Pharmacodynamics, Pharmacokinetics and Safety of Bay 1093884, an Antibody Directed Against Human TFPI, in Patients with Factor VIII or IX Deficiency (With and Without Inhibitors): A Phase 1 Study

Background: BAY 1093884 is a fully human monoclonal antibody directed against the K1 and K2 domains of human tissue factor pathway inhibitor (TFPI). TFPI regulates the initiation of coagulation and is important for regulation of normal hemostatic response. In factor (F)VIII and FIX deficiency (hemophilia A [HA] and B [HB], respectively), lack of adequate amplification (FXa generation) in tandem with intact regulation contributes to reduced clotting. Inhibition of TFPI has the potential to normalize coagulation in hemophilia even in the absence of FVIII or FIX and it is expected to result in normalization of thrombin generation and consequently improvement in clot formation. TFPI is an attractive therapeutic target as it is unaffected by FVIII or FIX inhibitors and can be targeted in both HA and HB. A potential advantage of targeting TFPI is that downstream mechanisms that protect from excessive coagulation (including antithrombin), remain intact.

Methods: BAY 1093884 was evaluated in a multicenter, open-label study in patients with severe HA or HB, with or without inhibitors. Safety, pharmacokinetics (PK) and pharmacodynamics (PD) were assessed after single intravenous (IV) (0.3 and 1 mg/kg) and subcutaneous (SC) (1, 3, and 6 mg/kg) doses. A multiple-dose (MD) cohort received BAY 1093884 150 mg SC once weekly for 6 weeks. Total and free TFPI protein levels and TFPI activity were measured. Thrombin generation assay (TGA) parameters were measured using a calibrated automated thrombogram. Clot formation was assessed using rotational thromboelastometry (ROTEM). A mechanistic physiology-based PK/PD (PBPK/PD) model was developed based on single-dose PK data observed in the study, and used to determine dosing in the MD cohort. Data for HA, HB, Japanese and inhibitor patient subgroups were analyzed separately.

Results: Thirty-two patients were enrolled; 30 with HA and 2 with HB, of which 4 (3 HA, 1 HB) were included in the MD cohort. At least 4 patients were treated in each administration method/dose group. Clot formation using ROTEM was assessed in 6 patients. Total and free TFPI-related parameters showed dose-dependent effects, with the concentrations being below the lower limit of quantification for longer duration at higher doses. Improvements in TGA parameters were observed following both IV and SC administration; duration of improvement was longer at higher doses. BAY 1093884 led to shortening of the clotting time in all 6 patients tested. Dose-dependent increases in plasma concentration exceeding dose proportionality were seen following IV and SC administration, consistent with expectations of target-mediated drug disposition for BAY 1093884. Moderate accumulation was observed after MD. Limited information on subgroups (patients with inhibitors [n=4], patients with HB [n=2] and Japanese origin [n=1]) suggests that the PK and PD data for these subgroups were consistent with those from non-inhibitor HA patients. The PBPK/PD model adequately described the concentration time course and total TFPI following single-dose IV and SC administration of BAY 1093884; the PK and PD data from the MD cohort were as predicted by the model. The model also supported that the PK/PD of BAY 1093884 are independent of inhibitor status. No serious adverse events, deaths, or adverse events leading to discontinuation were observed. Fibrinogen and platelet counts remained within normal range at most visits and no meaningful changes in levels of antithrombin, protein C, or FV were seen. The mechanisms for anticoagulation and fibrinolysis were intact.

Conclusions: Overall, BAY 1093884 gave no safety concerns in the dose range evaluated. Results suggest that BAY 1093884 shifts hemostatic balance towards greater thrombin generation and clot formation as intended. PK/PD characteristics were similar for inhibitor and HB patients as compared with HA patients, potentially supporting use in all of these patient types. The PBPK/PD model developed can be used for dose selection in future studies. Treatment efficacy will be evaluated in a future study.