(622d) Quantitative Assessment of Biofitness in Heterogeneous Bacterial Populations under Antibiotic Pressure

Tam, V. H. - Presenter, University of Houston

Background: Resistance of bacterial populations to antibiotics may be mediated by different mechanisms (porin loss, efflux pumps, kinases, etc.), but their relative prevalence is not evenly distributed. We postulate that various antibiotic resistance mechanisms may be associated with a different biofitness cost, and we use a mathematical model to assess their biofitness quantitatively.

Methods: The proposed mathematical modeling approach uses population dynamics to capture the effect of antibiotics on the biofitness of heterogeneous microbial populations consisting of bacteria spanning several degrees of resistance. The bacterial heterogeneity is manifest in terms of both differect kill-rate kinetics induced by the antibiotic and different growth rates resulting from biofitness adaptation. The proposed modeling approach was tested on a bacterial population of Pseudomonas aeruginosa in a levofloxacin (levo) environment. Experiments were conducted over 24 hours using levo at several time-invariant concentrations, and model parameters were estimated based on the experimental data. Subsequently, predictions were made about the effect of levo on similar bacterial populations, subject to antibiotic concentrations following realistic pharmacokinetic profiles that normally result from periodic antibiotic injection. Corresponding experiments were conducted over a 5-day period.

Results: Model predictions about the bacterial population response to pharmacokinetically realistic antibiotic profiles were confirmed by the 5-day experiments. The model estimated the adaptation of biofitness associated with resistant subpopulations.

Conclusions: The proposed mathematical modeling approach can use ordinary data to make useful predictions about the effect of pharmacokinetically realistic antibiotic concentration profiles on the biofitness of heterogeneous bacterial populations. This capability can have significant effects on both the development and therapeutic use of antibiotics, by drastically reducing the number of tests needed for the transition from pre-clinical to clinically relevant rests.