Treatment of gonorrhea has followed the same pattern since the 1930s: a first-line therapy is recommended for empirical treatment and when the prevalence of resistance in Neisseria gonorrhoeae to this therapy increases the recommendation is changed to an antimicrobial agent with no reported resistance in the patient population. The World Health Organization (WHO) recommends all first-line therapies for gonorrhea should successfully treat 95% of infections1,2 after reviews of available antimicrobials and reported clinical efficacy.3,4 As resistance to successive antimicrobials used for the treatment of gonorrhea has increased, such as penicillin and ciprofloxacin, a number of changes to guidelines have been made.
Antimicrobial surveillance programs are used globally to inform when treatment guidelines should change by monitoring the prevalence of resistance. There are various models in place, collecting differing amounts of microbiological and clinical data. In England and Wales, the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) is a sentinel surveillance program and has been central to decisions regarding changes to the first-line gonorrhea treatment in the United Kingdom, with a single treatment regimen advocated for all patient groups.5,6 In 2004, the United States, using data from the Gonococcal Isolate Surveillance Programme, took a different approach by creating separate recommendations for men who have sex with men (MSM) and heterosexual patients. It was recommended that ciprofloxacin be used for heterosexual patients only, due to the high proportion of isolates from the MSM population resistant to fluoroquinolones.7
The most recently used therapies have been the extended-spectrum cephalosporins, cefixime (oral), and ceftriaxone (injectable). However, isolates with decreased susceptibility to cefixime have rapidly developed in Japan, with reports of multidrug-resistant strains of N. gonorrhoeae emerging8–10 such that ceftriaxone has become the first-line drug of choice in many countries. Since this change in England and Wales, there has been a continued decline in isolates exhibiting decreased susceptibility to cefixime.11 More recently, decreased susceptibility to ceftriaxone has been reported in England as well as other countries,12,13 and so there is concern that if ceftriaxone-resistant strains become widespread, there are few options remaining for future treatment. Also, the delivery method of ceftriaxone may be a potential barrier to treatment, as it requires a deep intramuscular injection, which can be painful and is refused by some patients. There is a possibility that in some situations such as treatment in the general practice setting or in some countries, ceftriaxone is not readily available. New antimicrobials are in development, but these are several years away from clinical use even if they prove to be effective.14 Another option could be to adapt treatment guidelines to target use of previous first-line therapies to patients either on a basis of (a) being epidemiologically likely to have a susceptible isolate or (b) being known to have susceptible isolates either after susceptibility testing in the laboratory of by the use of a point-of-care susceptibility test. In this study, we explore the feasibility of the former approach by seeking to identify specific patient subgroups where 95% or more of isolates are susceptible to previously used antimicrobials and are independently associated with antimicrobial susceptibility. This may enable targeted therapy, reduce onward transmission of the infection, and prolong the use of current first-line treatments for other patients where necessary.
Data from GRASP covering 2007 to 2011 were used in this study. This surveillance has been described in detail elsewhere: briefly, over a 3-month period each year, isolates from consecutive patients with gonorrhea attending 1 of 26 participating genitourinary medicine (GUM) clinics are referred to the Sexually Transmitted Bacteria Reference Unit.15,16 Susceptibility testing results are then matched with patient demographic, behavioral, and clinical data submitted by GUM clinics. Isolates were defined as susceptible to penicillin if they exhibited a minimum inhibitory concentration (MIC) less than 1 mg/L and being β-lactamase negative, to ciprofloxacin based on an MIC less than 1 mg/L, and to cefixime based on an MIC less than 0.125 mg/L.
The proportions of isolates susceptible to each antimicrobial were reviewed for each patient group defined in terms of age, ethnicity (as defined by the UK Office for National Statistics), number of sexual partners in the past 3 months, sex abroad in the past 3 months, symptoms present, previous gonorrhea, concurrent chlamydia infection, and HIV status. Subgroups were defined as having a high proportion of isolates susceptible to previous first-line therapies with 95% as the primary reference point used as per WHO recommendations.1
Patient characteristics associated with isolates susceptible to penicillin, ciprofloxacin, and cefixime were assessed using univariate and multivariable analyses of odds ratios. Three models were created for each antimicrobial under investigation. In the univariate analysis, variables with a P value less than 0.05 from a Pearson χ2 test or, where appropriate, the Fisher exact test, were selected to take forward to the multivariable analysis. For the multivariable analysis, logistic regression models were built using a forward selection procedure to model the odds of antimicrobial susceptibility. Only records with susceptibility testing for all antimicrobial agents and demographic data were included. Year of isolation was also assessed and controlled for in the multivariable models.
Separate models were created for patients identifying as heterosexual (males and females grouped) and male patients who self-identified as homosexual or bisexual (MSM), due to differing prevalence and distribution of antimicrobial susceptibility in these groups. The GRASP data set contained few female patients that identified as homosexual (<1%), so no separate analysis was performed for this group.
Stata 12.0 (StataCorp LP, College Station, TX) was used for all statistical analysis.
A total of 6173 gonococcal isolates were tested for antimicrobial susceptibility as part of GRASP between 2007 and 2011. More than half (3415; 55%) of isolates were from heterosexuals and 2326 (38%) from MSM. The remaining 431 samples did not have sexual orientation information recorded so were excluded from this analysis. Overall 4684 (82%) isolates were susceptible to penicillin, 3899 (68%) to ciprofloxacin, and 5240 (91%) to cefixime.
Analysis of Isolates from MSM Patients
Fewer than 95% of isolates from all MSM patient subgroups were susceptible to penicillin, ciprofloxacin, or cefixime (Table 1). The univariate analyses identified characteristics in this population associated with isolates susceptible to penicillin or ciprofloxacin, but no patient characteristics were associated with infection with cefixime-susceptible isolates (Table 2). A multivariable model was constructed for infection with penicillin-susceptible isolates; however, after controlling for year of isolation, the significant associations identified in the univariate analyses did not remain. In addition, after controlling for year in the ciprofloxacin model, the association between patient with symptoms and penicillin-susceptible isolates did not remain. Isolates susceptible to penicillin were more likely to occur in later years compared with the first year of this study, which was the opposite found for cefixime (Table 2).
Analysis of Isolates From Heterosexual Patients
At least 95% of isolates from all heterosexual subgroups were susceptible to cefixime, whereas fewer than 95% of isolates from all these subgroups were susceptible to penicillin or ciprofloxacin (Table 3). Results from the univariate analysis showing the association between N. gonorrhoeae isolates susceptible to previous first-line therapies and patient variables in heterosexuals are presented in Table 4. Patient variables remaining significantly associated with infection with susceptible isolates in the multivariable model after controlling for year of isolation were (Table 5): patients aged 13 to 24 years compared with patients 25 years or older, patients of black ethnicity compared with patients of white ethnicity, and patients with concurrent chlamydia. Patients who self-identified as an ethnicity other than white, black, or Asian and patients who did not report sex abroad were more likely to be infected with a penicillin- or ciprofloxacin-susceptible isolate of N. gonorrhoeae. In contrast to the analyses of isolates infecting MSM, patients with symptoms were less likely to be infected with penicillin-, ciprofloxacin-, or cefixime-susceptible N. gonorrhoeae.
In this study, we have explored the feasibility of using N. gonorrhoeae antimicrobial susceptibility data linked to patient demographic and behavioral data to identify patient subgroups in England and Wales who could be successfully treated by previous first-line therapies. If treatment guidelines continue to follow the WHO-recommended 95% treatment threshold, then none of the previous first-line therapies could be recommended as treatment for any MSM subgroup diagnosed as having gonorrhea in GUM clinics. In addition, penicillin and ciprofloxacin would not be suitable therapies to use for any subgroup of heterosexual patients diagnosed in GUM clinics. However, cefixime could be used for heterosexual patients attending for sexual health care regardless of any other patient characteristic, as the prevalence of infection with cefixime susceptible isolates was greater than 95% in all patient subgroups.
The heterogeneity of the prevalence of infection with antimicrobial susceptible N. gonorrhoeae across patient subgroups is likely to be due to patterns of sexual mixing. Although the prevalence of penicillin- and ciprofloxacin-susceptible gonococcal infection in heterosexual subgroups was less than the recommended 95% treatment threshold, 3 patient subgroups (younger age, black ethnicity, and concurrent chlamydial infection) were strongly associated with infection with these susceptible isolates and the proportion of isolates susceptible was higher than 85%. This is the cautionary treatment threshold set by the WHO.2 Currently, the efficacy of using this lower threshold to control gonorrhea is unknown and use is likely to only be considered if all other alternatives for the treatment of gonorrhea have been exhausted.
There still remain logistical issues if targeted treatment guidelines were to be implemented, and questions remain about the longevity of this approach. The US attempted targeted treatment guidelines based on sexual orientation in 2004 when fluoroquinolones, such as ciprofloxacin, were recommended only for heterosexual patients.7 This guideline was a temporary solution, as the recommendations changed again in 2007 counseling against fluoroquinolones as treatment for any patient, due to increases in fluoroquinolone resistance among isolates from heterosexual males.17 In the United Kingdom, it is likely that this targeted approach would also be a temporary solution if used. A similar increase in resistance to previous first-line therapies could emerge in groups that have been identified as associated with susceptibility, therefore negating any targeted recommendation that may evolve from this research. In addition, there is some concern that continued use of cefixime may result in accelerated development of ceftriaxone-decreased susceptible isolates, which would have the unintended effect of reducing the available gonorrhea treatments further.18
There would be practical issues to consider when asking busy sexual health clinicians to target treatment based on the patient’s sexual orientation. Nevertheless, computer prompts and patient management pathways are already used in UK sexual health clinics to flag and remind clinicians to offer specific tests depending on patient information (such as sexual orientation which is routinely collected by clinicians), so these could be adapted to recommend specific treatments. Adding further complexity to treatment guidelines may work within sexual health clinics, as there is evidence that compliance with recommended treatments is high in this setting19; however, complex treatment guidelines may be more difficult to implement in other settings.
A final issue to consider both in terms of whether this approach could be implemented, and a limitation of this study, is geographic variation in prevalence and distribution of gonococcal antimicrobial susceptibility. This study used data from a sentinel network that covers only 26 of the 208 GUM clinics in England and 2 GUM clinics in Wales. Sexual health clinic case mix and sexual mixing patterns vary considerably across the country and could lead to localized clusters of patients with infection, with isolates exhibiting different susceptibility patterns, and these could be missed by a surveillance program such as GRASP. Further studies of the approach here could aim to collect this level of data at different locations, in addition to addressing logistical issues associated with this new approach, as described above. There has already been one small study in a metropolitan sexual health clinic to identify population subgroups that could be treated with antimicrobials other than cephalosporins, which had similar results to those presented here.20 If this targeted approach were to be implemented either nationally or at a sexual health clinic level, close monitoring of antimicrobial resistance in N. gonorrhoeae would be essential to ensure the guidelines remain fit for purpose. This would require laboratories local to sexual health clinics to increase the provision of precise antimicrobial susceptibility testing of gonococcal isolates, and sexual health clinics might need more administration and epidemiological support to collect and use this information effectively. This would increase the resources required, and a potential cost-benefit analysis study would be necessary to assess if this is a worthwhile use of funding.
Other future studies could investigate whether the 95% treatment threshold should be reduced to allow for other antimicrobials to be used more frequently for treatment. The evidence base for the 95% treatment threshold is obscure and originated at a time when there were more antimicrobials that met this criterion than are currently available.21 It would be important to assess the impact of different test of cure procedures, as lowering the treatment threshold would have implications for the number of patients successfully treated. The implications of unsuccessful treatment and the potential associated increase in the development of sequelae such as pelvic inflammatory disease in women are also an important aspect to consider. The model could also assess the impact of bridging between sexual networks, which could increase the prevalence of resistant infection in different patient subgroups. The addition of antimicrobial susceptibility point-of-care tests in the future could also help improve the targeting of therapies.22
Any future guideline changes should consider how to treat partners. It is conceivable that infected partners will have a similar antimicrobial susceptibility profile as the index patient; therefore, treatment with the same antimicrobial may be appropriate. However, tracking the treatment for partners would take additional resources and sexual partner networks can be complicated, meaning that this assumption may not always be correct. Future studies to investigate patterns of antimicrobial susceptibility within known sexual networks could provide evidence for or against this assumption.
This study demonstrated that of the previous first-line therapies investigated, cefixime would be the only antimicrobial suitable for use for infection in heterosexual patients alone. However, further work to establish the practicality, longevity, and advisability of this approach is required before any changes are made to gonorrhea treatment guidelines in the United Kingdom.
1. Tapsall JW. Antimicrobial Resistance in Neisseria gonorrhoeae
. 2001. Geneva: World Health Organisation, 2013: 2–10.
2. World Health Organisation. Guidelines for the Management of Sexually Transmitted Diseases. Geneva: World Health Organisation, 2003.
3. Bignell C. Antibiotic treatment of gonorrhoea—Clinical evidence for choice. Genitourin Med 1996; 72: 315–320.
4. Moran JS, Levine WC. Drugs of choice for the treatment of uncomplicated gonococcal infections. Clin Infect Dis 1995; 20 (suppl 1): S47–S65.
5. Bignell C, Fitzgerald M. UK national guideline for the management of gonorrhoea in adults, 2011. Int J STD AIDS 2011; 22: 541–547.
6. Bignell CJ. BASHH guideline for gonorrhoea. Sex Transm Infect 2004; 80: 330–331.
7. Centres for Disease Control and Prevention. Increases in fluoroquinolone-resistant Neisseria gonorrhoea
among men who have sex with men—United States, 2003, and revised recommendations for gonorrhea treatment, 2004. Morb Mortal Wkly Rep 2004; 53: 335–338.
8. Ito M, Yasuda M, Yokoi S, et al. Remarkable increase in central Japan in 2001–2002 of Neisseria gonorrhoeae
isolates with decreased susceptibility to penicillin, tetracycline, oral cephalosporins, and fluoroquinolones. Antimicrob Agents Chemother 2004; 48: 3185–3187.
9. Lahra M. Surveillance of antibiotic resistance in Nesseria gonorrhoeae
in the WHO Western Pacific and South East Asian Regions, 2010. 2013; 36: 2–10.
10. Unemo M, Nicholas RA. Emergence of multidrug-resistant, extensively drug-resistant and untreatable gonorrhea. Future Microbiol 2012; 7: 1401–1422.
11. Ison CA, Town K, Obi C, et al. Decreased susceptibility to cephalosporins among gonococci: Data from the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) in England and Wales, 2007–2011. Lancet Infect Dis 2013; 13: 762–768.
12. La RG, Goubard A, Bercot B, et al. Gonococcal infections and emergence of gonococcal decreased susceptibility to cephalosporins in France, 2001 to 2012. Euro Surveill 2014; 19: pii: 20885.
13. Public Health England. GRASP 2012 report: The Gonoccocal Resistance to Antimicrobials Surveillance Programme. 2013. Public Health England.
14. Ison CA, Deal C, Unemo M Current and future treatment options for gonorrhoea. Sex Transm Infect 2013; 89 suppl 4: iv52–iv56.
15. Chisholm SA, Alexander S, Desouza-Thomas L, et al. Emergence of a Neisseria gonorrhoeae
clone showing decreased susceptibility to cefixime in England and Wales. J Antimicrob Chemother 2011; 66: 2509–2512.
16. Delpech V, Martin IM, Hughes G, et al. Epidemiology and clinical presentation of gonorrhoea in England and Wales: Findings from the Gonococcal Resistance to Antimicrobials Surveillance Programme 2001–2006. Sex Transm Infect 2009; 85: 317–321.
17. Centres for Disease Control and Prevention. Update to CDC’s sexually transmitted diseases treatment guidelines, 2006: Fluoroquinolones no longer recommended for treatment of gonococcal infections. Morb Mortal Wkly Rep 2007; 56: 332–336.
18. Centers for Disease Control and Prevention. Update to CDC’s sexually transmitted diseases treatment guidelines, 2010: Oral cephalosporins no longer recommended treatment for gonococcal infections. MMWR Morb Rep Wkyl Rep 2012; 61; 590–594. 10-8.
19. Health Protection Agency. GRASP 2011: The Gonococcal Resistance to Antimicrobials Surveillance Programme. 2012. Health Protection Agency.
20. Karunaratne J, Phillips DR. Antibiotic choice for Neisseria gonorrhoeae
in the absence of culture results: Can demographic-risk group analysis be used to guide treatment? Int J STD AIDS 2011; 2: 59.
21. Newman LM, Moran JS, Workowski KA. Update on the management of gonorrhea in adults in the United States. Clin Infect Dis 2007; 44 (suppl 3): S84–101.
22. Sadiq ST, Dave J, Butcher PD. Point-of-care antibiotic susceptibility testing for gonorrhoea: Improving therapeutic options and sparing the use of cephalosporins. Sex Transm Infect 2010; 86: 445–446.