Identification of a Gene Signature for Exhausted CD8+ T Cells

CD8+ T lymphocytes are regarded as a critical component of antitumor immunity, while immune invasion often occurs during the development of T cell exhaustion, characterized by the progressive accumulation of coinhibitory checkpoints, including PD-1, PD-L1, CTLA-4, TIM-3, and LAG-3 (17). We defined a gene expression signature of exhausted CD8+ T cells with integrative bioinformatics through publicly available NSCLC data considering the data quality and availability. We obtained an RNA-seq dataset of intratumoral CD8+ T cells showing high or no PD-1 (PDCD1) expression in a published study (24), and we generated an upregulated PD-1-positive gene list from another previous study (25). Pearson correlation analysis was conducted using the upregulated PD-1-positive gene list in the TCGA (microarray+ RNA-seq cohort) and CGGA (microarray+ RNA-seq cohort) datasets with an adjusted P-value < 0.05 and |correlation efficiency| > 0.25 as the eligibility criteria. In total, a 5-gene signature was identified in the glioma database, and an exhausted CD8+ T cell (GET) score was quantified in a tumor by conducting ssGSEA to obtain the ssGSEA score. In combination with clinical and molecular profiles, the prognostic and predictive values of the GET score were determined through different immune phenotypes.