Cross-sectional study of HPV testing in self-sampled urine and comparison with matched vaginal and cervical samples in women attending colposcopy for the management of abnormal cervical screening

Objectives: Human papillomavirus (HPV) testing in cervical screening offers the potential for self-sampling to improve uptake among non-attenders. High-risk (HR) HPV detection in urine shows promise, but few studies have examined its sensitivity for cervical intraepithelial neoplasia (CIN2+) detection compared with standard cervical samples. The aims of this cross-sectional study were to optimise conditions for urine testing for HPV detection; to determine concordance for HR-HPV detection in matched urine, vaginal and cervical samples; to compare the sensitivity of HR-HPV testing for the detection of CIN2+ in matched samples; and to determine the acceptability of urine testing for cervical screening.

Design: Cross-sectional study.

Setting: Secondary care colposcopy clinic in North West England.

Participants: Women aged 25 years of age or older, attending colposcopy clinic for management of abnormal cervical screening results or a suspicious-looking cervix. In total, 104 women took part in the study. Triple matched samples were available for 79 and 66 women using Abbott RealTime (ART) and Roche Cobas 4800 (RC), respectively.

Intervention: Self-collected urine and vaginal samples and practitioner-obtained cervical samples were tested for HR-HPV by ART and RC assays, including comparison of neat and preservative-fixed urine. Colposcopic opinion was recorded and directed cervical biopsies taken if clinically indicated. The acceptability of self-testing was evaluated by questionnaire.

Primary outcome measure: The sensitivity of urine to detect underlying CIN2+.

Secondary outcome measures: The comparative sensitivity of vaginal and cervical samples to detect CIN2+; the acceptability of urine sampling.

Results: Preservative-fixed, but not neat urine, showed good concordance with vaginal samples for the detection of HR-HPV. The sensitivity for detecting CIN2+ was 15/18 (83%) for urine and 16/18 (89%) for cervical and vaginal samples by ART, and 15/17 (88%) for all samples by RC. Urine-based testing was broadly acceptable to women.

Conclusions: Urinary HR-HPV detection offers an alternative strategy of cervical screening. Larger studies to determine its clinical utility are warranted.