The clinical value of HPV genotyping in triage of women with high-risk-HPV-positive self-samples.

Cytology alone, or combined with HPV16/18 genotyping, might be an acceptable method for triage in hrHPV-cervical cancer screening. Previously studied HPV-genotype based triage algorithms are based on cytology performed without knowledge of hrHPV status. The aim of this study was to explore the value of hrHPV genotyping combined with cytology as triage tool for hrHPV-positive women. 520 hrHPV-positive women were included from a randomised controlled self-sampling trial on screening non-attendees (PROHTECT-3B). Eighteen baseline triage strategies were evaluated for cytology and hrHPV genotyping (Roche Cobas 4800) on physician-sampled triage material. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), referral rate, and number of referrals needed to diagnose (NRND) were calculated for CIN2+ and CIN3+. A triage strategy was considered acceptable if the NPV for CIN3+ was ≥98%, combined with maintenance or improvement of sensitivity and an increase in specificity in reference to the comparator, being cytology with a threshold of atypical cells of undetermined significance (ASC-US). Three triage strategies met the criteria: HPV16+ and/or ≥LSIL; HPV16+ and/or ≥HSIL; (HPV16+ and/or HPV18+) and/or ≥HSIL. Combining HPV16+ and/or ≥HSIL yielded the highest specificity (74.9%, 95% CI 70.5-78.9), with a sensitivity (94.4%, 95% CI 89.0-97.7) similar to the comparator (93.5%, 95% CI 87.7-97.1), and a decrease in referral rate from 52.2% to 39.5%. In case of prior knowledge of hrHPV presence, triage by cytology testing can be improved by adjusting its threshold, and combining it with HPV16/18 genotyping. These strategies improve the referral rate and specificity for detecting CIN3+ lesions, while maintaining adequate sensitivity.

Comparative performance of novel self-sampling methods in detecting high-risk human papillomavirus in 30,130 women not attending cervical screening.

We determined whether the participation rate for a brush-based cervicovaginal self-sampling device is noninferior to the participation rate for a lavage-based one for testing for hrHPV (high-risk human papillomavirus). Additionally, positivity rates for hrHPV, the detection rates for cervical intraepithelial neoplasia grades 2 and 3 or worse (CIN2+/3+), and user comfort were compared. A total of 35,477 non-responders of the regular cervical screening program aged 33-63 years were invited to participate. Eligible women (n = 30,130) were randomly assigned to receive either a brush-based or a lavage-based device, and a questionnaire for reporting user convenience. Self-sampling responders testing hrHPV-positive were invited for a physician-taken sample for cytology; triage-positive women were referred for colposcopy. A total of 5,218 women participated in the brush-based sampling group (34.6%) and 4809 women in the lavage-based group (31.9%), i.e. an absolute difference of 2.7% (95%CI 1.8-4.2). The hrHPV-positivity rates in the two groups were identical (8.3%, relative risk (RR) 0.99, 95%CI 0.87-1.13). The detection of CIN2+ and CIN3+ in the brush group (2.0% for CIN2+; 1.3% for CIN3+) was similar to that in the lavage group (1.9% for CIN2+; 1.0% for CIN3+) with a cumulative RR of 1.01, 95%CI 0.83-1.24 for CIN2+ and 1.25, 95%CI 0.92-1.70 for CIN3+. The two self-sampling devices performed similarly in user comfort. In conclusion, offering a brush-based device to non-responders is noninferior to offering a lavage-based device in terms of participation. The two self-sampling methods are equally effective in detecting hrHPV, CIN2+/CIN3+ and are both well accepted.