Effect of ciprofloxacin against Pseudomonas aeruginosa with different resistance mechanisms can be predicted with mechanism-based mathematical modelling where using PK/PD indices fails

Introduction: Pseudomonas aeruginosa has a large armamentarium of mutational resistance mechanisms enabling resistance emergence during therapy against almost all antibiotics. PK/PD indices are based on minimum inhibitory concentrations (MICs) and link bacterial response to antibiotic exposure. The index relevant for fluoroquinolone antibiotics is the ratio of free drug area under the concentration-time curve to MIC over 24h (fAUC/MIC).

Aim: To determine if the effect of ciprofloxacin on isogenic bacterial strains of Pseudomonas aeruginosa with different resistance mechanisms could be predicted by PK/PD indices, and to develop a mechanism-based mathematical model depended on resistance mechanisms.

Methods: The isogenic P. aeruginosa strains were: PAO1 (wild-type reference strain), PAΔAD (ampD knockout/ampC overexpression), PAOD1 (spontaneous oprD mutant/loss of proin OprD), PAΔmexR (mexR knockout/mexAB overexpression), PAΔDΔmxR, PAOD1mxR and PAOD1ΔD (combinations of the resistance mechanisms). Agar dilution MICs were determined in triplicate. The strains were exposed to ciprofloxacin (0.5-4mg/L), in static concentration time kill studies (SCTK) performed in biological replicates over 72h. Viable counts were determined at six times and mathematical modelling performed.

Results: PK/PD indices did not predict bacterial response with an fAUC/MIC of 48-384 needed to suppress regrowth depending on strain (Figure).

A mechanism-based mathematical model was developed that described the bacterial response to antibiotic based on the resistance mechanisms. It was built in a stepwise manner where single mutations were modelled and then used to predict the results of strains with two mutations (Figure). The model contained three subpopulations which were characterised as susceptible, intermediate or resistant to ciprofloxacin. Including controls and biological replicates, 132 treatments were modelled.

Conclusions: These studies indicate that PK/PD indices alone do not predict ciprofloxacin exposure required to suppress bacterial regrowth over 72h. In contrast, mechanism-based mathematical modelling could predict the impact of double mutations in response to treatment with ciprofloxacin.

Acknowledgements: Australian National Health and Medical Research Council (NHMRC) Ideas grant GNT1184428 to C.B.L., A.O. and R.L.N.

Research Training Program Scholarship.

Alice Terrill

  • Monash University