Deciphering longitudinal optical-density measurements to guide clinical dosing regimen design: A model-based approach

Area of Science:

• Pharmacodynamics and mathematical modeling in infectious diseases.
• Bacterial growth kinetics and antibiotic susceptibility testing.
• Computational biology and personalized medicine.

Background:

• Model-based analysis of bacterial growth curves offers insights into antibiotic treatment.
• Previous work established a mathematical framework for optimizing antibiotic dosing regimens.
• Longitudinal optical density measurements can inform individualized treatment strategies for bacterial infections.

Purpose of the Study:

• To predict the bactericidal efficacy of antibiotic exposures using longitudinal optical density measurements.
• To evaluate the effectiveness of ceftazidime and ceftazidime/amikacin combinations against Acinetobacter baumannii.
• To validate model predictions using an in vitro hollow-fiber infection model.

Main Methods:

• Collected longitudinal optical density measurements of Acinetobacter baumannii exposed to varying concentrations of ceftazidime and ceftazidime/amikacin.
• Converted optical density data to bacterial cell concentration (CFU/mL equivalent) over time.
• Employed a model-based analysis to predict bactericidal efficacy and validated predictions in a hollow-fiber infection model.

Main Results:

• Model predictions indicated low confidence (<62%) in bacterial suppression with ceftazidime monotherapy, even at high concentrations.
• High confidence (>95%) was predicted for bacterial suppression using ceftazidime/amikacin combinations (2:1 ratio).
• Experimental validation confirmed model predictions, showing 98% confidence for the combination therapy versus 38-59% for ceftazidime alone.

Conclusions:

• The proposed mathematical framework shows potential for clinicians to assess antibiotic utility against patient-specific bacterial isolates.
• The study underscores the superior efficacy of ceftazidime/amikacin combinations over ceftazidime monotherapy for bacterial suppression.
• Further research is necessary to broaden the application and validation of this predictive modeling approach.