Role of Human Papillomavirus DNA Load in Predicting the Long-term Risk of Cervical Cancer: A 15-Year Prospective Cohort Study in China.

Abstract

Biomarkers highly predictive of cervical cancer are urgently needed for triaging human papillomavirus (HPV)-positive women. A total of 1997 women aged 35-45 years in Shanxi, China, were recruited in 1999, and follow-up visits were conducted in 2005, 2010, and 2014. HPV load was measured by the Hybrid Capture 2 assay. Findings were determined by relative light units/cutoff (RLU/CO) and categorized into 4 groups: negative ( <1.0), low (range, 1.0 to <10.0), moderate (range, 10.0 to <100.0), and high (range, 100.0-∞). Cumulative incidence rates (CIRs) and adjusted hazard ratios (aHRs) for cervical intraepithelial neoplasia grade 2 or worse (CIN2+) were calculated for viral load subgroups, using survival analysis. Among 1739 women with normal or CIN1 pathological findings at baseline, 15-year CIRs for CIN2+ for those who were HPV negative and those with low, moderate, and high HPV loads groups were 3.1%, 8.4%, 19.9%, and 22.0%, respectively (Ptrend <.001). Compared with women who were negative for HPV from baseline through follow-up, those who had decreasing, increasing, or stable moderate/high loads had aHRs of 9.1, 38.7, or 379.7, respectively, for CIN2+. There was no significant difference between triage based on cytologic findings (for those with atypical squamous cells of undetermined significance or more-severe findings) and that based on a moderate/high HPV load for HPV primary screening (P = .343). A moderate/high HPV load may accelerate the progression of cervical precancers and potentially could be used as a triage indicator for HPV-positive women.