Exploring a 7-gene prognostic model based on ferroptosis for efficiently guiding immunotherapy in melanoma patients

Abstract

PurposeAlthough skin cutaneous melanoma (SKCM) is a relatively immunotherapy-sensitive tumor type, there is still a certain fraction that benefits less from treatment. Ferroptosis has been demonstrated to modulate tumor progression in many cancer types. This study focused on ferroptosis-related genes to construct a prognostic model for SKCM patients. Materials and methodsGene expression profiles of SKCM samples were obtained from public databases. Unsupervised consensus clustering was used to determine molecular subtypes related to ferroptosis. Least absolute shrinkage and selection operator (LASSO) and stepwise Akaike information criterion (stepAIC) were applied to construct a prognostic model based on differentially expressed genes between two molecular subtypes. ResultsC1 and C2 subtypes were identified with differential prognosis and immune infiltration. A 7-gene prognostic model was constructed to classify samples into high-FPRS and low-FPRS groups. Low-FPRS group with favorable prognosis had higher immune infiltration and more enriched immune-related pathways than the high-FPRS group. The two groups showed distinct sensitivity to immunotherapy, with the low-FPRS group predicted to have more positive response to immunotherapy than the high-FPRS group. A nomogram based on the FPRS score and clinical features was built for more convenient use. ConclusionsThe critical role of ferroptosis involved in SKCM development was further validated in this study. The prognostic model was efficient and stable to be applied in clinical conditions to support clinicians in determining personalized therapy for SKCM patients especially those with metastasis.