Marcel Mohr1, Kerstin Waese1, Ziyu Li2, Andreas Hohlbaum3, Friedemann Schmidt1
1Sanofi R&D Preclinical Safety, 2Sanofi R&D Integrated Drug Discovery, 3Sanofi R&D Preclinical Safety
Introduction: New Biological Entities (NBE), such as antibody-based biologics, have revolutionized the treatment of many neoplastic and inflammatory diseases over the last years. They are often less prone to off-target interactions than synthetic drugs. However, depending on the intended pharmacology of the NBE, healthy physiology of non-targeted cells or organs might be significantly altered. One key safety consideration is the occurrence of hematologic adversities that involve direct effects on the peripheral blood cells and their blood-forming precursor cells in the bone marrow or indirect effects e.g. on the bone marrow microenvironment. Understanding how the hematopoietic system is affected is crucial to assess and predict the potential for hematotoxicity such as anemia or thrombocytopenia caused by NBE. Methods: The hematologic system is regulated by complex biological processes, involving multiple progenitor cells, pathways, and anatomical compartments which are difficult to recapitulate with static in vitro assays. We adapted a mathematical model of human hematopoiesis that comprises cell proliferation and differentiation of the full hematopoietic system, with interacting feedback loops between lineages in homeostasis [Fornari et al. 2019]. Using the model, we studied the effect of a set of preclinical molecules on the red blood cell and platelet count in comparison to an antibody in clinical development with known effects on the hematopoietic system. In our present use-case, the molecules target a cytokine of the IL6 group leading to an inhibition of proliferation of megakaryocyte-erythrocyte progenitors. Clinical data of the comparator showed a risk of anemia and thrombocytopenia at the highest tested dose [Denton et al. 2023]. For each molecule, the inhibitory potential of the targeted cytokine was measured in vitro based on the proliferation of TF-1 erythroblast cells, which were isolated from bone marrow and used as a condition to simulate the inhibition of progenitor proliferation in relation to dosing schedule and treatment time. Ultimately, drug-specific changes in red blood cell and platelet counts could be calculated this way for a mean population of patients. For the benchmark, concentration-dependent simulations were compared to clinical observations. This clinical-computational correspondence was used to assess the simulations for the preclinical molecules with respect to blood cell count reduction over time. Results: A subset of the preclinical molecules and maximal tolerated doses could be identified as less prone to clinical occurrence of anemia and thrombocytopenia as compared to the benchmark. Conclusions: This in silico study could help to optimize the selection of potentially safer molecules in early research. The developed model framework is generalizable, since the underlying systems hematology model is neither tailored to a specific modality, nor to the precise mode of action of the new biological or chemical entity. Instead, drug-specific actions on the hematopoietic system are represented by in vitro data that are used to inform the mathematical model.
Fornari et al. Quantifying Drug-Induced Bone Marrow Toxicity Using a Novel Haematopoiesis Systems Pharmacology Model. CPT Pharmacometrics Syst. Pharmacol. (2019) 8, 858–868 Denton et al. Biological and clinical insights from a randomized phase 2 study of an anti-oncostatin M monoclonal antibody in systemic sclerosis. Rheumatology 62.1 (2023): 234-242
Reference: PAGE 33 (2025) Abstr 11581 [www.page-meeting.org/?abstract=11581]
Poster: Methodology - New Modelling Approaches