Monitoring methylation‑driven genes as prognostic biomarkers for cervical cancer

Aberrant DNA methylations are markedly associated with the development of cervical cancer (CC); however, only a limited number of studies have focused on identifying the DNA methylation‑driven genes in CC by integrative bioinformatics analysis to predict the prognosis of CC. In the present study, DNA methylation and transcriptome profiling data were downloaded from The Cancer Genome Atlas database. DNA methylation‑driven genes were obtained using the MethylMix R package. The Database for Annotation, Visualization and Integrated Discovery and ConsensusPathDB were used to perform Gene Ontology and pathway analyses, respectively. The survival R package was used to analyze overall survival rates associated with methylation‑driven genes. In total, data for 125 methylation‑driven mRNAs and 38 methylation‑driven long‑coding RNAs (lncRNAs) were obtained. Based on the univariate and multivariate Cox regression models, it was demonstrated that FLT1 (fms‑related tyrosine kinase 1) mRNA, MKI67 (marker of proliferation Ki‑67), PLEKHG6 (pleckstrin homology domain containing family G with RhoGef domain member 6) and POLE2 (DNA polymerase epsilon 2) lncRNAs were predictors of the overall survival of patients with CC. According to DNA methylation and gene expression, FLT1 mRNA, and the MKI67, PLEKHG6 and POLE2 lncRNAs functioned as independent biomarkers for the prognosis of CC. DNA methylation assays revealed that the promoter methylation levels of FLT1 were significantly upregulated in CC and cervical adenocarcinoma compared with normal controls. The results of immunohistochemical analysis revealed that the expression level of FLT1 in CC tissues was higher than that in normal tissues; however, the PLEKHG6 gene was expressed at high levels in normal tissues. On the whole, the present study demonstrates that methylation‑driven lncRNAs and mRNAs contribute to the survival of patients with CC, and FLT1 mRNA, and the MKI67, PLEKHG6 and POLE2 lncRNAs may be potential biomarkers for the prognosis of CC.