We measured the accuracy of p16(INK4a)-Ki67 (CINtec PLUS, Roche, Mannheim, Germany), and E6/E7mRNA (types 16/18/31/33/45 NucliSENS easyQ, bioMérieux, Boxtel, The Netherlands) as triage test, alone and combined with cytology.
Six thousand two hundred and seventy two women were recruited in a population-based screening using HPV DNA as primary test; 396 were positive and were tested for cytology and biomarkers. All tests were performed on the same sample. Cytology-positive women were referred to colposcopy; cytology-negative women were referred to one-year HPV re-testing. The endpoint was CIN2+ at baseline or follow up.
Sensitivity was 77.6% (95% confidence interval (CI) 65.3-86.7) and 53.2% (95%CI: 40.3-65.4) for cytology at atypical squamous cells of udetermined significance (ASC-US) and high-grade threshold, and 87.6% (95%CI:75.7-93.6), and 80.8% (95%CI: 67.6-89.8) for p16INK4a-Ki67, and E6/E7mRNA, respectively. Colposcopy referral was 36% (95%CI: 31.2-40.9) and 11.2% (95%CI: 7.8-14.1) for cytology at ASC-US and high-grade threshold, respectively, and 36.0% (95%CI: 29.9-29.6), and 47.5% (95%CI: 32.5-42.4) for p16(INK4a)-Ki67, and E6/E7mRNA, respectively. Strategies referring high-grade cytology or biomarker positive women to colposcopy reached sensitivity close to 100%, with modest increase in colposcopy referral.
The high sensitivity of combined strategies probably allows longer intervals in HPV-positive, triage-negative women.