Campylobacter and fluoroquinolones: a bias data set?

Abstract

There is no universally accepted standard method for the isolation of Campylobacter spp. and it is considered that currently available isolation media are not yet optimal for the recovery of Campylobacter spp. from a range of sample types. Almost all methods incorporate antibiotics into the isolation media to inhibit growth of other bacteria within the sample. It is established that the incorporation of such antibiotics into isolation media will inhibit the growth of some Campylobacter spp. as well as other bacteria. The results of the use of such suboptimal isolation methods are that the isolates which 'survive' the isolation procedure will be those which: (i) are able to 'out compete' the rest of the bacteria in the sample, i.e. they are able to grow faster; (ii) are resistant to the antibiotics used in the isolation media; and (iii) are randomly selected by the laboratory technician as being a 'typical'Campylobacter spp. It is clear that such a procedure is intrinsically biased and will mean that species resistant to the antibiotics used in the media will be isolated. This introduces real doubt that the bacteria isolated are truly representative of those initially found on the sample. It is also becoming clear that Campylobacter spp. are rather difficult to isolate as pure cultures and many are in fact mixtures of more than one strain. Again this introduces great uncertainty as to the prevalence and distribution of respective species from the different sample types. This is especially true when considering isolation of Campylobacter spp. causing disease in man as there is no certainty that the selected isolate is that which was responsible for disease. The incorporation of antibiotics into the isolation media not only introduces the issue of species bias but perhaps more importantly exposes the Campylobacter spp. to a cocktail of antibiotics thereby providing the potential for them to 'switch on' antibiotic resistance mechanisms. It might be argued that this has always been the case for isolation of Campylobacter spp., however, we know that the antibiotic cocktails used in media over the last 10 years have changed and indeed there was a time when the filtration protocol which didn't use antibiotics was more widely used. As most reports in the literature do not state what methods were used to isolate Campylobacter spp. it is not possible to quantify any relationship between antibiotics used in the isolation media and susceptibility data. An approved method for Campylobacter susceptibility testing was not available until May 2002, all data generated prior to this date will have been generated using non-standard methods. As tremendous variability in the reproducibility data for Campylobacter spp. was observed during the development of the standard agar dilution susceptibility method, data generated with disk diffusion and broth microdilution methods must be considered with caution. It has been shown that, compared with the conventional agar dilution method, the E-test tends to give rise to lower minimal inhibitory concentrations (MICs) for sensitive strains and higher MICs for resistant strains. There are no recommended antibiotic breakpoint concentrations for Campylobacter spp. A breakpoint is used to separate sensitive from resistant strains of bacteria and is thus crucial to any discussion of antibiotic resistance. This discussion is further complicated by introduction of the terms microbiological and clinical breakpoints. While a microbiological breakpoint can be a useful parameter with regard to identifying resistance factors it cannot on its own be used to predict whether that bacteria will respond to treatment from an appropriate antibiotic. Predicting clinical response is a function of the clinical breakpoint which considers the pharmacokinetic profile of the antimicrobial compound, i.e. the concentration of the antimicrobial compound in the body and the MIC. The National Committee for Clinical Laboratory Standards (NCCLS) uses microbiological, pharmacokinetic and clinical data to establish breakpoints, without c and clinical data to establish breakpoints, without such considerations it is not possible to consider what is truly clinically sensitive and resistant. There are no reported studies that have systematically determined appropriate breakpoints for Campylobacter, there are data however, which relate MICs to clinical outcome. It is without dispute that microbiological resistance in Campylobacter spp. occurs as a result of mutation in the gyrA gene with single point mutations most frequently causing a four- to eightfold shift in the MIC. What is also clear is that if a high enough concentration of antimicrobial relative to MIC of the infecting organism can be achieved not only will the parent organism be killed but also the 'resistant' mutant. Considering the above and the concentrations of ciprofloxacin achieved in the gastro-intestinal tract it is not surprising that clinical cure can be demonstrated for organisms with an MIC of 32 microg ml(-1).