Miriam S.R. Happ (1,2), Linda B.S. Aulin (1), Mélanie Mondemé (3), Mara Baldry (3), Wilhelm Huisinga (4), Christelle Faveeuw (3), Jean-Claude Sirard (3), Charlotte Kloft (1) and Robin Michelet (1)
(1) Department of Clinical Pharmacy and Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) Graduate Research Training Program PharMetrX, Germany, (3) Institut Pasteur de Lille, Center for Infection and Immunity of Lille, Inserm, France, (4) Institute of Mathematics, University of Potsdam, Germany
Introduction: Antibiotic resistance has been declared “one of the biggest threats to global health, food security, and development today” by the WHO [1]. Its global emergence, coupled with a lack of novel antibiotics in clinical development [2, 3], requires alternative strategies to improve antibacterial treatment outcomes. One promising strategy is to boost the natural immune response against the bacterial infection using immunomodulatory therapy in combination with antibiotic treatment [4]. The FAIR project [5] aims to leverage the Toll-like receptor 5 agonist flagellin, a derivative of bacterial flagellae, as a novel inhalative adjunct treatment for resistant bacterial pneumonia. Currently, a first-in-human (FIH) study in healthy volunteers (HVs) is being prepared for this combination. However, several challenges emerge when comparing HV data to the expected impact of this treatment modality in patients. While patients will receive flagellin as an adjunct therapy to a treatment with an antibiotic, HVs will only receive the novel compound. Additionally, due to the administration at the infection site and the involvement of the immune system in the mechanism of action, pneumonia patients are expected to show differences in drug pharmacokinetics (PK) and pharmacodynamics (PD) from HVs. Thus, it is important to investigate differences between infected and healthy systems and understand the influence of involved processes on therapy outcomes. Translational modelling of preclinical data was proposed to support further clinical trial design [6]: Amoxicillin was used as exemplary antibiotic treatment for preclinical studies of flagellin in mice super-infection models. In order to differentiate between antibiotic and immunomodulating effects, it is crucial to gain an understanding of the antibiotic PK and PD in the healthy and infected state first. This work aims to investigate potential differences in PK of amoxicillin between healthy and infected preclinical in vivo systems [7].
Methods: Amoxicillin serum concentrations were measured in samples collected from 76 C57BL/6J mice (superinfection model: N=50 [7]; naïve model: N=26). The superinfection model involved treatment with Influenza A virus (70 plaque-forming units intranasal), 7 days prior to inoculation with Streptococcus pneumoniae (5*104 colony-forming units intranasal). A single intragastric dose of 10 (N=2), 50 (N=22) and 350 µg (N=2) amoxicillin in naïve, and 150 µg in infected mice was administered, respectively. Infected mice received treatment 12 h after inoculation. Blood samples were collected up to 8 h after drug administration and serum was stored at – 80°C until analysis was performed using liquid chromatography-tandem mass spectrometry, employing electrospray ionisation (lower limit of quantification (LLOQ) = 0.05 µg/mL). An NLME model was developed (NONMEM 7.5.1) as part of a bigger translational and semimechanistic model framework to characterise the PK of amoxicillin, incorporating the M3 method [8] for handling <LLOQ measurements.
Results: Of 76 collected samples, 7 measurements (9.21%) were below the LLOQ. A classical two-compartment model with first-order absorption and elimination kinetics was developed based on infected and naïve mice serum amoxicillin concentrations. Covariate analysis revealed significant differences (p < 0.01) between amoxicillin clearance (CL) and central volume of distribution (Vc) in infected and naïve mice. CL in the infected state was 66.7% and Vc 204% of the respective estimates in naïve mice, translating to only one third of the elimination rate constant (k10), possibly caused by a reduced kidney function and increased vascular permeability. Visual predictive checks and goodness of fit plots confirmed good model performance.
Conclusions: The developed model successfully described the observed amoxicillin PK and revealed substantial differences between infected and naïve mice. It will be expanded by incorporating the effect on bacterial growth (PD). The PKPD model shall be linked to the innate immune system to include related interactions with the bacteria and effects of the immunostimulatory flagellin, in order to create a larger semimechanistic modelling framework that improves understanding of the underlying processes and elucidates differences in flagellin response between healthy and infected subjects.
References:
[1] World Health Organization (2020) Antibiotic resistance. In: World Health Organization. https://www.who.int/news-room/fact-sheets/detail/antibiotic-resistance. Accessed 7 Nov 2023
[2] Theuretzbacher U, Gottwalt S, Beyer P, et al (2019) Analysis of the clinical antibacterial and antituberculosis pipeline. The Lancet Infectious Diseases 19:e40–e50. https://doi.org/10.1016/S1473-3099(18)30513-9
[3] World Health Organization (2022) 2021 Antibacterial agents in clinical and preclinical development: an overview and analysis. Geneva
[4] Cattoir V, Felden B (2019) Future antibacterial strategies: From basic concepts to clinical challenges. J Infect Dis 220:350–360. https://doi.org/10.1093/infdis/jiz134
[5] FAIR HOME – FAIR Website. In: FAIR flagellin aerosol therapy. https://fair-flagellin.eu/. Accessed 29 Aug 2023.
[6] Michelet R, Ursino M, Boulet S, et al (2021) The use of translational modelling and simulation to develop immunomodulatory therapy as an adjunct to antibiotic treatment in the context of pneumonia. Pharmaceutics 13:601. https://doi.org/10.3390/pharmaceutics13050601
[7] Mondemé M, Zeroual Y, Soulard D, et al (2023) Amoxicillin treatment of pneumococcal pneumonia impacts bone marrow neutrophil maturation and function. Journal of Leukocyte Biology qiad125. https://doi.org/10.1093/jleuko/qiad125
[8] Beal SL (2001) Ways to fit a PK model with some data below the quantification limit. Journal of Pharmacokinetics and Pharmacodynamics 28:481–504. https://doi.org/10.1023/A:1012299115260
Reference: PAGE 32 (2024) Abstr 10809 [www.page-meeting.org/?abstract=10809]
Poster: Drug/Disease Modelling - Infection