Speaker
Description
Collateral sensitivity, where resistance to one antibiotic induces increased sensitivity to another, has been proposed as a strategy to combat antimicrobial resistance. Predicting the evolutionary outcomes of sequential antibiotic exposure, however, remains a major challenge. In this work we develop a mathematical framework that integrates collateral sensitivity networks with switched dynamical systems describing bacterial population dynamics. Within this framework, antibiotic therapies correspond to switching policies acting on phenotype populations.
Using invariant sets and network theory, we characterize therapeutic windows in which treatment schedules can maintain the infection within controllable regions of the state space. The approach embeds collateral interactions into a network of switched systems and provides a computational algorithm to identify effective sequential therapies. By leveraging experimental interaction data for Pseudomonas aeruginosa, the model predicts scenarios in which antibiotic cycling prevents evolutionary escape and maintains the population within a bounded containment region.
These results provide a control-theoretic interpretation of evolutionary therapy and illustrate how invariant-set analysis can guide the design of sequential treatment strategies to mitigate antimicrobial resistance.
