Identification of molecular subtypes based on tumor-specific genes in melanoma and development of a nomogram for prognostic prediction in patients with melanoma

Abstract

Abstract Purpose We aim to investigate the molecular characteristics of melanoma by analyzing tumor-specific genomic signatures and construct a nomogram for prognostic prediction in patients with melanoma. Methods The datasets of melanoma and normal skin tissue from the Genotype-Tissue Expression and the Cancer Genome Atlas databases were merged to conduct differential gene expression (DEGs) analysis. Cox regression and LASSO regression analysis were used to identify tumor-specific genes associated with prognosis and construct a risk score model. Melanoma patients in the training and validation cohorts were grouped into high-risk and low- risk groups according to the median value of risk score. K-M analysis, Gene Set Enrichment Analysis and single-sample Gene Set Enrichment Analysis were conducted to further analyze the differences of the survival, the enriched pathways, the tumor microenvironment, the hypoxia score, the tumor mutation burden and the gene mutation between high and low risk groups. A nomogram based on risk score and tumor stage was constructed for prognostic prediction in patients with melanoma. Results Seven DEGs including BCAN, CFAP61, CXCL13, KIT, KRT17, PLA1A, and PPP1R3C were identified to construct the risk score model. Patients in the high-risk group showed inferior survival compared with the low-risk group. The oxidative phosphorylation, the metabolism of xenobiotics by cytochrome P450, the tyrosine metabolism, the fructose and mannose metabolism, the glycolysis/gluconeogenesis and the melanogenesis pathways were significantly enriched in high-risk group patients. The high-risk and low-risk groups differed in terms of angiogenesis, anti-tumor and pro-tumor cytokines, B cells, T cells, M1 signature, myeloid cells traffic NK cells, Th1 and Th2 signature, Treg, tumor associated macrophage, coactivation, check point molecules, effector cells. The AUC values for the nomogram predicting the survival of melanoma patients at 12, 24, 36, and 60 months were 0.726, 0.791, 0.712, and 0.714, respectively. Conclusion This study provides a personalized assessment of molecular subtypes in melanoma. The nomogram based on the seven tumor-specific genes and tumor stage had favorable predictive value in melanoma.