Abstract
Immunotherapy is one of the main therapeutic approaches for non-small-cell lung cancer (NSCLC). Based on the poor response of immunotherapy, it is crucial to determine the most accurate and widespread predictive characteristics of NSCLC. We retrieved lung squamous cell carcinoma and lung adenocarcinoma gene expression profiles and clinical data from the cancer genome atlas database and classified them into 3 subtypes based on 29 immune gene sets. Combined with previous studies, the expression differences of related pathways and genes in different subtypes were analyzed. We classified them into 3 subtypes: Immunity High, Immunity Medium, and Immunity Low. Immunity High had the strongest immune cell infiltration and antitumor immune activity. Gene ontology enrichment analyses revealed enriched immune-related signaling pathways in lung squamous cell carcinoma. The hyperactivation of cancer-related pathways did not occur in any NSCLC. In addition, the Hippo signaling pathway was negatively correlated with immune signature, whereas epithelial-to-mesenchymal transition was positively correlated. In addition, we found significant differences in immune signatures between males and females; however, no correlation was observed with other clinical data. The identification of NSCLC subtypes based on immune signatures has potential clinical implications for NSCLC treatment.