Clinically validated HPV assays offer comparable long-term safety in primary cervical cancer screening: A 9-year follow-up of a population-based screening cohort

Molecular testing for human papillomaviruses (HPV) is gradually replacing cytology in cervical cancer screening. In this longitudinal population-based cohort study, 4140 women 20 to 64 years old attending organized screening were tested at baseline by five different screening methods and followed for 9 years. To assess long-term safety, the cumulative risks of CIN2+/CIN3+ were estimated after a negative baseline result obtained by conventional cytology and four clinically validated HPV assays: Hybrid Capture 2 (hc2), RealTime High Risk HPV assay (RealTime), cobas 4800 HPV Test (cobas_4800), and Alinity m HR HPV (Alinity). HPV-negative women at baseline had a substantially lower risk for CIN2+ compared to those with normal baseline cytology: 0.84% (95% CI, 0.46-1.22), 0.90% (95% CI, 0.51-1.29), 0.78% (95% CI, 0.42-1.15), and 0.75% (95% CI, 0.39-1.11) for hc2, RealTime, cobas_4800, and Alinity, respectively, compared to 2.46% (95% CI, 1.88-3.03) for cytology. No differences were observed between HPV assays in longitudinal sensitivity (range: 86.21%-90.36%) and negative predictive values (range: 99.54%-99.70%) for CIN2+ in women >= 30 years, but were significantly different from cytology (p < .05). The 9-year cumulative risk of CIN2+ differed significantly between HPV genotypes, reaching 32.1% (95% CI, 14.5-46.1) for HPV16, 24.9% (95% CI, 4.7-40.8) for HPV18/45, 27.2% (95% CI, 14.6-37.8) for HPV31/33/35/52/58, and 8.1% (95% CI, 0.0-16.7) for HPV39/51/56/59. Four clinically validated HPV assays showed comparable safety and better assurance against precancerous lesions than cytology, but some important differences were identified in the performance characteristics of HPV assays impacting the referral rate. Information about the HPV genotype is valuable for guiding further clinical action in HPV-based screening programs.