TEXT

Top Abstract Text References

Gonorrhea remains one of the most common sexually transmitted diseases in developing countries. Over the last decade, strains of Neisseria gonorrhoeae have been reported to develop high levels of resistance against several antimicrobial agents which have been used previously for the treatment of gonorrhea (5, 8). Since the discovery and subsequent increased incidence of penicillinase-producing N. gonorrhoeae (PPNG) isolates in 1976, there has been an increase in the emergence of N. gonorrhoeae strains resistant to tetracycline and spectinomycin (1, 2). These observations prompted health providers to replace these drugs with broad-spectrum cephalosporin and later with fluoroquinolones (3, 15).

Despite the rapid spread of gonococcal resistance throughout the Southeast Asia region, information on antimicrobial susceptibility in Indonesia is limited. Two recent studies from Surabaya (9) and Bandung (6), Indonesia, described substantial increases in resistance to several antibiotics of N. gonorrhoeae in a population at risk. However, antibiotic susceptibility of N. gonorrhoeae in Jakarta is not well documented. The aim of this study was to examine the antimicrobial susceptibility patterns of N. gonorrhoeae isolates from sex workers in Jakarta, Indonesia, especially their susceptibility against newer drugs such as ciprofloxacin and ceftriaxone.

Isolates of N. gonorrhoeae were obtained from female sex workers in North Jakarta in January 1996 during a survey conducted by the North Jakarta Municipal Health Service. Initial isolations were made on modified Thayer Martin agar (MTM; BBL Microbiology Systems, Cockeysville, Md.). Endocervical swabs obtained from the subjects were inoculated onto MTM plates and the plates were immediately placed in a candle extinction jar. Within 4 to 5 h the jars were transferred to an incubator, where they were held at 35°C for 24 to 48 h. Presumptive identification was made on the basis of colony morphology. Suspected colonies were Gram stained and examined microscopically, and oxidase and catalase tests were performed. Colonies of gram-negative diplococci positive for oxidase and catalase were transferred onto chocolate agar plates and incubated in a candle extinction jar at 35°C for 18 to 24 h. The overnight subcultures were tested for carbohydrate acid production (10). Confirmed N. gonorrhoeae isolates were placed in tryptic soy broth (Difco Laboratories, Detroit, Mich.) supplemented with 20% glycerol and frozen at 70°C until use for antibiotic susceptibility testing. Beta-lactamase production was tested by using nitrocefin disks (Cefinase; BBL Microbiology Systems), with Haemophilus influenzae ATCC 10211 as a negative control. Antimicrobial agents used in the tests were obtained from U.S. manufacturers. The antibiotics included penicillin G, tetracycline, cefotaxime, cefuroxime, ceftriaxone, kanamycin, gentamicin, thiamphenicol, chloramphenicol, and erythromycin (Sigma Chemical Company, St. Louis, Mo.), cefoxitin and norfloxacin (Merck Sharp and Dohme, West Point, Pa.), ciprofloxacin (Miles Pharmaceutical Inc., West Haven, Conn.), and spectinomycin (The Upjohn Co., Kalamazoo, Mich.). Antibiotic solutions were prepared according to the methods of NCCLS (13) and stock solutions were frozen at 70°C until use.

Antibiotic susceptibility testing was performed using the agar dilution method (13). The organisms used for quality control were Staphylococcus aureus ATCC 25923, Escherichia coli ATCC 25922, N. gonorrhoeae ATCC 49226, and two additional N. gonorrhoeae strains, F-28 (spectinomycin resistant) and F-45 (with chromosomally mediated resistance to penicillin and tetracycline) (Centers for Disease Control and Prevention, Atlanta, Ga.). Subcultures were incubated in a humidified atmosphere of 5% CO₂ for 24 h at 35°C. MICs were the lowest concentration of antibiotics that inhibited growth (14). For antibiotics for which NCCLS has no defined criteria for susceptibility or resistance, criteria reported in references 9, 11, and 12 were used. The criteria used for phenotypic characterization of N. gonorrhoeae, which were based on plasmid-mediated and chromosomally mediated resistance to penicillin and tetracycline as described previously (7, 8), are shown below in Table 3.

The data collected were evaluated using SAS statistical analysis software, restricted to a descriptive evaluation and intercorrelation of drug resistance findings. The Pearson correlation coefficient (r) for antibiotic resistance profiles was calculated at a 95% confidence level ( = 0.05).

N. gonorrhoeae was recovered from 122 (30.5%) of 400 subjects examined, and beta-lactamase was detected in 77 (63.1%). All beta-lactamase-positive isolates demonstrated a high level of penicillin resistance (MIC  2.0 µg/ml). Of the 45 beta-lactamase-negative isolates, 6 (13.3%) were resistant to penicillin (MIC  2.0 µg/ml), 20 (44.5%) demonstrated decreased susceptibility (MIC, 0.125 to 1.0 µg/ml), and the remaining 19 (42.2%) were still susceptible to penicillin (MIC  0.06 µg/ml). The MICs for 50% (MIC₅₀) and 90% (MIC₉₀) of isolates tested and the range of MICs are shown in Table 1 for each antibiotic tested, with the isolates separated into beta-lactamase positive and beta-lactamase negative. The MIC₅₀ and MIC₉₀ of penicillin for beta-lactamase-positive strains were 64- and 16-fold higher, respectively, than those for beta-lactamase-negative strains. The MIC₅₀s and MIC₉₀s of other drugs for beta-lactamase-positive and -negative strains did not show significant differences.

All N. gonorrhoeae isolates were susceptible to ciprofloxacin (MIC  0.06 µg/ml), cefuroxime (MIC  1.0 µg/ml), cefoxitin (MIC  2.0 µg/ml), cefotaxime (MIC  0.25 µg/ml), ceftriaxone (MIC  0.25 µg/ml), chloramphenicol (MIC  8.0 µg/ml), and spectinomycin (MIC  32 µg/ml). Four (3.3%) beta-lactamase-negative isolates were intermediately resistant to norfloxacin (MIC = 0.5 µg/ml) (Table 2), whereas beta-lactamase-positive isolates were all sensitive to norfloxacin. A total of 17 (13 beta-lactamase positive and 4 beta-lactamase negative) (13.9%) and 42 (32 beta-lactamase positive and 10 beta-lactamase negative) (34.4%) of the 122 N. gonorrhoeae isolates had decreased susceptibility against erythromycin (MIC  1.0 µg/ml) and thiamphenicol (MIC  1.0 µg/ml), respectively. Of the isolates less susceptible to thiamphenicol and erythromycin, 76.2% (32 of 42) and 76.4% (13 of 17), respectively, were found among the beta-lactamase-positive N. gonorrhoeae isolates. Thirty-eight (31.1%) isolates (15 beta-lactamase positive and 23 beta-lactamase negative) were resistant to kanamycin (MIC  16.0 µg/ml), and 36 (29.5%) isolates (25 beta-lactamase positive and 11 beta-lactamase negative) were resistant to gentamicin (MIC  16.0 µg/ml).

Based on the phenotypic categories for chromosome- and plasmid-mediated resistance of N. gonorrhoeae to penicillin and tetracycline (Table 3), only 1 (0.8%) of the 122 isolates was PPNG, 76 (62.3%) were PPNG and tetracycline-resistant N. gonorrhoeae (TRNG), 42 (34.4%) were TRNG, and 3 (2.5%) were chromosomally mediated tetracycline-resistant N. gonorrhoeae. Chromosomally mediated penicillin resistant and chromosomally mediated penicillin-and-tetracycline-resistant N. gonorrhoeae isolates were not detected.

The Pearson correlation coefficient analyses showed that isolates had parallel resistance profiles (PPNG and non-PPNG) to the different antibiotics tested. As an internal validity check, the resistance and susceptibility of N. gonorrheae isolates to penicillin were analyzed and found to be inversely correlated, as expected (r = 0.6, P < 0.05). Isolates susceptible to penicillin (non-PPNG strains) were generally resistant to kanamycin and tetracycline (r > 0.7, P < 0.05). Of this group, isolates resistant to kanamycin were also resistant to gentamicin (r > 0.7, P < 0.05). While resistance patterns of cefotaxime, ceftriaxone, and ciprofloxacin were similar (r = 0.9, P < 0.05), there was a correlation among the profiles of thiamphenicol, chloramphenicol (r = 0.9, P < 0.05), erythromycin (r = 0.6, P < 0.05), and cefoxitin (r = 0.8, P < 0.05).

While a report from Surabaya, East Java, Indonesia (9) described 84% PPNG-TRNG and 14% TRNG, we found a lower rate of PPNG-TRNG (62.3%) but a higher rate of TRNG (34.4%) in Jakarta. Although beta-lactamase-positive isolates are found at a higher frequency in Surabaya than in Jakarta, the numbers of N. gonorrhoeae isolates resistant to tetracycline are similar in the two places; 100% of the strains were highly resistant to tetracycline, which indicates selective pressure resulting from tetracycline's use for sexually transmitted disease prevention or treatment and for other illnesses.

No chromosomally mediated penicillin- or penicillin-and-tetracycline-resistant N. gonorrhoeae isolates were found in this study. These results are remarkable since regular administration of penicillin for prevention of syphilis is still a common practice among the sex workers in most places in Indonesia.

The broad-spectrum cephalosporins (cefotaxime, ceftriaxone, cefuroxime, and cefoxitin) demonstrated very high levels of activity (100% susceptibility) against N. gonorrhoeae isolates. Of these antibiotics, cefotaxime and ceftriaxone showed the highest activity, with MICs of 0.016 µg/ml, while cefuroxime and cefoxitin MICs were 0.5 and 2.0 µg/ml, respectively. Thiamphenicol is still one of the antibiotics used for treatment of gonorrhea in Indonesia; however, reports from Bandung and Surabaya (6, 9) indicate relatively high rates of less susceptible and resistant N. gonorrhoeae isolates. Lind (12) described penicillinase-producing strains that were significantly more susceptible to thiamphenicol than were non-penicillinase-producing strains, which is contrary to our findings that 76.2% of all N. gonorrhoeae isolates resistant to thiamphenicol (MIC  1.0 µg/ml) were beta-lactamase positive. It appears that strains of N. gonorrhoeae from Jakarta which were resistant to thiamphenicol have a greater correlation with PPNG or PPNG-TRNG than with chromosomally mediated resistance to penicillin or tetracycline.

In contrast to the report published by Joesoef et al. in the Surabaya survey (9), we documented a decreased susceptibility to erythromycin and gentamicin, but our results with respect to kanamycin susceptibility are in agreement with the earlier findings. With respect to spectinomycin susceptibility, our result is contradictory to that from Surabaya (9). All Jakarta N. gonorrhoeae isolates were susceptible to spectinomycin, whereas 18.1% of Surabaya isolates were resistant. Despite widespread use of the broad-spectrum cephalosporins to treat gonorrhea, 100% of N. gonorrhoeae isolates were susceptible to this group of drugs in this study (cefuroxime MIC  1.0 µg/ml; cefotaxime MIC  0.5 µg/ml; cefoxitin MIC  2.0 µg/ml; and ceftriaxone MIC  0.25 µg/ml). Although this finding is consistent with previous reports (3, 4, 9), a small number of N. gonorrhoeae isolates from Asia (3, 4, 15) reportedly have decreased susceptibility to some members of this group of drugs. All antibiotics of the quinolone group were highly active against N. gonorrhoeae except for norfloxacin, to which 3.3% of isolates demonstrated intermediate susceptibility (Table 2).

These data may be useful in predicting the antibiotic susceptibility of N. gonorrhoeae and suggest the use of certain drugs in the treatment of infection, particularly in the absence of laboratory facilities and the long time spans required for conventional microbiological analysis (symptomatic therapy).