• Antimicrobial resistance
• gonorrhoea
• gentamicin
• cephalosporin
• treatment
• PID
• genitourinary medicine services
• syndrome management
• resistance

The pattern of antimicrobial resistance in Neisseria gonorrhoeae is depressingly predictable. An antibiotic is chosen and used for a few years but resistance develops, so the dose is increased to maintain efficacy before the drug finally fails irrevocably. For penicillin, this process took a number of decades but with the subsequent use of tetracyclines, macrolides and fluoroquinolones, the cycle has shortened, and the utility of oral cephalosporins is now threatened within a few short years of their introduction as preferred therapy in many countries.1

Following the emergence of N gonorrhoeae strains with decreased susceptibility to oral cephalosporins, clinical treatment failures soon appeared in Japan and Hong Kong;2–4 more recently, similar outcomes have reported from other regions of the world, including Europe. There still exists some debate as to the minimum inhibitory concentration (MIC) breakpoint for oral cephalosporins that correlates with clinical failure. However, data from Deguchi et al suggest that clinical failures may occur with MICs of 0.125 mg/l or higher.3 Pharmacodynamic modelling with Monte Carlo simulations highlighted the fact that, for reliable efficacy, cephalosporins require a free drug level above the MIC for 20–24 h (compared to 7–10 h in the case of penicillin G).5 These analyses predicted that failures with standard doses of cefixime (400 mg) and ceftriaxone (250 mg) become likely around MICs of 0.125 mg/l and 0.25 mg/l respectively. The clinical practice of treating patients with gonorrhoea with an additional agent for presumptive Chlamydia trachomatis infection—particularly azithromycin and, to a lesser extent, doxycycline—may have prevented the appearance of treatment failures in those individuals infected with N gonorrhoeae strains possessing high cephalosporin MICs.

The mechanism of resistance to oral cephalosporins involves changes in the structure and function of a number of key proteins, notably the penA-encoded penicillin binding protein 2 and the MtrC-MtrD-MtrE multiple transferable resistance efflux pump. The latter is important as those strains with decreased susceptibility to oral cephalosporins also have high MICs to penicillins, tetracyclines and macrolides.2 Gonococci tend to retain resistance phenotypes and genotypes to those antibiotics used earlier for treatment, even when their use has been discontinued, which generally rules out reintroducing them in the current era of multidrug-resistant gonorrhoea.6 It is, therefore, not too surprising to note that the majority of circulating gonococcal strains with decreased susceptibility to oral cephalosporins are also resistant to fluoroquinolones.2

The world's first extensively drug-resistant N gonorrhoeae strain (H041), defined by criteria put forward by Tapsall et al, was recently described in detail.6 7 This H041 strain has a ceftriaxone MIC of 2–4 mg/l, which is fourfold to eightfold higher than previously reported, as well as a cefixime MIC of 8 mg/l. In addition, the H041 strain was determined to be resistant to β-lactams (with the possible exception of carbapenems and piperacillin-tazobactam, for which breakpoints are not available), tetracyclines, macrolides, fluoroquinolones, trimethoprim-sulphamethoxazole and chloramphenicol. The same strain was susceptible to both rifampicin and spectinomycin, and had low MICs to gentamicin and kanamycin, although these MIC values were difficult to interpret as breakpoints do not exist for these two antimicrobial agents.

If we are facing the imminent failure of cephalosporins for treating gonorrhoea, then what are the options for the future and how can we use antibiotics more effectively to prolong their efficacy?

Combining two or more antibiotics together is one logical approach, if slightly more expensive and with a greater potential to cause side effects than monotherapy. A combination of drugs generally provides wider antimicrobial coverage and is particularly valuable when giving empirical treatment before the resistance pattern for an organism is known, as is often the case when treating gonorrhoea. Unfortunately, the number of effective drugs available is now extremely limited, but azithromycin is frequently combined with ceftriaxone to cover possible coinfection with chlamydia8–10 and this combination may also be more effective against N gonorrhoeae.8

We have been fortunate in the past in having effective single dose treatments for gonorrhoea which were simple for patients to take and could be directly observed to ensure adherence, but the drug levels attained with this approach and the length of time that they are maintained above the organism's MIC are necessarily limited. Increasing the drug dose or giving multiple doses will increase the length of time that the drug concentration exceeds the MIC and will increase the peak MIC concentration, and thus may extend the utility of individual drugs. This was the rationale for the recent increase in the recommended dose of ceftriaxone to 500 mg in the UK's gonorrhoea treatment guidelines.8

Another option to prevent the emergence of resistance would be to regularly rotate the antibiotics which are recommended for empirical therapy, thus reducing the selective pressure arising from any single agent. However, this would present a number of practical problems around coordinated implementation, and remains of unproven value in preventing the emergence of resistance.

Whatever approach is taken, we need to identify alternative antibiotics which can be used in patients with gonorrhoea who either fail or are intolerant to cephalosporin treatment, and have coexisting resistance to penicillin, fluoroquinolones, tetracyclines and macrolides. Gentamicin has been used in many developing countries, and has the advantages of low cost and, at least anecdotally, high efficacy. What place should it have in our therapeutic repertoire?

Gentamicin is an aminoglycoside antibiotic with concentration-dependent bactericidal activity against Gram negative bacteria. It binds at the 30S bacterial ribosomal subunit, and results in mRNA misreading during protein synthesis, permeabilisation of the cell membrane, inhibition of the initiation of DNA replication and loss of cell viability.11 As aminoglycosides are hydrophilic and highly lipophobic, they are rapidly distributed after parenteral injection but are poorly absorbed from the gastrointestinal tract. The peak concentration of an intramuscular dose is seen in about 30–90 min. Gentamicin is excreted almost exclusively through the kidney and very little is reabsorbed after glomerular filtration. Gentamicin, like other aminoglycosides, exhibits a post-antibiotic effect (PAE) in terms of suppressing bacterial growth for a period of time after drug administration. During the period of PAE, bacteria are less likely to take up further amounts of the drug, which is one of the arguments in favour of dosing once a day.

Several mechanisms have been proposed for aminoglycoside resistance, including decreased antibiotic uptake and accumulation, modification of the ribosomal target, efflux and enzymatic modification.11 The major mechanism of resistance in most bacteria is enzymatic modification of aminoglycosides, with the resultant effect of impaired ribosomal binding. However, based on early studies, this does not appear to be the mechanism in N gonorrhoeae. Within gonococci, resistance has been observed in the case of kanamycin (aminoglycoside) and spectinomycin (aminocyclitol). Resistance appears to be due to mutations in linked loci which result in altered sensitivity of the 30S ribosomal subunit of N gonorrhoeae.12 To the best knowledge of the authors, there are no published reports of confirmed clinical resistance to gentamicin.

The efficacy of gentamicin has been reported in a number of small observational and controlled (although not randomised) trials. Most of the studies date from the 1970s and 1980s, and many are methodologically weak by modern comparison.13–22 The most common dose of gentamicin used was 240 mg given once by intramuscular injection, which equates to about 6 ml of fluid. This volume is near to the maximum that patients can tolerate as a single injection, so the choice of a higher dose would probably require two separate injections. Clinical and microbiological cure rates range from 89% to 100% following gentamicin, although, in one study, a lower cure rate of around 65% was reported in the gentamicin and comparator arms.20 Reassuringly, no adverse events were reported in these studies.

Perhaps the most important data concerning long-term use of gentamicin to treat gonorrhoea come from Malawi, where a single intramuscular dose of gentamicin (240 mg) was adopted in 1993 as the preferred antibiotic to treat presumptive gonorrhoea within the context of syndrome management treatment algorithms. Although regular gonococcal antimicrobial susceptibility surveys have not been undertaken on a nationwide basis in Malawi during the last 18 years, there are periodic data from Lilongwe for microbiological MIC determinations among N gonorrhoeae isolates and clinical response.13 23 The last published survey from 2007 confirms the continued susceptibility of gonococci to gentamicin, with all isolates having gentamicin MIC values of ≤4 mg/l. In comparison, slightly higher gentamicin MICs have been recently reported from a multinational European survey (MIC range: 1–16 mg/l; modal MIC: 8 mg/l).23 24 Brown et al reported the challenge they faced in trying to obtain clinical in vitro correlates within Malawi, although such an approach was not possible in the European survey as gentamicin is not used routinely for treatment of gonorrhoea in the participating countries.23 Such correlates are urgently needed for gentamicin and should be the focus of further research.

The main toxicity of gentamicin relates to the kidney and the inner ear, where gentamicin concentrations may be higher, but the potential for damage is usually linked to the duration of exposure above a toxicity threshold (traditionally measured by trough monitoring). The use of a relatively low single intramuscular dose reduces the potential for harm to occur. The risk of ototoxicity following aminoglycoside exposure is also higher in those with pre-existing hearing problems or impaired renal function, and a ribosomal RNA polymorphism (A1555G) has been associated with ototoxity, although the linkage is not strong enough to be used as a screening test.25

Further information on the efficacy and safety of gentamicin for the treatment of gonorrhoea will emerge from an ongoing randomised comparative trial of gentamicin plus azithromycin against gemifloxacin plus azithromycin (http://clinicaltrials.gov, accessed 25 August 2011) but this study is not expected to report fully for a further 2–3 years. In terms of a research agenda, especially given the importance of extragenital gonorrhoea in the emergence of novel gonococcal antimicrobial resistance determinants, it will also be important to determine the effectiveness of gentamicin to cure pharyngeal or rectal gonococcal infections.

In summary, the available options for treating multi-resistant N gonorrhoeae remain limited. However, the efficacy and safety data we currently possess for gentamicin, although several decades old, are encouraging and gentamicin's use as a single intramuscular dose of 240 mg would be appropriate when there is resistance, allergy or intolerance to penicillins, macrolides, tetracyclines and fluoroquinolones.