In vitro models in the study of antibiotic therapy of infections in neutropenic patients.

Most conventional methods for in vitro testing of antibiotics involve exposure of a bacterial inoculum to a constant, static concentration of drug. The in vivo concentrations of antibiotics change continually according to their pharmacokinetics. When two drugs are used, the ratios of their concentrations also change with time. The usual checkerboard tests for combined activity of two or more antibiotics do not consider the pharmacokinetic properties. An in vitro two-compartment pharmacokinetic model has been developed that presents changing concentrations of one or two antibiotics to isolated bacterial inocula. This model simulates the treatment of a bacterial infection in the absence of host defenses and thus mimics infection in a neutropenic patient. This model has been used to study the synergistic activity of beta-lactam/aminoglycoside combinations compared with conventional checkerboard and time-kill methods. Also, in this model, the addition of azlocillin or ceftazidime to netilmicin prevented the selection of resistant subpopulations of Pseudomonas aeruginosa that occurred with the aminoglycoside alone. In vitro pharmacokinetic models add kinetic parameters to conventional susceptibility testing and may prove useful in the design of trials of the optimal dosing and administration of antibiotics for infected neutropenic patients.