(418f) Experimental Investigation and Mathematical Modeling of the Inoculum Effect

Microbial resistance to antimicrobial agents has reached alarming proportions. Reduced killing by antimicrobials at higher bacterial inoculum has emerged as a major problem in recent years, reducing the efficacy of the available antimicrobials as well as rendering current dosing regimens ineffective. This phenomenon has been widely attributed to the so-called "inoculum effect" exhibited by dense bacterial populations. In this work we have present both experimental and modeling studies to test whether the observed inoculum effect is due to the formation of a biofilm and to predict what antimicrobial agent dosing regimen would be effective.

To study the inoculum effect time kill experiments were performed with 4 baseline inocula of E-coli ATCC 25922 at different initial concentrations, differing by an order of magnitude from each other. Time-kill data were collected upto 24 hrs for constant but escalating piperacillin concentrations ranging from 0.25?e to 256?e MIC (minimum inhibitory concentration). The inoculum effect was attributed to reduced effective concentration of the drug taking part in bacterial killing. Biofilm formation by dense bacterial population, subsequently creating a barrier for drug diffusion was found to be indeed the cause of such reduced effective concentration.

Following prior work by our group, a new mathematical model structure was developed to characterize the dynamics of heterogeneous microbial populations (i.e. populations with subpopulations of varying degrees of resistance to an agent) and predict effective dosing regimens. After fitting model parameters based on initial experimental data, model predictions were validated based on subsequent experiments that identified when a biofilm was present and verified model predictions.