Identification of candidate miRNA biomarkers correlated with the prognosis of cell carcinoma and endocervical adenocarcinoma via integrated bioinformatics analysis

Objectives: This study was designed to explore MicroRNAs (miRNAs) associated with the prognosis of cell carcinoma and endocervical adenocarcinoma (CESC) to search for biomarkers of CESC and provide guidelines for the clinical treatment. Methods: mRNAs of CESC patients were downloaded from The Cancer Genome Atlas (TCGA), and miRNA expression and clinical data of the patients were preprocessed. Key miRNAs associated with the prognosis of cervical cancer were identified by weighted gene co-expression network (WGCNA). The correspond-ing target genes were intersected with differentially expressed genes (DEGs) acquired from variation analysis, and the pathways and functional enrichment of genes were analyzed. Key genes were screened by Kaplan-Meier (K-M) survival analysis. Risk models were constructed using Cox proportional hazard regression model and the Least Absolute Shrinkage and Selection Operator (LASSO) method, and the predictive value of the models was evalu-ated by time-associated receiver operating characteristic (ROC) curves. Finally, independent prognostic factors were identified by COX analysis. Results: The hsa-miR-3150b-3p associated with the prognosis of CESC was identified by WGCNA. A total of 136 target genes were differentially expressed in CESC tissue and were associated with biologi-cal processes such as phylogeny, multicellular organism development and cell development. CBX7, ENPEP, FAIM2, IGF1, NUP62CL and TSC22D3 were associated with the prognosis of CESC, and a prognostic prediction model was constructed using these six genes, which had a good predictive value for the prognosis of cervical cancer within 1, 3 and 5 years (AUC: 0.784, 0.680 and 0.683, respectively). Among them, ENPEP (hazard ratio = 1.3996, 95% confidence interval: 1.0552-1.8565) was identified as an independent prognostic factor. Conclusions: In this study, a highly accurate prognostic model consisting of six gene signatures was developed to predict the prognosis of patients with cervical cancer, which provides a reference for developing individualized treatment plans for patients.