IV-026 Budi Octasari Susanto

An integrative translational framework for in vitro modelling for macozinone to support prediction of early human anti-tubercular efficacy

Budi O Susanto (1), Marie Sylvianne Rabodoarivelo (2), Jordana Galizia (2), Maxime R. Eveque-Mourroux (3), Alfonso Mendoza-Losana (4), Santiago Ferrer-Bazaga (4), Claudia Antoni (5), Stewart T Cole (5,6), Ainhoa Lucía Quintana (2), Santiago Ramón-García (2), Ulrika SH Simonsson (1)

(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden; (2) Department of Microbiology, University of Zaragoza, Zaragoza, Spain; (3) Univ. Lille, Inserm, Institut Pasteur de Lille, U1177 - Drugs and Molecules for Living Systems, F-59000 Lille, France; (4) Department of Bioengineering, Universidad Carlos III de Madrid, Madrid, Spain; (5) Innovative Medicines for Tuberculosis (iM4TB), Lausanne, Switzerland; (6) Institut Pasteur, Paris, France

Introduction/Objectives: 

Macozinone (PBTZ169) is a new drug candidate with novel mechanism of action for treatment of tuberculosis. This compound has undergone phase 1 (1–3) and phase 2a (4) clinical studies but the Phase 2a study was terminated due to low patient recruitment. In order to support the further development of this compound by understanding the exposure-response relationship and translation to early clinical efficacy, we aimed to quantify the in vitro pharmacokinetic-pharmacodynamic (PKPD) relationship and implement a previously developed translational framework based on the sparsely available preclinical information. Furthermore, we also wanted to validate the prediction from our translational framework based on published early clinical studies of macozinone.  

Methods: 

In vitro time-kill assay was performed with M. tuberculosis H37Rv strain in logarithmic phase exposed to different concentrations (0-30 ng/ml) of macozinone for 14 days. Colony-forming units (CFU) and most probable number (MPN) were used as biomarkers of bacterial count. Quantification of macozinone in the media was also performed and used as concentration driver of the effect. The multistate tuberculosis pharmacometric (MTP) model (5) was implemented to characterize the pharmacokinetic-pharmacodynamic (PKPD) relationship of macozinone in vitro. A translational pharmacometric framework was used to predict the early clinical trial efficacy of anti-tuberculosis drugs based on in vitro time-kill assay data (6,7). To account for the differences between in vitro and humans, several translational factors were included. A previously developed human population pharmacokinetic model (8) from the phase 1 study (2,3) was utilized to account for the change in drug concentrations over time in humans. An exposure matching strategy was done to mimic the human exposure after capsule formulation in a phase 2a study (4) after different doses using the earlier developed population PK model (8) based on native crystal powder formulation from the phase 1a study. Lung distribution based on a PK study in rat was utilized to account for the rate and extent of distribution of macozinone from plasma to lung which represented the drug concentration at the site of action. The variability of mycobacterial susceptibility in humans was accounted by sampling two-fold dilutions from in vitro MIC of macozinone. Published early bactericidal activity (EBA) of macozinone from phase 2a study was used to validate the predictions (4).

The First Order Conditional Estimation with Interaction (FOCE-I) method with log transform-both-sides was implemented and modelling of in vitro data was performed using NONMEM version 7.4.3. Model performance was evaluated based on likelihood ratio test, goodness of fit plots, individual plots and visual predictive check (VPC) obtained from PsN and ‘xpose4’ package in R. The simulation of phase 2a EBA was performed using ‘desolve’ package in R.

Results: 

Macozinone effects on the killing of fast-multiplying bacteria followed a sigmoidal Emax function and the killing on the slow-multiplying bacteria showed an on/off relationship. No statistically significant effect on the inhibition of fast-multiplying bacterial growth and killing of non-multiplying bacteria was found. This is also in line with previous experiments showing that macozinone has no effect on persister bacteria (9). The area under plasma concentration versus time curve from 0 to 24 hours (AUC0-24h) after the first dose of macozinone following daily dosing of 160 mg, 320 mg and 640 mg for 14 days in a phase 2a study was matched successfully using the previously developed population PK model. The predicted early clinical efficacy of macozinone at 7 days (EBA0-7 days) and 14 days (EBA0-14days) at different doses was well predicted, based on the translational framework. The simulations showed that increasing the dose above 160 mg did not result in an increase of the EBA.

Conclusions: 

Preclinical in vitro PKPD information with human exposure applied to our translational framework was used to predict early clinical efficacy which later was validated using clinical data. The validated model-informed drug development platform for macozinone can be used to support additional Phase 2a trials in monotherapy or combination. The translational framework applied in this work can also be implemented to support the early development of other new anti-tuberculosis drugs.

References:

  1. U.S. National Library of Medicine. Phase 1 Study of PBTZ169 [Internet]. [cited 2024 Feb 24]. Available from: https://clinicaltrials.gov/study/NCT03036163
  2. U.S. National Library of Medicine. Study to Evaluate the Safety, Tolerability, Pharmacokinetics and Ex-vivo Antitubercular Activity of PBTZ169 Formulation [Internet]. [cited 2024 Feb 22]. Available from: https://clinicaltrials.gov/study/NCT03423030
  3. U.S. National Library of Medicine. Study to Evaluate the Safety, Tolerability and Pharmacokinetics of PBTZ169 in Multiple Dosing [Internet]. [cited 2024 Feb 22]. Available from: https://clinicaltrials.gov/study/NCT03776500
  4. U.S. National Library of Medicine. Phase 2a Study of PBTZ169 [Internet]. [cited 2024 Feb 19]. Available from: https://classic.clinicaltrials.gov/ct2/show/NCT03334734
  5. 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(4):964–74.
  6. Wicha SG, Clewe O, Svensson RJ, Gillespie SH, Hu Y, Coates ARM, et al. Forecasting clinical dose–response trom preclinical studies in tuberculosis research: translational predictions with rifampicin. Clin Pharmacol Ther. 2018;104(6):1208–18.
  7. Susanto BO, Wicha SG, Hu Y, Coates ARM, Simonsson USH. Translational Model-Informed Approach for Selection of Tuberculosis Drug Combination Regimens in Early Clinical Development. Clin Pharmacol Ther. 2020;108(2):274–86.
  8. Susanto BO, Chtioui H, Ciullini L, Prod’Hom S, André P, Spaggiari D, et al. Population pharmacokinetics of macozinone (PBTZ-169) and active metabolites in healthy volunteers after different oral formulations. 31st PAGE Meet Abstr 10652 [Internet]. 2023; Available from: www.page-meeting.org/?abstract=10652
  9. Robertson GT, Ramey ME, Massoudi LM, Carter CL, Zimmerman M, Kaya F, et al. Comparative Analysis of Pharmacodynamics in the C3HeB/FeJ Mouse Tuberculosis Model for DprE1 Inhibitors TBA-7371, PBTZ169, and OPC-167832. Antimicrob Agents Chemother. 2021;65(11):e00583-21. 

Funding:
The project leading to this publication has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No 853989. The JU receives support from the European Union’s Horizon 2020 Research and Innovation Programme and EFPIA and Global Alliance for TB Drug Development Non-Profit Organisation, Bill & Melinda Gates Foundation, University of Dundee. http://www.imi.europa.eu. The computations were enabled by resources in projects NAISS 2023/5-492 and NAISS 2023/23-591 provided by the National Academic Infrastructure for Supercomputing in Sweden (NAISS) at UPPMAX, funded by the Swedish Research Council through grant agreement no. 2022-06725.

Reference: PAGE 32 (2024) Abstr 11089 [www.page-meeting.org/?abstract=11089]

Poster: Real-world data (RWD) in pharmacometrics