ANTIMICROBIAL RESISTANCE of Neisseria gonorrhoeae is an increasing problem worldwide. Inappropriate use of antibiotics, favored by uncontrolled sale and self-medication in many countries, enhances the emergence of resistant strains. The lack of coherent control strategies and the lack of tools to evaluate the efficacy of current antimicrobial therapies make the emergence of new forms of resistance a more serious problem in the developing world than in industrialized countries.
The speed with which penicillinase-producing N. gonorrhoeae (PPNG) and tetracycline-resistant N. gonorrhoeae (TRNG) may spread and disseminate in a community illustrates a major difference between the impact of plasmid and chromosomal antimicrobial resistance in the gonococcus. The emergence and spread of TRNG has many parallels with that of PPNG. Resistance to both penicillin and tetracycline results from the acquisition of part of a transposable element not found previously in N. gonorrhoeae.1,2 In both instances, spread has been occurring with remarkable rapidity in several countries.3,4
Several researchers have documented the high frequency and increasing prevalence of antimicrobial resistance of N. gonorrhoeae in Africa, especially to penicillin and tetracycline.5–9 The reported prevalences of PPNG in African countries vary between 15% and 80%.10,11 Tetracycline-resistant N. gonorrhoeae in Africa were first described in Zaire and, more recently, in Kenya and Tanzania.9,12,13 In female prostitutes from Kinshasa, the level of TRNG had increased from 11% to 45% within 30 months.14
The objectives of this study were to document the antimicrobial susceptibility of gonococcal isolates from three sites in West and Central Africa, to document the spread of PPNG and TRNG over the past years in these three sites, and to discuss the consequences of rising resistance to antibiotics in the management of gonococcal infection in Africa.
Materials and Methods
A total of 2,288 gonococcal isolates were obtained from three different study sites in Africa: 251 isolates were collected at Projet Retro-CI in Abidjan, Ivory Coast, between April 1992 and December 1993, from consecutive women attending an outpatient gynecology clinic and from female prostitutes enrolled in a survey on sexually transmitted diseases (STDs)15; 952 consecutive isolates were obtained between January 1988 and July 1993 from men and women consulting for STDs at the outpatient clinic of the Centre Hospitalier in Kigali, Rwanda; and 1,085 isolates were obtained between May 1988 and October 1990 from a cohort of female prostitutes examined in a special treatment clinic in Kinshasa, Zaire.16
The same standard laboratory procedure was used in the three sites. Clinical specimens were inoculated directly on modified Thayer Martin medium containing vancomycin, colistin, nystatin, and trimethoprim (Becton Dickinson, Cokeysville, MD) and incubated for 48 hours at 36°C in candle extinction jars. Isolates were stored in skim milk at −80°C (Kigali, Kinshasa) or in liquid nitrogen (Abidjan) before being shipped in dry ice or in liquid nitrogen to the Institute of Tropical Medicine, Antwerp, for susceptibility testing and identification. Overall, more than 90% of isolates survived storage and transport.
Antimicrobial Susceptibility Testing and Typing
Penicillinase-producing N. gonorrhoeae were detected with nitrocefin (chromogenic cephalosporin; Unipath, Basingstoke, UK). Minimum inhibitory concentrations (MICs) of penicillin, tetracycline, thiamphenicol, spectinomycin, ciprofloxacin, and ceftriaxone were determined with an agar dilution technique. Gonococcal reference strains World Health Organization (WHO) A, B, C, D, E, and American Type Culture Collection 49226 were included. Inocula of 104 colony-forming units were delivered on gonococcal agar base supplemented with 1% Iso VitaleX (Becton-Dickinson, Cockeysville, MD). Plates then were incubated at 36°C in 5% carbon dioxide with high humidity. The MICs were read after 20 hours. Tetracycline-resistant N. gonorrhoeae were identified presumptively by a MIC 63 ≥ 16 μg/ml of tetracycline and confirmed by the presence of the 25.2-MDa plasmid.12 Auxotyping and serotyping were performed by standard procedures.17,18
Data were analyzed using the software package Epi-Info 5.0 (Centers for Disease Control Epidemiology Program Office, Atlanta, GA; and World Health Organization Global Programme on AIDS, Geneva, Switzerland). Proportions were compared with the Yates corrected chisquare test or Fisher exact test when appropriate.
The antimicrobial susceptibilities of the gonococcal isolates, by geographical origin, are summarized in Table 1. All isolates were susceptible to spectinomycin (MIC ≤ 64 μg/ml), ciprofloxacin (MIC ≤ 0.06 μg/ml), and ceftriaxone (MIC ≤ 0.25 μg/ml). The MIC50, MIC90, and MIC ranges for the three antibiotics were similar in the three sites. Chromosomal resistance to thiamphenicol (MIC ≥ 2 μg/ml) was found in 25%, 44%, and 37% of the gonococci from Abidjan, Kigali, and Kinshasa, respectively. Chromosomal resistance to penicillin (MIC ≥ 2 μg/ml) was observed in only 1 isolate from Abidjan (0.4%), in 133 isolates from Kigali (14%), and in 73 isolates from Kinshasa (6.7%). Chromosomal resistance to tetracycline (MIC 2–8 μg/ml) was observed in 56 (22%), 377 (40%), and 412 (38%) isolates from Abidjan, Kigali, and Kinshasa, respectively.
Plasmid-mediated resistance to penicillin and to tetracycline was common at all three sites. The overall prevalences of PPNG and TRNG, and the frequency of TRNG among PPNG and non-PPNG, by geographical origin, are listed in Table 2. Tetracycline-resistant N. gonorrhoeae were significantly more frequent among PPNG than among non-PPNG at the three sites. Of the gonococcal isolates obtained during the study periods, 27%, 39%, and 26% were susceptible to penicillin, and 17%, 48%, and 32% were susceptible to tetracycline in Abidjan, Kigali, and Kinshasa, respectively.
The frequency of TRNG and PPNG over time is shown in Figure 1. The levels of TRNG increased gradually at all three sites. Among the isolates from Kinshasa obtained in 1988, 1989, and 1990, we found a prevalence of TRNG of 14%, 35%, and 41%, respectively. In Kigali, high-level resistance to tetracycline was not observed in 1988, but a single case of TRNG was seen in 1989; since then, a rapid spread of TRNG occurred, resulting in prevalences of 2%, 18%, 39%, and 64% in 1990, 1991, 1992, and 1993, respectively. In Abidjan, we found a prevalence of already 20% TRNG in 1992, and an increase to 65% in 1993. The prevalences of PPNG did not change in Kinshasa and in Abidjan during the observation periods. In Kigali, however, a significant increase of PPNG occurred over time: the prevalence was 44% for the period 1988 through 1991 and 57% during 1992 and 1993 (P = 0.001).
Auxotypes and Serovars
All gonococcal isolates from Abidjan and Kigali and 975 isolates from Kinshasa were auxotyped and serotyped. At the three places, the majority of isolates were prototrophic (Proto) or required proline (Pro-) for their growth. We identified 18, 23, and 22 different serovars in Abidjan, Kigali, and Kinshasa, respectively. Sixty-one (6%) isolates from Kinshasa were not typable. There were 57% and 36% of the isolates from Abidjan and Kigali, respectively, and 55% of the typable isolates from Kinshasa belonging to serogroup IA. The IA-6 was the most common serovar at all three places.
The association of serogroup and plasmid-mediated resistance to penicillin and tetracycline, by geographical origin, is listed in Table 3. The IB-specific serovars were significantly more common among PPNG than among non-PPNG in Kigali and in Kinshasa, but not in Abidjan. The IA-specific serovars were significantly more common among TRNG than among non-TRNG in Abidjan and in Kinshasa, and no association was observed in Kigali.
A total of 42, 69, and 46 different auxotype/serovar (A/S) classes were represented among the gonococcal isolates from Abidjan, Kigali, and Kinshasa, respectively. In Abidjan, 26 A/S classes were observed among TRNG and 27 among non-TRNG. The TRNG in Kigali belonged to 18 and non-TRNG to 67 different A/S classes. In Kinshasa, 17 A/S classes were represented among TRNG and 44 among non-TRNG. There was a significant correlation between several A/S classes and TRNG or non-TRNG, although Pro-/IB-1 was the only A/S class at all three places consistently associated with non-TRNG (detailed results not shown).
This study documented a very high level of chromosomally mediated and plasmid-mediated resistance in N. gonorrhoeae in different parts of Africa. It also documented the emergence of TRNG in Kigali and a rather rapid spread of TRNG in all three study sites. All isolates were susceptible to spectinomycin, ceftriaxone, and ciprofloxacin. It is difficult to conclude whether the rapid spread of TRNG was a consequence of the use of tetracycline, because the only antibiotics used in the three study sites for treatment of gonorrhea were quinolones (norfloxacin 800 mg, ofloxacin 400 mg, or ciprofloxacin 500 mg as a single dose) or spectinomycin (2 g intramuscularly immediately). However, tetracycline was used as standard treatment of genital Chlamydia trachomatis infection, a common condition among those patients. A more important contributing factor for the rapid spread of TRNG could be the uncontrolled use and automedication with tetracycline for a broad range of infectious diseases. Tetracycline, ampicillin, and trimethoprim sulfamethoxazole are inexpensive drugs and widely available in Africa.
A rapid spread of PPNG in Africa has occurred between 1981 and 1986.19 Most gonococcal isolates from Africa now carry the 24.5-MDa plasmid, coding for its own transfer as that of β-lactamase plasmids through conjugation to other strains of N. gonorrhoeae. In this study, the overall frequency of PPNG was rather high at the three places and remained quite stable in Abidjan and Kinshasa, whereas a significant increase of PPNG occurred in Kigali in 1992. The spread of TRNG in Africa started in the late 1980s, and TRNG is significantly more frequent among PPNG than among non-PPNG.9,14 In addition to plasmid-mediated resistance, chromosomal resistance to penicillin and to tetracycline was common at the three sites (except for penicillin in Abidjan), as it has been reported elsewhere.5–9,20,21 As a consequence, these antimicrobials can no longer be used for gonococcal treatment anywhere in Africa. A high level of resistance to thiamphenicol also was observed in the three study sites and also has been reported for other African countries.8,21 However, studies from 1990 reported virtually no resistance to thiamphenicol in Zimbabwe and Ethiopia.5,22 Those data indicate that resistance to this antibiotic probably may vary substantially among countries and that recommendation of thiamphenicol as first-line drug for treatment of gonorrhea should in any case be supported by local susceptibility data.
No resistance has ever been detected to ceftriaxone anywhere in Africa. The high price of the drug in Africa certainly has contributed to limit its use. However, even after several years of widespread use in Southeast Asia, no unquestionable resistance case of gonorrhea has yet been reported in the literature.
Ciprofloxacin still is 100% effective in Africa. However, decreased susceptibility to norfloxacin and ofloxacin has been described in Rwanda.8 The emerging resistance to ciprofloxacin in Southeast Asia makes us think that resistance may soon appear in Africa too, therefore requiring close surveillance.
A few resistant strains to spectinomycin have been documented in the past 15 years, essentially in Asia. In Africa, only two recent studies have reported resistant gonococcal isolates to spectinomycin, including those in Abidjan.23,24 These contradictory data might be explained by specimen collection among selected (referred) patients and by the use of disc diffusion testing, therefore highlighting the need for standard surveillance methodology, both in the field and in the laboratory.
The gonococcal populations observed at the three cities were rather heterogenous as determined by serotyping and auxotyping. At the three study sites, the most frequent serovar was IA-6, which is the most common serovar reported from Africa.10,25 Most serovars represented less than 5% of gonococcal isolates. None of the major A/S classes contained isolates uniformly resistant to penicillin, tetracycline, or thiamphenicol. The typing patterns in this study were complex, and the use of detailed discrimination of A/S classes has not much enhanced the information obtained by serotyping or auxotyping alone.
As a result of our findings as well as those of other African gonococcal susceptibility studies, some recommendations for gonorrhea treatment and surveillance can be made. First-choice antibiotics currently recommended by WHO for uncomplicated gonococcal infection are ceftriaxone, ciprofloxacin, and spectinomycin.26 However, in many African countries, these drugs are neither available nor affordable to patients with STD. Alternatives must be identified among less expensive antibiotics. From the WHO Model List of Essential Drugs,27 trimethoprim-sulfamethoxazole combines the advantages of oral route (although not single-dose), rock-bottom price and widespread availability throughout Africa. Because of methodologic difficulties, trimethoprim-sulfamethoxazole was not tested in this study. A few recent publications from Africa have reported high-resistance levels of gonococcal strains to trimethoprim-sulfamethoxazole; however, methodologic flaws make these data questionable.22,24,28 Resistance levels lower than 10% were reported in three recent studies from Zaire, Tanzania, and South Africa.9,12,29 Kanamycin, although not tested in this study and not listed in the WHO Model List, might be another alternative because it is generic, and the recommended regimen for uncomplicated gonococcal infection is a single 2-g intramuscular injection. Recent studies in Africa have shown its in vitro efficacy.5,8,29,30 However, kanamycin must be injected. Complete treatment is ensured, but the uncontrolled use of injectable antibiotics may further contribute to the spread of HIV and other blood-borne infections. In any case, because of country-to-country variability in resistance patterns, kanamycin and trimethoprim-sulfamethoxazole should not be recommended without baseline assessment and regular surveillance. These costs must be taken into account when estimating the overall cost of management of gonorrhea. There is a need for further research in cost-effectiveness analysis of the management of gonococcal infection. Such analysis should help determine beyond which antibiotic resistance level potential savings in drug expenses (by recommending less expensive drugs) become neutralized by the costs associated with complications and further transmission of gonococcal infection. There also is a need to develop standard surveillance guidelines throughout Africa. Only then can susceptibility data be compared between countries and recommendations drawn for the management of gonococcal infection. In the interim, no single recommendation can be made. Some African countries chose to stay with generic drugs (such as kanamycin or cotrimoxazole) at first-encounter level, keeping more expensive drugs for the reference level. Others chose to abide by WHO recommendations, at the risk of straining their budget or making services inaccessible to the poorest. Some programs, however, with the help of donors managed to purchase fluoroquinolones under favorable conditions. Only careful resistance monitoring will show the appropriateness of these strategies.
1. Roberts M, Elwell LP, Falkow S. Molecular characterization of two beta-lactamase-specifying plasmids isolates from Neisseria gonorrhoeae.
J Bacteriol 1977; 131:557–563.
2. Morse SA, Johnson SR, Biddle JW, Roberts MC. High-level tetracycline resistance in Neisseria gonorrhoeae
is result of acquisition of streptococcal tet M determinant. Antimicrob Agents Chemother 1986; 30:664–670.
3. McCutchan JA, Adler MW, Berrie JRH. Penicillinase-producing Neisseria gonorrhoeae
in Great Britain, 1977–81: alarming increase in incidence and recent development of endemic transmission. Br Med J 1982; 285:337–340.
4. Knapp JS, Zenilman JM, Biddle JW, et al. Frequency and distribution in the United States of strains of Neisseria gonorrhoeae
with plasmid-mediated, high-level resistance to tetracycline. J Infect Dis 1987; 155:819–822.
5. Mason PR, Gwanzura L, Latif AS, Marowa E. Antimicrobial susceptibility of Neisseria gonorrhoeae
in Harare, Zimbabwe. Sex Transm Dis 1990; 17:63–66.
6. Lind I, Arborio M, Bentzon MW, et al. The epidemiology of Neisseria gonorrhoeae
isolates in Dakar, Senegal, 1982–1986: antimicrobial resistance, auxotypes and plasmid profiles. Genitourin Med 1991; 67:107–113.
7. Ison CA, Pepin J, Roope NS, Demba E, Secka O, Easmon CSF. The dominance of a multiresistant strain of Neisseria gonorrhoeae
among prostitutes and STD patients in The Gambia. Genitourin Med 1992; 68:356–360.
8. Bogaerts J, Tello WM, Akingeneye J, Mukantabana V, Van Dyck E, Piot P. Effectiveness of norfloxacin and ofloxacin for treatment of gonorrhoeae and decrease of in vitro susceptibility over time in Rwanda. Genitourin Med 1993; 69:196–200.
9. West B, Changalucha J, Grosskurth H, et al. Antimicrobial susceptibility, auxotype and plasmid content of Neisseria gonorrhoeae
in Northern Tanzania: emergence of high level plasmid mediated tetracycline resistance. Genitourin Med 1995; 71:9–12.
10. Ison CA, Roope NS, Dangor Y, Radebe F, Ballard R. Antimicrobial susceptibilities and serotyping of Neisseria gonorrhoeae
in southern Africa: influence of geographical source of infection. Epidemiol Infect 1993; 110:297–305.
11. Osoba AO, Jonhston NA, Ogunbanjo BO, Ochei J. Plasmid profiles of Neisseria gonorrhoeae
in Nigeria and efficacy of spectinomycin in treating gonorrhoeae. Genitourin Med 1987; 63:1–5.
12. Van Dyck E, Rossau R, Duhamel M, et al. Antimicrobial susceptibility of Neisseria gonorrhoeae
in Zaire: high level plasmid-mediated tetracycline resistance in Central Africa. Genitourin Med 1992; 68:111–116.
13. Plourde PJ, Tyndall M, Agoki E, et al. Single-dose cefixime versus single-dose ceftriaxone in the treatment of antimicrobial resistant Neisseria gonorrhoeae
infection. J Infect Dis 1992; 166:919–922.
14. Van Dyck E, Laga M, Manoka AT, Behets F, Piot P. Epidemic spread of plasmid-mediated tetracycline resistant Neisseria gonorrhoeae
in Zaire. Int J STD AIDS 1995; 6:345–347.
15. Ghys P, Diallo M, Ettiègne-Traoré V, et al. Genital ulcers associated with human immunodeficiency virus-related immunosuppression in female sex workers in Abidjan, Côte d'Ivoire. J Infect Dis 1995; 172:1371–1374.
16. Laga M, Alary M, Nzila N, et al. Condom promotion, STD treatment leading to a declining incidence of HIV-1 infection in female Zairean sex workers. Lancet 1994; 344:246–248.
17. Hendry AT, Stuart IO. Auxanographic grouping and typing of Neisseria gonorrhoeae.
Can J Microbiol 1979; 25:512–521.
18. Knapp JS, Tam MR, Nowinski RC, Holmes KK, Sandström EG. Serological classification of Neisseria gonorrhoeae
with use of monoclonal antibodies to gonococcal outer membrane protein I. J Infect Dis 1984; 150:44–48.
19. Osoba AO. Overview of penicillinase producing Neisseria gonorrhoeae
in Africa. Afr J Sex Transm Dis 1986; 2:51–55.
20. Addy PAK. Susceptibility pattern of Neisseria gonorrhoeae
isolated at the Komfo, Anokye Teaching Hospital, Ghana, to commonly prescribed antimicrobial agents. East Afr Med J 1994; 71:368–372.
21. Abong BT, Fonkoua MC, Guibourdenche M, Riou JY, Ndayo WM, Garringue G. Analyse de 190 souches de Neisseria gonorrhoeae
isolées à Yaoundé de 1984 à 1987: auxotypes, contenus plasmidiques, sensibilité aux antibiotiques. Bull Soc Pathol Exot 1991; 84:136–144.
22. Geyid A, Tesfaye HS, Abraha A, Lemeneh Y, Desta S, Feleke W. Isolates of STD causative agents from sex workers Addis Ababa (a preliminary report). Ethiopian J Health Dev 1990; 4:155–162.
23. Dosso M, Kacou MD, Akona-Koffi, et al. Le niveau de résistance des souches de Neisseria gonorrhoeae
à Abidjan: un problème de santé publique. VIII International Conference on AIDS in Africa. Marrakech, December 12–16, 1993. Abstract M.O.P.30.
24. Haberberger RL, Fox E, Polycarpe D, Abbatte E, Salah S. Antibiotic susceptibility patterns of Neisseria gonorrhoeae
in Djibouti. Trans R Soc Trop Med Hyg 1990; 84:738.
25. Plummer FA, Simonsen JN, Chubb H, et al. Epidemiologic evidence for the development of serovar-specific immunity after gonococcal infection. J Clin Invest 1989; 83:1472–1476.
26. World Health Organization. Management of patients with sexually transmitted diseases. WHO Technical Report Series 810, 1991.
27. World Health Organization. The use of essential drugs. WHO Technical Report Series n114°850, 1995.
28. Roux R, Dosso M, Herouin P. Traitement de l'uréthrite aigue non compliquée à Abidjan. Médecine d'Afrique Noire 1991; 38:459–462.
29. Coovadia YM, Van Den Ende J, Hoosen A, Kharsany A. Susceptibility of penicillinase-producing and non-penicillinase-producing strains of Neisseria gonorrhoeae
isolated in Durban, South Africa, to 15 β-lactam antibiotics. Sex Transm Dis 1988; 15:30–34.
30. Brunham RC, Fransen L, Plummer F, et al. Antimicrobial susceptibility testing and phenotyping of Neisseria gonorrhoeae
isolated from patients with ophtalmia neonatorum in Nairobi, Kenya. Antimicrob Agents Chemother 1985; 28:393–396.