Objectives To use nucleic acid amplification techniques (NAAT) for detection of markers associated with gonococcal antimicrobial resistance (AMR) in non-cultured clinical samples to enhance surveillance of Neisseria gonorrhoeae AMR in New Zealand.
Methods A total of 198 clinical samples from patients living in two cities, Wellington and Auckland and the more rural region of Gisborne, New Zealand, which were positive for N. gonorrhoeae by the Cobas 4800 were tested for three markers that predict reduced susceptibility or resistance to three antibiotics. Residual DNA extracts from the Cobas 4800 NG/CT test were tested for a single-nucleotide polymorphism in the gyrA gene at codon 91 associated with quinolone resistance; a sequence on the plasmid in penicillinase-producing N. gonorrhoeae (PPNG) which confers resistance to penicillin and the mosaic penA sequence associated with reduced susceptibility to extended-spectrum cephalosporins in N. gonorrhoeae.
Results A total of 186/198 (94%) of the samples provided a valid result on gyrA genotyping, confirming the utility of N. gonorrhoeae DNA extracted by the Roche Cobas 4800 CT/NG test for subsequent detection of AMR markers. The NAAT results for Wellington, Auckland and Gisborne, respectively, showed that 77%, 33% and 32% of samples had the marker associated with quinolone resistance, while 4%, 15% and 0% were positive for the PPNG plasmid marker, and 9%, 5% and 0% samples were positive for mosaic penA sequence.
Conclusions The use of residual clinical DNA samples from the Cobas 4800 CT/NG test proved an efficient and effective method for performing AMR genotyping. These data also show for the first time the presence of gonococci with a mosaic penA sequence in New Zealand. Overall, the results further highlight the potential of molecular methods to aid N. gonorrhoeae AMR surveillance, particularly for regions where gonococcal culture is no longer performed.
- ANTIBIOTIC RESISTANCE
- NEISSERIA GONORRHOEA
Statistics from Altmetric.com
Neisseria gonorrhoeae has progressively developed resistance to all empiric treatments.1 Extended-spectrum cephalosporins (ESCs) are now considered the last bastion of empiric treatment, and no new drugs for the treatment of gonorrhoea to which no resistance has been reported are currently available for clinical use.2 Alarmingly, the evidence of ESC-reduced susceptibility (and even resistance), to the ESC ceftriaxone, has recently emerged.3 With more frequent international travel, geographical isolation is no longer an advantage that countries such as New Zealand can rely on to limit importation of novel resistant strains.
In recent years, diagnosis of gonorrhoea in New Zealand has moved away from predominantly culture-based methods to nucleic acid amplification techniques (NAAT).4 Unfortunately, this switch to NAAT-based diagnosis has decreased the availability of viable organisms for culture-based antimicrobial susceptibility testing (AST) which is of concern, given increasing reports of drug-resistant gonorrhoea.5 The potential for molecular methods to aid gonococcal AST has recently been reviewed and considered feasible.5 We conducted a preliminary examination of three genotyping assays to detect markers associated with gonococcal antimicrobial resistance (AMR), with the additional aim to better understand the distribution of these mutations in New Zealand.
Three real-time PCR methods were used and targeted: a single-nucleotide polymorphism in the gyrA gene (gyrA-PCR)6 associated with quinolone resistance; a sequence on the plasmid in penicillinase-producing N. gonorrhoeae (PPNG) which confers resistance to penicillin (PPNG-PCR)7 and the mosaic penA sequence associated with reduced susceptibility to cephalosporins (mosaic-PCR).2 All assays were based on previously described methods but were adapted to the LightCycler 480 (V.1.0, Roche) real-time PCR platform (details of primers, probes and cycling conditions are available in online supplementary material).
The three assays were initially applied to DNA extracted from 10 previously characterised clinical isolates and four N. gonorrhoeae strains (New Zealand Environmental and Science Research Limited (ESR) culture collection). All 10 control isolates provided genotypic results that correlated with those expected based on phenotypic data (data not shown). Representative strains were then used as controls for the clinical sample testing.
The methods were then applied to ‘residual DNA’ from de-indentified specimens testing positive for N. gonorrhoeae (n=198) on the Roche Cobas 4800 CT/NG test, which was performed as per the manufacturers’ protocol with modifications as previously described.8 The N. gonorrhoeae-positive samples comprised vaginal (n=26), cervical (n=52), throat (n=7), rectal (n=4), urethral/penile swabs (n=18), urine specimens (n=83) and urogenital or unspecified sites (n=8). Samples were from three regions; the rural area of Gisborne (n=32; 19 women, 13 men; collected between 24 September 2012 and 16 December 2012) and the urban regions of Auckland (n=88; 54 men, 33 women, one unknown sex; collected between 4 March 2013 and 17 April 2013) and Wellington (n=78; 33 men, 44 women, one unknown sex; collected between April 2013 and July 2013). All specimens were obtained from community-based healthcare providers such as general practice, sexual health clinics and family planning clinics.
All N. gonorrhoeae-positive samples from one Gisborne laboratory sent for validation of an in-house NAAT were included in this study as well as all NAAT N. gonorrhoeae-positive samples received by one Auckland and one Wellington laboratory during the specified periods. Patient duplicates were removed. All non-genital samples had previously been confirmed as N. gonorrhoeae positive by supplementary PCR assays for the porA pseudogene and opa gene.9 Overall, the residual DNA from the Cobas 4800 CT/NG test proved suitable for use in the AMR assays, with 94% (186) of samples successfully amplified by the gyrA-PCR indicating acceptable sample DNA quality (two Auckland samples, nine Wellington samples and one Gisborne sample failed to generate gyrA melt curves). The fact that 6% of the samples failed to amplify reflects the lower sensitivity of AMR PCRs compared with the commercial PCR for detection of N. gonorrhoeae and is consistent with previous reports.5 ,6 Nevertheless, these negative results were not viewed as a limitation; rather we consider 94% of positive samples to provide excellent coverage for surveillance purposes.
Of the 186 samples, a total of 92 (49%) carried the gyrA ser91phe mutation. When tested by the PPNG-PCR and mosaic-PCR methods, 16 (9%) and 10 (5%) samples, respectively, provided positive results (table 1). A total of 24 samples (13%) were positive for two resistance markers; 14 (8%) carried both the PPNG and the gyrA ser91phe markers, and 10 (5%) carried both the gyrA ser91phe and the mosaic penA markers. Regional differences in the prevalence of resistant genotypes and a comparison with phenotypic-based data from ESR10 are shown in table 1.
p Values were calculated using the http://www.graphpad.com online tool. Fisher's exact test and two-tailed p values were used to assess statistical significance.
The results of this testing showed that the three markers of reduced susceptibility or resistance were not uniformly distributed throughout the North Island of New Zealand. There were significantly more samples with the ciprofloxacin mutation in the Wellington region (77%) than in Auckland (33%) (p<0.0001) or Gisborne (32%) (p<0.0001). This could be linked to greater mobility of the population in larger cities or to different patterns of treatment. The data also suggest a significant increase in the percentage of isolates with resistance to ciprofloxacin in the Wellington region over time (p<0.0002). The 2011 ESR data showed 50.9% ciprofloxacin resistance10 compared with 77% reduced susceptibility predicted by genotype. However, the genotypic data may be an overestimation as it has been shown that isolates harbouring only the gyrA mutation ser91phe may have low to intermediate resistance to ciprofloxacin.1 Conversely, samples from Auckland showed a statistically significant decrease in ciprofloxacin resistance from 43.5% to 33% (as predicted by detection of the gyrA marker) between the 2011 ESR data and the results from this study (p=0.0447). This discrepancy could be explained by the absence of samples from sexual health clinics in the Auckland group or resistance by mechanisms not detected in this study. Further investigations are required. Unlike bacterial culture, the PPNG-PCR is unable to detect chromosomally mediated resistance to penicillin, and therefore it was not unexpected that the rates of penicillin resistance detected by PCR were lower than that of culture. Despite this limitation, a lower proportion of penicillin resistance in Wellington compared with Auckland was evident using both methods and is further suggestive of differences in predominant gonococcal strains circulating in these two cities. The fact that no PPNG strains were detected in Gisborne may suggest that these populations are more isolated in terms of sexual networks; however, these data need to be interpreted with caution as the numbers of samples were low.
The mosaic penA sequence is the key alteration of interest in reduced susceptibility to ESCs3 although other mutations are likely to explain the wide range of ceftriaxone minimum inhibitory concentrations (MICs) seen in gonococci positive for the mosaic penA gene2 Our data for the first time confirm that mosaic penA-harbouring gonococci are present in New Zealand. This is consistent with reports of N. gonorrhoeae isolates with reduced susceptibility to ceftriaxone in the Auckland region in recent years.10 The 0% resistance shown in table 1 is due to the fact that Clinical and Laboratory Standards Institute criteria are widely used in New Zealand with a breakpoint for ceftriaxone of >0.25 mg/L, whereas reduced susceptibility strains typically have MICs of 0.03–0.125 mg/L. This study shows that the penA marker was present in samples from equal numbers of men and women. Although relatively low in numbers (10/186; 5%), this shows that the spread of the mosaic penA-reduced susceptibility marker has not been limited to men who have sex with men. The limitations of the study include the relatively small number of samples from three laboratories examined by the three molecular methods. Other novel and/or untested mutations may influence the phenotype as has been shown for isolates positive for the gyrA ser91phe mutation and the mosaic penA gene which can vary in their resistance to ciprofloxacin6 and ceftriaxone,2 respectively. In addition to the penicillinase-producing plasmids, chromosomal mutations can affect resistance to penicillin.1 These factors may have biased the comparison with the larger number of samples tested 1–2 years earlier included in the ESR data.
As testing for N. gonorrhoeae in New Zealand has moved from culture-based methods to NAAT, there has been a reduction in antimicrobial susceptibility tests conducted, and thus surveillance. This preliminary work has shown that genotyping residual DNA samples from the Cobas 4800 CT/NG test may prove an effective means of enhancing gonococcal AMR surveillance. It is hoped that ultimately this approach may help slow the incursion and subsequent spread of antibiotic-resistant strains through improved surveillance and appropriate use of antibiotic therapies.
We wish to thank Nadika Liyanarachchy and staff of the molecular biology department, Aotea Pathology, Wellington, for their support. Acknowledgement is also due to Dr Arlo Upton and Liselle Bisessor, Labtests Auckland, and Dr Sam Chan, MedLab Central, for provision of samples.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Handling editor Jackie A Cassell
Contributors MN designed the study, performed the testing and data collection and wrote the paper. CB and MN were supervisors, advised on data collection and analysis and assisted with writing the paper. DW analysed discrepant data, advised on the data collection and analysis and assisted with writing the paper.
Funding This study was supported by Aotea Pathology Ltd, Massey University and Roche Diagnostics NZ Ltd.
Competing interests Roche funding.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.