Identification of Prognostic Signature and Immune Infiltration in the Skin Cutaneous Melanoma Microenvironment

Abstract Skin cutaneous melanoma (SKCM), characterized by high immunogenicity, has an increasing incidence in recent years. The development of immunotherapy recently offered a promising treatment for patients with SKCM. Unfortunately, not all patients derive benefit from such treatment, so we still face considerable challenges. Hence, it is imperative to develop novel prognostic signature and identify immunotherapeutic targets. In the present study, patients in high immune scores group presented a higher survival rate, while no statistical difference was observed in groups with different stromal scores. 493 DEGs were identified for functional analysis, which were enriched in function related to immune regulation such as lymphocyte activation and cytokine-cytokine receptor interaction. Subsequently, 84 DEGs intersected from Univariate Cox regression analysis and top 100 hub genes in PPI network were identified for the construction of prognostic model. Finally, a prognostic signature including 3 genes (HLA-DQB2, CD80 and GBP4) was established in TCGA training dataset, which was effectively validated in test dataset. Moreover, the model was considered as an independent prognostic factor via univariate and multivariate analysis. Besides, CIBERSORT and correlation analysis demonstrated that the expression level of risk scores was significantly correlated to infiltration levels of immune cells in SKCM. Above all, our study developed a novel prognostic signature, serving as potential prognostic biomarker for SKCM patients. A closely correlation between the prognostic model and tumor immune microenvironment was confirmed, offering a novel insight for the pathogenesis and potential immunotherapy for SKCM.