Analyzing the characteristics of immune cell infiltration in lung adenocarcinoma via bioinformatics to predict the effect of immunotherapy

Recent studies have shown that tumor immune cell infiltration (ICI) is associated with immunotherapy sensitivity and the prognosis of lung adenocarcinoma (LUAD). However, the immunoinfiltrative landscape of LUAD has not been elucidated. We propose two computational algorithms to unravel the ICI landscape to evaluate the efficacy of immunotherapy in LUAD patients. The raw data of LUAD patients from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases were analyzed. After merging these datasets and removing the batch differences, we used the Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) algorithm to obtain the immune cell content of all the samples. The unsupervised consistency clustering algorithm was used to analyze the ICI subtypes, and three subgroups were obtained. In addition, the unsupervised consistency clustering algorithm was used to analyze the differentially expressed genes (DEGs) of the ICI subtypes and obtain three ICI gene clusters. Finally, the ICI score was determined by using principal component analysis (PCA) for the gene signature. The ICI score of LUAD patients ranged from − 32.26 to 12.89 and represents the prognosis and the response to immunotherapy. High ICI scores were characterized by the T cell receptor signaling pathway, B cell receptor signaling pathway, and natural killer cell–mediated cytotoxicity, suggesting that some immune cells were activated and had increased activity, which may be the cause of the better prognosis for patients with high ICI scores. Additionally, patients with higher ICI scores showed a significant immune therapeutic advantage and clinical benefit. This study shows that the ICI score may be a potent prognostic biomarker and predictor of therapy with immune checkpoint inhibitors.