Population pharmacokinetics of the antitubercular drug TBAJ587 and its two main metabolites
Albin Leding (1), Paul Bruinenberg (2), Almari Conradie (2), Jerry Nedelman (2), Ulrika S.H. Simonsson (1)
(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (2) TB Alliance, New York, NY, USA.
Introduction: Diarylquinolines (DARQs) have become an essential part of regimens against Mycobacterium tuberculosis (TB)1. TBAJ587 is a next-generation DARQ being developed by TB Alliance. This DARQ is active against TB in vitro and in vivo, with a more promising efficacy and safety profile as compared to previously developed DARQs such as bedaquiline. From TBAJ587 two main metabolites are formed, M2 and M3. M3 shows higher efficacy in vitro compared to M2, while both of them show lower efficacy in vitro then TBAJ587.
Objectives: This work aimed to characterise the population PK of TBAJ587 and its two main metabolites (M2 and M3) in healthy volunteers and how food intake might impact the PK.
Methods: A population PK model was developed for TBAJ587 and its metabolites M2 and M3, utilising data from a partially blinded SAD phase 1 trial (TBAJ-587-CL001 [NCT04890535])2. From the SAD trial, PK data were available from six oral dose cohorts (in mg: 25 [n=6], 50 [n=4], 100 [n=5], 200 [n=8], 400 [n=4], 800 [n=6]) and one food effect cohort (200 mg [n=9]). The dataset included a total of 42 healthy adults providing a total of 5787 observations up to 126 days, with matching TBAJ587 and metabolite observations at each timepoint. The total of below-limit-of-quantification observations was 25% and for TBAJ587, M2 and M3 2%, 46% and 28%, respectively. After oral dosing, TBAJ587 was the most abundant, M3 was intermediate abundant and M2 was the least abundant. The compounds’ sub-models were treated sequentially in the modelling approach, first parent, then M3 and lastly M2 with the previous model fixed. Different metabolite modelling approaches were evaluated and the most suitable was chosen. Different structural models were evaluated for each compound including different absorption models for TBAJ587. After establishing each compound sub-model, the food and dose effects were evaluated. Stochastic models were tested throughout the model development. Covariate effects greater than 30% were considered clinically significant. Finally, all data were simultaneously fitted with the TBAJ587 model fixed. Model selection was based on likelihood ratio tests, visual predictive checks, parameter uncertainty and parsimony. NONMEM v7.5.0 and PsN v5.0.0 were utilised for model development and evaluation.
Results: The final model was a three-compartmental disposition model for both TBAJ587 and M3, and a two-compartment model for M2. Absorption of TBAJ587 was described with a transit compartment model with dose-dependent absorption rate. Formation of each metabolite was independent and described as a direct formation via the elimination of TBAJ587. Both metabolites M2 and M3 showed dose-dependent apparent fraction metabolised described with power functions. The elimination of TBAJ587 and M2 was dose dependent with decreasing clearance with higher dose given a power function for both. The elimination of M3 showed no dose dependency. Typical estimated apparent clearance for TBAJ587, M2 and M3 were 5.4 L/h (CL/F) after a 200 mg dose with a exponent of 0.298, 34 L/h (CL/F/Fm) after a 200 mg dose with a exponent of 0.146 and 19 L/h (CL/F/Fm), respectively. Administration of TBAJ587 with food increased the apparent bioavailability by 68% and doubled the mean transit time. In addition, the food intake decreased the apparent fraction metabolised by half for both metabolites. The final model was also compared to a full simultaneous fit, which increased the parameter uncertainties but had similar VPC.
Conclusions: The final population PK model described TBAJ587, M2 and M3 data well, including non-linearities in the PK and food effects. The PK modelling showed that both TBAJ587 and M2 have a decreased CL with increasing doses, while M3 exhibits linear elimination kinetics. The selected approach to handle multiple metabolites was best to describe the data. M3 was the most abundant metabolite, which was re-confirmed by the PK modelling.
Funding: This work has received support from the Innovative Medicines Initiatives 2 Joint Undertaking (grant No 853989). http://www.imi.europa.eu
References:
[1] Conradie, F., Diacon, A. H., Ngubane, N., et. al., Treatment of Highly Drug-Resistant Pulmonary Tuberculosis, NEJM 382:10, 893-902 (2020).
[2] Global Alliance for TB Drug Development, Evaluation of the Safety, Tolerability, PK of TBAJ-587 in Healthy Adults [internet], clinicaltrials.gov, NIH; 2021, [Nov 30, 2022; Feb 1, 2023]. Available from: https://clinicaltrials.gov/ct2/show/NCT04890535?term=TBAJ587&cond=Tuberculosis&draw=2&rank=1