Evaluation of the Cobas 4800 HPV Test for Detecting High-Risk Human Papilloma-Virus in Cervical Cytology Specimens

As new platforms for high-risk strains of human papillomavirus (HR HPV) testing are introduced into the clinical laboratory, it is important to verify their performance and agreement. In this validation study, post-aliquot cervical cytopathology specimens (n = 226) were used to analyze agreement between the Invader HPV ASR assay (Hologic) and the recently FDA-approved Cobas 4800 high-risk HPV assay (Roche). Residual sample from 92 Invader positive and 134 Invader negative samples were analyzed with the Cobas 4800 test. Discordant results were further analyzed by Linear Array HPV genotype testing (Roche). To assess intra- and inter-run precision, 31 Invader positive samples were run in duplicate on the Cobas 4800 by different operators over multiple days and purchased HR HPV DNA control was run in ten replicates. Cross-contamination during cytology processing was evaluated by spiking 6 Invader negative samples with different volumes of Acrometrix HPV High Risk Positive Control and analyzed on the Cobas with 4 negative samples in between. There was significant discordance between the assays (p < 0.001; exact McNemar X(2) test), with overall agreement of 82%. Of the 92 Invader positive samples, 58 (63%) were positive with the Cobas assay, while 34 (37%) were negative. Of the 134 Invader negative samples, 6 (4%) were positive with the Cobas while 128 (96%) were negative. The observed discordance may be attributed to the previously described false positive rate of the Invader ASR assay. The Cobas 4800 high-risk HPV assay is a viable new tool for use in the clinical setting to identify high-risk HPV.