Linda B.S. Aulin (1), E. van Ballegooie (1), P.H. van der Graaf (1,2), P. Valitalo (3), J.G.C. van Hasselt (1)
(1) Leiden Academic Centre for Drug Research, Leiden University, Leiden, the Netherlands. (2) Certara QSP, Canterbury, the UK, (3) Finnish Medicines Agency, Kupio, Finland
Introduction: Bacterial respiratory tract infections (RTIs) are associated with high mortality and are the leading cause of infection-related death worldwide[1]. Substantial differences between plasma concentrations and concentrations in the lung may exist for antibiotics, which may require adaptation of dose regimens for RTIs to reach optimal efficacy and to prevent emergence of antimicrobial resistance. The pharmacologically relevant target sites of antibiotics in the lung are the epithelial lining fluid (ELF) and alveolar macrophages (AMs), as these represent the focus of extracellular and intracellular bacterial lung infections respectively. While the concentrations in the ELF and AMs can be determined using bronchoalveolar lavage (BAL) sampling, there are significant limitations to this invasive technique. One approach which has been used in several therapeutic areas, including anti-invectives, is to in silico predict drug concentrations for different tissues by the use of physiologically based pharmacokinetic (PBPK) models[2].
Objectives: Although PBPK models describing the lung have previously been developed, no model explicitly including the clinically relevant compartments for RTIs has been published. To this end we aimed to develop a PBPK model for drug distribution to clinically important lung compartments, such as the ELF and AMs, with particular focus on anti-infective agents, and to evaluate its predictive performance.
Methods: A whole body PBPK model was expanded with a detailed multi-compartmental model structure representing the lungs. The lung section of the model was permeability-limited, while the remaining tissues were implemented according to a perfusion-limited model. For the lung model, we considered both lungs separately and considered three different lung zones, thus accounting for the spatial differences in perfusion and volume within the lung. The ELF and AMs were explicitly included in the model, thereby allowing for concentration predictions for theses clinically relevant compartments. Active transport was considered for the alveolar epithelium and AMs. As a proof-of-concept, we focused on prediction of lung distribution of the commonly used fluoroquinolone antibiotics ciprofloxacin, moxifloxacin, grepafloxacin, and levofloxacin. The model was used for predicting concentrations in plasma, ELF, and AMs, and the model predictions were subsequently compared to clinical BAL pharmacokinetic studies.
Results:
The model predicted plasma concentrations were in agreement with the clinical data for all the fluoroquinolones, with mean absolute percentage errors (MAPE) ranging between 7.11 and 23.4%. ELF concentrations of ciprofloxacin and levofloxacin were well-predicted (MAPE 29.7% and 40.15% respectively). Concentrations for moxifloxacin and grepafloxacin concentrations in ELF were under-predicted (MAPE 70.4% and 82.2% respectively), which is likely associated with limitation in in vitro assays quantifying active transport. AM concentrations were predicted for levofloxacin, ciprofloxacin, and grepafloxacin with MAPEs of 20.5%, 58.3%, and 45.7% respectively. No AM concentration data was available for moxifloxacin.
Conclusions: The developed lung PBPK model allows prediction of lung distribution into ELF and AMs of antibiotics based on their physiochemical properties, and could be of interest to investigate the effect of specific pathological lung conditions on antibiotic lung disposition. The PBKP model could thus constitute a tool to aid in improving antibiotic treatment of RTIs for existing drugs, with the possibility to consider special populations and pathophysiological changes, and inform drug development.
References:
[1]GBD 2016 Causes of Death Collaborators. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet (2017) 390, 1151-1210; doi 10.1016/S0140-6736(17)32152-9
[2]Gaohua, L. et al. (2015). Development of a Multicompartment Permeability-Limited Lung PBPK Model and Its Application in Predicting Pulmonary Pharmacokinetics of Antituberculosis Drugs. CPT Pharmacometrics Syst. Pharmacol. (2015) 4, 605–613; doi:10.1002/psp4.12034
Reference: PAGE 28 (2019) Abstr 8849 [www.page-meeting.org/?abstract=8849]
Poster: Drug/Disease Modelling - Absorption & PBPK