Background
High-risk Pseudomonas aeruginosa ST235-clone causes endemics worldwide, is commonly multidrug-resistant and hypervirulent, and associated with high mortality rates. Therefore, innovative and model-informed treatment options, such as novel β-lactam combinations, are urgently needed.
Aims
The aim of this study was to 1) evaluate the effects of meropenem and ceftazidime/avibactam, in monotherapy and combinations, on bacterial killing and resistance emergence in the in vitro dynamic hollow-fibre infection model (HFIM), and 2) develop a mechanism-based mathematical model (MBM) to describe the full time-course of bacterial killing and resistance emergence.
Methods
Four ST235-clone clinical isolates, showing diverse mutation-driven and horizontally-acquired resistance genotypes, were investigated in 240-h HFIM studies against clinically relevant regimens of meropenem, ceftazidime/avibactam, and their combinations. The developed MBM incorporated a) life-cycle bacterial growth dynamics, b) pre-existing bacterial subpopulations, c) estimation of drug concentration at the site of action, and d) the impact of resistance mechanisms on drug influx, efflux and degradation. Parameter estimation was performed using nonlinear mixed-effects modelling in S-ADAPT (version 1.57) with the importance sampling algorithm (pmethod=4) and facilitated by SADAPT-TRAN.
Results
The developed model successfully described simultaneously 113 bacterial time-course profiles across isolates and dosing regimens in monotherapies and combination therapies (including replicates) using shared drug-effect parameters (Figure). All standard errors of the estimated parameters were <20% indicating good precision. Resistance emergence was explained by amplification of pre-existing resistant subpopulations, supported by genomic data and modelled accordingly. Meropenem was modelled with a direct killing function whereas ceftazidime was modelled as causing inhibition of successful replication. Additionally, avibactam was described as decreasing the concentration of ceftazidime required to achieve 50% of the maximum inhibition of successful bacterial replication.
Conclusions
The developed modelling approach represents a key step toward model-informed and genomic-guided optimisation of antimicrobial treatment. This MBM is the first to characterise the different bacterial responses to these treatments concurrently, only by the resistance mechanisms present and their interplay.