Development of a Quantitative Systems Pharmacology Model to Explore Hemostatic Equivalency of Antithrombin Lowering

Abstract

Fitusiran is an investigational siRNA therapeutic that targets antithrombin (AT) to rebalance hemostasis in people with hemophilia A (PwHA) or B (PwHB), irrespective of inhibitor status. Phase 3 clinical studies have demonstrated that fitusiran prophylaxis significantly reduced bleeding events compared to on-demand or prophylaxis treatment with factor concentrates or bypassing agents (Srivastava et al. Blood 2021; Young et al. Blood 2021; Kenet et al. ISTH 2022). To better understand the hemostatic equivalency of fitusiran prophylaxis (i.e., AT lowering), a quantitative systems pharmacology (QSP) model was developed to investigate thrombin generation in PwHA and PwHB in the context of AT lowering. QSP modeling is a mechanistic approach that integrates clinical and nonclinical data of the coagulation pathway to simulate the results of multiple in vitro coagulation assays, including the thrombin generation assay (TGA) and to mechanistically understand the readouts of these assays. The QSP model builds on prior published models of the coagulation cascade (Attarwala et al. ISTH 2017; Sridharan et al. Blood 2017), updated to represent reported steady-state levels of coagulation factors and accessory proteins in the plasma of healthy individuals and PwHA and PwHB (Figure 1). In addition, the model was extended to describe the impact of α-2-macroglobulin, a key regulator of thrombin, with increased influence at reduced AT levels that is critical to understand the impact of fitusiran on thrombin generation (Kattula et al. Blood 2020). After representing coagulation dynamics from plasma from healthy individuals, PwHA, and PwHB, the QSP model was applied to simulate TGA and relevant metrics, including thrombin peak height (Figure 1). TGA simulations were performed across different percentages of AT and FVIII. These simulations reproduced experimental TGA data from AT- and FVIII-double depleted plasma that was spiked with different levels of AT and/or FVIII. For these in vitro experiments, Factor VIII (FVIII) and AT-double depleted plasma was generated from congenital FVIII-deficient plasma by immunodepleting AT using an anti-AT antibody and recombinant AT protein was spiked back in to achieve different levels of AT. The QSP simulations also reproduced the reported formation of α-2-macroglobulin complexes with thrombin in each condition. Subsequently, the model was validated using both internal and published TGA and activated partial thromboplastin time (aPTT) datasets that were not used during model development. Next, a large (n=1000) virtual population (VP) of severe PwHA (<1% FVIII) was generated based on calibration of the model to individual patient data from completed fitusiran clinical studies. This analysis considered individual patient pharmacokinetic and AT time courses to describe thrombin peak height associated with different levels of AT reduction from fitusiran prophylaxis within a representative population of severe PwHA. The target therapeutic AT range of fitusiran prophylaxis treatment at steady state is 15-35%. Based on this model, the peak thrombin generated at 15-35% AT was similar to peak thrombin observed at 10-20 IU/kg FVIII in PwHA. In summary, by considering the impact of α-2-macroglobulin, the QSP model provides a mechanistic representation of thrombin generation under conditions of AT lowering, consistent with internal TGA data from donor-derived spiked plasma and clinical TGA data from both PwHA and PwHB. The VP analysis of fitusiran prophylaxis provided hemostatic equivalency with FVIII in a representative population of severe PwHA, with a targeted therapeutic range of AT of 15-35% resulting in a thrombin peak profile which was comparable to 10-20 IU/kg FVIII. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal