Urine high‐risk human papillomavirus testing as an alternative to routine cervical screening: A comparative diagnostic accuracy study of two urine collection devices using a randomised study design trial

Jennifer C. Davies,Alexandra Sargent,Elisabeth Pinggera,Suzanne Carter,Clare Gilham,Peter Sasieni,Emma J. Crosbie
DOI: https://doi.org/10.1111/1471-0528.17831
2024-04-27
BJOG An International Journal of Obstetrics & Gynaecology
Abstract:Objective To evaluate the sensitivity of human papillomavirus (HPV) tested urine to detect high‐grade cervical precancer (cervical intraepithelial neoplasia grade 2+ [CIN2+]) using two urine collection devices. Design Randomised controlled trial. Setting St Mary's Hospital, Manchester, UK. Population Colposcopy attendees with abnormal cervical screening; a total of 480 participants were randomised. Matched urine and cervical samples were available for 235 and 230 participants using a first‐void urine (FVU)‐collection device and standard pot, respectively. Methods Urine was self‐collected and mixed with preservative – randomised 1:1 to FVU‐collection device (Novosanis Colli‐pee® 10 mL with urine conservation medium [UCM]) or standard pot. Matched clinician‐collected cervical samples were taken before colposcopy. HPV testing used Roche cobas® 8800. A questionnaire evaluated urine self‐sampling acceptability. Main outcome measures The primary outcome measured sensitivity of HPV‐tested urine (FVU‐collection device and standard pot) for CIN2+ detection. Secondary outcomes compared HPV‐tested cervical and urine samples for CIN2+ and evaluated the acceptability of urine self‐sampling. Results Urine HPV test sensitivity for CIN2+ was higher with the FVU‐collection device (90.3%, 95% CI 83.7%–94.9%, 112/124) than the standard pot (73.4%, 95% CI 64.7%–80.9%, 91/124, p = 0.0005). The relative sensitivity of FVU‐device‐collected urine was 0.92 (95% CI 0.87–0.97, pMcN = 0.004) compared with cervical, considering that all women were referred after a positive cervical HPV test. Urine‐based sampling was acceptable to colposcopy attendees. Conclusions Testing of FVU‐device‐collected urine for HPV was superior to standard‐pot‐collected urine in colposcopy attendees and has promising sensitivity for CIN2+ detection. General population HPV testing of FVU‐device‐collected urine will establish its clinical performance and acceptability as an alternative to routine cervical screening.
obstetrics & gynecology
What problem does this paper attempt to address?