Pharmacokinetic-Pharmacodynamic Modeling and Simulation for Bactericidal Effect in an In Vitro Dynamic Model

A pharmacokinetic (PK)/pharmacodynamic (PD) modeling strategy to explain the data from an in vitro dynamic model is proposed. Two carbapenem antibiotics, doripenem and meropenem, and three Pseudomonas aeruginosa strains were used as example drugs and strains. The PD model we originally developed to explain the in vitro time-kill data was modified by incorporating bactericidal activities and simulated in vivo PK profiles of the drugs. By employing only one parameter regarding the bactericidal activity from the data at a certain dosage regimen, the bacterial profiles at various dosage regimens could be well simulated for both antibiotics by the PK/PD model. Moreover, simulated bacterial counts for various dosage regimens correlated with time above minimum inhibitory concentration derived from free drug concentrations (fT > MIC) for doripenem. The predicted fT > MIC values to achieve PK/PD endpoints for three strains (static effect: 25.0%, 23.9%, and 39.8%, 2-log killing effect: 28.1%, 29.5%, and 49.6%, 90% maximum killing effect: 36.5%, 46.8%, and 80.7%) were similar to those estimated from free drug concentrations in animal infection models. The proposed in vitro PK/PD model would be useful for simulating bactericidal kinetics in the dynamic model and predicting the human therapeutic target for PK/PD indices estimated from animal infection models.