Rob C. van Wijk1, Elke H.J. Krekels1, Anita K. Ordas2, Astrid van der Sar3, Sharka M. Dijkema1, Dirk-Jan van den Berg1, Rida Bahi1, Jeremy Liu1, Theo Verboom3, Ulrika Simonsson4, Thomas Hankemeier1, Herman P. Spaink2, Piet H. van der Graaf1,5
1Systems Biomedicine and Pharmacology, Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands; 2Division of Animal Sciences and Health, Institute of Biology Leiden, Leiden University, Leiden, The Netherlands; 3Department of Medical Microbiology and Infection Control, VU University Medical Center, Amsterdam, The Netherlands; The Netherlands; 4Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden; 5QSP, Certara, Canterbury, UK
Objectives: In tuberculosis (TB) research, the zebrafish (1) proves a promising model organism, because the aquatic pathogen Mycobacterium marinum, shows similar pathology to M. tuberculosis (2). However, to date most studies using zebrafish larvae take drug concentrations in the incubation medium as exposure measure, neglecting the fundamental pharmacological principle that exposure at the site of action drives drug effects and should be used as the basis for between-species translation of pharmacological results(3). Recently, we developed a methodology to measure internal drug exposure in larval homogenates and used mixed-effects modelling approaches to quantify the internal exposure over time(4). Blood concentrations are however essential to quantify distribution volume and absolute clearance values, which is the basis for between-species PK translation. Here, a nanoscale blood sampling method is developed for zebrafish larvae.
To translate pharmacodynamics (PD) between species, differences in disease mechanisms and progression need to be understood and quantified. Here, in vitro natural growth curves of M. marinum are quantitatively analysed with a multistate tuberculosis pharmacometric (MTP) model that has been previously developed for M. tuberculosis and applied to in vitro, murine and clinical data(5–7). The effect of the antibiotic isoniazid can then be assessed based on the MTP model for M. marinum and internal exposure measures in zebrafish larvae, after which the extrapolation potential to higher vertebrates, including humans, can be assessed.
Methods: Blood sampling on a nanolitre scale was tested from different locations including the aorta and posterior cardinal vein, or by cardiac puncture in larvae of 5 days post fertilization (dpf). LC-MS measurements in pooled samples yielded isoniazid blood concentration. Non-linear mixed effects modelling using NONMEM 7.3(8) through interface Pirana 2.9.6(9) and PsN 4.7.0(10) was used to analyse these concentrations combined with internal amounts from larval homogenates exposed to 5 isoniazid doses to quantify a.o. distribution volume and absolute clearance values.
Natural growth curves of M. marinum over 221 days, were obtained from in vitro cultures of the E11 and M-USA strain. In the MTP model, which distinguishes fast, slow or non-replicating mycobacteria, Gompertz, logistic, and exponential growth functions were tested for the fast- and slow- multiplying sub-states. Constant, and time- and concentration dependent transfer functions were tested for the fast- to slow- multiplying sub-states, the other transfer rates were fixed to values found for M. tuberculosis(5). For the inoculum, different CFUs in the fast- and/or slow multiplying sub-state were estimated. These fits were compared to reported observed and MTP model predicted in vitro M. tuberculosis natural growth curves(5).
Results: Blood sampling from the posterior cardinal vein was most reproducible, yielding blood volumes of up to 1.76 nL. To reach quantifiable levels, 19-32 samples were pooled. Isoniazid blood concentrations were 10% of the external concentration and within the range of isoniazid concentrations in patients(11,12). A dose-linear increase in isoniazid internal exposure was observed, with steady states ranging from 30 – 100 pmole/larva. Absolute clearance and distribution volume were estimated at 0.31 uL/h and 2.1 L/kg, respectively.
Model fits with the MTP model showed different natural growths for the two M. marinum strains. Gompertz functions for growth with similar growth rates, but a 8-fold lower system capacity for M-USA were found. For M-USA an exponential time-dependent transfer between fast- and slow multiplying sub-states was found and an inoculum with all CFU in the fast multiplying sub-state, while for E-11 a constant transfer between fast- and slow multiplying sub- states and an inoculum with all CFU in the slow multiplying sub- state was found. As a result, M-USA remains in the fast-multiplying state for the majority of the 221 days, while E11 shows a profile more similar to M. tuberculosis, shifting to slow- and non-multiplying sub-states after 30-60 days. This suggests that studies using E11 might be more representative for M. tuberculosis behaviour.
Conclusion: We quantified for the first time PK and PD of isoniazid treatment of the zebrafish as TB model organism, essential for translation of pharmacological findings to higher vertebrates, including humans.
References:
[1] Van Der Graaf PH. Pharmacometrics and/or Systems Pharmacology. CPT Pharmacometrics Syst Pharmacol. 2018;1–2.
[2] Sar AM Van Der et al. A star with stripes?: zebrafish as an infection model. TRENDS Microbiol. 2004;12(10):451–7.
[3] Morgan P et al. Can the flow of medicines be improved? Fundamental pharmacokinetic and pharmacological principles toward improving Phase II survival. Drug Discov Today. 2012;17(9–10):419–24.
[4] Van Wijk R et al. Impact of post-hatching maturation on the pharmacokinetics of exogenous compounds in zebrafish larvae. Sci Rep. 2018;9:2149.
[5] Clewe O et al. A multistate tuberculosis pharmacometric model: A framework for studying anti-tubercular drug effects in vitro. J Antimicrob Chemother. 2016;71(4):964–74.
[6] Chen C et al. The multistate tuberculosis pharmacometric model: a semi-mechanistic pharmacokinetic-pharmacodynamic model for studying drug effects in an acute tuberculosis mouse model. J Pharmacokinet Pharmacodyn. 2017;44(2):133–41.
[7] Svensson R et al. Application of the Multistate Tuberculosis Pharmacometric Model in Patients With Rifampicin-Treated Pulmonary Tuberculosis. CPT Pharmacometrics Syst Pharmacol. 2016;5(5):264–73.
[8] Beal S et al (eds). NONMEM 7.3.0 Users Guides. (1989-2013). ICON Development Solutions, Hanover, MD, USA.
[9] Keizer R et al. Pirana and PCluster: a modeling environment and cluster infrastructure for NONMEM. Comput Methods Programs Biomed. 2011;101(1):72–9.
[10] Keizer RJ et al. Modeling and Simulation Workbench for NONMEM: Tutorial on Pirana, PsN, and Xpose. CPT Pharmacometrics Syst Pharmacol. 2013;2(e50).
[11] Hemanth Kumar AK et al. Pharmacokinetics of thrice-weekly rifampicin, isoniazid and pyrazinamide in adult tuberculosis patients in India. Int J Tuberc Lung Dis. 2016;20(9):1236–41.
[12] Ribera E et al. Pharmacokinetic interaction between rifampicin and the once-daily combination of saquinavir and low-dose ritonavir in HIV-infected patients with tuberculosis. J Antimicrob Chemother. 2007;59(4):690–7.
Reference: PAGE 28 (2019) Abstr 9133 [www.page-meeting.org/?abstract=9133]
Poster: Drug/Disease Modelling - Infection