First-void urine testing for high-risk human papillomavirus: A within-cohort comparison of three preanalytical approaches

To compare three preanalytical approaches for first-void urine (FVU)-based high-risk human papillomavirus (hrHPV) DNA detection against paired cervical swab (CS) specimens.

This cross-sectional observational study included 69 participants with a previous positive hrHPV screening result and paired CS and FVU specimens. From each FVU specimen, three urine preparations were generated: unpreserved urine, EDTA-treated urine for DNA stabilization, urine buffered with urine conservation medium (UCM; Novosanis, Wijnegem, Belgium). All specimens were analyzed using the QIAscreen HPV PCR Test (Qiagen, Hilden, Germany). Agreement with paired CS specimens, genotype-specific agreement, paired CS cycle threshold (Ct) values, and participant-reported perceptions of urine sampling were assessed.

Overall agreement between urine and CS specimens was moderate across all three approaches, with Cohen's κ values of 0.566 for unpreserved urine, 0.538 for EDTA-treated urine, and 0.581 for UCM-buffered urine. Paired CS Ct values were lower in participants with hrHPV DNA-positive urine specimens than in those with hrHPV DNA-negative urine specimens, consistent with greater urinary detectability at higher cervical hrHPV DNA levels. Genotype-specific agreement was stronger for HPV16 and HPV18 than for pooled non-16/18 hrHPV. Participant-reported data suggested good acceptability of urine sampling.

The three approaches showed broadly similar results to one another, while agreement with paired CS specimens was moderate across all approaches. These findings support further evaluation of urine-based hrHPV testing as a non-invasive sampling option that could help expand access to testing, particularly in low-income and other resource-limited settings.