Gonococcal antimicrobial resistance: perspectives from the African region
- 1Former Lead Specialist, STI Team, Department of Reproductive Health and Research, World Health Organization, Geneva
- 2National Institute for Medical Research, Mwanza Centre, Mwanza, Tanzania
- 3Ministry of Health and Child Welfare, AIDS and TB Programmes, Harare, Zimbabwe
- 4Département de Bactériologie-Virologie, Institut Pasteur de Côte d'Ivoire, Abidjan, Côte d'Ivoire
- 5Bacteriology Service, Centre Pasteur du Cameroun, Yaounde, Cameroon
- Correspondence to Dr Francis Ndowa, 6 Thames Road, Vainona, Harare, Zimbabwe;
- Received 24 May 2013
- Revised 4 September 2013
- Accepted 20 September 2013
Many countries in Africa have weak surveillance systems for data collection of sexually transmitted infections, and hardly any programmes for gonococcal antimicrobial susceptibility assessment. The widespread adoption of the syndromic approach to the diagnosis and management of sexually transmitted infections has also meant that the collection of a genital specimen for laboratory analysis is no longer routinely done when patients present with genital complaints, and clinical staff and laboratory technicians have lost the skill to collect genital specimens and processing them for culture and antimicrobial susceptibility testing. Following reports of gonococcal antimicrobial resistance to quinolones, WHO urged countries to monitor gonococcal antimicrobial resistance in a more systematic and regular manner. Although the response in Africa has been slow to take off, a number of studies have been conducted in a few countries and plans for implementation are in place in others. However, the number of isolates studied has been small in nearly all the countries except one, and the barriers to scaling up gonococcal antimicrobial resistance surveys seem overwhelming. In spite of the studies being few and of small sample sizes, enough information can be discerned to indicate that quinolones can no longer be a medicine of choice for the treatment of gonorrhoea in Africa and the threat of antimicrobial resistance developing in Neisseria gonorrhoeae to third-generation cephalosporins is real and imminent.
As of 2008, the 46 countries which come under the geographical and political region covered by WHO Regional Office for Africa had an estimated population of 384.4 million adults between the ages of 15 and 49 years. It also was estimated, in the same year, that this population had the highest incidence rates of gonococcal infections in the world, at about 50 new infections per 1000 women and 60 new infections per 1000 men.1 Gonorrhoea remains the most frequent cause of urethral discharge in men, accounting for up to 85% of cases in South Africa, and up to 83% in Zimbabwe.2 ,3
Early and effective treatment of gonococcal infections is an essential component of the control of gonorrhoea because treatment interrupts the chain of transmission of infection and prevents the occurrence of complications. However, the ease with which this infection was treated in the past few years, often with a single dose of antibiotic treatment, has come under serious threat by the development of multidrug resistance in Neisseria gonorrhoeae. The last class of antibiotics available to cure gonorrhoea reliably are third-generation cephalosporins, but there have been recent reports of treatment failures and reduced susceptibility of N gonorrhoeae to both oral and parenteral formulations of cephalosporins.4–7
To understand the scope and magnitude of the problem of antimicrobial resistance (AMR) in N gonorrhoeae, and put in place mechanisms to reduce the burden of gonorrhoea and the anticipated escalation of costs due to increasing AMR WHO launched the WHO Global Action Plan to Control the Spread and Impact of Antimicrobial Resistance in Neisseria gonorrhoeae.8 WHO Global Action Plan calls for strengthened surveillance for sexually transmitted infections (STIs) with particular emphasis on gonococcal AMR surveillance.
State of AMR in N gonorrhoeae in the African region
Many countries in Africa do not have a systematic data collecting systems for STIs, an indicator of weak national STI surveillance systems in general. The widespread adoption of the syndromic approach to the diagnosis and management of STIs has also meant that the collection of genital specimens for laboratory analysis is no longer routinely done when patients present with genital complaints. Nonetheless, the benefits of the syndromic approach have been seen in that some bacterial STIs have declined over the years, including urethritis, syphilis and chancroid.9 ,10 Following reports of gonococcal AMR to quinolones, WHO urged countries to monitor gonococcal AMR in a more systematic and regular manner.
In March 2011, WHO convened a 3-day workshop in Harare, Zimbabwe, to take stock of what countries had accomplished in implementing surveillance systems for gonococcal AMR at national level. The workshop was attended by national representatives from Benin, Cameroon, Ethiopia, Ghana, Kenya, Madagascar, Nigeria, South Africa, Tanzania, Uganda and Zimbabwe. Facilitators of the workshop included representatives from WHO, partners and collaborators from the South African Centre for HIV and STIs at the National Institute for Communicable Diseases, and the United States of America Centers for Diseases Control and Prevention (CDC-Atlanta, USA). Although the response to the call for monitoring gonococcal AMR in Africa has been slow to take off, a number of studies have been conducted in some countries, and plans for implementation are in place in others.
In Cameroon, the Centre Pasteur Cameroun has carried out some periodic analyses of AMR in N gonorrhoeae in 2004, 2008 and 2012. The samples were collected from referred cases from diverse health facilities including, primary healthcare clinics, healthcare services for sex workers, family planning and antenatal clinics. Since November 2012, 11 healthcare centres identified in Yaoundé have become involved in the collection of samples that are then sent to Centre Pasteur for isolation of N gonorrhoeae. A surveillance network which includes 20 laboratories of different levels (public and private) has been put in place in Yaoundé to enhance future gonococcal surveys. The isolates were inoculated directly onto chocolate agar base and identified by colony morphology, Gram stain, oxidase tests and carbohydrate degradation tests. The disc diffusion method for antimicrobial susceptibility and agar plate dilution were used for the determination of antimicrobial susceptibility and minimum inhibitory concentrations (MICs). The results showed a dramatic increase in penicillinase-producing N gonorrhoeae (PPNG) from 38.5% in 1985 to 87.3% in 2005. In 2004–2005, a study of 79 isolates (39 from women and 40 from men) from the venereology centre of Yaoundé, and the Pasteur Center of Cameroon, showed above 98% susceptibility of N gonorrhoeae to ceftriaxone, spectinomycin and ciprofloxacin, but some unpublished data showed an increase in AMR in N gonorrhoeae strains to nalidixic acid from 3.3% in 2008 to 9.5% in 2012. Since nalidixic acid is related to quinolones, this suggested that N gonorrhoeae may also have become resistant to quinolones, such as ciprofloxacin and ofloxacin. Consequently, 24 gonococcal isolates collected in 2012, three of which were resistant to nalidixic acid, were tested for MICs for ciprofloxacin. The results showed that all the isolates which were susceptible to nalidixic acid had ciprofloxacin MICs≤0.002 µg/mL whereas, of the isolates resistant to nalidixic acid, one had ciprofloxacin MIC of 6 µg/mL and two had MICs ≥ 32 µg/mL. This corroborated the good correlation between nalidixic acid and fluoroquinolone resistance in N gonorrhoeae.
In Central Africa Republic, 30 isolates were collected from March 2004 to June 2006 from two health centres and the Pasteur Institute of Bangui, from two women and 28 men, as part of a multisite study. The strains were evaluated for antimicrobial susceptibility, using the disc diffusion method and found to be all susceptible to ceftriaxone and spectinomycin and, at that time, also sensitive to ciprofloxacin. As part of the same study, 126 isolates were obtained from 31 women and 95 men attending three health centres specialised in STI care in Antananarivo, Madagascar. Of the 68 strains evaluated for MICs in Madagascar, all were sensitive to ceftriaxone, 67 (98.5%) were sensitive to spectinomycin and 64 (94.1%) were sensitive to ciprofloxacin.11 Ciprofloxacin was used as first-line treatment for gonorrhoea at that time in Madagascar.
In Côte d'Ivoire attempts to collect data for gonococcal antimicrobial susceptibility assessments were made difficult by a security crisis over the past 2–3 years. Only about 30 isolates were collected from 2009 to 2011. Though statistically too small to draw any conclusions from, the tests showed 3 (9.7%) isolates resistant to ciprofloxacin. No indication of decreased susceptibility to ceftriaxone was evident at that time. Thayer–Martin medium, agar dilution and E-test were used in Côte d'Ivoire to identify and determine the MICs for N gonorrhoeae. Plans are currently underway to organise a nationwide gonococcal antimicrobial programme with a view to increasing the number of samples since the security situation has improved in the country.
In a prospective cohort study from two sexual health and family planning clinics, in Bissau, Guinea-Bissau, to characterise N gonorrhoeae, Chlamydia trachomatis and Mycoplasma genitalium, 27 N gonorrhoeae isolates cultured on a modified Thayer–Martin medium between February 2006 and January 2008 were confirmed on rapid oxidase production, Gram-stain and use of PhadeBact Monoclonal GC test. The E-test was used for determination of antimicrobial susceptibility. Four isolates were added from a study conducted previously. Thus, 31 isolates were analysed and the levels of resistance were 10% to ciprofloxacin, 6% to erythromycin, 13% to rifampicin, 68% to ampicillin, 74% to tetracycline, and 68% to penicillin G. All the isolates were susceptible to cefuroxime, cefixime, ceftriaxone, spectinomycin and azithromycin, and the MIC of kanamycin (8–32 mg/L and gentamicin (0.74–6 mg/L) were low.12
In 2011, Kenya published a letter in which a small number of gonococcal isolates from high-risk men and women in the coastal area were tested for antimicrobial susceptibility. Seventeen of the 26 participants (16 men and 1 woman) had gonococcal infection. Using the E-test to determine MICs, 12 of the 17 isolates (71%) were resistant to ciprofloxacin, and there was also high resistance to tetracycline (100%) and penicillin (82%). All the isolates were susceptible to ceftriaxone.13 Although ciprofloxacin remains as an option for treatment for gonorrhoea, the new Kenya guidelines for the management of STIs (2013) recommend ceftriaxone and cefixime as first option.
In a different project in Kisumu, Kenya, antimicrobial susceptibility was evaluated in gonococcal isolates taken from men enrolled in a randomised trial of male circumcision to prevent HIV. From 2002 to 2009, 166 isolates were cultured directly onto Thayer–Martin chocolate agar N gonorrhoeae isolates were identiﬁed by colony morphology, Gram stain, and oxidase test. Antimicrobial susceptibilities were determined using the agar dilution method using breakpoints in accordance with the Clinical and Laboratory Standards Institute (CLSI). The results showed that the MIC90 for ciprofloxacin had increased from 0.004 µg/mL in the period 2002–2006 to 4 µg/mL in the period 2007–2009. There was a statistically significant trend of increasing proportions of quinolone-resistant strains from 9.5% in 2007, 27.3% in 2008 to 50% in 2009. Penicillin resistance was observed in 65% of the isolates. Also noted during the same periods, 2002–2006 to 2007–2009, were significant increases in MICs for cefixime (0.008–0.016 μg/L), ceftriaxone (0.008–0.016 μg) and azithromycin (0.125–0.25 μg/L). All isolates were susceptible to spectinomycin and, of the 51 isolates tested for susceptibility to gentamicin, one isolate (2%) had an MIC of 4 µg/mL, 49 (96%) had an MIC of 8 µg/mL, and only one other had an MIC of 16 µg/mL (the working breakpoint for resistance for gentamicin.)14
In Malawi, a study was conducted in 2007 in men with urethral discharge. From 126 urethral swabs, 108 had N gonorrhoeae from which 100 isolates were studied for MICs. N gonorrhoeae was diagnosed by Gram stain and culture. Antimicrobial susceptibility was determined by disk diffusion and E-test MIC determination and agar dilution MIC determination. At that time, all isolates were susceptible to ceftriaxone, gentamicin, spectinomycin and azithromycin. One isolate was resistant to cefixime. Kanamycin and ciprofloxacin were also highly effective, while only one isolate demonstrated reduced susceptibility to ciprofloxacin and another one isolate had reduced susceptibility to kanamycin.15
In Mozambique, a study was carried out to determine antimicrobial susceptibility profile of N gonorrhoeae in Maputo in 2005. New York City GC agar plates were used for culture and identification of presumptive N gonorrhoeae was confirmed by carbohydrate utilisation tests and MICs to spectinomycin, ciprofloxacin, ceftriaxone, cefixime, tetracycline and penicillin were determined using the CLSI methodology and the agar dilution method for MICs to kanamycin. N gonorrhoeae was cultured from 40 of 116 men and 15 of 154 women. Seven per cent of the isolates were resistant to kanamycin and all were susceptible to ciprofloxacin, ceftriaxone and cefixime at that time.16 Kanamycin was the first-line treatment for gonorrhoea at that time.
In Namibia, a microbiological survey was conducted and completed over 9 months in 2007 to guide revision of the STI treatment guidelines. Gonococcal culture was performed and MIC determinations by E-test showed the overall prevalence of ciprofloxacin resistance in N gonorrhoeae of 24%. As a result, the Ministry of Health and Social Services revised the national STI treatment guidelines the following year and instituted oral cefixime as the treatment of choice for gonococcal infections. In this study, 173 (87%) of the 199 men with urethral discharge had N gonorrhoeae (by real-time PCR), 18 (9%) had C trachomatis, 9 (5%) had Trichomonas vaginalis and 4 (2%) had M genitalium.17
In South Africa, AMR surveys were conducted in Gauteng (annually since 2007), the Northern Cape (in 2006), Mpumalanga (in 2006), the Western Cape (between 2006 and 2007), the Free State (in 2009), the Eastern Cape (in 2010), the North West Province (between 2010 and 2011) and Limpopo (2010–11). Additionally, there were other surveys conducted between 2003 and 2008 in Durban, Johannesburg and Pretoria to add on to the South African data which were published in peer reviewed journals. Widespread ciprofloxacin resistance in N gonorrhoeae was demonstrated by these studies, but cephalosporins remained effective.18 However, more recently, in 2012, two cases of cefpodoxime-resistant N gonorrhoeae isolates were reported in the Gauteng province. The two isolates also demonstrated decreased susceptibility to azithromycin and gentamicin and higher than usual MIC for ceftriaxone.4
In Zimbabwe, results from the gonococcal AMR study conducted in 2010 in 12 polyclinics in the City of Harare in men presenting with urethral discharge revealed some interesting findings despite the relatively small sample size. The project managed to recruit only 131 men with urethral discharge over 6 months, from whom only 69 gonococcal isolates were identified and tested for antimicrobial susceptibility. N gonorrhoeae culture was performed on New York City selective agar plate and presumptive gonococcal colonies were identified by Gram staining, oxidase and monoclonal antibody testing (PhadebactR monoclonal GC test, Bactus AB, Sweden). Antimicrobial susceptibility testing was conducted using E-tests for MICs, and breakpoints were in accordance with the CLSI for ciprofloxacin, cefixime and ceftriaxone. Good efficacy of ciprofloxacin was observed in 88% of the isolates, while four isolates (6%) showed decreased susceptibility, and another four were resistant to ciprofloxacin. This was such a close call for ciprofloxacin to be abandoned based on such borderline results. It would have been interesting to see how the values changed with larger sample sizes. Most of the gonococci were still highly sensitive to kanamycin and the cephalosporins, ceftriaxone and cefixime (table 1). Of concern, however, was that, there were a small number of strains showing reduced susceptibility to cefixime (three isolates) and ceftriaxone (two isolates), with MICs of >0.25 mg/L, even though cephalosporins were not used for the treatment of gonococcal infections in the country. Despite the small sample size, this observation must sound alarm bells and stress the need for regular gonococcal antimicrobial susceptibility monitoring in Zimbabwe. That gonococcal urethritis is still a significant problem in this part of Africa is borne out by the fact that in 128 of the recruited men in this survey, gonococcal infection was confirmed in just over 80% of them using the Multiplex PCR. The other causative agents found were C trachomatis in 12%, M genitalium in 5% and T vaginalis in 2%.3
In 2011, Tanzania embarked on a multisite gonococcal antimicrobial survey covering seven regions of Tanzania mainland. Progress reports and results are eagerly awaited from this undertaking.
Ostensibly, in the vast continent of Africa, gonococcal AMR surveys are implemented only in a handful of countries as specific research projects, or on an ad hoc basis as part of surveillance, except in South Africa where there seems to be either a regular monitoring system or frequent special studies. Where gonococcal AMR studies were conducted and monitored, sites failed to collect isolates in sufficient quantities from which decisions on changing treatment protocols can be made with confidence. For example, in Zimbabwe it took 6 months to enrol 131 men with male urethral discharge syndrome, and then only 69 gonococci were isolated from to conduct MICs on. The reasons for the low recruitment are numerous, including the healthcare-seeking behaviours of most men with urethral discharge who may prefer to see private doctors or to buy their medication from street merchants and, where samples have been collected, there are inadequate means of transportation of samples to a processing laboratory and delays by laboratory staff to process the isolates. A particularly peculiar occurrence in the multisite study in which were involved Cameroon, Central Africa Republic and Madagascar was the loss of all 79 isolates from Cameroon and 97 of the 126 isolates from Madagascar due to problems of electricity supply to the freezers.11 All these hurdles led to fewer than recommended sample sizes of gonococcal isolates being observed from almost every country that implemented the exercise between 2004 and 2012, except South Africa. Some countries did not manage even above 100 isolates for analysis. When the countries’ data are aggregated, 3453 isolates were tested for antimicrobial susceptibility in the past decade, but by far the biggest contributor to this figure was South Africa. Furthermore, in the same period, 2004–2012, the number of countries conducting gonococcal antimicrobial surveys each year has been small—hardly more than four countries in any 1 year. The largest number of isolates tested in any 1 year was in 2007 from three countries, Malawi, Namibia and South Africa, and by far the biggest contributor was South Africa (table 2).
Thus, although the resistance levels to ciprofloxacin and increasing MICs of cephalosporins observed in some cases need to be interpreted with caution because of the small sample sizes of the isolates tested, the results are leaning towards a problem of AMR, particularly with quinolones, and an impending risk of losing cephalosporins in the treatment of gonorrhoea.
The need for a systematic gonococcal AMR surveillance programme in Africa (GASP-in-Africa) is evident, but the needs to make it happen are many. Surveillance for gonococcal AMR is a public health necessity to ensure an effective response to the emergence of cephalosporin-resistant gonorrhoea heralded by increasing MICs and the recently reported treatment failures of the two patients in South Africa. The lack of surveillance in most countries in Africa probably means that what has been observed in South Africa is more widespread on the African continent and vigilance should be heightened.
A response is needed to the loss of skills by laboratory staff to process isolates for susceptibility testing. This will require training centres and laboratories for healthcare providers and laboratory staff. It will also be important to ensure quality assurance for each participating laboratory and have a number of centres in the region to provide backup to the monitoring of AMR. The funding needed to strengthen the capacity of healthcare providers and laboratories and laboratory staff seems unavailable at the national level. Most STI programmes cite lack of funds as one of the main reasons for weak STI surveillance programmes, including gonococcal antimicrobial surveillance. Invariably, the programmes immediately look to donors to come to their rescue.
To minimise cost and dependence on external funding, it will be imperative to make gonococcal antimicrobial surveys an integral part of national STI surveillance, and not be regarded as special research. These surveys are an essential component of patient care and should, therefore, be implemented as such on a regular basis. Without such studies, appropriate and effective medicines and regimen cannot be determined.
Recruitment of patients for AMR surveys is likely to improve if specimen collection becomes routine and presurvey visits are carried out to ensure staff motivation and verify that the sites selected have sufficient attendance rates. Since the syndromic management of STIs has been adopted in all the countries in Africa, gonococcal AMR studies will probably have to be conducted at intervals of 1–2 years, appropriately entrenched within the national STI surveillance system. Additionally, in between the periodic surveys, gonococcal isolates must continue to be obtained from selected sites, especially from patients with treatment failure and patients attending STI clinics and/or from selected populations at high risk of infection, such as sex workers or men who have sex with men. This will serve two purposes; first to ensure a regular monitoring of gonococcal strains and signal any resistance early and, second, to keep clinical staff and laboratory technicians familiarised with sample collection, processing of specimen and culturing of N gonorrhoeae. By doing so, the skills regained by healthcare providers and laboratory staff in conducting surveys for gonococcal antimicrobial susceptibility assessments will be maintained and passed on to future generations.
National governments will need to collaborate closely with each other across borders so that survey results can be shared rapidly, and will enable a discussion to ensue among technical experts who will recommend appropriate actions and responses to the menace of AMR in N gonorrhoeae in Africa. If GASP-in-Africa is not implemented in a systematic and coordinated manner, the return of high prevalence gonococcal infections will be inevitable, with serious health consequences for women, infants and young men who are at higher risk of sexually transmitted infections.
High rates of urethral discharge are still the case in Africa, commonly due to gonococcal infections.
As surveillance and monitoring of gonococcal antimicrobial resistance are not conducted routinely in the majority of countries in Africa, untreatable gonorrhoea is going to emerge as a silent infection with serious sexual and reproductive health consequences.
Countries in Africa need to strengthen STI surveillance and strengthen local laboratory capacity to culture N. gonorrhoeae and regularly monitor antimicrobial susceptibility of the pathogen.
Handling editor Jackie A Cassell
Contributors All the coauthors reviewed the first draft of the manuscript and added to it together with suggestions for references to be included. The coauthors also provided updated information about the situation of gonococcal antimicrobial resistance in their respective countries.
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.