Human papillomavirus type 16 infections and 2-year absolute risk of cervical precancer in women with equivocal or mild cytologic abnormalities

Background. The 2-year absolute risk for cervical precancer attributable to infection by human papillomavirus type 16 (HPV16), the most common and oncogenic HPV type, in the millions of women diagnosed annually with equivocal or mildly abnormal cytology has not been definitively evaluated. Methods: Baseline cervical specimens of 5060 women with equivocal (atypical squamous cells of undetermined significance [ASCUS]) or mildly abnormal (low-grade squamous intraepithelial lesion [LSIL]) cytology were tested for HPV DNA using Hybrid Capture 2 (HC2) and typespecific L1 consensus primer polymerase chain reaction. We calculated absolute risks with 95% confidence intervals (CIs) for cumulative diagnosis, during the 2-year study period, of cervical intraepithelial neoplasia grade 3 (CIN3) (n = 535) or cancer (n = 7) (collectively referred to as CIN3) and compared risk by HPV16 status and by other oncogenic HPV types using logistic regression. All statistical tests were two-sided. Results: The baseline prevalences of HPV16 in women with ASCUS or LSIL cytology were 14.9% and 21.1%, respectively. Women with ASCUS or LSIL cytology who were HPV16 DNA positive at baseline had 2-year cumulative absolute risks for >= CIN3 of 32.5% (95% CI = 28.4% to 36.8%) and 39.1% (95% CI = 33.8% to 44.7%), respectively. By comparison, women with ASCUS who were positive by HC2 for other oncogenic HPV types combined had an 8.4% (95% CI = 6.9% to 10.4%) risk for >= CIN3, which was similar to the risk posed by having ASCUS (risk = 8.8%, 95% CI = 7.9% to 9.8%) without knowledge of the oncogenic HPV DNA status. Women with LSILs who were positive by HC2 for other oncogenic HPV types combined had a 9.9% (95% CI = 8.0% to 12.0%) 2-year risk for >= CIN3, which was less than the risk posed by having LSILs (risk = 15.0%,95% CI = 13.3% to 16.9%) without knowledge of the oncogenic HPV DNA status. Together, women with ASCUS or LSILs who were HPV16-positive had the highest 2-year risk for >= CIN3 compared with women who were HPV-negative (odds ratio [OR] = 38, 95% CI = 22 to 68; P <.001), fivefold greater than the increased risk in women who were positive for other oncogenic HPV types (OR = 7.2, 95%CI = 4.2 to 13, P <.001). Conclusions: Distinguishing the high absolute risk for cervical precancer in HPV16-positive women from the lower risk posed by other oncogenic HPV types might have clinical implications.