A new marker based on risk stratification of human papillomavirus DNA and tumor size to predict survival of locally advanced cervical cancer

Abstract

ObjectiveTo assess the prognostic value of human papillomavirus (HPV) viral load in locally advanced cervical carcinoma treated with radical concurrent chemoradiotherapy.MethodsFrom January 2012 to October 2013, a total of 246 locally advanced cervical carcinoma patients were included in this retrospective study. HPV DNA status was tested by Hybrid Capture 2 assay. Tumor size was measured on T2WI. All the patients in the study received concurrent cisplatin-based chemoradiotherapy with intensity-modulated radiotherapy and three-dimensional brachytherapy. Survival rate was calculated by the Kaplan-Meier method, and a log-rank test was used to compare the survival. Multivariate analysis employed the Cox regression model.ResultsThe median follow-up time was 52 months. The median value of HPV DNA was 163.13 relative light unit/cut-off (RLU/CO) (range 1.65–2162.62 RLU/CO). The 5-year overall survival, distant metastasis-free survival of patients in the low HPV DNA group (HPV DNA ≤ 163.13 RLU/CO) and the high HPV DNA group (HPV DNA > 163.13 RLU/CO) were 46.3 % vs 58.5 % (p = 0.009) and 65.9 % vs 75.6% (p = 0.003), respectively. Multivariate analysis showed that the HPV DNA, tumor size, and International Federation of Gynecology and Obstetrics (FIGO) stage were independent prognostic factors for overall survival and distant metastasis-free survival. We choose the tumor size and HPV DNA as the risk stratification factors to build a new prediction marker which can better predict overall survival for locally advanced cervical cancer than can the FIGO stage.ConclusionsHPV DNA may be a useful biomarker for locally advanced cervical cancer. Low HPV load predicts a worse survival. The new marker based on risk stratification by combining HPV DNA and tumor size is better associated with overall survival of locally advanced cervical cancer treated with concurrent chemoradiotherapy.