A mechanistic approach to optimize combination antibiotic therapy

Antimicrobial resistance is one of the most significant healthcare challenges of our times. Multidrug or combination therapies are sometimes required to treat severe infections; for example, the current protocols to treat pulmonary tuberculosis combine several antibiotics. However, combination therapy is usually based on lengthy empirical trials, and it is difficult to predict its efficacy. We propose a new tool to identify antibiotic synergy or antagonism and optimize combination therapies. Our model explicitly incorporates the mechanisms of individual drug action and estimates their combined effect using a mechanistic approach. By quantifying the impact on growth and death of a bacterial population, we can identify optimal combinations of multiple drugs. Our approach also allows for the investigation of the drugs' actions and the testing of theoretical hypotheses. We demonstrate the utility of this tool with in vitro Escherichia coli data using a combination of ampicillin and ciprofloxacin. In contrast to previous interpretations, our model finds a slight synergy between the antibiotics. Our mechanistic model allows investigating possible causes of the synergy.