European harmonization of MIC breakpoints for antimicrobial susceptibility testing of bacteria

Abstract

The success or failure of antimicrobial therapy in bacterial and fungal infections is predicted ideally by antimicrobial susceptibility testing (AST), in which microorganisms are divided into treatable and non-treatable categories on the basis of MIC breakpoints. In Europe, the categorization was traditionally a clinical one and it was made irrespective of whether or not the organism harboured resistance mechanisms. MIC breakpoints generally divide bacteria into three categories of susceptibility: susceptible, intermediate or indeterminate, or resistant. These terms can be defined as susceptible (S—where the antimicrobial activity is associated with a likelihood of therapeutic success), intermediate or indeterminate (I—where the antimicrobial activity is associated with an indeterminate or uncertain therapeutic effect) and resistant (R—where the antimicrobial activity is associated with a higher than expected likelihood of therapeutic failure). MIC breakpoints are used either directly, as in MIC determination and ‘breakpoint’ susceptibility testing methods in broth or agar, or indirectly when converted into inhibition zone diameters in disc diffusion techniques. The last decade has demonstrated that breakpoints need a new function: to detect the biological phenomenon of phenotypical resistance, and to monitor the development of antimicrobial resistance. The need for this new function has shown us firstly, that breakpoints designed to guide therapy do not necessarily distinguish between bacteria with and without resistance mechanisms, and do not necessarily allow their early and/or reproducible detection. Secondly, it has demonstrated that the lack of harmonized breakpoints among methods in different countries, or even within the same country, often obviates meaningful comparison of resistance rates, monitoring of development of resistance in international surveillance systems and investigation of the effects of intervention strategies. Breakpoints have evolved to try to satisfy both the need to guide therapy and the need to detect biological resistance, often resulting in compromises that satisfy neither. MIC breakpoints may be defined by national breakpoint committees and/or regulatory authorities, including medicines’ agencies, such as the European Medicines Evaluation Agency (EMEA) in Europe, and the Food and Drug Administration (FDA) in the USA. In Europe, there are several active national breakpoint committees, including CA-SFM1 in France, the DIN2 in Germany, the CRG3 in the Netherlands, the NWGA4 in Norway, the SRGA5 in Sweden and the BSAC Working Party on Antimicrobial Sensitivity Testing6 in the UK. Each committee is long established, has 8–15 members from the fields of clinical microbiology, infectious diseases, pharmacology and, in some cases, from other fields of medicine. Each committee has devised a process for defining breakpoints. Several also have described AST methodologies to go with their breakpoints, but with few exceptions these methodologies are limited to their respective country. The equivalent committee in the USA is the NCCLS7 with many adherents worldwide, including European countries. It seems little more than chance that the committees occasionally recommend the same breakpoints and use the same terminology and/or methods. For example, the quite remarkable differences in cefotaxime and ciprofloxacin breakpoints for Enterobacteriaceae are shown in Table 1. The difference in terminology is exemplified by the different signs used for separating the intermediate and resistant categories, CA-SFM and CRG using R > X mg/L, whereas the others, including the NCCLS, use R ≥ X mg/L. Although MIC determination methodologies are similar, all having the same origin, there are some differences in the recommended media. For disc diffusion methods the differences are even greater. European committees normally recommend a lighter inoculum than the confluent inoculum recommended by NCCLS, and there are differences in media, disc contents, incubation conditions and other technical details. The lack of harmonization of breakpoints, terminology and methods creates difficulties in the comparison of results of antimicrobial surveillance, and in communication between the medical profession, the pharmaceutical industry and the regulatory authorities. Antimicrobial chemotherapy is generally similar in different countries and it is not logical for the same organism to be regarded as susceptible in some countries whereas resistant in others. MIC breakpoints are defined against a background of data, including intended or approved indications, clinical response data, dosing schedules, pharmacokinetics and pharmacodynamics, and microbiological data in a variety of formats. The process of setting breakpoints never was, and probably never will be, exact or strictly scientific; it contains elements of philosophy, ‘fairness’ and compromise. One might be sceptical at the introduction of a ‘shift factor’ as part of