Abstract
BACKGROUND: Cutaneous melanoma is the most lethal of all skin cancers. Recent studies suggested that miR-3127 is dysregulated in multiple tumor types and has important roles in tumorigenesis and cancer progression, giving it potential as a prognostic biomarker. The aim of this study was to use bioinformatic analysis to assess miR-3127 expression and correlate expression patterns with disease course in patients with cutaneous melanoma. METHODS: miRNA, mRNA sequencing, DNA methylation data, and clinical information of cutaneous melanoma cases were downloaded from the Human Cancer Atlas - Skin Cutaneous Melanoma (TCGA-SKCM). miR-3127 expression was classified into miR-3127-low and miR-3127-high clusters using maximally selected rank statistics. RESULTS: Clustering analysis showed that high expression of miR-3127 (≥20.3 reads per million) was associated with worse progression-free (p < .001) and overall (p = .011) survival compared to low miR-3127 expression. More than five thousand differentially expressed genes between the two miR-3127 sample groups encoded cell differentiation markers, cytokines, growth factors, translocated cancer genes, and oncogenes. Pathway analysis revealed that miR-3127-high samples related to activity of proliferation, DNA repair, and ultraviolet response. CONCLUSIONS: The expression level of miR-3127 could act as a prognostic indicator for patients with melanoma.