Prospective evaluation of the Hybrid Capture 2 and AMPLICOR human papillomavirus (HPV) tests for detection of 13 high-risk HPV genotypes in atypical squamous cells of uncertain significance

The use of high-risk human papillomavirus (hrHPV) testing as an adjunct to cervical cytology in population-based screening programs is currently based on DNA hybridization and PCR assays. The aim of this study was to prospectively assess the diagnostic performance of the Hybrid Capture 2 test (HC2; Digene Corporation) in comparison with that of the recently developed PCR-based AMPLICOR HPV test (Roche Molecular Systems) for the detection of 1.3 hrHPV types. A reverse line blot hybridization assay (Innogenetics) was used as an internal reference standard in discordant cases. Two hundred seventy-one patients with atypical squamous cells of uncertain significance (ASCUS) in cervical samples underwent hrHPV testing. The chi-square test was performed to compare respective proportions. Totals of 160/271 (59%) and 156/271 (58%) were found to be positive for hrHPV with HC2 and AMPLICOR, respectively. Concordant results were obtained for 235 (86.7%) of the 271 samples (kappa statistic, 0.73 +/- 0.04). Considering types 26, 53, and 66 as oncogenic types, negative predictive values (NPVs) of HC2 and AMPLICOR were 92.8% and 87.8%, respectively (difference was not significant), and their respective accuracies were 94.8% and 91.9% (difference was not significant). Considering types 26, 53, and 66 as not oncogenic, the respective HC2 and AMPLICOR NPVs were 92.8% and 97.4% (difference was not significant), and accuracy was significantly higher for the AMPLICOR assay (95.9% versus 90.8% for HC2) (P < 0.05). For ASCUS samples, the NPV was 92.8% for HC2 testing and might be compromised if the copy number of HPV DNA was low. The NPV was 97.4% for the AMPLICOR assay and might be compromised if HPV types 26, 53, and 66 were considered oncogenic. The accuracy of these two assays is good and is compatible with routine clinical use in the triage of ASCUS cases.