Abstract Background: Melanoma represents one of the most aggressive malignancies and has a high tendency to metastasize. Metastatic potential may not be predicted from the morphologic features alone, and the use of routine histologic prognostic factors is insufficient for melanoma staging. Because metastatic melanoma remains extremely difficult to cure, there has been an urgent need to find new markers to identify tumors capable of metastasizing. Here, we used a computational strategy based on transcriptome meta-analysis for melanoma staging. We aimed to identify biomarkers that differentiate between non-metastatic and metastatic melanomas. Methods: Affymetrix HU 133A Plus 2.0 microarray data were obtained from the GEO database for the meta-analysis, representing benign nevi (N), stages I, II and III melanoma (EI-III), stage IV melanoma from primary site (EIV) and stage IV melanoma from metastatic site (Mm). Differential expression was analyzed using linear models (118 files) and time course (64 files). Transcripts with a log fold change of <-1/>1 and B-statistic >3 using linear models and evaluate by MB-statistic for Time Course were included in the study. The analysis focused on genes that encode secreted proteins, whose expression was upregulated throughout the progression of melanomas, starting at N and increasing gradually from EI-III to EIV and finally to Mm. Results: We obtained 6137 differentially expressed genes with linear models (between EI-III vs. N, EIV vs. EI-III, Mm vs. EIV, EIV vs. N and Mm vs. N contrasts) and took the top 6137 genes ranked with Time Course. Our results show 17 possible biomarkers that grouped into three types: 1) genes capable of differentiating between EIV/Mm and EI-III/N (SPP1); 2) genes capable of differentiating between Mm and EIV/EI-III/N (CXCR4, FCGR1B, NGRN, PHTF1, RAB8B, ARHGEF2, ARMC9, BCL2A1, CALU, DLGAP5, DMXL2, FKBP11 and STARD3NL); and 3) genes capable of differentiating between Mm/EIV/EI-III and N (ADAM10, CLIC4 and STK36). Conclusions: Our analyses, in agreement with previously reported results, identify SPP1 and CXCR4 as biomarkers of metastatic disease and poor clinical outcome, supporting the quality of our analysis. We also identified new biomarkers that may be used to predict undetectable metastasis. In order to provide a more precise evaluation, experimental validation or functional exploration of the reported gene expression differences is warranted. Citation Format: Daniel Ortega-Bernal, Claudia Rangel-Escareño, Elena Arechaga-Ocampo, Claudia H. Gonzalez-De la Rosa. Biomarkers for staging melanoma, a search at transcriptome level [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2252.
Read full abstract