Methods: RNA sequence and clinical information of LUAD patients were taken from the TCGA database. Microarray datasets of GSE30219, GSE72094, and GSE37745 and their survival information were downloaded from the GEO database and the prognostic signature was verified through GEO external validation cohorts. Finally, TIDE, immune checkpoint molecules and two independent external immunotherapy validation cohorts (GSE91061 and Imvigor210) were analyzed to evaluate immunotherapy efficacy. Results: We constructed an immune-related diagnostic signature and verified the prognostic value in TCGA, GSE30219, GSE72094, and GSE37745. The gene sets of high-risk samples were correlated with cell cycle, DNA replication, ribosome, and steroid hormone biosynthetic pathway, SPTA1 mutation rate, and lower benefit from ICI therapy; whereas the low-risk patients were correlated with intestinal immune network for IgA production, FLGgene mutation rate, and higher benefit from ICI therapy. Conclusions: The immune-related gene signature for LUAD may have prognostic relevance in ICI therapy.