Comparison of HPV-positive triage strategies combining extended genotyping with cytology or p16/ki67 dual staining in the Italian NTCC2 study

Abstract

Background: Each high-risk HPV genotype has different oncogenic potential, and the risk of CIN3+ varies according to genotype. We evaluated the performance of different strategies of HPV-positivity triage combining cytology, p16/ki67 dual staining (DS), and extended genotyping.

Methods: Samples from 3180 consecutive women from the NTCC2 study (NCT01837693) positive for HPV DNA at primary screening, were retrospectively analyzed by the BD Onclarity HPV Assay, which allows extended genotyping. Genotypes were divided into three groups based on the risk of CIN3+. HPV DNA-positive women were followed up for 24 months or to clearance.

Findings: Combining the three groups of genotypes with cytology or DS results we identify a group of women who need immediate colposcopy (PPV for CIN3+ from 7.8 to 20.1%), a group that can be referred to 1-year HPV retesting (PPV in those HPV-positive at retesting from 2.2 to 3.8), and a group with a very low 24-month CIN3+ risk, i.e. 0.4%, composed by women cytology or DS negative and positive for HPV 56/59/66 or 35/39/68 or negative with the Onclarity test, who can be referred to 3-year retesting.

Interpretation: Among the baseline HPV DNA positive/cytology or DS negative women, the extended genotyping allows to stratify for risk of CIN3+, and to identify a group of women with a risk of CIN3+ so low in the next 24 months that they could be referred to a new screening round after 3 years.

Funding: Italian Ministry of Health (grant number RF-2009-1536040). Hologic-Genprobe, Roche Diagnostics, and Becton & Dickinson provided financial and non-financial support.

Keywords: Accuracy; Cervical cancer screening; Cervical intraepithelial neoplasia; HPV DNA testing; HPV genotyping; Human papillomavirus; Triage; p16/ki67 dual staining.

Fig. 1

Flowchart of the 3129 HPV DNA positive women triaged with cytology. All the baseline analyses are marked in black. All the analyses performed at 1 year are marked in red. As for the NTCC2 study design, cytology-positive women were referred to immediate colposcopy, while cytology-negative women were randomized to immediate colposcopy or to 1-year HPV DNA retesting. Those still HPV DNA positive were referred to colposcopy. The endpoints, CIN2+ and CIN3+ lesions, are reported according to the baseline genotyping results stratified based on the oncogenic potential of the genotype/s revealed.

Fig. 2

Proportion of women referred to colposcopy, to 1-year retesting, or to 3-year retesting, and proportion of CIN3+ lesions, detected at each step, in a cohort of 100 HPV DNA positive women applying, as triage strategy, three most representative combinations of genotyping and cytology among those described in Table 2. Cyto-1: 2-level protocol with partial genotyping; Cyto-3: 3-level protocol with partial genotyping; Cyto-5: 3-level protocol with extended genotyping.

Fig. 3

Predicted outcomes, the proportion of women referred to colposcopy, to 1-year retesting, or to 3-year retesting, and proportion of CIN3+ lesions, detected at each step, in a cohort of 100 HPV DNA positive women applying, as triage strategy, three most representative combinations of genotyping and dual staining described in Table 2. DS-1: 2-level protocol with partial genotyping; DS-3: 3-level protocol with partial genotyping; DS-5: 3-level protocol with extended genotyping.

Fig. 4

Summary graph of the overall colposcopy referral and 24-month CIN3+ risk, for triage negative women referred at 3-year retesting, considering all the 3-level triage strategies described in Table 2.