High-Risk Human Papillomavirus Screening of Gynecologic Specimens: Comparison of the Roche cobas 4800 With the Hologic Cervista Assays

Abstract

NorDxWith the commercial availability of new high-risk HPV screening assays, we sought to compare a real-time PCRbased assay with our current signal amplification assay. The Cobas 4800 HPV assay (Roche Molecular Diagnostics, Indianapolis, IN) recently received FDA approval for detection of 14 HR-HPV genotypes: HPV-16 and HPV-18 are each identified separately and the other 12 HR-HPV genotypes are identified in a pool. The Hologic Cervista HR-HPV reagents (Hologic, Bedford, MA) currently used by our laboratory detect 14 HR-HPV genotypes in 3 different groups or clades (A5/A6, A7, and A9). Samples positive for the HR-HPV screening assay are reflexed to the Hologic Cervista 16/18 genotype assay if requested by the clinician. The purpose of this study is to compare the Roche assay with our current method. We selected 272 gynecologic specimens collected in PreservCyt (Hologic) for this study. Cytology was performed, followed by evaluation of HR-HPV infection. Agreement between the 2 HPV assays was 83%. Of the 46 discrepant samples, the majority (52%) were negative by Roche and positive for all 3 HPV virus clades on the Hologic platform (>3.0 HPV-fold over zero, minimum fold over zero value for HPV positivity, 1.93). These "triple-positive" specimens all had a very high DNA content with an average DNA fold over zero between 18 and 25; the minimum valid value is 1.5-fold over zero for this assay. When the same specimens were diluted in PreservCyt and tested again on the Hologic platform, valid negative results were obtained. Some of these triple-positive samples were sent to a reference laboratory for confirmatory testing by a third method (Qiagen, Valencia, CA), and, again, all were negative. These results indicate that the triple-positive results obtained are falsepositives, confirming previous clinical observations showing high discrepancy rates between cytology and Hologic HPV results.

Nucleic acid‐based assays for the detection of high‐risk human papillomavirus: A technical review

“…In this case, specimens with very high total extracted DNA contents were associated with elevated FOZ ratios (average, 18-25) determined to represent a subset of samples with apparent false-positive results. 14 The Cervista HPV 16/18 test also uses Invader chemistry, and is approved by the FDA for the qualitative detection of DNA from hrHPV types 16 and 18 in cervical specimens. Approved indications include: 1) in women aged 30 years, the Cervista HPV 16/18 test can be used adjunctively with the Cervista HPV HR test in combination with cervical cytology to assess the presence or absence of hrHPV types 16 and 18, and 2) adjunctive use with the Cervista HPV HR test in patients with ASCUS cervical cytology results to assess the presence or absence of hrHPV types 16 and 18.…”

“…Another example of “triple‐positive” results has been reported by at least 1 study that sought to resolve discordant results obtained through comparison with an independent hrHPV test. In this case, specimens with very high total extracted DNA contents were associated with elevated FOZ ratios (average, 18‐25) determined to represent a subset of samples with apparent false‐positive results …”

Comparison of Commercial Hybridization and Automated Transcription-Mediated Amplification Modalities for Detection of High-Risk Human Papillomavirus Nucleic Acid

“…In an evaluation of Cervista and the DNA amplification-based COBAS 4800 HPV assay (Cobas; Roche), Webber et al (22) performed additional testing on 26 Cervista triple-positive specimens. The mean gDNA of this sample set was 20.75 (range of values, 13.48 to 25.47).…”