Background

Neisseria gonorrhoeae, a Gram-negative bacterium, is responsible for the sexually transmitted disease gonorrhea. In 2008, over 106 million new cases of N. gonorrhoeae were reported worldwide [1]. In China, infectious disease surveillance reports more than 100,000 new gonorrhea cases each year. In 2012, there were 95,263 gonorrhea cases reported in China with an incidence rate of 7.07 per 100,000 population. Gonorrhea ranked sixth in the most frequently reported infectious diseases in China [2]. There are currently no vaccines available for N. gonorrhoeae [3]. Antimicrobial treatment is essential to control this infectious disease. However, antimicrobial resistance (AMR) in N. gonorrhoeae is emerging as a significant public health problem worldwide [4],[5]. In most areas of the world, N. gonorrhoeae has developed resistance to sulfonamides, penicillins, tetracyclines, macrolides and fluoroquinolones [6]. Decreased susceptibility, sporadic resistant isolates and clinical treatment failure to oral or parental third-generation cephalosporins have also been observed [7]-[10]. Moreover, extensive/multiple-drug-resistant (MDR) N. gonorrhoeae isolates are commonly observed in various areas of the world [8],[10].

In China, the burden of gonococcal penicillin (PEN) and tetracycline (TET) resistance reached high levels [11]-[13]. In response, fluoroquinolones were introduced as a treatment option. Since 1995, orally administered ciprofloxacin (CIP) has been recommended for the treatment of gonorrhea in China. However, as a result, the proportion of CIP resistance drastically increased from 15% in 1996 to nearly 100% in the mid-2000s in China [11]-[13]. The use of CIP for the treatment of gonorrhea has since been terminated in China [14],[15]. Since 2007, ceftriaxone (CRO) has been recommended as the first line drug for the treatment of gonorrhea in China [14]. However, in Shanghai CRO has been used since 2000, and the presence and spread of gonococcal isolates with nonsusceptibility to CRO (minimal inhibitory concentration (MIC) = 0.125–0.25 mg/L) was reported in 11.9% of isolates in 2005 and 19.7% of isolates in this city in 2008 [11],[12].

In this study, we used the data of the Shanghai Gonococcal Antimicrobial Susceptibility Surveillance Program from N. gonorrhoeae isolates collected over the past 25 years to analyze trends in AMR development in Shanghai. The AMR data were aligned with the timeline of treatment guidelines for gonorrhea in China. We found that N. gonorrhoeae isolates in Shanghai were resistant to PEN, TET and CIP, which were historically used for the treatment of gonorrhea in this region. Plasmid-mediated N. gonorrhoeae antimicrobial resistance has been prevalent in the past decades. High proportions of N. gonorrhoeae isolates are nonsusceptible to the currently recommended treatment regime of CRO (MIC ≥ 0.125). In Shanghai, MDR associated with the CRO nonsusceptible phenotype was prevalent in N. gonorrhoeae isolates. Our report indicates the usefulness of regional surveillance data in the modification of treatment guidelines and provides insights that will be important in future strategic planning for the treatment of gonorrhea.

Methods

Collection of N. gonorrhoeaeisolates

The Shanghai Gonococcal Antimicrobial Susceptibility Surveillance Program was established in 1988 and has been led by the Shanghai Skin Disease Hospital (SSDH), Shanghai, China. The number of gonorrhea cases reported by the SSDH was approximately 14–16% of the reported cases from the entire Shanghai city over the years [16]. The number of gonococcal isolates tested for antimicrobial susceptibility accounted for about 10% of the SSDH reported cases or 2.1% of reported cases in Shanghai in 2012. All of the surveyed N. gonorrhoeae isolates were from the Sexually Transmitted Infection Clinic of the SSDH.

Each year in the SSDH, approximately 100–200 N. gonorrhoeae isolates were collected from individual patients for antimicrobial susceptibility testing. The first 10–20 N. gonorrhoeae isolates each month were collected from patients with culture positive gonorrhea [12]. During the period from 1988 to 2000, N. gonorrhoeae isolates were collected from both male and female patients, and during the period from 2001 to 2013, N. gonorrhoeae isolates were only collected from male patients. N. gonorrhoeae was isolated and cultured as previously described [12].

Gonococcal isolates were collected as part of standard patient care, and ethics approval was not required for their use.

Determination of the minimal inhibitory concentrations (MIC)

The antimicrobial susceptibility of N. gonorrhoeae isolates was determined using the agar dilution method [17] at the Laboratory Diagnostic Centre of the SSDH. WHO reference strains A, B, C, D and E were used in susceptibility tests [12]. Five antimicrobial agents were tested against N. gonorrhoeae isolates: PEN (0.002–32.0 mg/L), TET (0.002–32.0 mg/L), CIP (0.002–32.0 mg/L), spectinomycin (SPT; 2.0–128.0 mg/L) and CRO (0.002–2.0 mg/L). Antimicrobial agents were provided by the China GASP at the National Center for Sexually Transmitted Disease Control, China (1988–2002) or purchased from Sigma-Aldrich (Distributor: the Shanghai ANPEL Scientific Instrument Co. Ltd., Shanghai, China). The MIC testing was performed on GC agar base supplemented with 1% IsoVitalex as previously described [12]. Each MIC determination was performed in duplicate. Gonococcal isolates with inconsistent CRO MICs between the duplicates were re-tested.

Determination of plasmid-mediated resistance in N. gonorrhoeaeisolates

Beta-lactamase producing N. gonorrhoeae (PPNG) isolates were determined using the chromogenic cephalosporin test before 2003 and the nitrocefin test in 2003–2013 (Oxoid; distributed by GuangZhou LOSO Science Ltd.) [12]. All N. gonorrhoeae isolates were tested for β-lactamase production. Plasmid-mediated tetracycline-resistant N. gonorrhoeae (TRNG) isolates were defined as those having a TET MIC of ≥ 16.0 mg/L. PP/TRNG isolates were defined as those isolates that were β-lactamase positive and had a TET MIC of ≥ 16 mg/L. Therefore, the characterizations for presumed PPNG and TRNG were solely based on the detection of β-lactamase and the TET MIC, respectively.

Criteria for resistant phenotypes of N. gonorrhoeae

Breakpoints for the susceptibilities to PEN, TET, CIP and SPT were those described by the Clinical Laboratory Standards Institute (CLSI) [17]. The classifications of resistance phenotypes included PEN resistance (MIC ≥ 2.0 mg/L), TET resistance (MIC ≥ 2.0 mg/L), CIP resistance (MIC ≥ 1.0 mg/L) and SPT resistance (MIC ≥ 128.0 mg/L). Chromosomally-mediated resistant N. gonorrhoeae (CMRNG) isolates exhibited a PEN MIC of ≥ 2.0 mg/L, a TET MIC of ≥ 2.0 mg/L and were not classified as PPNG or TRNG. For CRO susceptibility, the European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint was followed (nonsusceptible, MIC ≥ 0.125 mg/L) [18]. In this study, the susceptibilities to CRO in N. gonorrhoeae were further categorized as susceptible (MIC ≤ 0.03 mg/L), reduced susceptibility (MIC = 0.06) and nonsusceptible (MIC ≥ 0.125 mg/L) [11]. The MDR phenotypes of N. gonorrhoeae were defined as various combinations of the above resistant phenotypes (Table 1).

Table 1 Criteria for the phenotypes of multiple drug resistance in N. gonorrhoeae
Full size table

Results

N. gonorrhoeaehas become less susceptible to CRO

During the period from 1995 to 2013, 3153 N. gonorrhoeae isolates were examined for their susceptibility to CRO (Table 2). Overall, 71.90% isolates were susceptible to CRO (MIC ≤ 0.03 mg/L); 20.30% had a reduced susceptibility phenotype (MIC = 0.06 mg/L) and 7.80% exhibited a CRO nonsusceptible phenotype (MIC ≥ 0.125 mg/L). The highest CRO MIC value for the N. gonorrhoeae isolates was 0.25 mg/L, which accounted for 0.92% of the isolates tested (29/3153).

Table 2 Susceptibility of N. gonorrhoeae isolates to ceftriaxone in Shanghai
Full size table

The proportions of CRO non-susceptible isolates were over or equal to 5% in all years since 1998, and were around 7%-13% in the last 3 years (2011 to 2013). The proportion of isolates showing reduced susceptibility to CRO ranged from 12% to 40% from 1998 to 2013. From 1998 to 2013, the MIC50 values for CRO were between 0.015 mg/L and 0.06 mg/L, and the MIC90 values were 0.06–0.125 mg/L (Table 2).

N. gonorrhoeaeis resistant to CIP, PEN and TET

During the period from 1996 to 2013, 3036 N. gonorrhoeae isolates were examined for susceptibility to CIP. Overall, 85.30% isolates were resistant to CIP. CIP resistant N. gonorrhoeae accounted for 16.50% of the N. gonorrhoeae isolates in 1996, compared with over 85% in 1999. CIP resistance among the isolates ranged from 88% to 100% in the period between 2000 and 2013 (Additional file 1: Figure S1). A total of 3663 N. gonorrhoeae isolates were examined for their susceptibility to PEN. Overall, 62.00% isolates were resistant to PEN. Proportions of PEN resistance increased from 20–40% before 1998, to 70–90% during the period from 2000 to 2013 (Additional file 1: Figure S1). During the period from 1995 to 2013, 3153 N. gonorrhoeae isolates were examined for their susceptibility to TET. Only one concentration of TET (16 mg/L) was tested during 1998–2003 (n = 1355), whereas the remaining 1798 N. gonorrhoeae isolates were tested for the full range of TET concentrations. For the 1798 isolates tested during 1995–1997 and 2004–2013, 50.61% (910/1798) were resistant to TET. The proportion of TET resistant isolates increased from 16% (1995–1997) to over 40% (2004–2012), and then further increased to 82.57% in 2013 (Additional file 1: Figure S1).

Susceptibility of N. gonorrhoeaeisolates to SPT

From 1995 to 2013, 3153 N. gonorrhoeae isolates were examined for their susceptibility to SPT. Only two resistant isolates (0.06%, MIC ≥ 128 mg/L) were identified, one in 1999 and the other in 2002. Overall, isolates with intermediate resistance (MIC = 64 mg/L) accounted for 6.76%, and the related proportion of susceptible isolates was 93.18%. During the period from 2009 to 2013, the proportion of isolates with an SPT MIC ≥ 64 mg/L was less than 5% (4.46% in 2009, 0% in 2010, 2.22% in 2011 and 0% in 2012 and 2013), with the SPT MIC50 ranging between 4 mg/L and 32 mg/L during this period. The majority of SPT MIC90 values were 16–32 mg/L, and the values of MIC90 in 2003 and 2007 were 64 mg/L (data not shown).

Plasmid-mediated and chromosomally-mediated resistant N. gonorrhoeae

In 1995, only 0.85% of isolates exhibited a presumed PPNG phenotype. Around 5% of isolates were presumed PPNG from 1996 to 1998. The proportion of presumed PPNG isolates increased rapidly from 13% in 1999 to 37% in 2000 and ranged from 37% to 57% during the period from 2001 to 2013 (Additional file 1: Figure S2). The proportion of presumed TRNG isolates was 0–1% for the period from 1995 to 1998, and increased to 7.69% and 6.67% for 1999 and 2000, respectively. During 2001–2005, the proportion of presumed TRNG was around 17% - 23%. This proportion again increased from 30% (2006) to 56% (2008), and remained above 30%, from 2009 to 2013 (Additional file 1: Figure S2). There were no PP/TRNG isolates detected during the first 4 years of surveillance (1995–1998). The proportion of presumed PP/TRNG isolates was 0.48% and 1.78% for 1999 and 2000, respectively. This proportion increased rapidly to 12–14% during the period from 2001 to 2005 and remained above 15% between 2006 and 2013, peaking at 33.8% in 2008 (Additional file 1: Figure S2).

The proportion of chromosomally-mediated resistance to PEN and TET among the 1569 CMRNG isolates from 2003 to 2013 ranged from 4.23% to 29.14%. In recent years, the proportion of CMRNG isolates detected has been 15.56% (2011), 6.06% (2012) and 23.85% (2013).

Multiple drug resistance (MDR) in N. gonorrhoeae

An MDR phenotype for N. gonorrhoeae was defined as resistance to two or more antimicrobial agents (Table 1). Double-resistant phenotypes accounted for 39–68% (PEN-TET), 61–92% (PEN-CIP) and 43–81% (TET-CIP) of N. gonorrhoeae isolates, and the triple-resistant phenotype (PEN-TET-CIP) accounted for 38–61% of N. gonorrhoeae isolates during the period from 2003 to 2013 (data not shown).

Between 2001 and 2013, the proportion of the CRO nonsusceptible phenotype combined with PEN, TET or CIP resistance fluctuated from 4.5% to 19.72% (PEN-CRO), 2.41% to 12.68% (TET-CRO) or 0% to 19.72% (CIP-CRO) (Table 3). CRO-associated triple-resistant phenotypes accounted for 2.41–12.68% (PEN-TET-CRO), 2.41–12.68% (TET-CIP-CRO) and 2.17–19.72% (PEN-CIP-CRO) of N. gonorrhoeae isolates (Table 3), and the proportion of quadruple resistant phenotypes (PEN-TET-CIP-CRO) accounted for 2.41–12.68% of N. gonorrhoeae isolates (Table 3).

Table 3 MDR associated with the CRO nonsusceptible phenotype in N. gonorrhoeae isolates in Shanghai
Full size table

Discussion

In China, antibiotic resistance of N. gonorrhoeae has become an increasing problem over the past few decades in the treatment of gonorrhea infection. Since the mid-1990s, a high proportion of N. gonorrhoeae isolates sampled in Shanghai were resistant to PEN, TET and CIP. Plasmid-mediated resistance has been highly prevalent since the early 2000s. The proportion of N. gonorrhoeae isolates exhibiting a CRO nonsusceptible phenotype (MIC ≥ 0.125 mg/L) has been greater than 5% since 1998. In 2013, 13.76% of isolates exhibited a CRO MIC of ≥ 0.125 mg/L, suggesting that the effectiveness of the currently recommended treatment regime for gonorrhea is diminishing. MDR phenotypes associated with PEN, TET and CIP resistance were prevalent. N. gonorrhoeae isolates exhibiting a CRO nonsusceptible phenotype associated MDR were also found to be prevalent in recent years. In Shanghai, N. gonorrhoeae isolates were found to be susceptible to SPT.

The CRO MIC of N. gonorrhoeae isolates has been extensively monitored. In the United States, the proportion of isolates with elevated CRO MICs (≥0.125 mg/L) ranged from 0.1% to 0.4% between 2006 and 2012 [19]. In Canada, 7.3% of N. gonorrhoeae isolates showed CRO MICs ≥ 0.125 mg/L in 2010 [20]. In England and Wales during 2007–2011, 0–0.4% of N. gonorrhoeae isolates showed CRO MICs ≥ 0.125 mg/L [4]. In Australia, no isolates of N. gonorrhoeae with CRO MIC values ≥ 0.125 mg/L were reported in 2011 [21],[22]. In Latin America and the Caribbean countries, 0.4% of N. gonorrhoeae isolates had CRO MICs ≥ 0.125 mg/L between 2007 and 2011 [23].

In the present study, we found that the proportion of N. gonorrhoeae isolates with CRO MICs ≥ 0.125 mg/L in Shanghai (1996–2013) greatly varied from year to year, ranging from 0% (1995) to 19.72% in 2008, and 13.76% in 2013. Our findings were consistent with those of a survey of CRO susceptibility in Nanjing, China [24],[25]. The proportion of CRO less susceptible N. gonorrhoeae isolates (CRO MIC ≥ 0.06 mg/L) was 17.86–50.5% during the period from 1999 to 2006 in Nanjing [24]. Chen et al. reported that the proportion of isolates with a CRO MIC = 0.125 mg/L was 39.4% in 2007 and 11.3% in 2012 [25]. The proportion of CRO nonsusceptible isolates (MIC ≥ 0.125 mg/L) ranged from 22% to over 48% in Nanning in 2000–2012 [26], much higher than that detected in Shanghai. Regional differences in gonococcal AMR within a country have also been demonstrated in other countries. In Canada, isolates of N. gonorrhoeae from the province of Saskatchewan retained a high prevalence of susceptibility to antibiotics, including CIP and PEN and CRO. Unlike reports of reduced susceptibility to third-generation cephalosporins (CRO or cefixime) from other regions of Canada and the rest of the world, all isolates collected during 2003–2008 from Saskatchewan were fully susceptible to CRO and cefixime [27]. Approximately 7% of the isolates (n = 23) showed a CRO MIC of 0.03–0.06 mg/L, and 93% of the isolates (n = 297) showed a CRO MIC of < 0.03 mg/L [27]. In Canada between 2000 and 2009 (n = 40,875), the proportions of PEN, TET or CIP resistant isolates were 11.8%, 21.3% and 13.2%, respectively; and the CRO MIC had shifted from 0.016 mg/L in 2000 to 0.063 mg/L in 2009 [28]. In British Columbia (2006–2011), 13% of isolates (227/1837) showed a CRO MIC of 0.064 mg/L and 1% of isolates (27/1837) showed a CRO MIC of 0.125 mg/L [29]. Indeed, cefixime treatment failure was found in 3.09% of the investigated cases (9/291) in Toronto between 2010 and 2011 [9]. Taking all of these data into account, it is clear that regional monitoring of N. gonorrhoeae AMR is as important as national surveillance programs.

Currently, the recommended therapy for gonorrhea is a single dose of oral or parental third-generation cephalosporin. The doses of CRO used for the treatment of uncomplicated gonorrhea vary among countries; typically 250–500 mg per dose is injected intramuscularly, but in Japan 1 g is administered intravascularly [14],[23],[30]-[34]. A landmark study suggested that increasing the dose of CRO is associated with a decreased proportion of CRO nonsusceptible N. gonorrhoeae isolates in the UK [4]. However, this decrease may not be sustainable [21]. We observed that the proportion of CRO nonsusceptible isolates transiently increased from 5% in 2007 when CRO became a recommended drug in China, to over 19% in 2008, 11% in 2009 and 13% in 2013. In Shanghai, CRO at a single low dose (250 mg) has been used to treat uncomplicated gonorrhea since 2000, and the proportion of CRO nonsusceptible N. gonorrhoeae isolates remains fairly stable. Based on our findings and reports from others in China [24]-[26], CRO remains an effective antimicrobial for the treatment of uncomplicated gonorrhea. Since CRO nonsusceptible N. gonorrhoeae isolates are prevalent in China, a higher dose of CRO (500 mg) is recommended, at least in some areas such as Shanghai, Nanjing and Nanning [24]-[26].

Detailed data on trends of MDR of N. gonorrhoeae in China are limited. However, with the consideration of the high resistant proportions of gonococcal isolates to PEN, TET, and CIP, non-susceptible CRO-associated MDR should not be ignored in China. We observed that over 5% of isolates were non-susceptible CRO-associated MDR in Shanghai in recent years.

One of the major purposes of antimicrobial susceptibility surveillance is to provide information that may indicate the need for modification of treatment guidelines. The World Health Organization recommended that when the prevalence of resistant gonococci to an antimicrobial agent rises to 5%, the antimicrobial should no longer be used [35]. In China, PEN had been the treatment of choice for gonorrhea until quinolones were introduced in 1993, PEN was removed from the treatment regime in 2000 [14],[15]. Before 2000, the recommended dose of PEN continually increased because of a progressive decline in susceptibility. Penicillin was removed from the treatment regime 10 years after it was discontinued in 1985 in most regions of the world, simply because of the availability of replacing antibiotics. In 1993, oral fluoroquinolones (CIP and ofloxacin) were recommended for gonorrhea treatment, and at that time CIP resistance was already 15% in China [12],[13]. Four years after (1997), CIP resistance reached over 80% (100% in 2000), while the drug remained in the treatment regime until 2007 [14].

Recently potential novel treatment strategies have been explored [6],[7]. A strategy of combining older drugs was discussed [36]-[38]. However, there is no evidence that this combination of older drugs produces synergistic effects. Considering the problems of MDR, this strategy should be used with caution in China. A high prevalence of MDR to almost all currently available drugs in Shanghai limits options for combined therapy. New antibiotics are needed but are largely unavailable [38]. It is therefore worth considering and evaluating SPT as an alternative first line drug for the treatment of uncomplicated urogenital gonorrhea in China. However, SPT is not suitable for the treatment of pharyngeal gonorrhea even though the prevalence of resistance is low in most areas of the world [39]. The MICs of SPT need to be closely monitored, and any signs of increased SPT MIC warrant a thorough investigation and close attention by the policy makers.

The current study has a number of limitations. 1) All of the N. gonorrhoeae isolates tested in this study were collected from the Sexually Transmitted Infection Clinic of the SSDH, and only a small number of isolates were tested, accounting for 2.1% of the total reported cases in 2012 in Shanghai. 2) A lack of demographic and clinical data prevented any comparisons of N. gonorrhoeae antimicrobial resistance profiles with patient characteristics; 3) Molecular subtyping of the isolates was not performed. This would provide insight into whether there is clonal spread of resistant isolates and should be the focus of future studies. 4) The agar dilution MIC testing method has an intrinsic error rate of plus/minus one dilution. While using a breakpoint of 0.125 mg/L for CRO nonsusceptible isolates, some isolates may actually have MICs of 0.06 mg/L.

Conclusions

In summary, antimicrobial resistance of N. gonorrhoeae to PEN, TET and CIP is highly prevalent in Shanghai. A high proportion of N. gonorrhoeae isolates were found to be CRO nonsusceptible (MIC ≥ 0.125 mg/L). Plasmid-mediated resistance and MDR are also prevalent in China. N. gonorrhoeae isolates are susceptible to SPT and this antibiotic needs to be further evaluated as a treatment option for urogenital gonococcal infections. Our findings suggest that the CRO dose should be increased to 500 mg in areas with a high prevalence of CRO nonsusceptible isolates. We propose that combination therapy using older antibiotics should be used with caution unless drug synergistic effects can be confirmed. Finally, both national and regional antimicrobial susceptibility surveillance programs are important to inform policy makers in a timely manner on the appropriate modifications of treatment guidelines.

Additional file