Objectives: The objectives of this study were to assess the prevalence of ciprofloxacin-resistant gonorrhoea in two South African cities and to investigate the association between the isolation of ciprofloxacin-resistant Neisseria gonorrhoeae and the HIV serostatus of patients.
Methods: Gonococci were cultured from endourethral swabs taken from consecutive men with urethritis attending clinics in Johannesburg and Cape Town. Minimum inhibitory concentrations (MIC) for ciprofloxacin and ceftriaxone were determined with E-tests. Isolates with a ciprofloxacin MIC of 1 mg/l or greater were defined as resistant and isolates with a ceftriaxone MIC of 0.25 mg/l or less were defined as susceptible. Rapid tests were used to screen and confirm the presence of HIV antibodies. Survey data from 2004 were used as a baseline to assess trends in gonococcal resistance to ciprofloxacin.
Results: In 2004, the prevalence of ciprofloxacin resistance was 7% in Cape Town and 11% in Johannesburg. In 2007, 37/139 (27%) Cape Town isolates and 47/149 (32%) Johannesburg isolates were resistant to ciprofloxacin; in comparison with 2004 data, this represents 2.9-fold and 1.9-fold increases, respectively. All isolates were fully susceptible to ceftriaxone. There was a significant association between HIV seropositivity and the presence of ciprofloxacin-resistant gonorrhoea among patients (p = 0.034).
Conclusions: Johannesburg and Cape Town have witnessed significant rises in the prevalence of ciprofloxacin-resistant gonorrhoea among men with urethritis. The resistant phenotype is linked to HIV seropositivity. There is now an urgent need to change national first-line therapy for presumptive gonococcal infections within South Africa.
Statistics from Altmetric.com
Since 2004, the relative prevalence of ciprofloxacin-resistant gonococci among male urethritis cases has escalated significantly in both Johannesburg and Cape Town.
A change in first-line therapy for presumptive gonorrhoea is urgently needed in South Africa.
The observation that HIV stropositivity is associated with ciprofloxacin-resistant gonococcal infections has public health implications regarding HIV transmission.
Gonorrhoea remains the most frequent cause of the male urethritis syndrome (MUS) in South Africa and may account for up to 85% of MUS cases in provincial aetiological surveys (STI Reference Centre, unpublished data). The current South African national guidelines for the management of sexually transmitted infections (STI) recommend ciprofloxacin 500 mg as a single oral dose for the treatment of presumptive gonococcal infections among men with MUS, scrotal swelling syndrome and women with either vaginal discharge syndrome or lower abdominal pain syndrome. In addition, sexual partners of patients presenting with these four syndromes will also require similar treatment.1 The current second-line antigonococcal therapy within South Africa consists of a single intramuscular 250 mg dose of ceftriaxone. Drugs for use in the country’s public primary healthcare facilities are contained on the Essential Drugs List, which is revised on a regular basis.1
Within South Africa, STI are managed syndromically.2 Integral to this approach is the requirement for periodic surveillance to assess both the microbiological causes of the various STI syndromes (aetiological surveillance) and antimicrobial resistance testing for key bacterial STI pathogens, notably Neisseria gonorrhoeae. Ciprofloxacin-resistant gonorrhoea was first reported as an emergent problem in South Africa’s KwaZulu Natal Province in 2003.3 Subsequent to that, the National Department of Health and National Institute for Communicable Diseases, in conjunction with a number of South African university laboratories, undertook a survey in 2004 to assess the prevalence of ciprofloxacin-resistant gonorrhoea across the country.4 5
This paper reports the results of a repeat survey undertaken in 2007 to determine the relative prevalence of ciprofloxacin-resistant gonorrhoea among men with urethral discharge in Cape Town and Johannesburg and investigates the association between the presence of ciprofloxacin-resistant gonorrhoea and HIV serostatus within the study population.
Consecutive patients presenting to Salt River Clinic (Cape Town) and Alexandra Health Centre (Johannesburg) with urethral discharge were asked to participate in the survey. Patients were recruited from Salt River Clinic between January and February 2007 (2 months) and from Alexandra Health Centre between January and April 2007 (4 months). After patients gave informed consent, limited clinical data relating to previous antibiotic use and travel outside the province were collected (data not shown). Anonymous genital and blood samples were collected from men whose STI syndrome was treated in accordance with national STI management guidelines, namely ciprofloxacin 500 mg as a single oral dose and a one week’s course of doxycycline 100 mg twice a day. All patients were offered access to voluntary counselling and testing for HIV at the facility. Every patient was asked to re-attend one week after treatment to assess clinical cure. Endourethral cultures were not taken to assess microbiological cure at the follow-up visits.
For men with urethral discharge, an endourethral swab was used to collect purulent urethral material, which was then placed in Amies transport medium (Cape Town) or applied to a New York City selective agar plate for the culture of N gonorrhoeae (Johannesburg). The inoculated plate was placed in a candle jar and all inoculated agar plates and swabs in transport media were delivered to the laboratory on a daily basis. In addition, a 10 ml blood sample was collected for anonymous HIV testing from each patient.
In Cape Town, the swab was removed from the Amies transport medium and applied to a New York City selective agar plate for the culture of N gonorrhoeae. In Johannesburg, the inoculated agar plates were removed from the candle jar each afternoon at the laboratory. All plates were incubated for 24–48 h in a 37°C incubator with a 5% carbon dioxide-containing humidified atmosphere. Presumptive gonococci were identified by Gram staining, oxidase, catalase and BBL Crystal biochemical strip testing (Becton, Dickinson and Co, USA). After inoculating plates containing GC agar base (Oxoid, UK) and 1% defined growth supplement with a suspension of gonococci equivalent to a MacFarlane standard of 1.0, antibiotic susceptibility testing was performed using E-tests for ciprofloxacin and ceftriaxone (AB Biodisk, Sweden). No other relevant agents, such as tetracycline, were tested in this antimicrobial resistance surveillance study. The breakpoints for ciprofloxacin were as follows: resistant 1 mg/l or greater; intermediate susceptibility 0.12–0.5 mg/l; susceptible 0.06 mg/l or less. Strains were defined as susceptible to ceftriaxone if they had a minimum inhibitory concentration (MIC) of 0.25 mg/l or less. Breakpoints were in accordance with the Clinical and Laboratory Standards Institute.6 Control strains used included ciprofloxacin-susceptible strains World Health Organization (WHO) A–E and ATCC 49226, as well as an internal gonococcal control strain MAL-61 known to be resistant to ciprofloxacin on repeated agar dilution testing. Antimicrobial susceptibility testing took place at the National Health Laboratory Service’s Division of Medical Microbiology, Tygerberg Hospital, University of Stellenbosch (Cape Town isolates) and the STI Reference Centre, National Institute for Communicable Diseases (Johannesburg isolates).
Serological testing for HIV was performed on all sera from patients with gonorrhoea at the STI Reference Centre, National Institute for Communicable Diseases, using the Determine rapid test (Abbott Laboratories, Japan). Positive Determine test results were confirmed using a second rapid test (Unigold; Trinity Biotech, Ireland).
Data were available for historical comparison from the 2004 National Gonococcal Antimicrobial Resistance Survey. In 2004, MIC for ciprofloxacin and ceftriaxone were determined for gonococci isolated from the same two cities by agar dilution using GC agar base with 1% defined growth supplement. In addition, during the same 2004 survey, the Johannesburg isolates were also tested for ciprofloxacin susceptibility with E-tests and the same susceptibility testing media. Unpublished data analysis from this survey, using Pearson’s correlation coefficient, indicated a statistically significant linear relationship between the agar dilution and E-test MIC results for resistant strains (r = 0.62, p<0.001). As in the 2007 survey, WHO strains A–E and ATCC 49226 were again used as controls. In the 2004 survey, 17/232 (7%) Cape Town and 34/302 (11%) Johannesburg isolates were resistant to ciprofloxacin. All isolates were susceptible to ceftriaxone.
Data were entered, cleaned and analysed using a Microsoft Office Excel 2003 database. A χ2 test was used to determine associations between HIV and the ciprofloxacin resistance phenotype with the level of significance set at p = 0.05.
Overall, 171 men with urethral discharge were enrolled at Cape Town (2 months) and 216 men in Johannesburg (4 months). Gonorrhoea, as determined by a positive bacteriological culture, accounted for 82% (141/171) of urethral discharge cases in Cape Town and 70% (151/216) of cases in Johannebsurg. Cultures were not performed for two men in Johannesburg.
MIC determinations and response to therapy
A total of 139 Cape Town-derived and 149 Johannesburg gonococcal isolates were tested for antimicrobial susceptibility to ciprofloxacin and ceftriaxone. Four isolates died before MIC determination could take place. In Cape Town, 37 (27%) isolates were resistant to ciprofloxacin; four (3%) isolates exhibited intermediate susceptibility and 98 (71%) were fully susceptible. In Johannesburg, 47 (32%) isolates were resistant to ciprofloxacin; six (4%) isolates exhibited intermediate susceptibility and 96 (64%) were fully susceptible. All isolates tested from both cities were susceptible to ceftriaxone.
Response to treatment
Follow-up was poor as only 16% (28/170) of men in Cape Town and 2% (5/213) of men in Johannesburg returned at one week for an assessment of the efficacy of therapy; there were no data for the remaining four men. Ciprofloxacin susceptibility data were available for 29 of these 33 men; the persistence of gonorrhoea was significantly associated with a ciprofloxacin resistance phenotype (p<0.001).
Comparison with historical 2004 data
Using the historical 2004 data as a baseline, it was observed that the relative prevalence of ciprofloxacin-resistant gonorrhoea in Cape Town had risen from 7% in 2004 to 27% in 2007 (fig 1). This represents a 2.9-fold increase over the 3-year period. Likewise, for Johannesburg, the relative prevalence of ciprofloxacin-resistant gonorrhoea rose from 11% in 2004 to 32% in 2007, a 1.9-fold increase (fig 1). The rising trend in the relative prevalence of ciprofloxacin-resistant gonorrhoea was significant for both cities (p<0.001). A similar trend was observed in terms of the detection of isolates with intermediate susceptibility (p<0.001 for both cities). The relative prevalence of gonococci with an intermediate susceptibility phenotype rose from 0.9% (2/232) in 2004 to 2.9% (4/139) in 2007 in Cape Town (p = 0.14) and from 0.7% (2/302) in 2004 to 4% (6/149) in 2007 in Johannesburg (p = 0.01) (fig 1).
Association of ciprofloxacin-resistant gonorrhoea with HIV serostatus
Sera were available for 138 (99%) of 139 gonorrhoea patients from Cape Town and 144 (97%) of 149 gonorrhoea patients from Johannesburg. The prevalence of HIV infection among the Cape Town patients was 21% (29/138) compared with 43% (62/144) among the Johannesburg patients. In Cape Town, 30% (11/37) of patients with ciprofloxacin-resistant gonorrhoea were HIV seropositive compared with 18% (17/97) of men with ciprofloxacin-susceptible gonorrhoea. Likewise, HIV seropositivity was associated with ciprofloxacin-resistant gonorrhoea among the Johannesburg patients (23/44, 52% versus 38/94, 40%). Although there was no significant association between HIV serostatus and ciprofloxacin resistance among the gonococcal isolates at each city, due to low numbers of resistant isolates, there was a statistically significant association between ciprofloxacin-resistant gonococci and HIV seropositivity when the data from the two cities were combined (table 1). Among 81 men with ciprofloxacin-resistant gonococcal isolates, the seroprevalence of HIV infection was 42% compared with an HIV seroprevalence of 29% among men with ciprofloxacin-susceptible gonorrhoea (p = 0.034).
Gonococcal resistance to ciprofloxacin is now well established in the far-east, north America and Europe.5 7 Until recently, it was reported that gonococci isolated in Africa remained susceptible to this antimicrobial agent.3 7 8 Uncertainty still exists as to whether quinolone-resistant N gonorrhoeae (QRNG) strains were imported into South Africa or arose de novo.9 The 2007 surveillance study findings from Cape Town and Johannesburg demonstrate significant increases in the relative prevalence of ciprofloxacin-resistant gonorrhoea in these two cities. The factors responsible for the further spread of ciprofloxacin-resistant gonococci within South Africa remain to be elucidated. Given that gonococcal populations are highly volatile and may vary rapidly over time and place, further detailed studies to investigate the spread of ciprofloxacin-resistant gonococci, utilizing a combination of demographic, clinical, epidemiological and molecular typing approaches, are required.
WHO recommends that first-line therapy for gonorrhoea be changed if the first-line agent fails to cure 5% or more of cases.2 The surveillance data presented in this paper demonstrate that 27% of Cape Town isolates and 32% of Johannesburg isolates exhibit microbiological resistance that is highly likely to lead to clinical failure with single-dose ciprofloxacin therapy. Doxycycline cannot be relied upon to treat QRNG isolates as the prevalence of tetracycline resistance among gonococci in African antimicrobial resistance surveys is high, often exceeding 90%.10 11
All isolates tested were susceptible to ceftriaxone, which is available on the Essential Drugs List, and this antimicrobial agent would be the obvious candidate for a switch in first-line therapy for presumptive gonococcal infections.5 Oral cephems, such as cefixime and cefpodoxime, are not currently available in South Africa but would offer an operational advantage over intramuscular ceftriaxone in terms of ease of administration. This is particularly important given both the HIV risk associated with occupational needlestick injuries in countries with high HIV prevalence among STI patients and the overtreatment, inherent in the syndromic management approach, of both index patients with MUS, scrotal swelling, vaginal discharge or lower abdominal pain syndromes and their sexual partners. However, this operational benefit requires to be balanced against concerns of generating antimicrobial resistance to oral cephems, such as cefixime, as has recently been observed in Japan.5 12
The effect of continuing to offer ciprofloxacin as first-line therapy to patients attending with gonococcal infections is to select for QRNG strains and enhance their transmission in the community. Those patients with QRNG infections who receive ineffectual ciprofloxacin therapy are at an increased risk of gonococcal complications, including epididymo-orchitis and pelvic inflammatory disease, with the associated sequelae of infertility and ectopic pregnancy, as well as an increased risk of HIV acquisition and transmission.13 14 Bacterial STI, such as gonorrhoea, exhibit epidemiological synergy and may enhance the transmission of HIV by three to five times.14 The presence of HIV DNA in urethral and semen specimens is significantly associated with gonococcal infection.15 16 Effective treatment for gonorrhoea significantly reduces the infectious HIV inoculum.16 Men with gonorrhoea are therefore an important group to target with effective antimicrobial therapies as part of national HIV prevention strategies.
South Africa is at the peak of a major HIV epidemic, with approximately 30% of antenatal clinic attendees testing HIV seropositive in the national antenatal clinic HIV surveillance programme.17 In the current study, 22% of men in Cape Town and 43% of those in Johannesburg with gonorrhoea were co-infected with HIV. This study demonstrated that there is a statistically significant association between infection with ciprofloxacin-resistant gonorrhoea and an HIV-positive serostatus (table 1); the reasons underlying this observation require further investigation.
In summary, this study has demonstrated a significant temporal rise in the relative prevalence of ciprofloxacin-resistant gonorrhoea in men with urethral discharge in two South African cities. The delay in effective treatment of ciprofloxacin-resistant gonorrhoea through the inappropriate use of ciprofloxacin as a first-line agent may thus enhance the presence and increase the HIV infectiousness of both urethral discharge and semen by several weeks, with obvious implications for the onward transmission of HIV. There is thus an urgent need to change first-line therapy for presumptive gonococcal infections in South Africa.
Since acceptance of this paper for publication and based largely on the data presented here, the South African National Department of Health has changed the national guideline for the management of presumptive gonococcal infection. Oral cefixime will now replace ciprofloxacin as the first line anti-gonococcal agent in the male urethritis and vaginal discharge syndromic management algorithms.
The authors acknowledge the support and help of Mrs G Sifanelo and colleagues at the western sub-district, City of Cape Town Health Department, Sr Amila Latif and colleagues at the Salt River Clinic, and Dr M Simba, Sr D Ntsanwisi and Mrs Y Lefakane and colleagues at the Alexandra Health Centre. They also acknowledge Professor H Koornhof and colleagues for their work in undertaking the 2004 gonococcal survey, which provided the historical data for the current study.
Funding: This study was supported by cooperative agreement no U62/CCU022901 from the Centers for Disease Control and Prevention (CDC), Atlanta, USA. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Department of Health and Human Services at the CDC and the National Centre for Global AIDS Prevention.
Competing interests: None.
Ethics approval: Ethics approval for the study was granted by the Human Research Ethics Committee (Medical) of the University of the Witwatersrand Ethics Committee, South Africa (protocol no M051024).
Contributors: DL, NdP, FR and EW set up the operational aspects of the study; AvZ and MS were responsible for collection of the specimens at the clinics; LS, MS and FR were responsible for laboratory testing of the gonococcal isolates and sera; SM entered the data and DL and SM undertook the data analysis. DL wrote the manuscript with contributions from all the authors.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.