Breakpoints: current practice and future perspectives.

Abstract

Much of the discussion over the past decades on the value and setting of breakpoints has been due to the fact that the breakpoint was used in two ways; as an indicator to predict the probability of clinical success and also to detect resistant (sub) populations. It is apparent that these two meanings have lead to a different approach to setting, interpretation and use of breakpoints based on clinical efficacy on the one hand and breakpoints based on detection of resistance on the other. Nevertheless, several of the current guidelines make no perceptible distinction between these two meanings. A case is therefore strongly made to recognize that there is a difference between clinical and microbiological breakpoints. The microbiological breakpoint may be used to detect organisms that do not belong to the natural bacterial population, but somehow have acquired resistance and might be useful in recognizing emergence of resistant subpopulations and may lead to subsequent measures to be taken. Alternatively, the clinical breakpoint is of principal value to the clinician in that it results in a classification of S (susceptible), I (intermediate susceptible) and R (resistant) and is used in clinical practice and correlate with a measure of clinical efficacy. Methods developed during the last few years to arrive at meaningful clinical breakpoints are discussed, such as CART analysis and Monte Carlo simulation. In discussing future developments, it is suggested that current reports containing S, I, and R be at least supplemented with the MICs measured and, using current techniques available such as Monte Carlo simulation, provide the probability of successful eradication of the micro-organism and successful treatment based on population pharmacokinetics and Minimal Inhibitory Concentration (MIC) distributions.