In vivo infection models in the pre-clinical pharmacokinetic/pharmacodynamic evaluation of antimicrobial agents

Highlights

• The neutropenic murine thigh and lung models are useful antibiotic PK/PD.

• Infection models are useful for defining the PK/PD index and magnitude.

• Host and microbe study design features impact the PK/PD target.

• Murine PK/PD targets correlate with patient outcome.

• Murine PK/PD target can be used to design clinical trial regimens.

Animal infection models serve a critical role in the pre-clinical development of antimicrobials. Thoughtful use of these tools can be useful to design and de-risk subsequent clinical trials. Specifically, pharmacokinetic/pharmacodynamic (PK/PD) evaluation of antimicrobials can define the PK/PD driver and target magnitude. In doing so they provide guidance for dosing regimen design and forecast the likelihood of success against target pathogens at the infection site of interest. This review outlines the key design features to consider for successful assessment of experimental output.

Introduction

The goal of both preclinical and clinical antimicrobial PK/PD investigation is to improve the probability of a positive therapeutic outcome. The premise underlying the PK/PD field of study is that there is an optimal drug exposure for efficacy and safety. Antimicrobial PK/PD traditionally links drug pharmacokinetics and a measure of potency in vitro (the minimum inhibitory concentration or MIC) to efficacy [1, 2•]. Animal infection models have been used to answer two key questions regarding pharmacokinetic optimization. First, which pharmacokinetic index is the strongest driver of efficacy, or simply put, how often do I need to administer the drug? Second, what is the PK/PD target, or how much drug do I need for effect? These PK/PD questions have been addressed using several animal infection models. However, the neutropenic mouse thigh and lung infection models are the traditional ‘work-horse’ models in the PK/PD field. The models represent relatively faithful mimics of soft tissue infection/sepsis and pneumonia, respectively [3]. When immunocompromised via neutropenia, most bacteria are pathogenic in the infection models. Organism burden, measured as number of colony forming units (CFU), at the site of infection provides a relatively reproducible measure of antibiotic effect that has accurately forecasted efficacy in patients. A variety of host, pharmacokinetic, and microorganism factors impact model performance and data interpretation. We discuss these study design variables and output assessment below.