• Neisseria gonorrhoeae
• antibiotic resistance
• point-of-care technology
• penicillin
• fluoroquinolone
• antibiotic sensitivy
• diagnosis
• gonorrhoea

Antimicrobial therapy for Neiserria gonorrhoeae is a major public health concern with high rates of resistance to penicillin, tetracycline and fluoroquinolones detected in England and Wales in the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) (GRASP 2008 Report: Trends in Antimicrobial Resistant Gonorrhoea. Health Protection agency. http://www.hpa.org.uk/Publications/InfectiousDiseases/HIVAndSTIs/0906GRASP2008/ last accessed 26 July 2010). Currently, third-generation oral cephalosporins, such as cefixime or injectable ceftriaxone, are recommended as first-line therapy. However, low-level increases in minimum inhibitory concentrations (MIC) for cephalosporins are now widely reported,1 and reports of azithromycin resistance2 further highlight reduced therapeutic options. Recommendations to limit spread of cephalosporin resistance may well include combination and even prolonged therapy, but these undermine clinical desirability for a single observable dose.

However, there is still a significant local variation in resistance rates, and most circulating gonococcal strains continue to be susceptible to fluoroqunolones and penicillin in most settings. With the emergence of rapid diagnostic technologies for STIs and other infections, we ask whether point-of-care (PoC) antimicrobial susceptibility testing might soon be ready to exploit this aspect of gonococcal epidemiology. The answer will depend on: (1) current knowledge of the molecular basis of gonococcal resistance; (2) availability of molecular diagnostic platforms supporting detection of multiple molecular resistance markers; and (3) validation of these assays.

Recent progress in understanding mechanisms of gonococcal resistance has been significant. Known chromosomal mutations associated with penicillin resistance include those altering the structure and function of penicillin-binding proteins (PBPs) and efflux pump activity genes, including penA, ponA, penB, mtrR regulator and possibly penC.3–5 The other classic resistance mechanism involves TEM-1 type penicillinase production, encoded by the plasmid mediated bla_TEM-1 gene. Tetracycline resistance may involve acquisition of the tet(M) gene,6 or a combination of mtrR, penB and rpsJ gene mutations,7 whereas fluoroquinolone resistance is mediated by mutations in gyrA8 9 and parC. Molecular mechanisms that have given rise to cephalosporin resistance include multiple mutations in PBP2, acquired most often through mosaic gene exchange with other Neisserial species, as well as single-point mutations.

A comprehensive validation analysis of multiple resistance markers for penicillin, tetracycline and fluoroquinolones including penA, penB, mtrR and bla_TEM1, rpsJ, tet(M), gyrA and parC, provides insight into their value as a diagnostic test for resistance and susceptibility.10 Resistance rates to antimicrobials (with defined MICs) in 464 gonococcal isolates were: penicillin, 30% (1 mg/ml); fluoroquinolone, 45% (1 mg/ml) and tetracycline, 38% (2 mg/ml). The presence of specific resistant-gene combinations predicted phenotypic resistance in 40%, 90% and 54% of isolates, respectively. However, in isolates where resistance-associated genes were completely absent (‘susceptible genotypes’) susceptibility to penicillin (MICs <0.5 mg/l), fluoroquinolones (MIC <0.25 mg/l) and tetracycline (MIC <2 mg/l) was 100% (95% CI 94% to 100%), 100% (CI 98% to 100%) and 98% (CI 94% to 100%), respectively. When more stringent MIC values for low-level resistance were used (0.12 mg/l, 0.12 mg/l and 0.5 mg/l) these predictive values for susceptibility fell to 90% (CI 80% to 96%), 99% (CI 96% to 100%) and 83% (CI 74% to 89%). The rates of penicillin, ciprofloxacin and tetracycline susceptible genotypes in this study were 11.6%, 32.2% and 19.6%, respectively.

Clearly more work is required to establish the diverse molecular mechanisms of resistance, particularly for cephalosporins and for second-line drugs such as azithromycin, where mechanisms may be diverse and complex.11 However, these data allow us to speculate on possible benefits of PoC antibiotic susceptibility tests. For example, of 1276 men and women with gonorrhoea in GRASP, June to August 2008, 28% had ciprofloxacin-resistant strains. Using a susceptibility genotype prevalence of 32.2% as in the study above (which had a higher phenotypic resistance rate than that in London), PoC susceptibility testing based on absence of markers might have identified 410 patients treatable by ciprofloxacin. Significant numbers of individuals with penicillin susceptible gonorrhoea may also have been identified, depending on the degree of overlapping resistance. Immediate, directly observable fluoroquinolone and penicillin therapy could then become available as a therapeutic option for significant numbers of patients, alongside other strategies for managing cephalosporin resistance such as using combination therapy.

It is unclear whether PoC susceptibility testing would contain spread of cephalosporin resistance, as the main drivers such as community prescribing of antimicrobials and importation of resistant strains from resource-poor settings, are likely to be unaffected. In addition, PoC susceptibility tests would need to become available before resistance becomes more widespread, particularly as increasing cephalosporin use is likely to result in increased chromosomally mediated resistance to tetracyclines and penicillins12 and because fluoroquinolone resistance may persist, possibly because of fitness advantages of resistance strains. This requires industry, academia and regulators to collaborate closely to bring innovative diagnostics to market, and it is encouraging that this may soon be the case for a number of novel, affordable microengineered rapid STI tests following recent investment to support such collaborations (http://www.innovateuk.org/ourstrategy/innovationplatforms/detectionandidentificationofinfectiousagents.ashx; last accessed July 2010). These models of support could be used to accelerate the development of PoC nucleic acid amplification test (NAAT)-based susceptibility tests, which have been made possible by advances in microengineering.13

Laboratory-based techniques for detecting molecular resistance markers such as spectrometric methods, real-time PCR,14 sequencing and microarrays may be useful as gonococcal culture is replaced by NAATs for diagnosis. However, the turnaround times for these methods (hours rather than minutes) is still too long for providing immediate treatment to patients with gonorrhoea. Furthermore, centralised tests will be less cost-effective in settings of low gonococcal prevalence. PoC susceptibility tests will have to be accurate and flexible, so that markers for new resistance mechanisms can be rapidly added as they are discovered and validated. This will require collaboration with ongoing national surveillance and research programmes. Performance of tests will be challenged by low levels of DNA from genital and non-genital samples, although one real-time PCR assay for the gyrA resistant genotype in first void female urine samples was able to give a result in 70% of cases.14 Nanotechnology platforms have the potential for single molecule detection15 and may thus overcome the poor sensitivity of the assay. Ultimately, however, molecular-based testing, whether centralised or PoC, will not be able to detect emergence of new resistance strains and should not be seen as an alternative to strong phenotypic antimicrobial resistance surveillance programs. These will need to remain in place to monitor for the ongoing threat of resistance.