Antimicrobial susceptibility of Neisseria gonorrhoeae in Barcelona during a five-year period, 2013 to 2017

Introduction Increasing rates of antimicrobial resistance in Neisseria gonorrhoeae cause problems for treating gonorrhoea. Aim This observational study aimed to describe isolates from all patients found infected with N. gonorrhoeae, in Barcelona, Spain, between 2013 and 2017, and with available antimicrobial susceptibility data. Methods Minimum inhibitory concentrations (MICs) of penicillin (PEN), cefixime (CFM), ceftriaxone (CRO), azithromycin (AZM), ciprofloxacin (CIP), spectinomycin (SPT), fosfomycin (FOF) and gentamicin (GEN) were determined by E-test. Susceptibility was assessed using clinical breakpoints from the European Committee on Antimicrobial Susceptibility Testing. Time trends for PEN, CFM, AZM and CIP were investigated using logistic regression. Results Of 1,979 patients with infection (2,036 isolates), 1,888 (95.4%) were men. Patient median age was 32 years. The proportions of isolates resistant to extended-spectrum cephalosporins were low, with 0.3% (5/1,982) resistant to CRO and 4.9% (98/1,985) to CFM. AZM resistance prevalence was 2.7% (52/1,981), including 16 isolates detected in 2016 and 2017, with high-level resistance. For CIP, 51.3% (1,018/1,986) of isolates were resistant, and for PEN, 20.1% (399/1,985). All isolates were susceptible to SPT. MIC50 and MIC90 values of GEN were 4 and 6 mg/L and of FOF 12 and 24 mg/L, respectively. Between 2013 and 2017, PEN and CFM resistance rates each decreased from 28.1% (92/327) to 12.2% (70/572) and from 8.3% (27/327) to 4.4% (25/572) (p ≤ 0.0073). In contrast, AZM resistance prevalence appeared to increase from 1.5% in 2014 (5/340) to 3.0% (17/572) in 2017. No trend was identified for CIP. Conclusion Antimicrobial susceptibility surveillance is important to timely detect new phenotypes and trends.


Introduction
Gonorrhoea is the second most reported bacterial sexually transmitted infection (STI), after chlamydia. Untreated gonorrhoea can lead to pelvic inflammatory disease and infertility in women and epididymitis and orchitis in men. In the European Union/European Economic Area countries, a total of 89,239 confirmed cases of gonorrhoea were reported in 2017 with an overall rate of 22.2 cases per 100,000 population [1]. This represented a 17% increase over the previous year, which was particularly striking in certain groups, such as men who have sex with men (MSM), who represented almost half of the cases in 2017, and the 25-34-year-old population [1].
In Spain, 8,722 cases of gonococcal infection were reported in 2017 (rate: 18.74 per 100,000 inhabitants), with a very wide range of infection incidence among the different regions of the country, from 2.44 and 48.50 cases per 100,000 inhabitants. The highest rates were registered in Catalonia (48.50), Balearic Islands (41.79) and Madrid (28.48) [2]. Of the 3,622 cases reported in Catalonia in 2017, men accounted for 82% of the diagnoses. Data from 1,136 patients (31%) could be collected: 44% were MSM followed by heterosexual men (22%) and women (20%) [3].
Neisseria gonorrhoeae, the bacterial species responsible for gonorrhoea, has developed resistance to the different families of antibiotics used in the past, challenging future treatment. Currently, extended-spectrum cephalosporins (ESC) are the last-line treatment option for N. gonorrhoeae infection. Unfortunately, strains showing resistance to ESC have been reported worldwide [4][5][6]. In 2017, the European Centre for Disease Control and Prevention (ECDC) reported a resistance rate to cefixime of 1.9% within the European Union/European Economic Area countries and no isolates showing resistance to ceftriaxone [7]. The resistance rate to azithromycin was 7.5%; and highlevel resistance (minimum inhibitory concentration (MIC) ≥ 256 mg/L) was detected in seven isolates [7]. This pathogen has moreover been able to acquire or develop nearly all known mechanisms of antimicrobial resistance: target modification, inactivation of the antimicrobial by enzymatic means, decreased influx of antimicrobials, and increased efflux of antimicrobials [8]. This situation highlights the importance of carrying out resistance monitoring programmes, in order to update therapeutic guidelines and to timely detect the emergence of multidrug-resistant strains.
Nowadays, most of the treatment guidelines recommend dual antimicrobial therapy (500 mg intramuscular ceftriaxone + 1 g or 2 g oral azithromycin) [9][10][11][12]. The rationale for gonococcal combination therapy using different antimicrobials with different mechanisms of action is to potentially mitigate the spread of antimicrobial resistance [13]. Combination therapy has been reported as possibly related to the decline in prevalence of N. gonorrhoeae isolates with decreased susceptibility to ceftriaxone (DSC), defined as a MIC > 0.032 mg/L [14,15].
The aim of this study is to describe antimicrobial susceptibility of N. gonorrhoeae in isolates collected between 2013 and 2017 in Barcelona, Spain.

Study population
The study population consisted of all patients who were diagnosed with a N. gonorrhoeae infection and with isolates for which antimicrobial susceptibility data were available. When patients tested positive for gonorrhoea on multiple sites and more than one culture was obtained, only the genital culture was included in the study.

Antimicrobial susceptibility
AST was performed on fresh colonies (<24 hours). The MICs of penicillin, ceftriaxone, cefixime, ciprofloxacin, azithromycin, spectinomycin, gentamicin, and fosfomycin were determined by means of the E-test method (bioMérieux, France), as described by the Clinical and Laboratory Standard Institute (CLSI) [16]. Interpretation was performed using clinical breakpoints from the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [17], except for gentamicin and fosfomycin, since cut-off points for these antibiotics are not established by EUCAST (nor by CLSI). N. gonorrhoeae ATCC 49226 was used as a reference strain for antimicrobial susceptibility testing.

Descriptive and statistical analysis
Descriptive analyses of the study population were performed. Statistical analyses were conducted using Stata (StataCorp, College Station, Texas, US). Trends of penicillin, cefixime, azithromycin and ciprofloxacin resistance over the study period were calculated using logistic regression analyses.
Determinants for resistance for penicillin (MIC > 1 mg/L), cefixime (MIC > 0.125 mg/L), azithromycin (epidemiological cut-off (ECOFF) > 1 mg/L) and ciprofloxacin (MIC > 0.06 mg/L) were identified using logistic regression analyses. Univariable and multivariable analyses were performed. Differences with p < 0.05 were considered statistically significant. As there were very few strains that reached the 0.125 mg/L threshold of ceftriaxone resistance, we did not determine associations between potential factors for ceftriaxone resistance and resistance to this antibiotic.  According to EUCAST, the MIC breakpoints for resistance to ceftriaxone and cefixime are identical and are MIC > 0.125 mg/L. The epidemiological cut-off for azithromycin is > 1mg/L.

Ethical statement
As this was a retrospective study, no ethical approval was needed. All data related to patients were coded to maintain confidentiality.

Results
Between 2013  Antimicrobial susceptibility data during the study period are shown in Table 2.
The distribution of MIC values by year for ceftriaxone, cefixime and azithromycin is shown in the Figure ( The average percentage of isolates with cefixime resistance (MIC > 0.125 mg/L) at 4.9% was higher than the average percentage with ceftriaxone resistance at 0.3% (  Table 2). The MIC 50 and MIC 90 values for gentamicin were 4 and 6 mg/L, while for fosfomycin these were 12 and 24 mg/L, respectively ( Table 2).

Discussion
This study describes antimicrobial surveillance data from 2,036 N. gonorrhoeae isolates in Barcelona during a 5-year period. To our knowledge, this is the first report in Spain describing antimicrobial resistance of such a considerable number of N. gonorrhoeae isolates.
Other studies have been published previously [18,19], but the amount of strains included was much lower. Our findings show that the rate of ceftriaxone resistance remains low and stable. These results agree with data from the ECDC [7]. From 2013 to 2015 the number of isolates with MIC ≤ 0.016 mg/L of ceftriaxone and cefixime seemed to increase progressively. However, in 2016 and 2017 we observed an important decrease in the percentage of strains with MIC ≤ 0.016 mg/L and an increase in strains with higher MICs, which concur with a report that isolates with DSC have been emerging in Europe in recent years [15].
In our work, we were able to observe a decrease in the resistance rate of both ESC throughout the study. The reason that this decrease is statistically significant only for cefixime may be due to the small number of ceftriaxone resistant strains.
Since dual therapy (ceftriaxone 500 mg + azithromycin 1 g) was implemented in Barcelona in 2012, resistance to ceftriaxone has started to decrease: 2.8% in 2012 (data not shown) to 0.0% in 2017. This could be explained by the change in treatment regimen whereby, until 2011, gonorrhoea was treated with ceftriaxone 250 mg. On the other hand, the resistance rate of azithromycin has increased slightly from 1.5% in 2014 to 3.0% in 2017 and high-level resistant isolates (MIC ≥ 256 mg/L) were first detected in 2016 and 2017. An increase in azithromycin rate has been documented in several European countries [15].
Although the European and Spanish guidelines still recommend dual therapy [12,20], the increase in azithromycin resistance detected worldwide, not only in N. gonorrhoeae but also in other STIs such as Mycoplasma genitalium [21], brings into question the advisability of this therapeutic strategy. In fact, since 2019, the British Association for Sexual Health and HIV (BASHH) has recommended monotherapy with ceftriaxone 1 g [22]. However, some studies support that the selection/induction of azithromycin resistance of N. gonorrhoeae by the use of the current dual therapy is limited [23], and it may be associated with the general use of azithromycin for respiratory infections or the treatment of non-gonococcal urethritis [13].
The overall azithromycin resistance rate in our study was 2.6% based on the EUCAST breakpoint (ECOFF is 1 mg/L). This percentage is lower than that reported by other Spanish groups such as Cobo et al. [18], who showed a resistance percentage of 13.8% in Almería.
On the other hand, Fuertes de Vega et al. reported a resistance rate of 5.2% in Barcelona [19], which is more similar to our results. We must bear in mind that the cut-off point used by these authors is that of prior EUCAST report versions (0.5 mg/L). If we analyse the azithromycin resistance rate of our study according to this cut-off, the percentage of resistance strains would be 4.1%.
The difference in the percentage of azithromycin resistance between different regions of Spain could be due to divergence in treatment patterns or the circulation of different genotypes with greater ability to develop antimicrobial resistance.
On other hand, the occurrence of extensively drugresistant (XDR) strains of N. gonorrhoeae reported in countries such as the United Kingdom and Australia in 2018 is concerning [24]. These strains show a high level of azithromycin resistance and they are also resistant to ceftriaxone, resulting in resistance to the first line dual therapy for gonorrhoea (ceftriaxone intramuscularly and azithromycin orally) recommended by European, Australian and World Health Organization guidelines. The appearance and dissemination of these type of strains compromises the successful treatment of this infection. This highlights the need to keep the capability to culture N. gonorrhoeae, in order to monitor antibiotic susceptibility and rapidly detect the emergence of XDR strains.
Although the resistance rate to ciprofloxacin observed in this study is similar to that of neighbouring countries [25], it is higher than that reported in eastern European countries such as Ukraine [26].
The breakpoints of gentamicin are not yet established by EUCAST. In this study it appears that N. gonorrhoeae does not have high MIC values for gentamicin (MIC 50 = 4mg/L and MIC 90 = 6mg/L). These values are similar to those reported previously and suggest that gentamicin can be a low-cost and efficacious alternative therapy [27]. In addition, other studies support the in vitro activity of gentamycin, used in combination with cefixime or ertapenem, to control the spread of multidrug-resistant (MDR) and XDR N. gonorrhoeae strains [28].
Similar to gentamicin, no clinical breakpoints for fosfomycin exist. The review conducted by Tesh et al. suggests that fosfomycin can be an alternative treatment for gonococcal infection [29]. In this review, isolates with a MIC < 16 mg/L were considered as susceptible, while those with MICs between 32 and 64 mg/L were moderately susceptible. These MICs are similar to those found by our study, but it is necessary to establish susceptibility breakpoints in order to understand how to appropriately dose fosfomycin to treat N. gonorrhoeae infections [29].
It is surprising that the resistance rates of almost all antimicrobials tested were higher in isolates from primary healthcare than in those from the STI unit. One reason that might explain this is that 70% of the Drassanes-Vall d'Hebron STI Unit patients are MSM. Although we do not have access to this information, we hypothesise by deduction that the majority of patients who attend primary healthcare units are men who have sex with women (MSW) and if so, this might suggest that there are different N. gonorrhoeae populations circulating in the two populations.
One limitation of this study is that antimicrobial susceptibility tests of N. gonorrhoeae were performed mainly in symptomatic patients, since they are the ones from whom a sample is collected for culture. Asymptomatic patients are diagnosed by nucleic acid amplification tests (NAAT). Another limitation is that when N. gonorrhoeae was isolated in multiple sites in one patient, antimicrobial susceptibility was only performed on genital samples. This could lead to the loss of strains with higher MIC values in pharyngeal carriers and consequently to underestimate antimicrobial resistance. Moreover, due to the retrospective nature of the study, there is a lack of demographic and epidemiological information.
Although our results are only representative of Barcelona, the patients included in the current study represent 62.1% of those reported in Catalonia [30].
In conclusion, we analysed the susceptibility of 2,036 isolates in Barcelona from 2013 to 2017. We were able to observe that susceptibility to ceftriaxone remains high and resistance has decreased since dual therapy was implemented in 2012. However, azithromycin resistance increased during the study period and highlevel azithromycin resistant strains were isolated for the first time in 2016.
Our study highlights the need to monitor antibiotic susceptibility and to perform molecular typing studies at a national level, which would allow identifying temporal and geographical changes, to detect the emergence and dissemination of new strains, and to maintain therapeutic guidelines updated.