Abstract
BACKGROUND: Advanced melanoma exhibits high primary resistance to immunotherapy and limited long-term response, suggesting the need for sensitive biomarkers and novel therapeutic targets to enhance treatment efficacy. We aimed to identify prognostic biomarkers and therapeutic targets for melanoma immunotherapy. METHODS: Bulk RNA-sequencing data from public melanoma cohorts were analyzed to classify tumor samples into two clusters based on distinct expression patterns of immune-related hallmarks. Differential gene expression and least absolute shrinkage and selection operator (LASSO) Cox regression analysis identified nine core genes, which were integrated into a prognostic risk model. Single-cell RNA-sequencing analysis was then performed to characterize the immune cell heterogeneity and functional interactions in the tumor microenvironment. RESULTS: The single-cell analysis revealed that CST7 (cystatin F)-expressing macrophages/monocytes were enriched in an M1 macrophage signature, indicative of an antitumor phenotype. These immune cells were significantly more frequent in immunotherapy responders than in non-responders. Further investigation identified a signaling interaction between CST7(+) macrophages/monocytes and a cytotoxic T-cell subset via the ICOSL (inducible T-cell costimulator ligand)-ICOS (inducible T-cell costimulator) axis, which was more prominent in therapy responders. CONCLUSIONS: These findings suggest that CST7(+) macrophages/monocytes are associated with immunotherapy response by correlating with cytotoxic T-cell activation through the ICOSL/ICOS pathway. The observed interaction highlights the potential for combining immune checkpoint inhibitors (ICIs) with ICOS/ICOSL agonistic antibodies to enhance and sustain long-term immunotherapy efficacy in patients with melanoma.