Health care policy trial of primary human papillomavirus-based cervical screening in Denmark: Comparison of three triage algorithms

Women 30-59 years were allocated to either HPV-based screening or cytology-based screening in this Danish health care policy trial. The optimal triage of HPV-positive women could be a combination of cytology triage with HPV genotyping or p16/Ki67 staining. We report number of screen positives, colposcopies, and cervical lesions of three different triage algorithms (p16/Ki67, HPV16/18, or HPV16/18/31/33/52) in HPV-positive women with low-grade cytological abnormalities. We included 178,317 women with a sample in 2021 of which 91,517 were screened with HPV and 86,800 with cytology. All women were followed for 18 months. Almost three times as many women screened positive with HPV-based screening compared to cytology-based screening (RR 2.99, 95% 2.93-3.05) and colposcopies derived from the screening program were also more common (RR 1.68, 95% 1.63-1.73). p16/Ki67 triage resulted in more colposcopies (RR 1.86, 95% 1.76-1.95) than HPV16/18 (RR 1.54, 95% 1.44-1.65) and HPV16/18/31/33/52 (RR 1.63, 95% 1.55-1.71). The excess in colposcopy referrals was reduced when non-screening-derived colposcopies were included (intention-to-treat). Nevertheless, more women with CIN2 or worse were detected in the HPV group than in the cytology group per screened woman; in the p16/Ki67 triage group (RR 1.65, 95% 1.54-1.77), in the HPV16/18 group (RR 1.36, 95% 1.23-1.50), and in the HPV16/18/31/33/52 group (RR 1.48, 95% 1.37-1.59). HPV-based screening, as compared with cytology screening, resulted in more screen positives, but all three triage algorithms substantially reduced the excess number of referrals to colposcopy. p16/Ki67 compared to triage with HPV16/18 may detect more cervical lesions.