Optimal switching strategies in multidrug therapies for chronic diseases

Area of Science:

• Mathematical modeling
• Computational biology
• Public health

Background:

• Antimicrobial resistance (AMR) is a major global health threat, causing millions of deaths annually.
• Current strategies like combination therapy and therapy switching face challenges due to the unpredictable nature of pathogen mutation.

Purpose of the Study:

• To develop a two-scale stochastic model for optimizing antimicrobial therapy strategies.
• To analyze the impact of drug switching frequency and combination therapy on delaying AMR.
• To identify cost-effective drug-switching protocols.

Main Methods:

• A two-scale stochastic model simulating therapy evolution in a multidimensional efficacy space.
• Incorporation of stochastic resets for therapy switches and boundary conditions for host recovery or resistance.
• Derivation of analytical expressions for average absorption times using Langevin and master equations.
• Inclusion of constraints on therapy number and cost, and extension to single-cell heterogeneity.

Main Results:

• Analytical expressions for average absorption times were derived, considering continuous and discrete genomic changes.
• The model identified optimal drug-switching protocols that balance resistance delay with therapy costs.
• The relevance of drug-switching intervals and the number of drugs was evaluated against AMR development timescales.
• Single-cell heterogeneity was incorporated to model individual mutation effects.

Conclusions:

• The proposed model provides a framework for optimizing antimicrobial therapy strategies to combat resistance.
• Nontrivial optimal protocols were discovered, suggesting that careful consideration of switching times and drug combinations can mitigate AMR.
• The model highlights the importance of accounting for genomic changes and host-pathogen dynamics in therapy design.