Blocking hemophilic arthropathy

In this issue of Blood, Magisetty and colleagues show that administration of an antibody (MAPC1591) inhibiting the anticoagulant activity of murine activated protein C (aPC), while preserving its anti-inflammatory and cell signaling activity, prevents joint bleeding–induced arthropathy in a mouse hemophilia model.¹ 

This contrasts with another antibody, MAPC1609, that inhibits aPC’s anticoagulant and anti-inflammatory effects but has no impact. These observations shed light on the function of endogenous aPC in the pathogenesis of hemophilic arthropathy and raise the prospect that function-selective modulation of the endogenous protein C pathway could be developed into a clinically viable option to ameliorate the development of hemophilic arthropathy.

Arthropathy is the most common clinical sequelae of bleeding in severe hemophilia A (factor VIII [FVIII] deficiency), and, to a lesser extent, in hemophilia B (FIX deficiency). It is caused by recurrent bleeds within joints and evolves into a complex pathology of synovial hyperplasia, macrophage infiltration, neoangiogenesis, cartilage degeneration, and chondrocyte apoptosis that leads to decreased joint function, chronic pain, and severely reduced quality of life.^2-5 The precise pathogenic mechanisms by which recurrent joint bleeding develops into hemophilic arthropathy are not fully understood.

FVIII protein–replacement therapy is effective in treating hemophilia A; however, even in hemophilic adults with access to factor replacement prophylaxis since childhood, arthropathy remains prevalent. In addition, ∼30% of patients receiving factor replacement prophylaxis develop neutralizing anti-FVIII antibodies (inhibitors) that reduce the effectiveness of FVIII treatment and increase the morbidity of arthropathy. Innovative therapeutics, such as emicizumab, a humanized bispecific antibody mimicking FVIIIa function by binding simultaneously to FIXa and FX, have been shown to restore hemostasis and reduce recurrent joint bleeds in patients with hemophilia A, even with inhibitors.⁶ The inhibition of natural anticoagulants, such as antithrombin, TFPI, and aPC, has emerged as an additional alternative approach to sustain hemostasis in hemophilia A (reviewed by Shapiro et al⁷).

Given that aPC exerts anticoagulant effects, in addition to well-established anti-inflammatory and cytoprotective effects, the impact of inhibiting all of aPC’s functions on hemophilic arthropathy remained unclear. Magisetty et al found that MAPC1591 and MAPC1609 antibodies inhibited aPC anticoagulant activity in an ex vivo thrombin-generation assay and an in vivo acute saphenous vein injury model. However, only MAPC1591 protected FVIII^−/− mice from the development of arthropathy following a needle puncture–induced knee joint injury. The initial bleeding within the first 5 hours after injury was not improved in FVIII^−/− mice whether they were treated with MAPC1591 or MAPC1609. However, the arthropathy phenotype was significantly improved 2 weeks later in the MAPC1591 group, with less edema, visual bleeding, synovial hyperplasia, macrophage infiltration, neoangiogenesis, cartilage degeneration, and chondrocyte apoptosis in the injured joint.

These results provide a first proof of principle that nonanticoagulant anti-inflammatory effects of endogenous aPC almost completely prevent arthropathy in a mouse hemophilia model. These findings dovetail with similar observations about the differential effects of these antibodies on the inflammatory response in a mouse model of LPS-induced septicemia.⁸ Therefore, the unknown targets of aPC that are responsible for this effect must regulate critical mechanisms driving the pathogenesis of hemarthropathy in this mouse model. Because blocking the function of the aPC/protein C receptor EPCR in the same mouse model also ameliorates hemarthropathy, the relevant candidate mechanisms of aPC are likely EPCR independent.⁹ 

Some questions remain regarding the ability of each antibody to improve hemostasis in the saphenous vein injury model, given the lack of any evident hemostatic effect in the first 5 hours after joint needle injury. These discrepancies are most likely related to model-specific differences in the severity of the bleeding insult, the type of injury and vessel, and differences in the quality of a clot formed in each model secondary to potential differences in shear rates and flow. Nevertheless, this suggests that limitation of initial bleeding after needle injury is not critical for the beneficial effects of MAPC1591 and supports experiments testing the mitigating effects of antibody dosing after joint bleeding has been induced.

Notably, a conceptually identical approach was taken in earlier studies to improve hemostasis in a nonhuman primate model of hemophilia A (antibody-mediated depletion of FVIII).¹⁰ As in the current study, an exosite-directed humanized antibody selectively inhibiting aPC’s anticoagulant functions significantly prolonged aPC’s half-life in plasma, was well tolerated in normal monkeys, and showed dose-dependent efficacy in hemophilic monkeys.

Together, these studies in 2 species affirm that anti-aPC antibodies selectively inhibiting aPC’s anticoagulant function are a viable option for rebalancing hemostasis in patients with hemophilia A and presumably, but yet to be shown, in hemophilia B. The future role of aPC-selective antibodies in the rapidly evolving armamentarium of bypassing drugs for hemophilia⁷ remains to be seen. The current work by Magisetty et al should certainly incentivize investigations into whether aPC-targeted antibodies may be particularly useful for boosting the anti-inflammatory effects of endogenous aPC (eg, via half-life prolongation¹⁰) in the prevention of arthropathy, potentially in the setting of combination therapies.