Evaluation of Mechanism Based PKPD Model for Antibiotics
David Khan (1), Elisabet I. Nielsen (1), Pernilla Lagerbäck (2), Cao Sha (3), Christer Malmberg (2), Ulrika Lustig (3), Erik Gullberg (3), Otto Cars (2), Diarmaid Hughes (3), Dan I. Andersson (3), Lena E. Friberg (1)
(1) Department of Pharmaceutical Biosciences, (2) Department of Medical Sciences, (3) Department of Medical Biochemistry and Microbiology, Uppsala University, Sweden
Objectives: PKPD models based on in vitro time-kill curve data are increasingly used in drug development for antibacterial drugs [1]. Often new PKPD models are developed to fit each situation such as different bacteria strains, increased inocula sizes and other experimental conditions. For successful forecasts of clinical effects, predictability outside traditionally used experimental conditions is vital. The aim of the current study was to perform an extensive model evaluation of a mechanism-based PKPD model built for data on ciprofloxacin and E. coli. The model’s predictive ability was assessed on bacterial killing following; a range of start inoculum, different isogenic strains, clinical strains, and mixed population experiments with mutant bacteria competing with wild type bacteria. Its capacity to predict the earlier identified PKPD index in mice and human studies was also evaluated.
Methods: The mechanism-based PKPD model for E. coli exposed to ciprofloxacin was developed with static time-kill curve data for wild type and six well characterized mutants of E. coli (start inocula 106 CFU/ml) (2). The model structure includes susceptible growing and resting non growing bacteria as well as pre-existing resistant bacteria and non-plateable bacteria. The earlier estimated model parameters were used for predictions of experiments with the same strains as in the model development. When predicting isogenic and clinical strains, the EC50 values were taken from the MIC-EC50 correlation estimated from the original strains.
Results: The model successfully predicted the lower rate and extent of bacterial killing in the experiments with high start inocula without re-estimation of model parameters. The model also predicted the time-course of bacterial killing for different isogenic strains, competition experiments and identified the earlier determined PKPD index (fAUC/MIC). Without re-estimation, the killing of the clinical strains were over predicted at high concentrations, however, allowing for both susceptible and resting bacteria at the start of the experiment improved the fit.
Conclusions: We have shown that an earlier developed mechanism-based PKPD model can predict experiments outside the traditional settings. This study provides an extensive framework on evaluation for mechanism based PKPD models based on in vitro data.
References:
[1] Nielsen, E.I. and L.E. Friberg, Pharmacokinetic-pharmacodynamic modeling of antibacterial drugs. Pharmacol Rev, 2013. 65(3): p. 1053-90.
[2] Khan, D., et al., In Silico Model Characterizing Time-Kill Curves for E. coli MG1655 Wild Type and Six Well-Characterized Mutants Exposed to Ciprofloxacin, ICAAC 2011 Poster A1-17.
Acknowledgement:The research leading to these results has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement n° 115156, resources of which are composed of financial contributions from the European Union's Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution. The DDMoRe project is also financially supported by contributions from Academic and SME partners.