Marinda van de Kreeke1, Anh Duc Pham1, M. Mehciz1, C.A.J. Knibbe1,2, E.H.J. Krekels1,3, L.B. Zwep1, J.G.C. van Hasselt1
1Division of Systems Pharmacology and Pharmacy, LACDR, Leiden University, 2Department of Clinical Pharmacy, St. Antonius Hospital, 3Certara Inc
Introduction The role of immune status is currently scarcely considered in antibiotic dose selection and treatment [1–3]. It has been suggested that the level of immune suppression may alter PK/PD target exposures [4]. Therefore, integrating immune status into antibiotic treatment guidelines may be essential for improving antibiotic treatment outcomes, in particular for severe infections such as bacteraemia. A quantitative understanding of the potential effects of immune status on antimicrobial pharmacodynamics is currently lacking. Here, we developed and applied a comprehensive, mechanism-based model for drug-host-pathogen interactions in order to characterize the joint effect of the innate immune response and antibiotic pharmacodynamics on bacterial clearance, with a focus on bacteraemia. To this end, we first derived parameters of this from literature-derived studies, and subsequently calibrated the model to in vivo data, after which the model was applied to evaluate the impact of various immune deficiencies (e.g., healthy, neutropenic, monocytopenia) and classes of antibiotics on expected changes in PK/PD targets. Methods A mathematical ordinary differential equation-based model was built using digitized literature in vitro studies which characterized isolates relationships related to host immune-pathogen interactions [5–8], focusing on monocytes and macrophages. The model captured immune cell and bacterial dynamics through sequential steps of phagocytosis and digestion. Using bacteraemia as a case study, the model parameters were further calibrated to expected in vivo values. First, we performed a global sensitivity analysis to further assess the impact of parameter values chosen within physiologically plausible ranges. Parameters non-sensitive toward CFU changes were fixed. Remaining influential parameters were then fitted to infection studies in mice [9]. The calibrated model was extended with a PK/PD component for multiple hypothetical antibiotics, including time vs. concentration dependent antibiotics and bacteriostatic vs. bactericidal antibiotics [10,11]. Using in silico dose fractionation studies under various immune conditions, including immune competent, varying levels of neutropenia and monocytopenia, we then derived PK/PD target exposures of 2 log CFU reduction [12]. Results Model parameters were estimated separately from specific literature-reported experiments. For the developed mathematical, key parameters estimated included a phagocytosis rate was 10% higher for neutrophils compared to monocytes (8.35 h?¹ vs. 7.78 h?¹), while monocytes had a slightly higher digestion rate (6.23 h?¹ vs. 5.61 h?¹). Both rates were constrained by cell concentration and finite ingestion capacity. Sensitivity analysis revealed that bacterial growth dynamics were particularly sensitive to the maximum phagocytosis rates of both neutrophils and monocytes. To be able to describe immunocompetent mice data, the in vitro values for the maximum phagocytosis rate for both neutrophils and monocytes required a 78% reduction. Dose fractionation simulations suggested increased daily antibiotic doses are required to compensate for decreasing immune capabilities to reach the PK/PD target. Specifically, in comparison with immune competent, the compromised with severe neutropenia and monocytopenia, 8.7 and 5.8 times higher doses for bacteriostatic and vs. bactericidal time-dependent antibiotics are needed. For concentration-dependent antibiotics, the number required a 3.1 and 2.1-fold increased respectively. Conclusion The developed mechanistic model successfully captured bacterial-immune cell interactions dynamics. The model enabled evaluation of immune responses under different immune states (e.g., healthy, neutropenia, monocytopenia) to assess how immune function influences antibiotic efficacy and optimize PK/PD targets. This framework helps optimize antimicrobial treatment in vulnerable populations, with future potential to include adaptive immunity and cytokine signalling
1 de la Court, J. R. et al. Infect Dis Ther 11, 2063–2098 (2022) 2 von Itzstein, M. S. et al. Cancers (Basel) 14, (2022) 3 Snow, T. A. C. et al. Antibiotics (Basel) 13, (2024) 4 Veiga, R. P. et al. Crit Care 22, (2018) 5 Thorsted, A. et al. CPT Pharmacometrics Syst Pharmacol 12, 1972–1987 (2023) 6 LEIJH, P. C. J. et al. Scand J Immunol 13, 159–174 (1981) 7 Nagl, M. et al. Clin Diagn Lab Immunol 9, 1165 (2002) 8 Bjerknes, R. et al. Gastroenterology 89, 1139–1146 (1985) 9 Martin, M. D. et al. Cell Rep 42, (2023) 10 Regoes, R. R. et al. Antimicrob Agents Chemother 48, 3670–3676 (2004) 11 Aulin, L. B. S. et al. Nat Commun 12, (2021) 12 Minichmayr, I. K. et al. Int J Antimicrob Agents 60, (2022)
Reference: PAGE 33 (2025) Abstr 11628 [www.page-meeting.org/?abstract=11628]
Poster: Drug/Disease Modelling - Infection