Early Post-Therapy Clearance of Human Papillomavirus and Treatment Response in Cervical Carcinoma

Abstract

Persistent high-risk human papillomavirus (HR-HPV) infection is known to be a causative factor in the development of cervical cancer. However, less is known about the significance of persistent infection following definitive radiation (RT). The aim of this study was to evaluate the association between HR-HPV clearance and patient outcome in the post-therapy setting. We hypothesized that early clearance of HR-HPV expression following definitive RT would be associated with improved outcomes. An institutional database of patients with cervical cancer treated with definitive intent RT was reviewed. Patients were eligible for inclusion if pre-treatment testing indicated HR-HPV infection and post-treatment HPV testing using a cervical swab with RNA-based HPV assays was conducted within 6 weeks of RT completion. Post RT, patients were stratified into two cohorts based on HPV RNA detection. Early clearance (EC) was defined as no HPV RNA detection and persistent infection (PI) as detectable HPV RNA. The Student’s t-test and Fisher's exact test were used to compare baseline clinical and demographic covariates. The Kaplan-Meier method and log-rank test were used to estimate and compare pelvic failure (PF), distant failure (DF), progressive disease (PD), and overall survival (OS) between cohorts and Cox regression to identify predictors of these outcome metrics. Ninety-one patients were eligible for the analysis, and 76 (84%) experienced EC while 15 (16%) had PI. Age at diagnosis, FIGO stage, histology, pre-treatment lymph node status, and use of chemotherapy were similar between the groups, but PI was more common among African-Americans (7/22 patients, 32%) compared to other ethnicities (8/69 patients, 12%). HPV genotype was available for 71 patients, and on subset analysis 51 (72%) had HPV-16 disease while 20 (28%) had other strains of HR-HPV. EC occurred more often in patients with HPV-16 (45/51 patients, 88%) compared to other strains (13/20 patients, 65%) (p=0.03). For the whole cohort, mean follow-up time was 12 months (range 1-52 months). At one year follow up, actuarial PF rates were 12% for patients with EC and 40% for patients with PI (p=0.02), and DF rates were 12% in the EC cohort and 29% in the PI cohort (p=0.03). Comparing EC and PI patients, rates of PD were 16% and 49% respectively (p=0.02). Notably, OS at one year was higher in patients experiencing EC compared to those with PI (96% v. 84%, p<0.01). On multivariate analysis (MVA), EC remained a significant predictor of mortality (HR=0.13, 95% confidence interval (CI): 0.03 – 0.71, p=0.01) as did increasing FIGO stage (HR=2.02, 95% CI: 1.08-4.15, p=0.04). Similarly, EC (HR=0.23, 95% CI: 0.07-0.81, p=0.02) and FIGO stage (HR=1.54, 95% CI: 1.04-2.51, p=0.04) were identified as predictors of PD on MVA. Early post-therapy clearance of HR-HPV within 6 weeks of treatment completion is associated with higher rates of disease control and improved overall survival in patients with cervical cancer.