Anitha Suram 1, Carter L. Johnson 2, Chimnons0 Onuoha 1, Ellen R. Swanson 2, Nisha Perez 1, Alexandra Grassian 1, Diana H. Marcantonio 2, Ronda Rippley 1, Wendy Ankrom 1
1 Blueprint Medicines Corporation (Cambridge, MA, USA), 2 Certara Predictive Technologies (Sheffield, UK)
Objective: Mast cells (MCs) serve an important role in inflammatory and allergic reactions. KIT, a receptor tyrosine kinase, regulates MC survival, proliferation, activation, and differentiation, as well as hematopoiesis. KIT inhibitors, including monoclonal antibodies (mAbs; e.g. barzolvolimab and briquilimab) and small molecules (e.g. THB-001 and BLU-808), are promising therapeutics for MC-mediated disorders such as chronic urticaria. To support early quantitative decision-making in drug development, we developed a quantitative systems pharmacology (QSP) model to characterize the relationship between KIT signaling inhibition and downstream biological effects, including MC inhibition (as measured by reduction in serum tryptase) and MC depletion. This model could inform dose selection and trial design by simulating the effects of novel KIT inhibitors, such as BLU-808.
Methods: The QSP model was developed using in vitro, clinical, and literature data from anti-KIT mAbs and small molecules. The model framework integrates: (a) drug-specific pharmacokinetic (PK) components (intravenous, oral, or subcutaneous route of administration); and (b) a pharmacodynamic model of MC dynamics incorporating key physiological processes including MC growth and death, KIT receptor activity, and tryptase release. The model was calibrated using publicly available clinical data from barzolvolimab (anti-KIT mAb), followed by THB-001 (small molecule KIT inhibitor), and subsequently validated/benchmarked against briquilimab (anti-KIT mAb) data.
Results: The model successfully reproduced the available clinical data for PK and serum tryptase (all three compounds) and MC depletion (both mAbs). After initial model calibration using barzolvolimab, the model predicted the reduction in serum tryptase (40% to 80%) and MC depletion (60% to 80%) for briquilimab based on its potency and PK characteristics. By substituting the mechanism of action (MoA) from mAb-mediated blocking of stem cell factor (SCF)-KIT binding to small molecule inhibition via a half-maximal inhibitory concentration (ICâ‚…â‚€) function, the model captured THB-001 responses without requiring modification of core biological parameters.
Conclusion: This QSP model provides a robust tool for predicting on-target KIT effects on MC through successful integration of MoA for both anti-KIT mAbs and small molecules. General applicability and mechanistic soundness were confirmed by capturing data from three distinct KIT inhibitors using drug-specific PK and potency parameters. The model enables the identification of effective molecule characteristics such as binding affinity and drug potency that drive pharmacological effects and can be applied to new emerging KIT inhibitors, including small molecules such as BLU-808. Direct measurement of tissue MCs is difficult and seldom captured in trials; this model’s MC centric design enables estimation of MC dynamics where clinical data are limited. Additionally, optimal dosing strategies and clinical trial design can be identified.
While uncertainty remains in predicting MC recovery post-KIT inhibition due to a lack of clinical data and competing hypotheses regarding MC maturation rate, the model enables exploration of a plausible biological range to inform dose selection and clinical study design. Published estimates of MC maturation vary by nearly a factor of 10, suggesting different recovery kinetics; this variability introduces structural uncertainty into model predictions. Additionally, biological assumptions – such as fixed SCF-KIT kinetics and steady progenitor pools – may differ across molecules or indications, potentially affecting model predictions. Acquisition of additional clinical data on these parameters would further enhance the model’s predictive accuracy.
Reference: PAGE 34 (2026) Abstr 11865 [www.page-meeting.org/?abstract=11865]
Poster: Methodology - New Modelling Approaches