The fluoroquinolones represented an important advance in antigonococcal therapy. They have a unique mechanism of action and therefore have little cross-resistance with other antibiotic agents. Single oral doses of these drugs are effective against uncomplicated gonorrhea of any anatomical site, are well tolerated, are relatively inexpensive, and were initially active against almost all gonococci, including those with previously documented types of resistance [2, 4]. From the beginning, however, theoretical concerns were raised about the potential of gonococci to develop resistance to fluoroquinolones. Resistance of N. gonorrhoeae and other pathogens could be readily induced in the laboratory, and clinical resistance of some previously susceptible pathogens was documented almost immediately [5, 6]. Gonococcal strains with reduced susceptibility to one or more fluoroquinolones were first isolated from patients in the late 1980s. The minimal inhibitory concentrations (MICs) of ciprofloxacin typically were 0.125 to 0.5 µg/mL [7, 8], in contrast to fully susceptible strains, for which the MICs are 0.001 to 0.06 µg/mL [1, 2, 9]. Subsequently, strains with such "low-level" resistance were reported in the Far East, Europe, Australia, Africa, North America, and Japan [9, 10], but treatment failures remained rare, as predicted by the activity and pharmacokinetics of the recommended drugs. More recently, strains of N. gonorrhoeae with high-level resistance to fluoroquinolones (characterized by MICs of ciprofloxacin of 1 to 16 µg/mL) have been documented in several geographic areas [9-13], and treatment failures have been reported [12, 13]. Although controlled therapeutic trials have not been done, data from these case reports and the pharmacokinetics characteristics of even the most active agents suggest that no fluoroquinolone is likely to be reliable as single-dose therapy for infection due to gonococci with high-level resistance.

The Gonococcal Isolate Surveillance Project, conducted by the CDC in 24 U.S. cities, has documented an increasing prevalence of gonococci that have decreased susceptibility to fluoroquinolones; the percentage of isolates that were susceptible increased from 0.3% in 1991 to 1.3% in 1994 (P < 0.001) [11]. In this issue, Gordon and colleagues [14] describe a large outbreak of infections caused by N. gonorrhoeae with low-level resistance to fluoroquinolones (MIC, 0.125 to 0.25 µg/mL) in Cleveland, Ohio, where the percentage of all gonococci that were composed of such strains increased from 2% to 16% during a 3-year period. The investigators also showed that the outbreak resulted from the spread of a single gonococcal strain or of closely related strains; this confirmed results of their earlier preliminary analysis [15]. Few, if any, cases of infection were treated with fluoroquinolones, and no treatment failures were documented.

The most important implication of the Cleveland experience is that the potential may exist for the evolution and spread of gonococci with high-level resistance to fluoroquinolones. Experimental data suggest that second and subsequent mutations for relative resistance are additive and can increase the MIC of ciprofloxacin to 4 µg/mL or more [16]. Thus, the risk for the development of high-level resistance may be proportional to the prevalence of low-level resistance. Another recent report [11] has documented the first cases of infection due to N. gonorrhoeae that had high-level resistance to fluoroquinolones in the United States. In Seattle, Washington, a single strain accounted for eight reported cases (the true total was undoubtedly higher), and in Denver, Colorado, a patient with a highly resistant isolate remained infected after being treated with 400 mg of ofloxacin.

The epidemiology of fluoroquinolone-resistant gonorrheae is reminiscent of that seen when penicillinase-producing N. gonorrhoeae (PPNG) first appeared. Soon after the first cases of infection with such strains were recognized in Africa and the Philippines in the mid-1970s, sporadic cases and small clusters of cases occurred in the United States [17]. It took several years for PPNG to become permanently established, but thereafter, spread was rapid. By 1990, PPNG accounted for 12% of cases of gonorrhea in the United States [18]. It therefore seems likely that gonococci with high-level resistance to fluoroquinolones will eventually become prevalent in the United States and that fluoroquinolones may cease to be useful for routine treatment.

What, then, are the "take home" messages for clinicians and public health authorities? First, fluoroquinolones should not yet be abandoned as therapy for gonorrhea. To date, high-level resistance to fluoroquinolones remains rare, and the Gonococcal Isolate Surveillance Project should give ample warning before such strains appear in substantial numbers. Prudence suggests that cefixime, 400 mg given orally, or ceftriaxone, 125 mg given intramuscularly [3], should be considered the regimens of choice, but the lower cost of ciprofloxacin is a valid consideration. Some might argue that fluoroquinolones should be held in reserve to delay the selection of resistance. However, this might have little effect because widespread use of fluoroquinolones for other infections may create more selection pressure for resistance in N. gonorrhoeae than does their use as therapy for gonorrhea. Fluoroquinolones should not be used to treat gonorrhea in travelers from Asia or Western Pacific countries, where high-level resistance may be most prevalent [10].

Second, doses of fluoroquinolones smaller than the recommended doses (500 mg of ciprofloxacin or 400 mg of ofloxacin) should not be used to treat gonorrhea. Some authors have advocated using a 250-mg dose of ciprofloxacin, and, indeed, the cure rates with this dose to date appear to be as high as those seen with the 500-mg dose [19]. However, few patients with gonorrhea caused by strains of N. gonorrhoeae with low-level resistance to fluoroquinolones were included in the reported treatment studies, and the lower dose may enhance the risk for treatment failure in such patients. This, in turn, may increase selection pressure for resistance. Moreover, 500-mg ciprofloxacin tablets cost only slightly more than the 250-mg tablets and appear to be equally well tolerated, so there is no compelling reason to use the lower dose.

Third, clinicians should routinely follow single-dose treatment for gonorrhea (whether with a cephalosporin or a fluoroquinolone) with azithromycin, 1.0 g in a single dose, or doxycycline, 100 mg twice daily for 7 days [3]. This is advised mainly because many patients with gonorrhea are also infected with Chlamydia trachomatis. However, administering a second antibiotic agent with a mechanism of action different from that of the primary drug may help retard the selection of resistant gonococci. Although neither azithromycin nor doxycycline is recommended as sole treatment, the antimicrobial susceptibilities reported by Gordon and colleagues [14] and others [2, 4] suggest that either drug would cure most patients infected with gonococci that are resistant to the primary drug.

Clinicians who frequently treat patients with gonorrhea should be aware of patterns of susceptibility in their communities, and health departments must be prepared to rapidly disseminate the information when high-level resistance to fluoroquinolones appears. Clinical trials are needed to assess the efficacy of fluoroquinolone therapy against infection with strains of N. gonorrhoeae that have low-level or high-level resistance. Finally, basic and applied research into new classes of antimicrobial drugs for gonorrhea are greatly needed. History shows that the gonococcus always wins, which makes it crucial to stay a step ahead.