MAA868 locks factor XIa in a zymogen-like state

In this issue of Blood, Koch et al show that MAA868, a novel humanized monoclonal antibody that locks activated factor XI (FXIa) in a zymogen-like state, produces long-lasting antithrombotic effects without disrupting hemostasis.¹  If these preclinical and phase 1 findings can be replicated in clinical trials, MAA868 and other FXI-directed anticoagulant strategies have the potential to change practice by offering safer anticoagulant therapy to tackle the burden of thrombosis.

Thrombosis is responsible for 1 in 4 deaths worldwide.²  Anticoagulation therapy is a mainstay for prevention and treatment of arterial and venous thrombosis. Although the direct oral anticoagulants are more convenient to administer and are associated with less intracranial bleeding than vitamin K antagonists such as warfarin, bleeding remains the major adverse effect.³  Therefore, development of safer anticoagulants continues to be a priority.

FXI has emerged as an attractive target for development of safer anticoagulants because basic and epidemiologic studies suggest that it modulates thrombosis with little disruption of hemostasis. FXI can be activated by FXIIa or by thrombin. Thrombin-mediated activation of FXI, which is enhanced by polyphosphates released from activated platelets, augments FXI activation and amplifies thrombin generation. Thrombosis is attenuated if this amplification step is blocked. Thus, FXI knockdown with antisense oligonucleotides attenuated thrombosis in various animal models without increasing bleeding.^4-6  Furthermore, in patients undergoing knee arthroplasty, FXI knockdown reduced the risk of postoperative venous thromboembolism more effectively than enoxaparin without increasing the risk of bleeding.⁷  However, antisense oligonucleotides take 3 to 4 weeks to lower FXI levels into the therapeutic range. Therefore, faster-acting FXI inhibitors are needed.

Koch et al describe MAA868, a novel humanized monoclonal antibody that binds the catalytic domain of FXI and FXIa with similar affinity (K_d values of 1.3 pM and 4.7 pM, respectively). Analysis of the x-ray crystal structure of MAA868 in complex with the catalytic domain of FXIa suggests that the antibody induces conformational changes in loops on the side of the catalytic domain opposite the active site cleft. Movement of these loops occludes the active-site cleft and transforms FXIa into a zymogen-like state. Consequently, once bound to FXI, MAA868 blocks FXIa activity regardless of whether FXI is activated by FXIIa or thrombin. Therefore, MAA868 is a potent inhibitor of FXIa.

In keeping with its specificity for FXI, MAA868 prolongs the activated partial thromboplastin time (aPTT), but not the prothrombin time, in human plasma. The antithrombotic effect of MAA868 was examined in a mouse model. Because MAA868 does not bind murine FXI, FXI-deficient mice were reconstituted with human FXI. When MAA868 was administered intravenously to these mice before ferric chloride–induced carotid artery injury, it prolonged the aPTT, reduced free FXI levels, and attenuated thrombosis in a concentration-dependent manner. The antithrombotic dose response was steep, and efficacy required molar concentrations of MAA868 equal to or exceeding that of FXI. A more than 95% reduction in free FXI was required for MAA868 to exert its maximal effect. The dual binding of MAA868 to FXI and FXIa seems to endow it with an antithrombotic advantage over DEF, an FXIa-specific antibody of comparable affinity, because MAA868 was more effective than DEF at each concentration tested in this model. Whether the unique mode of action of MAA868 will translate into an advantage over FXIa-specific inhibitors in clinical practice remains to be seen.

The anticoagulant activity of intravenous or subcutaneous MAA868 was examined in cynomolgus monkeys. With either route of administration, MAA868 prolonged the aPTT and reduced the concentration of free FXI in a concentration-dependent manner with effects lasting up to 8 weeks. Even with weekly dosing of 100 mg/kg for 13 weeks, there was no evidence of bleeding or toxicity. With this information in hand, a phase 1 single-ascending-dose study was undertaken.

MAA868 was administered subcutaneously to 49 healthy volunteers in doses ranging from 5 to 240 mg/kg; an additional 12 participants received placebo. The 240-mg/kg dose of MAA868 was given to 16 volunteers who had a body mass index above 35 kg/m². MAA868 was well tolerated with no hypersensitivity reactions and no overt bleeding. Although the time to maximum MAA868 concentrations ranged from 7 to 21 days, levels sufficient to prolong the aPTT by twofold or more were achieved within 24 hours with MAA868 doses of 150 or 240 mg/kg. The terminal elimination half-life of MAA868 ranged from 20 to 29 days and was independent of dose. FXI levels increased by about 50% and 70% with the 150 and 240 mg/kg doses of MAA868, respectively, likely because FXI is cleared more slowly when MAA868 is bound to it. Overall, MAA868 was well tolerated and produced rapid and sustained reductions in free FXI and prolongation of the aPTT for 3 to 4 weeks.

Where does MAA868 fit in the armamentarium of FXI inhibitors? Strategies to target FXI include an antisense oligonucleotide that reduces hepatic synthesis of FXI, monoclonal antibodies that suppress FXI activation and/or inhibit FXIa activity, aptamers that bind FXI or FXIa and inhibit its activity, and small molecules that block the active site of FXIa (see table).³  Several of these agents are already undergoing evaluation in phase 2 studies. For example, the FXI antisense oligonucleotide is being studied in hemodialysis patients with end-stage renal disease (NCT02553889 and NCT03358030). BAY1213790, a monoclonal antibody that binds and inhibits FXIa, has completed phase 2 evaluation in patients undergoing elective knee arthroplasty, but the study results have not yet been disclosed (NCT03276143). Finally, BMS-986177, an oral FXIa inhibitor, is being compared with placebo for secondary prevention in patients with high-risk transient ischemic attack or small stroke (NCT03766581). These studies indicate that there is burgeoning activity in the FXI space.

In summary, with rapid onset, the potential for monthly administration, and a unique mechanism of action, MAA868 is a promising FXI inhibitor. How it will stack up against the other agents in this class remains to be determined. Regardless of outcome, the next few years will be exciting as the potential of FXI inhibitors unfolds.