Transcriptome profiles of fatty acid metabolism-related genes and immune infiltrates identify hot tumors for immunotherapy in cutaneous melanoma

Background: Immunotherapy with checkpoint inhibitors usually has a low response rate in some cutaneous melanoma (CM) cases due to its cold nature. Hence, identification of hot tumors is important to improve the immunotherapeutic efficacy and prognoses of CMs. Methods: Fatty acid (FA) metabolism-related genes were extracted from the Gene Set Enrichment Analysis and used in the non-negative matrix factorization (NMF), copy number variation frequency, tumor mutation burden (TMB), and immune-related analyses, such as immunophenoscore (IPS). We generate a risk model and a nomogram for predicting patient prognoses and predicted the potential drugs for therapies using the Connectivity Map. Moreover, the NMF and the risk model were validated in a cohort of cases in the GSE65904 and GSE54467. At last, immunohistochemistry (IHC) was used for further validation. Results: Based on the NMF of 11 FA metabolism-related DEGs, CM cases were stratified into two clusters. Cluster 2 cases had the characteristics of a hot tumor with higher immune infiltration levels, higher immune checkpoint (IC) molecules expression levels, higher TMB, and more sensitivity to immunotherapy and more potential immunotherapeutic drugs and were identified as hot tumors for immunotherapy. The risk model and nomogram displayed excellent predictor values. In addition, there were more small potential molecule drugs for therapies of CM patients, such as ambroxol. In immunohistochemistry (IHC), we could find that expression of PLA2G2D, ACOXL, and KMO was upregulated in CM tissues, while the expression of IL4I1, BBOX1, and CIDEA was reversed or not detected. Conclusion: The transcriptome profiles of FA metabolism-related genes were effective for distinguishing CM into hot–cold tumors. Our findings may be valuable for development of effective immunotherapy for CM patients and for proposing new therapy strategies.