Search

Background Emicizumab (HEMLIBRA®) is a factor (F) VIII function-mimetic therapeutic bispecific antibody (BsAb) to FIXa and FX able to prevent bleeding in persons with hemophilia A (PwHA) when injected subcutaneously once every 1, 2 or 4 weeks. To develop a next generation version, we sought an agent able to keep hemostatic potential in non-hemophilic range with more convenient dosing regimen (dosing frequency/volume). We successfully created the emicizumab-based engineered four-chain BsAbs, NXT series. Among these, we selected NXT007 as a clinical candidate. Objectives The aim of this study is to clarify the in vitro and in vivo properties of NXT007 and predict its therapeutic potency non-clinically. Methods We evaluated the pharmacological activities of NXT007 in vitro using a thrombin generation assay (TGA) with FVIII-deficient patient plasma, and in vivo by inducing bleeding in FVIII-neutralizing antibody-treated acquired hemophilia A cynomolgus monkey (cyno) model. To clarify the FVIII-cofactor activity of NXT007, we performed an enzymatic kinetics analysis of FIXa-catalyzed FX activation with and without NXT007, as well as surface plasmon resonance analysis to determine the dissociation constant (KD) of NXT007 to FIX, FIXa, FX and FXa. We obtained its pharmacokinetic (PK) profile in non-human primates in a single dose SC/IV study. Results In vitro addition of NXT007 at 30 μg/mL increased the peak height of TGA in FVIII-deficient plasma to the same levels achieved by recombinant human FVIII at 40-100 IU/dL (FXIa-triggering) or 100-150 IU/dL (tissue factor-triggering). A single bolus intravenous injection of NXT007 (0.075 mg/kg) ameliorated bleeding symptoms in the cyno model to similar as a twice daily intravenous injection of recombinant porcine FVIII (20 U/kg). The in vitro and in vivo results were roughly concordant. NXT007 increased the turnover rate (kcat) of FIXa-catalyzed FX activation by approximately 4,000-folds compared to the condition without cofactor. The impact of NXT007 on the kcat was similar to that of emicizumab. As for binding affinities, the KD values of NXT007 to FIX, FIXa, FX and FXa were 1.08, 0.728, 0.0538 and 0.0231 μM, respectively in buffer solution. Compared to emicizumab, NXT007 bound more strongly to FX/FXa and with similar affinity to FIX/FIXa. This means that NXT007 would have an ability to form more FIX-BsAb-FX ternary complex than emicizumab. Calculated using the above KD values, at 30 μg/mL of BsAb the estimated concentration of FIX-NXT007-FX ternary complex in plasma is approximately 10-fold higher than that of the FIX-emicizumab-FX ternary complex which is roughly concordant with the difference in their FVIII equivalent thrombin generation activity. Prothrombin time (PT) was not clearly prolonged suggesting minimal impact on FX function by in vitro addition of NXT007 at up to 30 μg/mL, which was enough to induce sufficient thrombin burst in FVIII-deficient plasma as described above. A half-life of NXT007 was 19.6 to 24.4 days (0.02-2 mg/kg, SC) and SC bioavailability was 84.4% (2 mg/kg) in the in vivo cyno PK study, in which no obvious change in plasma FIX or FX levels was observed after 0.02-2 mg/kg single SC administration. Conclusions Based on the nonclinical results, NXT007, delivered in every-4-week SC injections, will keep a non-hemophilic range of equivalent FVIII thrombin generation in PwHA, Compared with emicizumab, NXT007's improved cofactor activity may be attributed to its more efficient ternary complex formation while keeping turnover rate with minimal impact on FX function suggested by PT value and antigen accumulation. A phase 1/2 clinical study of NXT007 is now on-going (NXTAGE; JapicCTI-194919). Disclosures Yamaguchi: Chugai Pharmaceutical Co., Ltd: Current Employment. Soeda:Chugai Pharmaceutical Co., Ltd.: Current Employment. Sato:Chugai Pharmaceutical Co., Ltd.: Current Employment. Shibahara:Chugai Pharmaceutical Co., Ltd.: Current Employment. Koga:Chugai Pharmaceutical Co., Ltd.: Current Employment. Ichiki:Chugai Pharmaceutical Co., Ltd.: Current Employment. Joyashiki:Chugai Pharmaceutical Co., Ltd.: Current Employment. Teranishi:Chugai Pharmaceutical Co., Ltd.: Current Employment. Nishimura:Chugai Pharmaceutical Co., Ltd.: Current Employment. Shiraiwa:Chugai Pharmaceutical Co., Ltd.: Current Employment. Kitamura:Chugai Pharmaceutical Co., Ltd.: Current Employment. Igawa:Chugai Pharmaceutical Co., Ltd.: Current Employment. Konishi:Chugai Pharmaceutical Co., Ltd.: Current Employment. Kitazawa:Chugai Pharmaceutical Co., Ltd.: Current Employment.

ACE910 is a humanized anti-factor IXa/X bispecific antibody mimicking the function of factor VIII (FVIII). We previously demonstrated in nonhuman primates that a single IV dose of ACE910 exerted hemostatic activity against hemophilic bleeds artificially induced in muscles and subcutis, and that a subcutaneous (SC) dose of ACE910 showed a 3-week half-life and nearly 100% bioavailability, offering support for effective prophylaxis for hemophilia A by user-friendly SC dosing. However, there was no direct evidence that such SC dosing of ACE910 would prevent spontaneous bleeds occurring in daily life. In this study, we newly established a long-term primate model of acquired hemophilia A by multiple IV injections of an anti-primate FVIII neutralizing antibody engineered in mouse-monkey chimeric form to reduce its antigenicity. The monkeys in the control group exhibited various spontaneous bleeding symptoms as well as continuous prolongation of activated partial thromboplastin time; notably, all exhibited joint bleeds, which are a hallmark of hemophilia. Weekly SC doses of ACE910 (initial 3.97 mg/kg followed by 1 mg/kg) significantly prevented these bleeding symptoms; notably, no joint bleeding symptoms were observed. ACE910 is expected to prevent spontaneous bleeds and joint damage in hemophilia A patients even with weekly SC dosing, although appropriate clinical investigation is required.

Modulating the complement system is a promising approach in drug discovery for overcoming uncontrolled complement activation, which is known to cause various disorders. While these disorders can be effectively treated by inhibiting complement C5 with a therapeutic antibody, the high concentration of C5 in plasma requires a huge dosage and frequent intravenous administration of the antibody. We have generated SKY59, an engineered humanized anti-C5 monoclonal antibody that shows long-lasting neutralization of C5 in cynomolgus monkeys, which enables subcutaneous delivery or less frequent administration. We found that, although the effective duration of the antibody was prolonged by engineering a pH-dependent C5 binding, it was not sufficient to achieve subcutaneous delivery with less frequent administration. Here, we further improved the effectiveness of the pH-dependent anti-C5 antibody by applying a novel approach: surface charge engineering. Intravenous injection of 40 mg/kg CFA0322, a high affinity anti-C5 monoclonal antibody that is not pH-dependent, caused about 3-fold increase of total C5 concentration in the plasma of cynomolgus monkeys. This is because the C5-antibody immune complexes (ICs) have slower clearance than C5 alone. The increase in total C5 concentration then requires an even higher dosage of the antibody to neutralize the increased amount of C5. To reduce the required dosage of anti-C5 antibody, pH-dependent antibodies were identified from rabbit immunization. A humanized lead antibody, 305LO1, showed a pH-dependent binding property and reduced the plasma total C5 concentration compared to CFA0322. However, we still observed an increase in total C5 plasma concentration over the baseline. We assumed that this was due to a slow uptake rate of the C5-antibody ICs into the cell. We hypothesized that the surface charges of the ICs partially contributed to the slow uptake rate. Surface-charge engineered antibodies were generated by engineering both the variable and the constant regions. Both approaches successfully reduced the increase of plasma C5 in cynomolgus monkey and the combination of the two completely suppressed the accumulation. Interestingly, surface-charge engineering did not affect the pharmacokinetics of the antibody. The effect of surface-charge engineering on C5 reduction correlated clearly with the retention time of the C5-antibody ICs in cation exchange chromatography, indicating the contribution of positive surface charges of the ICs. After further engineering to optimize various properties such as immunogenicity and physicochemical properties, SKY59 was successfully generated. Our surface-charge engineered pH-dependent antibody suppressed the increase of total C5 concentration in vivo by accelerating the rate of IC uptake into cells, C5 release from the antibody in endosomes, and salvage of the antigen-free antibody. Thus, SKY59 can be expected to provide significant benefits for patients with complement-mediated disorders. Disclosures Kitazawa: Chugai Pharmaceutical Co., Ltd: Employment, Equity Ownership, Patents & Royalties: Patents related to emicizumab.

Abstract 1126 Exogenous factor VIII (FVIII) is used to reduce bleeding complications in patients with severe hemophilia A. However, there are two drawbacks of current routine prophylaxis by FVIII. One is the requirement of frequent intravenous administration due to its short half-life and low subcutaneous bioavailability of FVIII. Second is the development of anti-FVIII antibodies (inhibitors) in approximately 30% of the severe patients which deprives the patients from routine prophylaxis by FVIII. To overcome these drawbacks, bispecific IgG antibody against activated factor IX (FIXa) and factor X (FX), which mimics the cofactor function of FVIII by placing these two factors into spatially appropriate positions, was screened from approximately 40,000 bispecific antibodies recognizing FIXa by the one arm and FX by the other arm. The therapeutic potential of the bispecific antibody identified from the screening was marginal due to insufficient FVIII-mimetic activity and poor pharmacokinetics, and moreover, large scale purification of recombinant bispecific IgG antibody was challenging. Therefore, the lead bispecific antibody, after humanization, was subjected to multidimensional optimization process in order to improve both the therapeutic potential and the manufacturability of the bispecific antibody. FVIII-mimetic activity was improved by modifying its binding properties to FIXa and FX, and the pharmacokinetics was improved by engineering the charge properties of the variable region. Difficulty of manufacturing bispecific antibody was overcome by identifying common light chain for anti-FIXa and FX heavy chain through framework/complementarity determining region shuffling, and by isoelectric point engineering of the two heavy chain variable regions to facilitate ion exchange chromatography purification of the bispecific antibody. Engineering to overcome low solubility and deamidation was also performed to enable stable high concentration liquid formulation for clinical use. ACE910, multidimensionally optimized bispecific antibody, exhibited potent FVIII-mimetic activity in human FVIII deficient plasma (more than 10% of FVIII activity at 300 nM in thrombin generation assay), and half-life of approximately 3 weeks with high subcutaneous bioavailability in cynomolgus monkey, enabling effective prophylaxis by subcutaneous administration with long dosing interval. In silico immunogenicity prediction analysis suggested that ACE910 was minimally immunogenic in human, in contrast to high immunogenicity of FVIII in human. Importantly, the activity of ACE910 was not affected by the presence of inhibitors, while polyclonal anti-ACE910 antibody did not inhibit FVIII activity, allowing the use of ACE910 without considering the development or presence of inhibitors. Furthermore, ACE910 could be purified in a large scale manufacturing, and formulated into patient-friendly subcutaneously injectable liquid formulation for clinical use. We believe that ACE910, with its multidimensionally optimized profile, would significantly improve the quality of life of hemophilia A patients by reducing not only bleeding but also the burden on the patients themselves, their parents, and all medical staff. Disclosures: Igawa: Chugai Pharmaceutical Co.,Ltd: Employment. Sampei:Chugai Pharmaceutical Co., Ltd.: Employment. Soeda:Chugai Pharmaceutical Co.,Ltd: Employment. Okuyama-Nishida:Chugai Pharmaceutical Co., Ltd.: Employment. Moriyama:Chugai Pharmaceutical Co.,Ltd: Employment. Wakabayashi:Chugai Pharmaceutical Co.,Ltd: Employment. Tanaka:Chugai Pharmaceutical Co.,Ltd: Employment. Muto:Chugai Pharmaceutical Co., Ltd.: Employment. Kojima:Chugai Pharmaceutical Co.,Ltd: Employment. Kitazawa:Chugai Pharmaceutical Co., Ltd.: Employment. Yoshihashi:Chugai Pharmaceutical Co.,Ltd: Employment. Harada:Chugai Pharmaceutical Co.,Ltd: Employment. Funaki:Chugai Pharmaceutical Co.,Ltd: Employment. Haraya:Chugai Pharmaceutical Co.,Ltd: Employment. Tatsuhiko:Chugai Pharmaceutical Co.,Ltd: Employment. Suzuki:Chugai Pharmaceutical Co.,Ltd: Employment. Esaki:Chugai Pharmaceutical Co.,Ltd: Employment. Nabuchi:Chugai Pharmaceutical Co.,Ltd: Employment. Hattori:Chugai Pharmaceutical Co., Ltd.: Employment.

Abstract 42 Background: Hemophilia A is treated by intravenous replacement therapy with factor VIII (FVIII), either on demand to resolve bleeding or as a prophylactic to prevent bleeding. Recently, routine prophylactic treatment is recommended to effectively prevent bleeding and to reduce bleeding-related chronic joint damage. However, the need for frequent intravenous injections of FVIII negatively affects patients' quality of life and their adherence to the routine prophylactic regimen. More importantly, approximately 30% of severe hemophilia A patients develop inhibitory antibodies toward the injected FVIII, rendering the replacement therapy ineffective. To overcome these drawbacks, we generated a bispecific antibody (termed ACE910) against activated factor IX (FIXa) and factor X (FX), which mimics the cofactor function of FVIII. Objectives: The aims of the present study were to examine the FVIII-mimetic cofactor activity of ACE910 in vitro and its hemostatic activity in vivo. Methods: The FVIII-mimetic cofactor activity of ACE910 was evaluated by a thrombin generation assay in human FVIII-deficient plasma as well as by an enzymatic assay using purified coagulation factors. For in vivo studies, an acquired hemophilia A model was established in cynomolgus monkeys by a single intravenous injection of mouse monoclonal anti-FVIII neutralizing antibody, which was cross-reactive to cynomolgus monkey FVIII but not to porcine FVIII. After artificial bleeding had been induced, ACE910 or porcine FVIII was intravenously administered in a single dose or in twice-daily repeated doses, respectively. Bleeding symptoms, including anemia and skin bruising, were monitored for three days. A pharmacokinetic study of ACE910 was also performed with a single intravenous or subcutaneous administration to cynomolgus monkeys. Results: ACE910 concentration-dependently showed FVIII-mimetic cofactor activity in the enzymatic assay and improved thrombin generation parameters in human FVIII-deficient plasma. Intravenous administration of ACE910 (a single dose of 3 mg/kg) significantly reduced the bleeding symptoms in the acquired hemophilia A model of a non-human primate. This hemostatic effect was comparable to twice-daily intravenous administration of porcine FVIII (repeated doses of 10 U/kg). The half-life of ACE910 was approximately three weeks for both single intravenous and subcutaneous administrations. The subcutaneous bioavailability of ACE910 was nearly 100%. Conclusion: The bispecific antibody against FIXa and FX, ACE910, exerted FVIII-mimetic cofactor activity in vitro. Furthermore, a single dose of ACE910 demonstrated hemostatic activity comparable to twice-daily repeated doses of 10 U/kg porcine FVIII in vivo. Moreover, ACE910 exhibited high subcutaneous bioavailability and approximately three-week half-life in a non-human primate. Our bispecific antibody against FIXa and FX is a subcutaneously injectable, long-acting agent that removes the need to consider the induction or presence of FVIII inhibitors and may establish a novel principle for the prophylactic treatment of hemophilia A patients. Disclosures: Muto: Chugai Pharmaceutical Co., Ltd.: Employment. Kitazawa:Chugai Pharmaceutical Co., Ltd.: Employment. Yoshihasi:Chugai Pharmaceutical Co., Ltd.: Employment. Takeda:Chugai Pharmaceutical Co., Ltd.: Employment. Soeda:Chugai Pharmaceutical Co., Ltd.: Employment. Igawa:Chugai Pharmaceutical Co., Ltd.: Employment. Sampei:Chugai Pharmaceutical Co., Ltd.: Employment. Sakamoto:Chugai Pharmaceutical Co., Ltd.: Employment. Okuyama-Nishida:Chugai Pharmaceutical Co., Ltd.: Employment. Saito:Chugai Pharmaceutical Co., Ltd.: Employment. Kawabe:Chugai Pharmaceutical Co., Ltd.: Employment. Shima:Chugai Pharmaceutical Co., Ltd.: Consultancy, Honoraria, Research Funding. Hattori:Chugai Pharmaceutical Co., Ltd.: Employment.

Thrombopoietin (TPO, the ligand for c-mpl) is a key factor for megakaryopoiesis. Several clinical trials of TPO have been conducted for thrombocytopenia without much success due to, in part, the production of neutralized antibodies against endogenous TPO, which causes thrombocytopenia. To overcome this problem, we previously demonstrated that mouse type minibody against c-mpl, with an amino acid sequence totally different from TPO, showed megakaryopoiesis and increased platelet numbers in monkey. This time, using CDR grafting, we generated a humanized sc(Fv)2VB22B minibody (huVB22B) against c-mpl for therapeutic use. The new minibody showed almost the same activity in vitro as TPO and the mouse type minibody, confirmed by both a human megakaryocyte cell (CD41+) differentiation assay and a proliferation assay with TPO-dependent cell line, M-07e. Single sc or iv administration of huVB22B to cynomolgus monkeys showed a dose-dependent increase in platelet numbers. Pharmacokinetic analysis showed that the plasma half-life (T1/2) of huVB22B at iv and sc administration to cynomolgus monkeys was 7–8 h and 11–16 h, respectively. After administration of huVB22B, the platelets of these monkeys increased and showed functional aggregation in response to ADP in vitro. Repeated administration of huVB22B (0.2, 2 and 20mg/kg/week) revealed that the increase in platelet level in cynomolgus monkeys was maintained for a month. Very slight reticular fibers in bone marrow were detected in a high dose group (20mg/ kg). No overt changes were detected by toxicity evaluations including clinical pathology and histopathology in 0.2 and 2mg/kg groups. No neutralized activities in plasma were observed during these experiments. Next, we examined the activities of huVB22B on human bone marrow-derived CD34-positive cells (BM-CD34+) and umbilical cord blood-derived CD34-positive cells (UCB-CD34+) in vitro. BM-CD34+ and UCB-CD34+ cells were cultured with huVB22B in serum free medium. HuVB22B induced differentiation of CD41+ cells from BM-CD34+ or UCB-CD34+ cells in a similar dose-dependent manner. However, UCB-CD34+ cells showed greater proliferation in response to huVB22B compared to BM-CD34+ cells. We then examined the in vivo activities of huVB22B on UCB CD34+ cells by treating NOD/SCID mice transplanted with human UCB-CD34+ cells with huVB22B and examining the bone marrow cells of the mice. The results showed that, compared with the control, administration of huVB22B showed an increase in the number of human hematopoietic progenitor cells (CD34+), lymphoid lineage cells (CD19+), and myeloid lineage cells (CD33+) in addition to human CFU-Meg cells (CD41+). These results suggest that c-mpl stimulation in vivo after transplantation might increase engraftment of progenitor cells in the bone marrow microenvironment and subsequently induce differentiation to multilineage cells. Umbilical cord blood transplantation faces frequent complications including a low-level stem/progenitor cell engraftment and delayed platelet recovery. Our results suggest that c-mpl stimulation might be used to increase the engraftment of UCB stem/progenitor cells and shorten the time of platelet recovery following UCB transplantation.