Predictors of effect of ampicillin-sulbactam against TEM-1 beta-lactamase-producing Escherichia coli in an in vitro dynamic model: enzyme activity versus MIC

Abstract

The clinical outcome in patients treated with ampicillin-sulbactam may not always be predictable by disc susceptibility testing or with the MIC as determined with a constant level (4 micrograms/ml) of the beta-lactamase inhibitor (MIC1). The enzyme activities (EA) and the MICs estimated at a constant ratio of ampicillin to sulbactam of 2:1 (MIC2) for 15 TEM-1 beta-lactamase-producing strains of Escherichia coli were examined as alternatives to MIC1 as predictors of the antibacterial effects of this combined drug as studied in an in vitro model which simulates ampicillin-sulbactam pharmacokinetic profiles observed in human peripheral tissues. Integral parameters describing the area under the bacterial count-time curve (AUBC), the area between the normal growth curve, and the killing curve of bacteria exposed to antibiotic (ABBC), and the second parameter expressed as a percentage of its maximal hypothetical value (ABBC/ABBCmax) were calculated. All three parameters correlated well with EA (AUBC, r = 0.93; ABBC, r = -0.88; ABBC/ABBCmax, r = -0.91) and with MIC2 (r = 0.94, -0.94, and -0.95, respectively) but not with MIC1. Both EA and MIC2 can be considered reliable predictors of the antibacterial effect of ampicillin-sulbactam in an in vitro model. These correlations suggest that in vitro kinetic-dynamic models might be useful to reexamine established susceptibility breakpoints obtained with data based on the MIC1 (MICs obtained with constant levels of beta-lactamase inhibitors). These data also suggest that quantitative determinations of bacterial beta-lactamase production and MICs based on the component concentration ratio observed in vivo might be useful predictors of the effect of ampicillin-sulbactam and other beta-lactam-inhibitor combinations.