Jose M Calderin 1, Sean Wasserman 2,3, Rosleine Antilus-Sainte 4, Melissa Cristaldo 4, Narineh M Odjourian 4, Faye Lanni 4, Firat Kaya 4, Noha Abdelgawad 1, Matthew Zimmerman 4, Veronique Dartois 4,5, Martin Gengenbacher 4,5, Paolo Denti 1
1 Division of Clinical Pharmacology, Department of Medicine, University of Cape Town (Observatory, South Africa), 2 Wellcome Discovery Research Platforms in Infection, Centre for Infectious Diseases Research in Africa, Institute of Infectious Disease and Molecular Medicine, University of Cape Town (Observatory, South Africa), 3 Institute for Infection and Immunity, City St George’s, University of London (London, United Kingdom), 4 Center for Discovery and Innovation, Hackensack Meridian Health (Nutley, USA), 5 Hackensack Meridian School of Medicine (Nutley, USA)
Objectives:
Tuberculous meningitis (TBM) is the most severe form of tuberculosis (TB) and is associated with high mortality and long-term neurological impairment. Current recommended regimens for TBM are largely extrapolated from pulmonary TB and are not optimised for central nervous system (CNS) infection, where drug penetration is restricted by highly selective barriers, potentially leading to subtherapeutic exposure at the site of disease [1]. Treatment optimisation in TBM is further hindered because CNS sampling in clinical settings is largely limited to lumbar cerebrospinal fluid (CSF), which may not accurately reflect drug concentrations at the site of infection. Therefore, preclinical models that closely recapitulate human TBM are crucial for improving understanding of drug pharmacokinetics in CNS tissues.
Bedaquiline is a potent anti-TB drug that improves outcomes when added to combination regimens for pulmonary multidrug-resistant TB and is increasingly being considered for evaluation in TBM [2]. However, data on bedaquiline penetration into CNS tissues remain limited. We therefore characterised the pharmacokinetics of bedaquiline in CNS tissues using a rabbit model of TBM.
Methods:
Rabbits infected with Mycobacterium tuberculosis received human-equivalent doses of bedaquiline after displaying signs of TBM [3]. Blood was intensively sampled on days 1 and 3 after treatment initiation at pre-dose and 0.5, 2, 4, 6, and 24 h post-dose. On day 3, animals were anaesthetised and euthanised 24 h after dosing, and terminal blood, CSF, and CNS tissue samples (meninges, brain, cervical cord, spinal cord) were collected. Total bedaquiline concentrations were quantified using a validated LC–MS/MS method. Data were analysed in NONMEM (v7.5.1) using the first-order conditional estimation with interaction method. Plasma pharmacokinetics were described by testing one-, two-, and three-compartment disposition models with first-order elimination and first-order absorption, with or without a lag-time or transit compartments. Residual unexplained variability was described using a combined proportional and additive error model. Allometric scaling of disposition parameters was included using each animal’s weight.
CSF and CNS tissue data were modelled using a stepwise, non-simultaneous approach: individual plasma pharmacokinetic parameters were fixed, and CSF and CNS tissue data were described using hypothetical effect compartments linked to the plasma central compartment [4,5]. This method characterises drug penetration by estimating the plasma-to-CSF or plasma-to-CNS tissue equilibration half-life and CSF- or CNS tissue-to-plasma exposure ratios. A previously reported pharmacokinetic model of bedaquiline in humans was used to simulate plasma concentration–time profiles in a typical 55 kg male TB patient [6]. Plasma-to-CSF and plasma-to-CNS tissue equilibration rate constants estimated in rabbits were scaled across species to the 55 kg reference individual using a power-law relationship with weight (exponent −0.25) [7], and applied to simulate human CSF and CNS tissue concentration–time profiles.
Results:
Eight rabbits with a median weight (min–max) of 2.9 (2.3–3.4) kg provided 84 plasma samples and one set of CSF and CNS tissue sample each. Four CSF samples were excluded due to blood contamination. Bedaquiline plasma pharmacokinetics were best described with a two-compartment disposition model with lag-time delayed absorption.
Bedaquiline concentrations in CSF were very low, with an estimated CSF-to-plasma ratio of 0.010 and a long equilibration half-life (10.1 h). In contrast, penetration into the brain and meninges was markedly higher, with penetration ratios of 1.11 and 6.34 and faster equilibration with half-lives of 1.55 and 4.91 h, respectively.
Bedaquiline penetration into the cervical and lumbar cord was also extensive, with penetration ratios of 1.42 and 2.54, and fast equilibration with half-lives of 0.62 and 0.78 h, respectively. Simulations of human concentration–time profiles indicated that bedaquiline concentrations exceeded the critical concentration in all CNS tissues but not in CSF [8].
Conclusions:
Low bedaquiline CSF concentrations are likely due to its high plasma protein binding (>99.9%) [2], which limits the free, diffusible fraction. In contrast, higher CNS tissue concentrations suggest that bedaquiline’s high lipophilicity favours penetration and accumulation in lipid-rich tissues. The marked discrepancy between bedaquiline concentrations in CSF and CNS tissues confirms that CSF is a poor surrogate for site of disease exposure in TBM. CSF penetration observed in the rabbit TBM model is consistent with previous reports in humans [9], supporting its translational relevance. The high relative CNS exposures of bedaquiline support its further clinical evaluation for TBM therapy.
References:
1. Davis A, Meintjes G, Wilkinson RJ. Treatment of Tuberculous Meningitis and Its Complications in Adults. Curr Treat Options Neurol 2018; 20:5. Available at: https://doi.org/10.1007/s11940-018-0490-9.
2. Khoshnood S, Goudarzi M, Taki E, et al. Bedaquiline: Current status and future perspectives. J Glob Antimicrob Resist 2021; 25:48–59.
3. Lanni F, Antilus Sainte R, Hansen , Mark, et al. A preclinical model of TB meningitis to determine drug penetration and activity at the sites of disease. Antimicrob Agents Chemother 2023; 67.
4. Sheiner LB, Stanski DR, Vozeh S, Miller RD, Ham J. Simultaneous modeling of pharmacokinetics and pharmacodynamics: Application to d ‐tubocurarine. Clin Pharmacol Ther 1979; 25:358–371. Available at: https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt1979253358.
5. Zhang L, Beal SL, Sheiner LB. Simultaneous vs. Sequential Analysis for Population PK/PD Data I: Best-Case Performance. J Pharmacokinet Pharmacodyn 2003; 30:387–404.
6. Svensson E, Dosne A, Karlsson M. Population Pharmacokinetics of Bedaquiline and Metabolite M2 in Patients With Drug‐Resistant Tuberculosis: The Effect of Time‐Varying Weight and Albumin. CPT Pharmacometrics Syst Pharmacol 2016; 5:682–691.
7. Boxenbaum H. Interspecies scaling, allometry, physiological time, and the ground plan of pharmacokinetics. J Pharmacokinet Biopharm 1982; 10:201–227.
8. World Health Organization. Technical Report on critical concentrations for drug susceptibility testing of medicines used in the treatment of drug-resistant tuberculosis. Geneva: World Health Organization; 2018 (WHO/CDS/TB/2018.5). Licence: CC BY-NC-SA 3.0 IGO. 2018;
9. Upton CM, Steele CI, Maartens G, Diacon AH, Wiesner L, Dooley KE. Pharmacokinetics of bedaquiline in cerebrospinal fluid (CSF) in patients with pulmonary tuberculosis (TB). Journal of Antimicrobial Chemotherapy 2022; 77:1720–1724. Available at: https://academic.oup.com/jac/article/77/6/1720/6543975.
Reference: PAGE 34 (2026) Abstr 12125 [www.page-meeting.org/?abstract=12125]
Poster: Drug/Disease Modelling - CNS