Time to Develop Standardized Molecular Diagnostics for the Simultaneous Detection of Neisseria gonorrhoeae and Its Antimicrobial Resistance

This is the authors' abstract. We don't add key points for this paper.

A key factor in controlling the transmission of Neisseria gonorrhoeae is the identification of infected individuals coupled with their effective treatment.1 For effective treatment, knowledge of the prevalence of gonococcal susceptibility to recommended antimicrobials is essential, because N. gonorrhoeae has developed resistance to all classes of antibiotics.1,2 To do this, many countries have established national antimicrobial susceptibility (AMS) surveillance programs.2,3 The determination of AMS currently requires culturing of N. gonorrhoeae from clinical specimens.1 However, in resource-rich areas, culture is becoming ever-more rare because the diagnosis of gonococcal infections is mostly based on nucleic acid amplification testing (NAAT); for example, in Canada, more than 70% of gonococcal infections are detected using NAATs.4 Presently, there are no molecular AMS tests based on NAATs. In resource-limited areas, syndromic management of patients, coupled with the absence of laboratory infrastructure for the culture of gonococcal isolates, precludes up-to-date information on the AMS of N. gonorrhoeae.2 Both the shift from the use of cultures to NAATs for the diagnosis of gonorrhea and the absence of any AMS testing in resource-limited areas have resulted in underestimations of gonococcal antimicrobial resistance (AMR).2 Under these circumstances, the development, evaluation, and implementation of molecular tests to identify N. gonorrhoeae and its AMS from nonculture specimens could provide the information required to develop evidence-based treatment guidelines.5 A further complication of molecular approaches for gonococcal AMS testing is their effective application to an assortment of clinical specimens, including remnants of swabs used for culture, urines or other biological specimens, and remnant specimens from NAAT testing. The research by Melendez et al.,6 described in this issue, reported on using residual swabs, collected for the culture-based identification of N. gonorrhoeae, for the molecular identification of N. gonorrhoeae, and the molecular characterization of AMR in a given specimen. The study used published, real-time polymerase chain reaction (PCR) assays. Interestingly, the authors reported that N. gonorrhoeae DNA was detected only in 67.3% of N. gonorrhoeae–positive swabs using a porA/opa real-time PCR multiplex assay for gonococcal identification.7,8 This low detection rate was attributed to low concentrations of N. gonorrhoeae DNA in the remnant specimens, or to the degradation of the DNA during different storage conditions. There are many hurdles to develop reliable, uncomplicated, and rapid molecular tests for gonococcal AMR. One is the sheer number of genes implicated in AMR that could, and should, be investigated and then interpreted, to provide a complete and nuanced spectrum of the AMS of any gonococcal specimen. Various mutations in least 10 genes are associated with chromosomal resistance to β-lactam antibiotics, tetracycline, macrolides, fluoroquinolones, or aminocyclitols (spectinomycin). These genes include penA (encoding penicillin-binding protein 2; PBP2), mtrR (encoding the multiple transfer resistance repressor; MtrR), porB (encoding the outer-membrane protein PorB), ponA (encoding the penicillin-binding protein 1; PBP1), rspJ (encoding the ribosomal protein S10 in 30S rRNA), gyrA (encoding DNA gyrase A), parC (encoding topoisomerase IV), rpsE (encoding the ribosomal protein S5), 16s rRNA, 23S rRNA, and genes implicated in plasmid-mediated resistance to penicillin and tetracyline.9,10 Melendez et al.6 evaluated 4 gonococcal AMR determinants using real-time PCR including a single gene (penA) implicated in resistance to extended-spectrum cephalosporins (ESCs), 2 genes implicated in resistance to fluoroquinolones, and DNA sequences identifying plasmid-mediated resistance to penicillin.9,11–15 Mutations in gyrA and parC have been validated in several studies to predict the susceptibility of N. gonorrhoeae to fluoroquinolones.16–19 Quinolone resistance in N. gonorrhoeae is caused by single-nucleotide polymorphisms (SNPs) in DNA gyrase A (S91 and D95) and topoisomerase IV (S88 and E91), the 2 genes evaluated in the study by Melendez et al.6 A multitarget PCR test, detecting penicillinase-producing N. gonorrhoeae from nonculture specimens, which is routinely used for the molecular surveillance of penicillin resistance in Western Australia and the Northern Territory, Australia, was also used by Melendez et al.20 A number of additional assays could have been used to detect and interpret possible resistance to other antibiotics, including SNPs to detect resistance to penicillin (porB—PorB at G120; A121 and ponA—PBP1 at L421) and chromosomal and plasmid-mediated resistance to tetracycline (ribosomal protein S10 [V57] or tetM).9,10 Single-nucleotide polymorphisms in the MtrR promoter (adenine deletion [A−] or dinucleotide insertion [TT+]), its DNA binding domain (A39T, G45D) and multimeric region (H105Y), play a role in conferring gonococcal resistance to macrolides, penicillin, ESCs, tetracycline, and floroquinolones.9,10,21 PBP2 and MtrR can carry multiple SNPs, which together can confer different levels of resistance to penicillin and ESCs.22,23 Mutations in PBP2 (penA) can range from an insertion at position 345a (designated as nonmosaic, pattern I) to more than 60 mutations (e.g., mosaic patterns, pattern X, XXXIV, etc).24 Mosaic penA patterns have been implicated in resistance to ESCs, but nonmosaic mutations in penA have been also characterized as mediating decreased susceptibility to ESCs.9,10,24 Thus, it is a challenge to develop diagnostic PCR assays for ascertaining gonococcal AMS profiles involving multiple determinants (e.g., penA, mtrR, porB, and ponA for β-lactam antimicrobials). The cost associated with multiple PCR tests for the determination of N. gonorrhoeae AMR is high but might be relieved through real-time multiplex PCR assays, which would target multiple genes.25 A limitation of current PCR assays is that they use a variety of primers for the same gene that produce amplicons of varying sizes and that different methodologies that have not been standardized or validated in different settings are used. For example, 10 different sets of primers have been used to amplify penA, a major determinant in gonococcal resistance to β-lactam antibiotics, yielding PCR products of 501 to 2039 base pairs.26–33 Because of these variations, penA resistance determinant data are difficult to compare across different laboratories. Furthermore, most of these published assays are based on the evaluation of cultured specimens, and not noncultured specimens in which the amount and quality of the DNA available for testing are variable. Melendez et al.6 suggested that the study should be replicated with “urine NAAT specimens,” the most commonly used specimen for the NAAT diagnosis of N. gonorrhoeae infections. The issue with using such specimens is standardization for easy and rapid clinical diagnosis. The low detection of gonococcal DNA in the remnant specimens indicates the need to develop and evaluate methods for extracting DNA from different clinical specimens including urine NAATs. Various commercial kits have been used to extract DNA from urine NAATs for determining AMR in N. gonorrhoeae, and a comparative analysis should be undertaken to ascertain best methods.20,34 Individual gene-based PCR assays for molecular testing of AMR limit the amount of information that could be obtained on gonococcal AMS to known mutations in a limited number of genes. A whole genome sequencing (WGS) approach to processing clinical specimens would provide comprehensive information on all genes implicated in gonococcal AMR and extend the possibility of discovering novel mutations implicated in AMR.5,35 The WGS approach is currently being used to assess AMR in N. gonorrhoeae using DNA primarily obtained from cultures; purifying DNA for WGS is challenging when nonculture specimens, such as Aptima urines, are used.5 Several reports describe the limitation of Aptima collection buffer when used for WGS analysis of N. gonorrhoeae or Chlamydia trachomatis.36,37 In addition, such NAAT specimens may contain large concentrations of human or other bacterial DNA, which would interfere with gonococcal WGS. Other methods have been evaluated to purify bacterial DNA from various noncultured specimens for WGS analysis, but the adaptation of such methods to clinical laboratories would be expensive and time consuming.38 Thus, the development of a universal methodology for the efficient extraction of bacterial DNA from various clinical, non–culture-based specimens, to predict AMS based on WGS, would be a significant technological advancement. N. gonorrhoeae Sequence Typing for Antimicrobial Resistance, a sequence-typing tool based on 7 genetic determinants in N. gonorrhoeae, is available to characterize and type AMR in N. gonorrhoeae isolates.39 This tool is ideally suited for characterizing gonococcal AMR profiles obtained through WGS. Furthermore, to predict AMS based on molecular data, large, accessible databases, which link minimum inhibitory concentrations with mutation patterns in various genes for the same isolate, would be important for the development of machine learning algorithms that might be applied to successfully predict AMS. In conclusion, Melendez et al.6 have highlighted both the importance and limitations of molecular approaches for AMR detection in N. gonorrhoeae. Gonococcal characterization was limited by several factors, including low concentrations or low quality of DNA, the limited number of AMR determinants characterized, and the use of multiple PCR tests. Because the selection of PCR tests for specific resistance determinants was limited, the results underestimate the prevalence of resistance. Even so, critical information on the prevalence of resistance in a situation where absolutely no information might have been obtained was provided. More studies need to be undertaken using assays targeting all N. gonorrhoeae AMR determinants and linking minimum inhibitory concentrations to genetic determinants of prediction.