Antimicrobial resistance: pharmacodynamic approaches for explicating PD changes for bacterial strains that are susceptible and resistant to the antimicrobial drug treatments of choice

Abstract

Bacterial pathogens of humans and animals are becoming increasingly resistant to an array of antimicrobial drugs of different classes. This is not easily remedied because the timeline for development of new and effective antimicrobials is uncertain. One way to extend the useful life of existing antimicrobial drugs could be by optimizing the treatment regimens, originally designed for fully susceptible bacterial pathogen strains, to be efficacious for less susceptible strains. The optimized regimens can be designed through the pharmacokinetic-pharmacodynamic (PK-PD) mathematical modeling of the probable antimicrobial drug concentrations (the PK component) and its effects on the pathogen population dynamics (the PD component) in vivo. This approach requires the initial in vitro PD data and data modeling to project the PD responses of strains that have acquired resistance to the antimicrobials of choice. This thesis involves two studies to analyze and compare the PD of existing antimicrobials against different bacterial species depending on the pathogen population characteristics such as density, and within a species against strains with and without acquired resistance to the drugs. In both cases, we conducted in vitro microbiological experiments to generate the data and mathematically modelled the data to describe the antimicrobial PD. In our first study, we investigated the relationships between the antimicrobial’s minimum inhibitory concentration (MIC) and the bacterial pathogen density for Gram-negative Escherichia coli and nontyphoidal Salmonella enterica subsp. enterica and Gram-positive Staphylococcus aureus and Streptococcus pneumonia (for n=4 strains per (sub)species and across the densities 1 to 8 log₁₀(colony forming units (CFU)/mL)), for antimicrobial classes with bactericidal activity against the (sub)species. This study was focused on bacterial strains susceptible to the studied antimicrobials. We fit six candidate mathematical models to the log₂(MIC) vs. log₁₀(CFU/mL) curves but did not identify one model best capturing the relationships across the pathogen-antimicrobial combinations. Gompertz and logistic models (but not a previously proposed Michaelis-Menten model) most often captured the relationships. Based on the study results, we have reported for the first time that bacterial density after which the MIC sharply increases and the intra-(sub)species between-isolate range of that density may depend on the antimicrobial’s mechanism of action. We termed that density the MIC advancement-point. Capturing these dependencies could help determine using the MICs for which bacterial densities is most informative for designing effective antimicrobial treatment regimens. In our second study, we investigated whether there are predictable changes in the PD parameter values among nontyphoidal Salmonella enterica subsp. enterica strains that are susceptible and those that have reduced susceptibility (due to acquiring specific genes encoding resistance) to the first-line treatment choice antimicrobials for serious salmonellosis in adults. The antimicrobials are the fluoroquinolone ciprofloxacin and cephalosporin ceftriaxone. We generated the antimicrobial PD data, and used a combination of the data PD modeling and statistical analysis of the estimated PD parameter values to test whether the acquired resistance genotype or imposed phenotype (the drug’s MIC) provide insight pertaining to the direction and degree of changes in the PD parameter values. The study results suggest there are statistically significant trends in the PD parameter values between the drug susceptible and resistant nontyphoidal Salmonella strains for both ciprofloxacin and ceftriaxone. With further studies, this research could continue towards designing modifications of the treatment regimens to achieve efficacy against the strains with reduced susceptibility.