High Prevalence of High-Level Ciprofloxacin Resistance in Neisseria gonorrhoeae in Tel Aviv, Israel: Correlation with Response to Therapy

Abstract

A survey of the drug susceptibilities of gonococcal isolates from 100 consecutive patients attending clinics in the Tel Aviv area of Israel during the period from February to September 2000 has shown a 61% rate of ciprofloxacin resistance (MICs, ≥1 μg/ml); 51 isolates were highly resistant (MICs, ≥4 μg/ml). Ciprofloxacin-resistant strains were more prevalent among isolates with chromosomally mediated resistance to penicillin and tetracycline. Therapeutic failures with ciprofloxacin correlated with MICs of ≥1 mg/ml (P < 0.01). The high rate of ciprofloxacin resistance coinciding with a sharp rise in the incidence of gonorrhea precludes the use of fluoroquinolone drugs for the empirical treatment of gonococcal infections acquired in Israel.

Gonorrhea was rarely reported in Israel in the 1990s, with an average annual incidence of 0.9 cases per 100,000 population (14). As most laboratories did not carry appropriate media, diagnosis was usually made on the basis of Gram-stained smears, and susceptibility testing of Neisseria gonorrhoeae was not routinely performed. More recently, however, the incidence of gonorrhea has been increasing and was up to 13 reported cases per 100,000 population in 2001 (10). The documentation of a case of gonococcal urethritis with clinical and laboratory failure of ciprofloxacin therapy (9) prompted a study of the antimicrobial susceptibilities of N. gonorrhoeae isolates in the Tel Aviv area of Israel.

Gonococcal strains with reduced in vitro susceptibilities to fluoroquinolones (MIC of ciprofloxacin, 0.125 to 0.5 μg/ml) were first described in the mid-1980s (5) and since then have been identified in the Far East (7, 8, 20), Australia (21, 22), Africa (3), Europe (12, 13, 23), and the United States (4). Fluoroquinolone-resistant N. gonorrhoeae (ciprofloxacin MICs, ≥1 μg/ml) emerged during the early 1990s and became well established in several Asian countries (16). Strains for which ciprofloxacin MICs were ≥8 μg/ml were first isolated in 1994 (15) and have mostly been detected in the Far East (1, 24). Two cases of gonorrhea caused by strains for which the MIC was 16 μg/ml were recently reported in the United States (5). These changes in the susceptibility patterns of N. gonorrhoeae may lead to the reassessment of the current recommendations for empirical therapy of gonococcal infection (6). Continuous surveillance of the epidemiology of N. gonorrhoeae based on type characterization and detailed in vitro testing of the antimicrobial susceptibilities of isolates are imperative for establishing empirical treatment guidelines.

MATERIALS AND METHODS

Source of isolates.

N. gonorrhoeae strains were isolated from clinical specimens (mostly urethral discharges from male patients) at the Central Laboratory of Maccabi Health Services, Rishon-le-Zion, Israel. The Central Laboratory serves a population of 400,000 in the Tel Aviv area, situated in central Israel. All isolates recovered from February 2000 and on were transferred to the Infectious Diseases Research Laboratory at the E. Wolfson Hospital, Holon, Israel. The first 100 successive isolates were included in the study.

Isolation and identification procedures.

Swabs containing urethral discharge were streaked onto New York City agar (Hy Laboratories, Ltd., Rehovot, Israel). The plates were incubated in 5% CO₂ for up to 72 h at 35°C. N. gonorrhoeae was identified by colony morphology, Gram stain, oxidase and catalase tests, and sugar fermentation tests. The identities of the isolates were further confirmed with the API-NH system for the identification of Neisseria and Haemophilus (bioMerieux, Lyon, France). Isolates were kept at −70°C in Trypticase soy broth with 15% glycerol until they were studied further.

Susceptibility determination.

The MICs of penicillin, tetracycline, ceftriaxone, and ciprofloxacin were determined by the Etest method (AB Biodisk, Solna, Sweden), as specified in the product's package insert provided by the manufacturer. The Etest constitutes a reliable alternative to the agar dilution method for antimicrobial susceptibility testing of gonococci (2). An inoculum density of a 0.5 McFarland standard was prepared in Mueller-Hinton broth directly from a fresh overnight subculture and was applied to the entire surface of the culture plates containing GC agar base and 1% defined growth supplement (Hy Laboratories, Ltd.). After the Etest strips were applied, the plates were incubated in 5% CO₂ at 35°C for 24 h. N. gonorrhoeae ATCC 49226, for which the MICs of the antimicrobial agents tested were known, was used as the reference strain. The antimicrobial susceptibility was judged by use of the breakpoint criteria defined by the National Committee for Clinical Laboratory Standards (NCCLS) (17). Susceptibility to spectinomycin was determined by the disk diffusion method with 100-μg disks (Becton Dickinson Microbiology Systems, Sparks, Md.).

β-Lactamase testing.

Isolates were tested for β-lactamase production by the cefinase test (BBL, Becton Dickinson Microbiology Systems).

Phenotypic characterization.

The criteria used for phenotypic characterization of the N. gonorrhoeae isolates based on plasmid- and chromosomally mediated resistance to penicillin and tetracycline were as follows: for penicillinase-producing N. gonorrhoeae (PPNG) isolates, tetracycline MICs were <16 μg/ml and the isolates were β-lactamase positive; for N. gonorrhoeae isolates with plasmid-mediated tetracycline resistance (TRNG), tetracycline MICs were ≥16 μg/ml and the isolates were β-lactamase negative; for PPNG-TRNG isolates, tetracycline MICs were ≥16 μg/ml and the isolates were β-lactamase positive; for N. gonorrhoeae isolates with chromosomally mediated penicillin resistance (CMRNG^P), penicillin MICs were ≥2 μg/ml, tetracycline MICs were < 2 μg/ml, and the isolates were non-PPNG; for N. gonorrhoeae isolates with chromosomally mediated tetracycline resistance (CMRNG^T), penicillin MICs were <2 μg/ml, tetracycline MICs were 2 to 8 μg/ml, and the isolates were non-PPNG; for N. gonorrhoeae isolates with chromosomally mediated resistance to both penicillin and tetracycline (CMRNG^PT), penicillin MICs were ≥2 μg/ml, tetracycline MICs were ≥2 μg/ml, and the isolates were non-PPNG and non-TRNG; and for N. gonorrhoeae isolates not resistant to both penicillin and tetracycline (nonresistant^PT), penicillin and tetracycline MICs were <2 μg/ml (19).

Response to therapy.

Data on the therapeutic regimens administered and the responses to therapy were obtained from the treating physicians and the patients. The statistical significance of the difference in outcomes between infections caused by ciprofloxacin-susceptible strains and those caused by ciprofloxacin-resistant strains treated with a fluoroquinolone was calculated by Fisher's exact test.

RESULTS

The susceptibilities of 100 gonococcal isolates are shown in Table 1. On the basis of the susceptibility criteria established by NCCLS (17), 16 isolates were resistant (MICs, ≥2 μg/ml) to penicillin and 8 were penicillinase producers (PPNG) for which the penicillin MICs ranged from 1 to 64 μg/ml; 8 isolates, none of which was TRNG, were resistant (MICs, ≥2 μg/ml) to tetracycline; and 61 isolates were resistant (MICs, ≥1 μg/ml) and 3 isolates had reduced susceptibilities (MICs, 0.125 to 0.5 μg/ml) to ciprofloxacin. The distribution of isolates according to their susceptibilities to ciprofloxacin is depicted in Fig. 1. According to the NCCLS breakpoints (17), 36% of the isolates were susceptible to ciprofloxacin (MICs, ≤0.06 μg/ml), 3% were intermediate (MICs, 0.125 to 0.5 μg/ml), and 61% were resistant (MICs, ≥1 μg/ml). Fifty-one isolates demonstrated high-level resistance to ciprofloxacin (MICs, ≥4 μg/ml). No isolate was resistant to ceftriaxone or spectinomycin. Twenty-two and 5% of the ciprofloxacin-susceptible isolates were resistant to penicillin and tetracycline, respectively, while 36 and 38% of ciprofloxacin-resistant organisms were also resistant to penicillin and tetracycline, respectively.

TABLE 1.

Antimicrobial susceptibility patterns of 100 N. gonorrhoeae isolates in Tel Aviv, Israel

Antimicrobial agent	MIC (μg/ml)a			Distribution (%)
	50%	90%	Range	Susceptible	Intermediate	Resistant
Penicillin	1.0	2.0	0.016-64.0	11	73	16
Tetracycline	1.0	2.0	0.032-2.0	26	66	8
Ciprofloxacin	4.0	8.0	0.002-16.0	36	3	61
Ceftriaxone	0.016	0.032	0.016-0.032	100	0	0
Spectinomycin	NAb	NA	NA	100	0	0

^a50% and 90%, MICs at which 50 and 90% of isolates are inhibited, respectively.

^bNA, not applicable; the disk diffusion test was used.

FIG. 1.

Distribution of gonococcal isolates according to ciprofloxacin MICs.

The phenotypic distributions of the N. gonorrhoeae isolates and the ciprofloxacin resistance rates among the isolates in the different phenotypic categories are shown in Table 2. All isolates with chromosomally mediated resistance to penicillin and/or tetracycline and only 1 (12.5%) of the penicillinase-producing isolates were resistant to ciprofloxacin.

TABLE 2.

Distribution of 100 consecutively isolated gonococcal strains according to phenotypic categories and correlation with ciprofloxacin resistance

Phenotypic category	No. of Isolates	No. (%) of isolates resistant to ciprofloxacin
PPNG	8	1 (12.3)
TRNG	0
PPNG-TRNG	0
CMRNGP	5	5 (100)
CMRNGT	5	5 (100)
CMRNGPT	3	3 (100)
NonresistantPT	79	47 (59.5)

The patient population consisted of 98 men and 2 women with a mean age of 32.5 ± 8.5 (standard deviation) years. Information on empirical therapy was available for 81 patients; 32 individuals were treated with a fluoroquinolone (ciprofloxacin or ofloxacin), and information on the clinical response to therapy was available for 21 of them. The gonococcal isolates from 15 patients were resistant to ciprofloxacin (MICs, ≥1 μg/ml), and clinical failure was documented in all patients. The N. gonorrhoeae isolates from six patients were susceptible to ciprofloxacin (MICs, ≤0.06 μg/ml), and all but one of the patients responded favorably to fluoroquinolone therapy. In the only clinical failure in this group, coinfection with Chlamydia trachomatis cannot be excluded since the patient subsequently responded to doxycycline therapy. The difference in the therapeutic outcomes between the two groups was statistically significant (P < 0.01).

DISCUSSION

Although susceptibility testing was not routinely performed in Israel in the 1990s, limited unpublished data indicate that isolates were invariably susceptible to fluoroquinolones, and clinical failure following fluoroquinolone therapy was not reported. The present survey of gonococcal isolates from 100 consecutive patients attending clinics in the Tel Aviv area during 2000 has shown a very high rate of high-level ciprofloxacin resistance coinciding with a sharp rise in the incidence of gonorrhea. Reduced susceptibility (or low-level resistance) was rare (3%), and the distribution of the isolates according to the ciprofloxacin MICs showed two peaks in the MICs, at 0.002 to 0.008 μg/ml and at 2 to 8 μg/ml (Fig. 1). The grouping of the MICs for the isolates within a narrow range may suggest a single pattern of mutations that developed or that was introduced over a short period of time rather than a sequential accumulation of point mutations. A similar pattern of close relatedness between isolates with reduced susceptibility to ciprofloxacin was observed by Kilmarx et al. (15), who investigated the spread of nonsusceptible gonococci in Cleveland, Ohio. Resistance to ciprofloxacin is chromosomally mediated and affects susceptibility to all other fluoroquinolones (1). Our findings confirm previous observations (1, 11, 16) that ciprofloxacin-resistant strains are more prevalent among isolates with chromosomally mediated resistance to penicillin and tetracycline. This association remains unexplained because the mechanism of resistance to ciprofloxacin is different from the mechanisms of resistance to penicillin and tetracycline. Although our data on the correlation between the MIC and the therapeutic response were not the product of a planned intervention trial, they are nevertheless highly demonstrative and confirm the association of therapeutic failure and infection with isolates for which ciprofloxacin MICs are ≥1 μg/ml. Because of the low number of patients infected with intermediate strains, none of whom was treated with a fluoroquinolone, we could not evaluate the response to treatment in this group. Very similar results were obtained in a recent prospective study from Bangladesh (18).

Since the introduction of sulfonamides in the 1930s, N. gonorrhoeae isolates have developed resistance to many of the antibiotics used in the treatment of gonococcal infections, and the fluoroquinolones are the latest to join the list. As all isolates in the present study were highly susceptible to ceftriaxone and also to spectinomycin, the former is now recommended together with cefixime as the first-line drug for the empirical treatment of gonorrhea acquired in Israel. Spectinomycin is the best alternative for patients allergic to cephalosporins.