Performance of urine samples compared to cervical samples for detection of precancer lesions among HPV-positive women attending colposcopy clinic in Mexico City

Abstract

Abstract Background High-risk human papillomavirus (hrHPV) detection in self-collected urine samples (SeCUS) may be a promising alternative for cervical cancer screening because of its greater acceptability, as long as it can offer comparable sensitivity to clinician-collected cervical samples (CCoS) for detecting precancer lesions. Objective To evaluate the performance of the SeCUS compared to that of the CCoS for cervical intraepithelial neoplasia grade 3 (CIN3) detection among hrHPV-positive women receiving colposcopy in Mexico City using different specific extended HPV typing procedures: HPV16/18, HPV16/18/35/39/68 or HPV16/18/35/39/68/31. Methods From March 2017 to August 2018, 4,158 female users of the cervical cancer screening program at Tlalpan Sanitary Jurisdiction in Mexico City were invited to participate in the FRIDA-Tlalpan study. All participants provided ≥ 30 mL of SeCUS, and then a CCoS was obtained with Cervex-Brush®, which was used for hrHPV typing. Participants who tested positive for hrHPV in CCoS were referred for colposcopy for diagnostic confirmation, and all SeCUS of these women were also tested for hrHPV typing. Results In total, 561 hrHPV-positive women were identified by CCoS via colposcopy, and 82.2% of the SeCUS of these women were also hrHPV positive. From both CCoS and SeCUS, 7 cases of CIN3 were detected. Considering HPV16/18 typing, CCoS and SeCUS detected 4 cases of CIN3, but after HPV16/18/35/39/68/31 extension typing, both CCoS and SeCUS detected all 7 of the CIN3 cases among the hrHPV-positive women. Conclusions Using extended hrHPV typing based on HPV16/18/35/39/68/31, our results suggest that the performance of SeCUS may be equivalent to that of CCoS for detecting CIN3 lesions. Although our results are inconclusive, they support the hypothesis that SeCUS may be an attractive alternative worthy of further research.