Digital twin-enhanced three-organ microphysiological system for studying drug pharmacokinetics in pregnant women.
Katja Graf, Jos�� Martin Murrieta-Coxca, Tobias Vogt, Sophie Besser, Daria Geilen, Tim Kaden, Anne-Katrin Bothe, Diana Maria Morales-Prieto, Behnam Amiri, Stephan Schaller, Ligaya Kaufmann, Martin Raasch, Ramy M Ammar, Christian Maass
Author Information
Katja Graf: Dynamic42 GmbH, Jena, Germany.
Jos�� Martin Murrieta-Coxca: Placenta Lab, Department of Obstetrics, Jena University Hospital, Jena, Germany.
Tobias Vogt: Dynamic42 GmbH, Jena, Germany.
Sophie Besser: Dynamic42 GmbH, Jena, Germany.
Daria Geilen: Dynamic42 GmbH, Jena, Germany.
Tim Kaden: Dynamic42 GmbH, Jena, Germany.
Anne-Katrin Bothe: Dynamic42 GmbH, Jena, Germany.
Diana Maria Morales-Prieto: Placenta Lab, Department of Obstetrics, Jena University Hospital, Jena, Germany.
Background: Pregnant women represent a vulnerable group in pharmaceutical research due to limited knowledge about drug metabolism and safety of commonly used corticosteroids like prednisone due to ethical and practical constraints. Current preclinical models, including animal studies, fail to accurately replicate human pregnancy conditions, resulting in gaps in drug safety and pharmacokinetics predictions. To address this issue, we used a three-organ microphysiological system (MPS) combined with a digital twin framework, to predict pharmacokinetics and fetal drug exposure. Methods: The here shown human MPS integrated gut, liver, and placenta models, interconnected via the corresponding vasculature. Using prednisone as a model compound, we simulate oral drug administration and track its metabolism and transplacental transfer. To translate the generated data from MPS to human physiology, computational modelling techniques were developed. Results: Our results demonstrate that the system maintains cellular integrity and accurately mimics drug dynamics, with predictions closely matching clinical data from pregnant women. Digital twinning closely aligned with the generated experimental data. Long-term exposure simulations confirmed the value of this integrated system for predicting the non-toxic metabolization of prednisone. Conclusion: This approach may provide a potential non-animal alternative that could contribute to our understanding of drug behavior during pregnancy and may support early-stage drug safety assessment for vulnerable populations.
