Identification of tumor microenvironment-associated immunological genes as potent prognostic markers in the cancer genome analysis project HOPE.

Project High-tech Omics-based Patient Evaluation (HOPE), which used whole-exome sequencing and gene expression profiling, was launched in 2014. A total of ~2,000 patients were enrolled until March 2016, and the survival time was observed up to July 2019. In our previous study, a tumor microenvironment immune type classification based on the expression levels of the programmed death-ligand 1 (PD-L1) and CD8B genes was performed based on four types: A, adaptive immune resistance; B, intrinsic induction; C, immunological ignorance; and D, tolerance. Type A (PD-L1 and CD8B) exhibited upregulated features of T helper 1 antitumor responses. In the present study, survival time analysis at 5 years revealed that patients in type A had a better prognosis than those in other categories [5 year survival rate (%); A (80.5) vs. B (73.9), C (73.4) and D (72.6), P=0.0005]. Based on the expression data of 293 immune response-associated genes, 62 specific genes were upregulated in the type A group. Among these genes, 18 specific genes, such as activated effector T-cell markers (CD8/CD40LG/GZMB), effector memory T-cell markers (PD-1/CD27/ICOS), chemokine markers (CXCL9/CXCL10) and activated dendritic cell markers (CD80/CD274/SLAMF1), were significantly associated with a good prognosis using overall survival time analysis. Finally, multivariate Cox proportional hazard regression analyses of overall survival demonstrated that four genes (GZMB, HAVCR2, CXCL9 and CD40LG) were independent prognostic markers, and GZMB, CXCL9 and CD40LG may contribute to the survival benefit of patients in the immune type A group.