Back to search

National and Kapodistrian University of Athens

Applications of physiologically based pharmacokinetic models for the development of drug treatments for tuberculosis

Abstract

dc:description

Background: Tuberculosis (TB) is one of the most infectious diseases globally, affecting various organs but mostly the lung tissue (pulmonary TB). The current standard treatment for tuberculosis relies on a combination of four drugs that have been in use for over 60 years. The treatment duration typically extends for at least six months and can last up to two years in cases of drug resistance. The existing medications are not efficient by contemporary standards, necessitating the development of a new treatment that is faster, safer, and capable of shortening therapy duration while addressing the challenge of drug-resistant strains. Objectives: The main objective of the present thesis was to develop physiologically based pharmacokinetic (PBPK) models for commonly used drugs against tuberculosis, specifically rifampicin, pyrazinamide, isoniazid, ethambutol, and moxifloxacin, focusing on lung pharmacokinetics (PK) in both healthy and TB-infected tissues (cellular lesions and caseum) in preclinical species and then extrapolate the findings to humans. Furthermore, to apply the developed PBPK methodology to bedaquiline, which is a newer, antitubercular compound with notable PK characteristics and, also, available lung PK data only in preclinical species. The final objective is to predict the drug’s lung PK behavior in humans. A second objective of the thesis, also within the general theme of PBPK in anti-TB therapies, was to develop a PBPK model of bedaquiline for adolescents and children and, more specifically, to offer additional insight into the drug's dosing for very young children aged 0 – 2 years of age. Methods: Empirical models were applied for plasma PK of each species and linked to multicompartment, permeability-limited lung models using a middle-out PBPK approach, incorporating suitable physiological parameters that could be scaled across different species. The lung's extracellular water (EW) served as the connection between healthy and infected tissue, with passive diffusion presumed for drug transport between cellular lesions and caseum. Furthermore, the individual characteristics of each drug were also considered for lung distribution. These features included pKa values, whether the compound is a substrate of transporters present on membranes within lung tissue, lysosomal trapping when drug-induced phosholipidosis is observed as well as anomalous diffusion kinetics of some drugs in the caseum of TB-infected tissue.Moreover, a whole-body PBPK model for plasma PK for adolescents and children was developed, with a focus on those under 2 years old,. This model was based on a previously available model for adults by keeping the drug-related parameters the same across populations and appropriately altering their physiology. It is noted that a CYP3A4 intrinsic clearance was used in both the pediatric and adult PBPK models, utilizing an 18% fraction of metabolism in adults mediated by the CYP3A4 enzyme based on an in vivo clinical study. Additionally, the enzyme’s abundance ontogeny was applied to very young children since these have immature enzyme function. The remaining clearance was assumed to be non-CYP3A4 hepatic metabolism. Additionally, an advanced dissolution, absorption and metabolism model was used to describe the absorption of bedaquiline in the various age-populations, since bedaquiline is a known Biopharmaceutical Classification System (BCS) class II compound (low solubility, high permeability). Results: In rabbits, the optimized unbound fractions of rifampicin, moxifloxacin, and ethambutol in intracellular water (IW) were 0.015, 0.056, and 0.08, respectively. For pyrazinamide and isoniazid in mice, the optimized unbound fractions in EW were 0.25 and 0.17, respectively. In humans, the average extrapolated daily tissue AUC (Area Under the Curve) and Cmax (maximum Concentration) values across different lung regions were within twofold of the observed values. Unbound concentrations in the caseum were lower than unbound plasma concentrations for both rifampicin and moxifloxacin. For rifampicin, unbound concentrations in the cellular rim were slightly lower, whereas for moxifloxacin, they were significantly higher than unbound plasma concentrations. Additionally, for bedaquiline in mice, the optimized permeability value of healthy lung cells was 5.57x104 cm/h, while the ratio of lysosomes to IW partition coefficient divided by fraction unbound in IW (KpLYS:IW/fuIW) was 3.48x106, taking into consideration lysosomal trapping to account for the drug’s lung distribution. In rats and dogs, the average extrapolated healthy lung tissue AUC, Cmax and single concentration-time points were all within two-fold of the respective observed values. Regarding the infected tissue, concentrations in the caseum were much lower than the WCC90 value (minimum bactericidal concentration achieving a one-log kill of non-replicating bacteria under hypoxic, nutrient-rich conditions - Wayne Cidal Concentration), while only unbound drug concentrations within lysosomes of foamy macrophages were above the MIC50 (lowest Minimum Inhibitory Concentration at which 50% of the isolates are inhibited).Furthermore, regarding the pediatric model of bedaquiline, predicted exposure agreed well with the observed exposure of children between 5 – 12 years of age from a clinical study. For children between 2 and 5 years the predicted geometric mean bedaquiline AUC168h values were 144,083 ng*h/mL at week 12, which are within 60-140% (86,200-201,000 ng*h/mL) of adult target values (dose in children 2-5 years old: 8 mg/kg qd for the first 2 weeks, followed by 4 mg/kg tiw for the next 22 weeks). However, the respective geometric mean AUC values for children below 2 years, when the same dose was administered as in children between 2 – 5 years old, were much lower (AUC168h values at week 12 were 99,129 ng*h/mL). Moreover, the exposure of bedaquiline in various age subgroups below 2 years old was examined. Through sensitivity analysis it was found that the main parameters that reduce the exposure of the drug in children below 1 year of age are absorption-related, since neonates have higher duodenal pH in the fed state (6.4) compared to infants (5.5) and lower bile salts concentrations are observed during the first months of life. Conclusions: The use of PBPK models in the development of drug treatments for TB may be crucial. More specifically, the developed PBPK approach was able to extrapolate lung PK from preclinical species to humans and to predict unbound concentrations in the various TB-infected regions, unlike empirical lung models. Τhe final findings suggest that free drug pharmacokinetics in plasma do not consistently serve as an accurate proxy for unbound PK in TB-infected tissue, especially in the hard-to-reach caseum compartment. Also, differences can arise when different ionization exists between EW and cellular granuloma compartments as well as when steady state in the very slowly diffusing caseum is achieved after long times. Finally, unbound tissue concentrations can be utilized as input in suitable software systems to connect PK with the pharmacodynamic effects of both existing and new anti-TB drugs.Additionally, PBPK modeling revealed a potential new mechanism of action for bedaquiline, a cationic amphiphilic drug that induces phospholipidosis and exhibits a minimal fraction unbound in plasma. The current PBPK approach estimated unbound tissue pharmacokinetics above the drug's MIC50 value in lysosomes of foamy macrophages, which could be crucial for the drug's effectiveness against bacteria located within phagolysosomes. A phagolysosome is a cytoplasmic body formed by the fusion of a phagosome with a lysosome.Regarding bedaquiline’s pediatric dosing, the PBPK model developed suggests that the dose (per kg) for children aged 1-2 years should remain consistent with the dosage administered to children aged 2-5 years. However, the dosage for children younger than one year should be increased, given the fact that bedaquiline is a BCS class II compound. In addition, the fraction of bedaquiline absorbed in neonates during the first month of life is expected to be significantly lower compared to older infants, potentially reducing their exposure to the drug. Conclusively, these findings underscore the essential role of physiologically based pharmacokinetic modeling in the advancement of antituberculosis drug development.

Degree

thesis:*
Grantor dc:publisher
National and Kapodistrian University of Athens
Year dc:date
2024

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Karakitsios, Evangelos
  • Καρακίτσιος, Ευάγγελος

Subjects

dc:subject × 14

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/57181

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Karakitsios, Evangelos; Καρακίτσιος, Ευάγγελος. Applications of physiologically based pharmacokinetic models for the development of drug treatments for tuberculosis. National and Kapodistrian University of Athens, 2024. http://hdl.handle.net/10442/hedi/57181