A large audit of Cobas CT/NG screening and Xpert CT/NG supplemental testing reveals low confirmatory rates in oropharyngeal samples with evidence of commensal cross-reactivity with an isolate of Neisseria lactamica.

Abstract

Sensitive and high-throughput testing for Neisseria gonorrhoeae (NG) is widely underpinned by automated commercial solutions using nucleic acid amplification tests. We outline our experience of using Roche Cobas 8800 CT/NG (Cobas NG) for primary testing of NG at scale. Over 360000 specimens were tested over a three-year period. A disproportionate number of 'indeterminate' oropharyngeal swabs (25%) tested positive by Cobas NG but failed to confirm by Xpert CT/NG, significantly higher than genital samples (4%) (chi-square p-value <0.05). Such samples are highly problematic, as they fail to provide a definitive result for the patient. To investigate this non-specificity, publicly available sequences for Cobas NG DR-9 targets were aligned with Neisseria genome databases. We found a strain of Neisseria lactamica (PubMLST 49718), isolated from a throat swab, which was predicted to cross-react with Cobas NG. We subsequently confirmed a Cobas NG false-positive result with this isolate. BLAST analysis showed the target gene NG DR-9 was rare among available non-NG Neisseria whole genome sequences. This may suggest the existence of an as-yet unknown and difficult to cultivate oral commensal that contains NG DR-9. Our findings further highlight the need for NG confirmatory testing of oropharyngeal samples screened with Cobas NG.