Question
What are the performances of nucleic acid amplification assays (NAATs) for the detection of N. gonorrhoeae when tested on known bacterial strains and on Cobas Amplicor PCR assay N. gonorrhoeae positive and negative clinical samples?  

Design
The performances of two real-time PCR assays, developed on the LightCycler and based on amplification of a nested region of the cppB gene of the multicopy 4.2 kb cryptic plasmid of N. gonorrhoeae, were compared with the Abbott LCx ligase chain reaction and the Roche 16S rRNA PCR assays for the confirmation of samples positive for N. gonorrhoeae by the Roche Cobas Amplicor PCR assay.   

Participants
Known bacterial strains tested by the NAATs included 120 ATCC and clinical strains of N. gonorrhoeae isolated from 5 continents, 5 groups of N. meningititis, 7 other Neisseria species, 9 serotypes of C. trachomatis, 3 other Chlamydia species, 5 enteric bacterial species, S. epidermidis and T. vaginalis.  Clinical specimens included 122 testing positive and 50 testing negative for N. gonorrhoeae by the Roche Cobas Amplicor PCR assay.  The positive specimens comprised stored patient samples at the Royal Women’s Hospital in Melbourne, Australia, and included 99 tampon, 14 urine, 4 urethral, and 7 cervical samples.  

Description of Tests and Diagnostic Standard
DNA was extracted from each sample using the automated MagNA Pure LC (Roche Diagnostics) instrument.  The DNA was tested by 5 amplification assays including the Cobas Amplicor PCR (Roche Diagnostics) and 16S rRNA PCR (Roche Diagnostics) assays as previously described, the LCx ligase chain reaction assay (Abbott Diagnostics) as described by the manufacturer, and two real-time PCR assays performed on the LightCycler (Roche Diagnostics).  Both real-time assays targeted the N. gonorrhoeae cppB gene. In one assay, primers amplified a 273 bp fragment, which was detected using fluorescence resonance energy transfer (FRET) probes.  The presence of an amplification curve and a melting curve peak at 67^oC confirmed the presence of the N. gonorrhoeae cppB gene sequences in the sample.  The second real-time PCR assay contained primers that targeted a 90 bp fragment that was detected using a molecular beacon (MB) probe.  Samples that yielded linear increases in their fluorescence reading relative to the negative control sample were considered positive.  Detection of beta-globin gene sequences was performed in a separate real-time Sybr Green PCR reaction as a positive internal control.  Consensus positivity among the clinical samples was defined as any sample positive by at least two of the four methods used other than the Cobas Amplicor PCR.  

Main Outcome Measures
The real-time PCR assays using the LightCycler with the FRET and MB probes were compared to the three other methods for detection of N. gonorrhoeae, using the consensus results as the gold standard.  

Main Results
All 120 ATCC and clinical N. gonorrhoeae strains and none of the other bacterial stains tested were positive by the FRET and MB real-time PCR assays.  Each of the 172 clinical samples previously tested by the Cobas Amplicor assay were positive for the human beta-globin gene.  Among the 122 Cobas positive specimens, 73, 68, 71, and 72 were positive by the 16S, LCx, FRET, and MB assays, respectively.  Seventy-two (59.0%) of 122 Cobas positive and 1 (2%) of 50 Cobas negative specimens tested by the other 4 assays were positive by the consensus criteria.  The one consensus-positive sample among the 50 Cobas negative samples was also negative by the16S rRNA PCR.  The confirmation rate was proportional to the optical density of the Cobas PCR, with the highest confirmation of 93.5% obtained with samples having optical density readings >3.5.  Using the consensus results as the gold standard, the FRET and MB probe assays had a sensitivity of 99% and 100% respectively, and a specificity of 98% and 100%, respectively.  

Authors’ Conclusions
More than 40% of the Cobas Amplicor positive samples were not confirmed as positive by the consensus criteria and are presumed false positive results.  This high false positive rate for the Cobas Amplicor N. gonorrhoeae PCR assay highlights the importance of confirmation of all positive results by a supplemental assay.  

Source of funding:  Abbott Diagnostics for providing diagnostics kits.  Support from molecular microbiology research funds.

For correspondence:  Sepehr Tabrizi, Department of Molecular Microbiology, The Royal Women’s Hospital, 132 Grattan Street, Carlton, Victoria 3053, Australia.  E-mail address:  sepehr.tabrizi@wch.org.au.