Test Accuracy of Human Papillomavirus in Urine for Detection of Cervical Intraepithelial Neoplasia.

The objective was to assess the diagnostic test accuracy of high-risk human papillomavirus (hrHPV) testing of self-collected urine and cervicovaginal samples for the detection of cervical intraepithelial neoplasia grade 2 or higher (CIN2+). We recruited a convenience sample of women 25 to 65 years of age who were undergoing clinically indicated colposcopy at two medical centers in North Carolina between November 2016 and January 2019. Women with normal cytology results and positive hrHPV results were also recruited. Urine samples, self-collected cervicovaginal samples, provider-collected cervical samples, and cervical biopsy samples were obtained from all enrolled women. Samples were tested for hrHPV DNA using the Onclarity assay (Becton Dickinson, Sparks, MD). Biopsy samples were histologically graded as CIN2+ or <CIN2. We calculated the sensitivity and specificity for detection of CIN2+ and assessed agreement between sample collection methods. We included 307 women (median age, 36 years) with valid histology results and triple-matched urine, self-collected cervicovaginal, and provider-collected cervical hrHPV results; 83 women (27%) had CIN2+. Urine-based hrHPV testing correctly identified 80% of CIN2+ cases (95% confidence interval [CI], 71 to 88%) using the PCR cycle threshold (CT ) established for provider-collected cervical samples, but sensitivity remained below the estimates for self-collected cervicovaginal and provider-collected cervical samples (both 94% [95% CI, 89 to 99%]). Using a higher CT cutoff value of ≤40, 90% sensitivity was achieved for urine-based hrHPV testing. Agreement between results for urine samples and self-collected cervicovaginal samples (kappa = 0.58) or provider-collected cervical samples (kappa = 0.54) was moderate. Urine-based hrHPV testing may be a promising approach to improve cervical cancer screening coverage, especially among women with limited access to health care.