A predictive parameter of antibacterial efficacy in vivo, based on efficacy in vitro and pharmacokinetics

A method is described to predict the efficacy of antibiotics at changing concentrations in vitro or in an experimental thigh infection in granulocytopenic mice from the activity at constant concentrations in vitro, using pharmacokinetic parameters. The combinations studied were erythromycin against Staphylococcus aureus (in vitro and in vivo), four cephalosporins against two Gram-negative rods (in vivo), netilmicin against Pseudomonas aeruginosa (in vitro) and vancomycin against S. aureus (in vivo). The effect in vitro (ER) was defined as the difference between the growth rate without antibiotic and the growth rate in the presence of an antibiotic. The relationship between concentration and ER was described with the Hill equation. Using pharmacokinetic parameters of exponentially declining concentrations in vitro or of a bi-exponential concentration course in vivo together with the parameters of the Hill equation, the corresponding time course of ER was calculated. By integration with respect to time, an estimate, called ERt, was obtained of the effect on bacterial numbers, called EN, defined as the difference between the number of bacteria in the controls and the number in the presence of the antibiotics. For all antibiotics studied the value of ERt was significantly correlated with the actual values of EN.