Meng Gu1, Simone Corbetta1, Tingjie Guo1, Coen van Hasselt1
1Systems Pharmacology and Pharmacy, LACDR, Leiden University
Background and Objectives: Bacteriophages, or phages, are viruses that infect and replicate within bacterial cells, ultimately leading to bacterial lysis. The use of phage-antibiotic combination (PAC) therapies is essential for the successful clinical application of phage therapies and to delay or avoid resistance emergence. In this context, collateral sensitivity (CS) is an evolutionary trade-off wherein resistance to one agent increases susceptibility to another, which may be exploited for treatment combination regimens with enhanced efficacy. Phage resistance has been associated with the development of antibiotic hypersensitivity, i.e., CS [1]. It remains unclear if and how CS can best be exploited towards design of optimal PAC regimens. This study aimed to investigate the effect of CS on therapeutic outcomes given different dosing strategies using a mathematical modeling framework. Methods: Model definition: A theoretical model was developed in R(deSolve) to describe bacteria-phage-antibiotic interactions and resistance dynamics using hypothetical antibiotic and phages with realistic properties. The model tracked the densities of uninfected bacteria (Bu), infected bacteria (Bi), phage-resistant bacteria (BrP), antibiotic-resistant bacteria (BrAB), double-resistant bacteria (Br) and phage (P), and antibiotic (AB) concentration. Bacterial growth followed a capacity-limited kinetics. Mutation rates for phage and antibiotic resistance were defined as first order process. Antibiotic killing followed a sigmoidal Hill function, using MIC as antibiotic sensitivity parameter. Phage infection dynamics included the number of phages released per lysing bacterial cell (b), general phage decay (dP), and the adsorption rate: ß=ßmax*(1+P/P50), where ßmax was the maximum adsorption rate, and P50 represented the phage density corresponding to half of the maximum adsorption rate. The phage lysis delay(t) was described using ten transit compartments allowing antibiotics to act on Bi before lysis, thus altering phage production. CS was modeled as a unidirectional effect, where BrP had increased antibiotic susceptibility via reduced MIC values. Simulation scenarios: To quantify CS-mediated dose reduction, we simulated the bacterial concentration profile for a wide range of antibiotic dose (0-5mg) under varying CS strengths. Specifically, we examined the reduction in antibiotic dose required to rescue treatment failure (CFU72>10) and compared across conditions with no CS, with strong CS(MICrP=0.1×MICWT) and weak CS (MICrP =0.5×MICWT). Furthermore, we evaluated the treatment outcome of different dosing schedules, both in the presence and absence of CS. These schedules included simultaneous and sequential regimens in which phage and antibiotic were administered separately as QD, BID, TID, or infusion. Sequential regimens followed a phage-first approach, with antibiotic administered after a time delay of 0.2-6 hours. This dosing pattern was repeated daily for 3 days, assuming an initial bacterial inoculum of 107CFU/mL Results: Evaluation of CS-based phage-antibiotic dosing strategies: Both strong and weak CS reversed treatment failure. For weak CS, the required antibiotic dose for treatment success (CFU24 <10) was reduced by 1/3 (from 1.5 mg to 1 mg) across all phage doses. Stronger CS decreased the required antibiotic dose by 2/3 (0.5 mg). Delayed antibiotic administration reduces phage requirement: Fractionated dosing did not improve treatment outcomes even with CS. In contrast, QD dosing achieved complete bacterial eradication by fast suppressing double-resistant bacteria through single high antibiotic dosing. However, sequential dosing reversed treatment failure for low phage dose combinations. With over 1h delay, the phage dose needed for eradication drop to 10 PFU/mL from 105 PFU/mL. When delays were shorter than t (0.4h), no effect was observed, while longer delays(>4h) worsened the treatment outcomes possibly because of the overgrowth of double-resistant bacteria. Conclusions: We conclude that presence of CS can lead to reduced antibiotic dosages while maintaining efficacy. Sequential dosing under CS conditions appeared to be a more optimal strategy. Insight from this analysis can be used to support rational design of CS-based PAC treatment regimens, which is especially relevant for antibiotics with narrow therapeutic windows.
Hasan M, Dawan J, Ahn J. Assessment of the potential of phage-antibiotic synergy to induce collateral sensitivity in Salmonella Typhimurium. Microb Pathog. 2023 Jul;180:106134. doi: 10.1016/j.micpath.2023.106134. Epub 2023 May 5. PMID: 37150310.
Reference: PAGE 33 (2025) Abstr 11457 [www.page-meeting.org/?abstract=11457]
Poster: Drug/Disease Modelling - Infection