Accuracy of different triage strategies for human papillomavirus positivity in an Italian screening population

Abstract

How to manage human papillomavirus (HPV)-positive women in cervical cancer screening remains debated. Our study compared different strategies to triage HPV positivity in a large cohort of women participating in a population HPV-based screening program. Women were tested for HPV (Cobas 4800; Roche), and those positive were triaged with cytology; cytology-positives were referred to colposcopy, while negatives were referred to 1-year HPV retesting. All HPV-positive women were also evaluated with p16/ki67 dual staining (Roche). All lesions found within 24 months of follow-up were included in the analyses. Of the 70 146 women tested, 4757 (6.8%) were HPV-positive. Of these, 1090 were cytology-positive and were referred to colposcopy. Of the 2958 HPV-positive/cytology-negative women who presented at 1-year retesting, 1752 (59.9%) still tested positive. Cumulatively, 532 CIN2+ (including 294 CIN3+) were found. The sensitivity of cytology, HPV16/18 and p16/ki67 as triage test for CIN3+ was 67.9%, 56.0% and 85.0%, respectively. The positive predictive value (PPV) for immediate colposcopy referral was 21.0%, 15.8% and 22.9%, respectively. Combining cytology with typing increased sensitivity to 83.9% and lowered PPV to 14.8%, while combining p16/ki67 and typing increased sensitivity to 91.1%, lowering the PPV to 15.9%. Women negative to p16/ki67 triage presented a cumulative 1-year CIN3+ risk of about 1%. In conclusion, when triaging HPV positivity, p16/ki67 performed better than cytology with or without HPV16/18 genotyping. The strategies that included dual staining achieved sensitivity and low 1-year risk for CIN3+ sufficiently high enough to permit considering extending the surveillance interval to 2 to 3 years for HPV-positive/triage-negative women.

Keywords: accuracy; biomarkers; cervical cancer; cervical intraepithelial neoplasia; screening.