Accuracy of Liferiver HarmoniaHPV and VenusHPV Assays on Urine and Vaginal Self-Samples

In this report, the clinical performance of Liferiver HarmoniaHPV and Liferiver VenusHPV was evaluated under the VALHUDES framework. Five hundred and twenty-three women collected first-void urine (FVU) with Colli-Pee and vaginal samples with Evalyn Brush or Qvintip. Cervical samples were taken with the Cervex Brush by a clinician. Both vaginal and cervical samples were resuspended in 20 mL ThinPrep. Triplet samples from 499 women were tested with HarmoniaHPV and VenusHPV tests. The clinical accuracy of HarmoniaHPV did not differ in FVU and vaginal self-samples versus cervical samples. The relative sensitivity for CIN2+ on FVU and vaginal samples was 0.95 [95% CI 0.89-1.02] and 0.95 [95% CI 0.88-1.02], respectively. Relative specificity for < CIN2 was 0.95 [0.86-1.04] on FVU and 0.93 [0.86-1.01] on vaginal samples. VenusHPV demonstrated lower sensitivity on both self-sample types, whereas the specificity was similar to cervical samples. Post-hoc adjustment of the VenusHPV C-values improved sensitivity (ratio FVU/cervical = 0.94 [95% CI 0.88-1.00]; ratio vaginal/cervical = 0.96 [95% CI 0.92-1.01]) without compromising specificity (ratio FVU/cervical = 1.00 [0.92-1.09]; ratio vaginal/cervical = 0.95 [95% CI 0.88-1.02]) on both self-samples. In conclusion, HarmoniaHPV and VenusHPV tests demonstrated similar clinical accuracy on FVU and vaginal self- versus cervical samples, although VenusHPV test required cut-off optimization.