To compare the high-risk (hr) HPV status of self-collected vaginal fluid and first void urine with that of physician-sampled cervical scrapes (CS) collected at the same visit 6 months post first life-time treatment of high grade CIN from 500 women using a clinically validated PCR-based test for cervical cancer screening and assesses correlation of results to post-treatment cytology.
Triplets of pre-cytology physician-sampled CS (PreservCyt LBC, Hologic), vaginal fluid (Qvintip, Aprovix) and urine were collected from 200 women enrolled at date (pre-treatment histology: 45 CIN2, 145 CIN3, 6 CIN3/AIS, 4 AIS; 30% with unclear cone margins). Qvintip brush heads (air dried; stored at room temperature) were transferred into Cervi-Collect Tubes (Abbott) prior to testing. Urine samples were transferred into Cervi-Collect Tubes within 30 minutes from collection and stored frozen until testing. HrHPV DNA testing was performed with the clinically validated RealTime High Risk HPV assay (Abbott) using the manufacturer´s standard protocol on the m2000 System. RealTime High Risk HPV is a multiplex real-time PCR detecting 14 hrHPV types and simultaneously differentiating HPV16 and HPV18 from 12 non-HPV 16/18 types; its internal control identifies a human ß-globin sequence to ensure sample cellularity.
Valid test results were obtained from all physician-sampled CS samples and the vast majority of self-collected specimens (Qvintip: 199; urine: 192). Evaluation of results from 191 matched triplets available for evaluation revealed high overall-concordance of hrHPV DNA test results between self-and physician-sampled specimens (Qvintip 88.0%, k=0.7 urine 90.1%, k=0.7 and across ascending cytological categories (N=155 WNL: Qvintip 87.1%, urine 91.6%; N=27 LSIL: Qvintip 92.6%, urine 81.5%; N=9 HSIL: Qvintip and urine 88.9%). Partial typing pattern in women with HSIL were almost identical among all three sample types, while a trend towards higher discrepancy of partial typing pattern was observed in women with LSIL and WNL, the latter mainly driven by the presence of non-HPV16/18 in self-collected materials.
Preliminary results from this study show good correlation of hrHPV DNA results from self-collected vaginal fluid and urine with those from physician-sampled CS collected in a test-of-cure setting. Self-collection may have potential for post-surgical follow up of high-grade cervical lesions and support reducing unnecessary procedures on healthy women by physicians or midwifes. However, larger studies are required to confirm these findings based on post-treatment outcome and evaluate potential limitations of the approach.