Urinary and Vaginal HPV Testing as a Novel Cervical Cancer Screening Tool: a Diagnostic Test Accuracy Study

Paired urine, vaginal, and clinician-collected cervical samples will be obtained from approximately 330 women aged 23-64 years referred for colposcopy and cervical biopsies due to abnormal screening result or referred for cervical excision due to treatment of CIN2+. Upon informed consent, the women will collect a first-void urine sample and a vaginal self-sample. Before colposcopy or excision, a clinician will collect a cervical sample. Among women referred for colposcopy, four cervical punch biopsies will be taken. For women referred for excision, a cone biopsy will be taken using a loop electrosurgical excision procedure. Upon arrival in the laboratory, the paired urine, vaginal and cervical samples will be HPV tested using the Cobas 4800 HPV DNA assay. Histological results will served as reference and be grouped as ≤CIN1 (normal tissue, CIN1) versus CIN2+. The performance of the six individual methylation markers and combinations thereof will be evaluated using the histological results as reference.

The primary endpoint will be the relative sensitivity and specificity of HPV testing in first-void urine, vaginal self-sample versus clinician-collected cervical samples to detect CIN2+. Non-inferiority of HPV testing between sample types will be evaluated against a margin of 90% for sensitivity and 98% for specificity. The performance of each methylation marker will be illustrated by ROC curves and assessed by the area under the curve.