The European Centre for Disease Prevention and Control (ECDC) has been coordinating the epidemiological surveillance of sexually transmitted infections (STIs) in the European Union and the European Economic Area (EU/EEA) since 2009. After chlamydial infection, gonorrhea is the second most common bacterial STI reported and cases have increased by 31% since 2008 to 39,179 cases reported from 28 EU/EEA Member States in 2011.1 The ECDC also coordinates the European Gonococcal Antimicrobial Surveillance Programme (Euro-GASP), which performs antimicrobial resistance (AMR) surveillance twice each year using a dual model of centralized and decentralized susceptibility testing.2 This surveillance of AMR in Neisseria gonorrhoeae is essential due to the extraordinary ability of the organism to develop resistance to antimicrobials,3 so appropriate treatments can be recommended using reliable and robust laboratory data. It is well established that first-line therapies should be changed once resistance levels of 5% have been detected in the population, so 95% therapeutic success is maintained.4
In 2011, Euro-GASP identified a high proportion of gonococcal isolates (7.6%) that had resistance to cefixime, which is a significant increase from the 5.1% of isolates detected in 2009.2 Ten isolates displayed resistance to ceftriaxone in 2011, 7 of which also had resistance to cefixime. All isolates with resistance to cefixime or ceftriaxone were additionally resistant to ciprofloxacin. Rates of ciprofloxacin and azithromycin resistance have both continued to decrease since 2009, but still remain high (48.7% and 5.3%, respectively). Because of the emergence of resistance and treatment failures to the last remaining options for first-line empiric monotherapy, that is, cefixime and ceftriaxone,3,5 in 2012 ECDC published a response plan to control and manage the threat of multidrug-resistant N. gonorrhoeae in Europe to maintain gonorrhea as a treatable infection.6 However, no thorough data regarding risk factors associated with antimicrobial-resistant N. gonorrhoeae across Europe have been available.
This study used epidemiological and behavioral data collected as part of Euro-GASP combined with the gonococcal susceptibility profiles to determine risk factors associated with harboring resistant gonococci in Europe. The representativeness of isolates tested in Euro-GASP was also explored.
MATERIALS AND METHODS
Euro-GASP has been previously described in detail.2 Briefly, in 2009, 17 countries (Austria, Belgium, Denmark, France, Germany, Greece, Italy, Latvia, Malta, the Netherlands, Norway, Portugal, Slovakia, Slovenia, Spain, Sweden, and the UK) participated in Euro-GASP with the inclusion of 4 additional countries in 2010 and 2011 (Cyprus, Hungary, Ireland, and Romania). Laboratories submitted consecutive isolates for centralized susceptibility testing or, if laboratories fulfilled the quality assurance criteria,2 partook in decentralized testing. To represent different episodes of infection laboratories were requested to collect only 1 isolate per patient from those who were infected multiple times within a 4-week period or at multiple sites; pharyngeal followed by rectal, then urethral specimens were preferred from males, and pharyngeal followed by cervical then other anogenital specimens from females. For centralized testing, isolates were tested in 1 of 3 locations (Public Health England, London; Statens Serum Institut, Copenhagen or Örebro University Hospital, Örebro) by Etest or agar dilution for cefixime, ceftriaxone, ciprofloxacin, azithromycin, spectinomycin, and gentamicin. For decentralized testing, laboratories performed their own antimicrobial susceptibility testing.
All countries participating in Euro-GASP report the results through The European Surveillance System (TESSy), a Web-based reporting system managed by the ECDC. Besides susceptibility testing data, countries also separately report epidemiological surveillance data on gonorrhea cases through TESSy on an annual basis. These case-based or aggregated data are analyzed and presented in an annual ECDC report.1
For the purpose of the analysis, patient data linked to the gonococcal isolates’ susceptibility profiles were extracted from TESSy. The data set included age, sex, site of infection, previous gonorrhea, concurrent STI, sexual orientation, clinical service type, and year of isolate collection. Clinical service type was collected from the second collection period in 2010. In addition to the Euro-GASP patient data, gonorrhea epidemiological surveillance data were extracted covering the same period. The data extracted included age, sex, and sexual orientation.
Data and Statistical Analysis
The representativeness of Euro-GASP data when compared with the epidemiological surveillance data was assessed by comparing the distributions of age, sex, and sexual orientation in the 2 surveillance systems.1 Overall percentages were calculated from known values only. The significance of the difference between the proportions of the collected Euro-GASP and epidemiological surveillance variables was established using the Z test. A probability value less than 0.05 was used to define significance. Details on which countries report which variables are available elsewhere (see Table, Supplemental Digital Content, http://links.lww.com/OLQ/A89, which displays the number of variables reported from each participating Euro-GASP country).2,7,8
The risk factors associated with resistant isolates were assessed by analyzing Euro-GASP data. Categorical patient variables used to identify risk groups were as follows: age (<25 and ≥25 years), previous gonorrhea (yes and no), concurrent STI (chlamydial infection and no chlamydial infection), sexual orientation and sex (men who have sex with men [MSM], male heterosexuals, and all women), year of isolate collection (2009, 2010, and 2011), anatomical site of infection (genital, pharyngeal, and anorectal), and clinical service type (STI and sexual health clinics [including dermatology-venereology and youth clinics], outpatient clinics [including hospital emergency, gynecology, infectious disease, and urology clinics], and primary care [including general practitioners]).
Geometric means for ceftriaxone and cefixime minimum inhibitory concentrations (MICs) were calculated for all isolates over the 3 years (2009–2011) and compared for patients of different sexual orientation and sex. Linear regression was used adjusting for years to study the relationship between the geometric mean of cefixime and ceftriaxone MICs over time, sexual orientation and sex, and anatomical site of infection. For establishing the geometric mean any cefixime or ceftriaxone MICs with “less than” values were converted to “equal to” values; for example, all cefixime MIC values of 0.016 or less (n = 1 764) were categorized as 0.016 and all ceftriaxone MIC values of 0.002 or less (n = 639) were categorized as 0.002.
Data from the 3 years, 2009 to 2011, were combined, and patient variables associated with resistance to cefixime (MIC > 0.125 mg/L), ciprofloxacin resistance (MIC > 0.5 mg/L), and azithromycin resistance (MIC > 0.5 mg/L) were identified using univariate and multivariable logistic regression analyses of odds ratios (ORs). Observations missing data from the variables used in the univariate or multivariable model were not included in the analysis (see Table, Supplemental Digital Content, http://links.lww.com/OLQ/A89, which displays the number of variables reported from each participating Euro-GASP country). The OR and 95% confidence intervals (CIs) were calculated for the univariate analysis, and a Pearson χ2 test was used to test if these ORs were significantly different from 1.0. A P value less than 0.05 was used to choose whether a variable should be included in the multivariable logistic regression to model the odds of resistance controlling for other variables. Using a forward stepwise approach, the most significant and strongest associations from the univariate analysis were added to the multivariable model sequentially. Further analysis using gentamicin, spectinomycin, and ceftriaxone was not performed with patient variables because of the absence of resistance to gentamicin (no defined breakpoints) and spectinomycin (MIC > 64 mg/L) and low numbers of isolates with resistance to ceftriaxone (10 isolates only in 2011). High-level penicillin resistance (penicillinase) data were not included in any analysis as chromosomal penicillin resistance was not tested.
Anatomical site of infection and clinical service type were analyzed separately in a univariate model only as they are clinic dependent variables rather than patient variables (i.e., clinics may have different strategies for which patient sites are sampled). Statistical analysis was performed in Stata v12 (StataCorp LP, College Station, TX).
Ethical approval for the study was not required because all examined gonococcal isolates were cultured and stored as part of routine diagnostics (standard care). Patient data were reported as part of a surveillance program (EU Decisions 2119/98/EC and 1082/2013/EU) in an anonymized manner with no patient-identifiable information.
Comparison of STI Epidemiological Surveillance Data and Euro-GASP Patient Characteristics
A total of 5034 gonococcal isolates, representing different episodes of infection, were tested in Euro-GASP from 2009 to 2011 (see Table, Supplemental Digital Content, http://links.lww.com/OLQ/A89, which displays the number of variables reported from each participating Euro-GASP country). Completeness of patient variables was highest for sex (97.7%) and lowest for previous gonorrhea (38.2%; Table 1). A comparison of the epidemiological surveillance data and Euro-GASP patient characteristics (Table 2) showed that the proportions of the reported patient characteristics of age and sexual orientation between the 2 surveillance systems were significantly different (P < 0.002), except for the proportion of MSM transmission (P = 0.0656).
Cefixime and Ceftriaxone Resistance Trends
The geometric mean MICs decreased significantly for cefixime and ceftriaxone (P < 0.001) from 2009 to 2011 (Table 3). When the data from 3 years were combined (2009–2011), isolates from infection in male heterosexuals had significantly higher geometric mean MICs of cefixime (P < 0.001) than isolates from infection in MSM and women (Table 3). However, for ceftriaxone, gonococcal isolates from male heterosexuals had a significantly higher geometric mean MIC (P < 0.001) than those from women only (Table 3). In the most recent year, 2011, isolates from male heterosexuals had higher geometric means for both cefixime and ceftriaxone MICs than those from MSM (P < 0.001) and women (cefixime: P = 0.014, ceftriaxone: P = 0.025; Table 3).
Looking at all the years combined, there was no significant difference between the geometric mean MICs of ceftriaxone when anatomical site of infection was compared; however, the geometric mean MICs for cefixime were lowest in isolates from anorectal specimens when compared with those from genital specimens (P = 0.031). When controlling for year, pharyngeal specimens had a higher ceftriaxone geometric mean MIC than genital specimens (P = 0.031), but there was no significant difference for cefixime.
Risk Factors for Antimicrobial Resistant Gonorrhea
In the univariate analysis, year of isolation, male heterosexuals, women, having no concurrent chlamydial infection, previous gonorrhea infection, or those older than 25 years were associated with infection with isolates showing resistance to cefixime (Table 4). In the multivariable analysis, the association with isolates with resistance to cefixime remained significant in male heterosexuals and women (male heterosexuals: adjusted OR [aOR], 2.39; 95% CI, 1.58–3.61; P < 0.001; women: aOR, 2.75; 95% CI, 1.68–4.5; P < 0.001), those older than 25 years (aOR, 2.07; 95% CI, 1.36–3.13; P = 0.001), and patients with no concurrent chlamydial infection (aOR, 1.87; 95% CI, 1.1–3.16; P = 0.021).
For gonococcal isolates that were ciprofloxacin resistant, there was a significant association in the univariate analysis with infection in male heterosexuals, individuals with no concurrent chlamydial infection, being older than 25 years, or those with no history of gonorrhea, whereas year of isolation (with 2009 as the reference) was negatively associated with ciprofloxacin resistance (Table 4). In the multivariable analysis, a significant association remained with infection with ciprofloxacin-resistant gonococci with those patients who were male heterosexual (aOR, 1.47; 95% CI, 1.2–1.81; P < 0.001), older than 25 years (aOR, 1.7; 95% CI, 1.39–2.08; P < 0.001), or did not have concurrent chlamydial infection (aOR, 2.14; 95% CI, 1.68–2.71; P < 0.001). Year of isolation remained negatively associated with ciprofloxacin resistance (2010: aOR, 0.73; 95% CI, 0.57–0.92; P = 0.01; 2011: aOR, 0.6; 95% CI, 0.47–0.76; P < 0.001).
Male heterosexuals, no concurrent chlamydial infection, or previous gonorrhea (Table 4) were significantly associated with infection with azithromycin-resistant gonorrhea in the univariate analysis, whereas year of isolation was negatively associated with azithromycin resistance. Only no concurrent chlamydial infection (aOR, 2.18; 95% CI, 1.28–3.71; P = 0.004) and year of isolation (2010: aOR, 0.68; 95% CI, 0.46–1; P = 0.01; 2011: aOR, 0.44; 95% CI, 0.29–0.67; P < 0.001) remained significant in the multivariable model.
Isolation from the pharynx or anorectum was not significantly associated with azithromycin, cefixime, or ciprofloxacin resistance (Table 4) using univariate logistic regression. The only significant association was a negative association between isolation from the anorectum and with cefixime resistance and ciprofloxacin resistance (Table 4).
Most of the patients reported through Euro-GASP attended an STI clinic (74%; 1523/2052), with a smaller number attending primary care (18%; 368/2052) or an outpatient clinic (8%; 161/2052). Isolates from patients attending outpatient clinics were more likely to be associated with ciprofloxacin resistance and resistance to cefixime (Table 4).
Linked gonococcal antimicrobial susceptibility and patient data from Euro-GASP have shown that isolates exhibiting resistance to cefixime and ciprofloxacin were significantly associated with infection in heterosexuals (males only for ciprofloxacin), older patients (>25 years of age), or those without a concurrent chlamydial infection. These findings, along with the geometric mean of the cefixime MICs being highest in isolates from male heterosexuals, suggest that the burden of resistance has been highest among heterosexuals. This situation requires careful monitoring and recent treatment failures in heterosexuals show that this group is at risk for acquiring MDR gonorrhea.9–11
Gonococcal isolates with resistance to cefixime were shown to be circulating in both heterosexual and MSM networks across Europe,2,12 whereas in the United Kingdom13 and in the Netherlands (for cefotaxime),14 the burden was more focused in the MSM community. A study in Shanghai revealed an association between higher ceftriaxone MICs and men, and patients older than 45 years.15 For ciprofloxacin, men and those older than 30 years were associated with resistance in Ontario, with heterosexuals appearing to significantly contribute to the ciprofloxacin resistance.16 In The Netherlands, MSM and women older than 35 years17 were associated with ciprofloxacin resistance, whereas MSM association with ciprofloxacin resistance was clinic dependent in California.18 It is difficult to compare these studies that used multivariable analyses to identify risk factors as the study populations, variables, time, and location vary greatly. However, the overall differences most probably reflect the different sexual transmission networks in different regions.
The reduction in the prevalence of isolates with resistance to cefixime2 and the fall in the geometric mean MICs of cefixime and ceftriaxone over the years have occurred before the publication of the response plans6,19 and the updated European treatment guideline, which now recommends 500 mg ceftriaxone intramuscularly combined with 2 g of azithromycin orally.20 The decrease in the geometric mean MIC should, however, be viewed with caution because this decrease runs in parallel to a bimodal MIC distribution of a “more susceptible” and “less susceptible” gonococcal population, which seems to be emerging (in heterosexuals and patients >25 years of age).2 The decreasing resistance levels in MSM and the development of bimodal populations for cefixime and ceftriaxone MICs have also been seen in other surveillance studies such as the Gonococcal Resistance to Antimicrobials Surveillance Programme, which is the national surveillance program for England and Wales.13,21 The greater susceptibility of anorectal isolates to cefixime and ciprofloxacin demonstrated in this study additionally supports the lower resistance levels in MSM. The decreasing resistance in isolates from MSM could be due to the use of more sensitive molecular tests (in several countries), particularly from extragenital sites, which has improved the detection of infection22 and subsequently allowed the administration of more appropriate therapies (i.e., ceftriaxone),13 which in turn can disrupt the dissemination of strains such as those that display resistance to cefixime.
Pharyngeal gonococcal infection is considered to be a potential reservoir for resistant isolates,3 but the present data show that isolates from the pharynx were not more likely to display azithromycin, cefixime, or ciprofloxacin resistance when compared with genital specimens. The geometric mean MICs of cefixime were not significantly higher in isolates from the pharynx, when compared with isolates from the genital and anorectal sites; however, pharyngeal specimens did have a higher ceftriaxone geometric mean MIC than genital specimens. It is postulated that cephalosporin resistance emerged in the pharynx through the acquisition of DNA in N. gonorrhoeae from commensal Neisseria species, which could have been selected for by the lower efficacy of cephalosporins in the pharynx,23,24 and any emerging resistance in this anatomical site should be monitored carefully.
It is difficult to establish if patients from the various health settings are sufficiently represented in Euro-GASP because the completeness of the reporting of clinic service setting data is low in the STI epidemiological surveillance data1 and many different diagnostic settings are available across Europe.1,25 However, most patients in Euro-GASP attended an STI clinic or sexual health service (74%), so the specialist STI health care settings were represented by Euro-GASP. The association of ciprofloxacin- and cefixime-resistant isolates from patients attending outpatients clinics is most probably due to most patients attending outpatient clinics (88/161) being from one country, Slovakia, which has had some of the highest levels of ciprofloxacin resistance (71%–91%) and resistance to cefixime (15%–36%) in 2010 and 2011 (when clinical service setting was reported).2,7
There are some limitations of this study. Because of differences in diagnostic and clinical procedures, not all EU/EEA countries participate in the Euro-GASP and the numbers of available isolates from some countries are low. Missing patient data and the different proportions of reported patient characteristics between the 2011 ECDC STI epidemiological surveillance and the Euro-GASP data, and between different countries, may bias the results. Male heterosexuals are overrepresented in the Euro-GASP data set, possibly due to undeclared or misclassified homosexuality or due to the increased sensitivity of culture in symptomatic men who may present more frequently to clinics. Underreporting of patient data in Euro-GASP can be due to ethical or juridical restrictions around linking patient and isolate data in some countries, and it is difficult to obtain evidence of sexual orientation misclassification or to adjust for it, so working toward more complete reporting is a priority for Euro-GASP. The underrepresentation of patients younger than 25 years in Euro-GASP may be due to chlamydia screening programs targeting this group and, subsequently, diagnosing dual N. gonorrhoeae infection by molecular tests instead of by culture.26
The differences in the number of countries participating, isolates, and reporting of patient variables suggest that weighting of the data might be needed to provide better estimates of resistance levels.27 However, the high level of heterogeneity in the European STI surveillance systems, case definitions,1,25 and laboratory diagnostics,28 along with the heavy influence of data submitted from the United Kingdom (59% of all EU/EEA gonorrhea cases in 2011),1 emphasizes that the European STI surveillance system needs to be more complete and comprehensive to underpin weighting for Euro-GASP. A more complete European picture from both surveillance systems would allow us to determine resistance levels across the EU/EEA region more accurately to then inform the European treatment guidelines. Euro-GASP performs high quality AMR surveillance and, although incomplete, does offer a snapshot at gonococcal-resistant profiles in the EU/EEA. The average number of 1678 isolates per year (over the years 2009–2011) from a “region” provides adequate evidence that can be used for the review of treatment guidelines.29
The linking of epidemiological and behavioral data to the susceptibility profiles of gonococcal isolates allows those at higher risk for acquiring AMR N. gonorrhoeae to be identified. Euro-GASP needs improved data numbers and representativeness before evidence-based risk groups can be identified and subsequent focused treatments or public health intervention strategies can be initiated with confidence.
3. Unemo M, Nicholas RA. Emergence of multidrug-resistant, extensively drug-resistant and untreatable gonorrhea. Future Microbiol 2012; 7: 1401–1422.
5. Van de Laar M, Spiteri G. Increasing trends of gonorrhoea and syphilis and the threat of drug-resistant gonorrhoea in Europe. Euro Surveill 2012; 17.
9. Ison CA, Hussey J, Sankar KN, et al. Gonorrhoea treatment failures to cefixime and azithromycin in England, 2010. Eur Surveill 2011; 16.
10. Unemo M, Golparian D, Hestner A. Ceftriaxone treatment failure of pharyngeal gonorrhoea verified by international recommendations, Sweden, July 2010. Euro Surveill 2011; 16.
11. Unemo M, Golparian D, Syversen G, et al. Two cases of verified clinical failures using internationally recommended first-line cefixime for gonorrhoea treatment, Norway, 2010. Euro Surveill 2010; 15.
12. Chisholm SA, Unemo M, Quaye N, et al. Molecular epidemiological typing within the European Gonococcal Antimicrobial Resistance Surveillance Programme reveals predominance of a multidrug-resistant clone. Euro Surveill 2013; 18.
13. 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.
14. Heymans R, Bruisten SM, Golparian D, et al. Clonally related Neisseria gonorrhoeae
isolates with decreased susceptibility to the extended-spectrum cephalosporin cefotaxime in Amsterdam, the Netherlands. Antimicrob Agents Chemother 2012; 56: 1516–1522.
15. Trecker MA, Waldner C, Jolly A, et al. Behavioral and socioeconomic risk factors associated with probable resistance to ceftriaxone and resistance to penicillin and tetracycline in Neisseria gonorrhoeae
in Shanghai. PLoS One 2014; 9: e89458.
16. Ota KV, Jamieson F, Fisman DN, et al. Prevalence of and risk factors for quinolone-resistant Neisseria gonorrhoeae
infection in Ontario. CMAJ 2009; 180: 287–290.
17. Koedijk FD, van Veen MG, de Neeling AJ, et al. Increasing trend in gonococcal resistance to ciprofloxacin in the Netherlands, 2006–8. Sex Transm Infect 2010; 86: 41–45.
18. Bauer HM, Mark KE, Samuel M, et al. Prevalence of and associated risk factors for fluoroquinolone-resistant Neisseria gonorrhoeae
in California, 2000–2003. Clin Infect Dis 2005; 41: 795–803.
19. Ndowa F, Lusti-Narasimhan M, Unemo M. The serious threat of multidrug-resistant and untreatable gonorrhoea: The pressing need for global action to control the spread of antimicrobial resistance, and mitigate the impact on sexual and reproductive health. Sex Transm Infect 2012; 88: 317–318.
20. Bignell C, Unemo M. On behalf of the European STI Guidelines Editorial Board 2012 European guideline on the diagnosis and treatment of gonorrhoea in adults. Int J STD AIDS 2013; 24: 85–92.
23. Chisholm SA, Mouton JW, Lewis DA, et al. Cephalosporin MIC creep among gonococci: Time for a pharmacodynamic rethink? J Antimicrob Chemother 2010; 65: 2141–2148.
24. Moran JS. Treating uncomplicated Neisseria gonorrhoeae
infections: Is the anatomic site of infection important? Sex Transm Dis 1995; 22: 39–47.
25. Lowndes CM, Fenton KA. Surveillance systems for STIs in the European Union: Facing a changing epidemiology. Sex Transm Infect 2004; 80: 264–271.
26. Skidmore S, Copley S, Cordwell D, et al. Positive nucleic acid amplification tests for Neisseria gonorrhoeae
in young people tested as part of the National Chlamydia Screening Programme. Int J STD AIDS 2011; 22: 398–399.
27. Hughes G, Nichols T, Ison CA. Estimating the prevalence of gonococcal resistance to antimicrobials in England and Wales. Sex Transm Infect 2011; 87: 526–531.
Supplemental Digital Content
© Copyright 2014 American Sexually Transmitted Diseases Association