An 8-gene signature predicts the prognosis of cervical cancer following radiotherapy

Gene expression and DNA methylation levels affect the outcomes of patients with cancer. The present study aimed to establish a multigene risk model for predicting the outcomes of patients with cervical cancer (CerC) treated with or without radiotherapy. RNA sequencing training data with matched DNA methylation profiles were downloaded from The Cancer Genome Atlas database. Patients were divided into radiotherapy and non-radiotherapy groups according to the treatment strategy. Differently expressed and methylated genes between the two groups were identified, and 8 prognostic genes were identified using Cox regression analysis. The optimized risk model based on the 8-gene signature was defined using the Cox's proportional hazards model. Kaplan-Meier survival analysis indicated that patients with higher risk scores exhibited poorer survival compared with patients with lower risk scores (log-rank test, P=3.22×10(−7)). Validation using the GSE44001 gene set demonstrated that patients in the high-risk group exhibited a shorter survival time comprared with the low-risk group (log-rank test, P=3.01×10(−3)). The area under the receiver operating characteristic curve values for the training and validation sets were 0.951 and 0.929, respectively. Cox regression analyses indicated that recurrence and risk status were risk factors for poor outcomes in patients with CerC treated with or without radiotherapy. The present study defined that the 8-gene signature was an independent risk factor for the prognosis of patients with CerC. The 8-gene prognostic model had predictive power for CerC prognosis.