Comparison of different strategies for the triage to colposcopy of women tested high-risk HPV positive on self-collected cervicovaginal samples

Objective To identify the optimal strategy for the triage of women who test high-risk (hr) HPV positive on self-collected cervicovaginal samples. Methods This is a diagnostic accuracy sub-analysis of the GRECOSELF study, which reported on HPV-DNA testing with self-sampling in Greece. More than 13,000 women, 25–60 years old, who resided in rural areas of Greece, provided a self-collected cervicovaginal sample. Samples were tested for HPV-DNA and HPV16/18 genotyping with the cobas® HPV test (Roche® Molecular Systems, Pleasanton, CA, USA). HrHPV positive women were referred for colposcopy. Prior to colposcopy a physician-collected sample was obtained for cytology. After colposcopy/biopsy, women were classified as having cervical disease or not, and treated accordingly. Results Out of 1070 hrHPV positive women, 773 were subjected to colposcopy. Seventeen triage strategies, combining HPV16/18 genotyping and cytology, were investigated. The strategy referring to colposcopy women positive for HPV16 regardless of the cytology report, and women positive for other hrHPVs, in case of a subsequent atypical squamous cells of undetermined significance or worse (ASCUS+) cytology report, presented optimal trade-off; sensitivity 96.36% [(95%CI: (91.41–100.0)], positive predictive value (PPV) 27.46% [95%CI: (21.16–33.76)], and number of colposcopies required to detect one case of Cervical Intraepithelial Neoplasia grade-2 or worse (CIN2+) 3.64. Conclusions The optimal strategy for the triage to colposcopy of hrHPV positive women, detected with the cobas® HPV test on self-collected cervicovaginal samples, is referring all HPV16 positive women directly to colposcopy, and women positive for HPV18 or other hrHPVs only after an ASCUS or worse cytology report.