Translational Model-Informed Selection of Tuberculosis Drug Combination Regimens for Early Clinical Development
Budi O. Susanto (1), Sebastian G. Wicha (2), Yanmin Hu (3), Anthony R. M. Coates (3), Ulrika S. H. Simonsson (1)
(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (2) Department of Clinical Pharmacy, Institute of Pharmacy, University of Hamburg, Hamburg, Germany, (3) Institute for Infection and Immunity, St. George’s University of London, London, United Kingdom
Objectives: There is still a gap between pre-clinical and clinical phase in tuberculosis (TB) drug development. New innovative tools are needed to streamline TB drug development. Since TB treatment contains several drugs, pharmacodynamic (PD) interactions can be a challenge for developing optimal regimens. Pre-clinical information about PD interactions needs to be used more optimally when designing early drug development clinical trials i.e. the so-called Early Bactericidal Activity (EBA) studies. The General Pharmacodynamic Interaction (GPDI) model is a novel model-based assessment of PD interactions [1], which has successfully been combined with the Multistate Tuberculosis Pharmacometric (MTP) model, a semi-mechanistic model developed to characterize drug effects on different growth states of the TB bacteria [2]. The MTP-GPDI approach has been used to identify PD interactions both in vitro [3] and in vivo [4], but no approach has yet been presented of how to predict human response given the identified pre-clinical interactions. A framework for clinical dose-response forecasting in EBA studies based on pre-clinical in vitro information using the MTP model has successfully been developed [5] but this far, only EBA studies after monotherapy have been predicted. The aim of this study was to develop a pre-clinical model-informed translational approach to guide dose selection of TB drug combinations in EBA trials using the MTP-GPDI model approach using rifampicin and isoniazid as an example.
Methods: Longitudinal colony-forming unit (CFU) data from rifampicin and isoniazid in different concentrations in monotherapy and combination based on in vitro static time-kill curve in Mycobacterium tuberculosis H37Rv strain were used to estimate exposure-response relationships as well as PD interactions at different mycobacterial growth states with the MTP-GPDI model approach [1-2]. The modelling of the in vitro data was performed using NONMEM 7.3. Clinical trial simulations of EBA studies were done using the final MTP models describing exposure-response in monotherapy with the GPDI model assessment of PD interactions coupled to the earlier developed model-informed MTP translational approach [5] with the earlier identified translational factors to account for differences between in vitro and human (post-antibiotic effects (PAE), mycobacterial susceptibility, bacterial growth phase and inoculum effect). To account for the change in drug concentration in humans and at the target site concentration, population plasma and epithelial lining fluid (ELF) pharmacokinetics of rifampicin [6-7] and isoniazid [8-9] were incorporated into the predictions. The simulations were performed using ‘deSolve’ package in R. The results of the predicted EBA trials were compared to different observed EBA trials in order to externally evaluate the approach against clinical data.
Results: Our MTP-GPDI model approach was able to predict EBA0-2 days, EBA0-5 days, and EBA0-14 days from different EBA studies of rifampicin and isoniazid in monotherapy as well as isoniazid-rifampicin in combination. In addition, the GPDI model identified bidirectional antagonism with rifampicin and isoniazid being perpetrator on each other. Rifampicin acted as a perpetrator on isoniazid on the inhibition of fast-multiplying bacteria (F) growth, killing of F state and killing of slow-multiplying bacteria (S) state. On the other hand, isoniazid became the perpetrator on rifampicin only on the killing of F state. No interaction was identified on the killing of non-multiplying bacteria (N) state. The simulations demonstrated that increasing the rifampicin dose will results in higher efficacy both in monotherapy as well as together with isoniazid. In contrast, the simulations showed that increasing the isoniazid dose will not give a significant improvement of the EBA in monotherapy and in combination. This finding indicated that maximum efficacy has reached for isoniazid in monotherapy and increasing isoniazid dose will not contribute to increase in efficacy for the combination.
Conclusions: The approach described in this study may help to inform decision making for dose selection and identification of optimal regimens in order to allow for early testing of drug combinations in TB clinical trials and contributing to closing the gap between pre-clinical and clinical drug development.
References:
[1] Wicha SG, Chen C, Clewe O, Simonsson USH. A general pharmacodynamic interaction model identifies perpetrators and victims in drug interactions. Nat Commun 2017; 8: 1–11.
[2] Clewe O, Aulin L, Hu Y, Coates ARM, Simonsson USH. A multistate tuberculosis pharmacometric model: a framework for studying anti-tubercular drug effects in vitro. J Antimicrob Chemother 2016; 71: 964–74.
[3] Clewe O, Wicha SG, de Vogel CP, de Steenwinkel JEM, Simonsson USH. A model-informed preclinical approach for prediction of clinical pharmacodynamic interactions of anti-TB drug combinations. J Antimicrob Chemother 2018; 73: 437–47.
[4] Chen C, Wicha SG, de Knegt GJ, Ortega F, Alameda L, Sousa V, de Steenwinkel JEM, Simonsson USH. Assessing pharmacodynamic interactions in mice using the multistate tuberculosis pharmacometric and general pharmacodynamic interaction models. CPT Pharmacometrics Syst Pharmacol 2017; 6: 787–97.
[5] Wicha SG, Clewe O, Svensson RJ, Gillespie SH, Hu Y, Coates ARM, Simonsson USH. Forecasting clinical dose–response trom preclinical studies in tuberculosis research: translational predictions with rifampicin. Clin Pharmacol Ther 2018; 104: 1208-18.
[6] Svensson RJ, Aarnoutse RE, Diacon AH, Dawson R, Gillespie SH, Boeree MJ, Simonsson USH. A population pharmacokinetic model incorporating saturable pharmacokinetics and autoinduction for high rifampicin doses. Clin Pharmacol Ther 2018; 103: 674–83.
[7] Clewe O, Goutelle S, Conte JE, Simonsson USH. A pharmacometric pulmonary model predicting the extent and rate of distribution from plasma to epithelial lining fluid and alveolar cells - using rifampicin as an example. Eur J Clin Pharmacol 2015; 71: 313–9.
[8] Wilkins JJ, Langdon G, Mcilleron H, Pillai G, Smith PJ, Simonsson USH. Variability in the population pharmacokinetics of isoniazid in South African tuberculosis patients. Br J Clin Pharmacol 2011; 72: 51–62.
[9] Conte JE, Golden JA, McQuitty M, Kipps J, Duncan S, McKenna E, Zurlinden E. Effects of gender, AIDS, and acetylator status on intrapulmonary concentrations of isoniazid. Antimicrob Agents Chemother 2002; 46: 2358–64.
