L.B.S. Aulin (1), H. Taghfavard (1), P.H. van der Graaf (1), A. Liakopoulos (2), D.E. Rozen (2), J.G.C. van Hasselt (1)
1. Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands. 2. Institute of Biology Leiden, Leiden University, Leiden, The Netherlands.
Background: Antimicrobial resistance (AMR) is a serious threat to public health. Over the last decades the development of novel antibiotics has been largely unsuccessful. Thus, innovative treatment strategies using existing antibiotics to prevent AMR are urgently needed. The phenomenon of collateral sensitivity (CS), where resistance to one antibiotic increases sensitivity to another one, could be utilized as a strategy to prevent or revert AMR. CS has been extensively explored in vitro [1,2], often using summary metrics such as minimal inhibitory concentration (MIC) to assess the magnitude of CS. However, it is currently not known in which situations CS-based dosing regimens will be able to effectively suppress AMR while also effectively treating the infection.
Objectives: We evaluate the clinical relevance of CS-based dosing regimens in their ability to suppress AMR. Specifically, we investigate the effect of the magnitude of CS on resistance development in relation to different combination treatment designs.
Methods: We consider a situation involving treatment with two hypothetical antibiotics, for which the bacteria display reciprocal CS. We are developing a four-state ordinary differential equation model. The four states represent one antibiotic-sensitive, two single-resistant, and one double-resistant bacterial sub-population, where resistance is implemented as an increase in MIC. Each state includes capacity-limited growth and antibiotic-mediated killing defined according to a sigmoidal relationship. Evolution of resistance between states was modelled by a stochastic process. Single- and double-resistant mutants were assumed to have 10% loss in fitness per mutation. One-compartmental pharmacokinetic models were used for both antibiotics. We simulate a set of treatment scenarios including both sequential and simultaneous combination treatment regimens for a duration of 300 h, varying parameters for drug effect (Emax and Hill-factor), and magnitude of CS. The impact of CS was quantified by calculating the time until a resistant population reach a density of 106 CFU/mL.
Results: A clear impact of CS on resistance development was identified for all sequential regimens. In absence of CS, resistance occurred for the majority of the sequential regimens. For theses regimens the time until single- and double-resistant sub-populations became established was between 31-168 h and 103-277 h, respectively. When CS was include as a two-fold reduction in MIC, complete suppression of resistance was observed for all investigated 12-hour and one-day cycling treatments. Simultaneous dosing resulted in full bacterial eradication without any resistance development.
Discussion and conclusion: Using our developed framework we gain insight regarding the ability of CS to suppress AMR development for different dosing strategies. The model serves as a flexible framework to aid in the design of experiments to further investigate the clinical relevance of CS to improve treatment efficacy and suppress AMR.
References:
[1] Podnecky et al. (2018). Conserved collateral antibiotic susceptibility networks in diverse clinical strains of Escherichia coli. Nature Communications, 9(1)
[2] Imamovic & Sommer (2013). Use of collateral sensitivity networks to design drug cycling protocols that avoid resistance development. Science Translational Medicine, 5(204).
Reference: PAGE () Abstr 9275 [www.page-meeting.org/?abstract=9275]
Poster: Drug/Disease Modelling - Infection