Identification of a Robust Methylation Classifier for Cutaneous Melanoma Diagnosis

Abstract

Early diagnosis improves melanoma survival, yet the histopathological diagnosis of cutaneous primary melanoma can be challenging, even for expert dermatopathologists. Analysis of epigenetic alterations, such as DNA methylation, that occur in melanoma can aid in its early diagnosis. Using a genome-wide methylation screening, we assessed CpG methylation in a diverse set of 89 primary invasive melanomas, 73 nevi, and 41 melanocytic proliferations of uncertain malignant potential, classified based on interobserver review by dermatopathologists. Melanomas and nevi were split into training and validation sets. Predictive modeling in the training set using ElasticNet identified a 40-CpG classifier distinguishing 60 melanomas from 48 nevi. High diagnostic accuracy (area under the receiver operator characteristic curve ¼ 0.996, sensitivity ¼ 96.6%, and specificity ¼ 100.0%) was independently confirmed in the validation set (29 melanomas, 25 nevi) and other published sample sets. The 40-CpG melanoma classifier included homeobox transcription factors and genes with roles in stem cell pluripotency or the nervous system. Application of the 40-CpG melanoma classifier to the diagnostically uncertain samples assigned melanoma or nevus status, potentially offering a diagnostic tool to assist dermatopathologists. In summary, the robust, accurate 40-CpG melanoma classifier offers a promising assay for improving primary melanoma diagnosis.

Clinical Implications of Primary Cilia in Skin Cancer

“…Although different combinations of mutations in a variety of genes underlie the diversity of melanocytic neoplasms, they all share the common denominator of activating the mitogen-activated protein kinase (MAPK) pathway. Beyond DNA mutations, deletions and amplifications, DNA methylation [81], microRNAs [82], transcription [83], and translation [84] also play important roles in determining the malignant potential of a particular melanocytic neoplasm [85]. Finally, the tumor's microenvironment and interaction with an individual's immune system also contribute to the clinical course of melanoma [86].…”

Peer Review #1 of "Identification of differentially expressed methylated genes in melanoma versus nevi using bioinformatics methods (v0.1)"

“…Four data sets from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) of The National Center for Biotechnology Information (NCBI) were selected. GSE3189 [7] and GSE46517 [8] (platform: GPL96 Affymetrix Human Genome U133A Array) relate to gene expression, and the DNA methylation microarrays are GSE86355 [9] and GSE120878 [10] , respectively (platform: GPL13534 Illumina Infinium HumanMethylation450 BeadChip array) ( Manuscript to be reviewed melanoma and 9 nevi samples were included in GSE46517. An additional melanoma and nevi samples were enrolled in GSE86355 while 89 melanoma and 73 nevi were enrolled in GSE120878 (detailed information is contained in Table 1).…”

“…tool, was used to analyze the function and pathway enrichment of obtained DEGs [10] . The Fisher exact test P-value was calculated and a P-value < 0.05 was regarded as being statistically significant.…”

Research Techniques Made Simple: Feature Selection for Biomarker Discovery

“…These methods include random forest and other decision tree-based models, as well as regularized methods like LASSO. In a recent study, an embedded method was used to identify a 40-CpG classifier for distinguishing primary invasive melanoma from nevi, with sensitivity of 96.6% and specificity of 100% upon validation (Conway et al, 2019). The investigators used a method called elastic net.…”

“…For diseases where early diagnosis greatly improves survival rates, such as melanoma, the identification of improved biomarker-based predictive models that are quantitative, reliable, economic, and easy to interpret would substantially improve patient well-being. Consequently, a myriad of reports connecting patient diagnosis or outcome to quantitative measurements-ranging from gene expression level or methylation status to clinical images analyzed by artificial intelligence-have emerged in recent years (Conway et al, 2019;Esteva et al,2017;Shen et al, 2018).…”