Design principles of collateral sensitivity-based dosing strategies

Abstract

Collateral sensitivity (CS)-based antibiotic treatments, where increased antibiotic resistance to one antibiotic leads to increased antibiotic sensitivity of second antibiotic, could constitute a strategy to limit emergence of antibiotic resistance. However, it is unclear how to design CS-based dosing schedules that effectively suppress resistance. Here, we use a mathematical modelling approach incorporating pharmacokinetic and pharmacodynamic features to simulate bacterial population dynamics for different combination treatment designs. We study how differences in pathogen- and drug-specific factors influence the probability of resistance at end of treatment for different dosing strategies. We show that drug administration sequence is critical, whilst surprisingly, reciprocal CS was not essential to suppress resistance. Overall, we find that one-day cycling or simultaneous treatment schedules were most effective to supress the probability of resistance. In conclusion, our analysis provides insight into key design principles that contribute to the success of CS-based treatment strategies in suppressing resistance.