Making sense of factor VIII’s breakdown Free

In this issue of Blood, Morris et al¹ characterize the mechanisms of factor VIII (FVIII) inactivation by activated protein C (APC) and A2 domain dissociation and how activated FIX and FX affect these processes.

FVIII is a complex multidomain protein (A1-A2-B-A3-C1-C2) that plays a key role during hemostasis. FVIII acts as a coenzyme to activated FIX (FIXa), increasing its enzymatic activity to FX in the tenase complex, leading to thrombin generation. Regulation of FVIII is important, as when this is absent due to a pathogenic F8 variant (severe hemophilia A), this results in bleeding and high FVIII levels are associated with increased risk of thrombosis.² Regulation occurs at all stages in the FVIII life cycle. FVIII is secreted from the liver as an inactive heterodimer which binds von Willebrand factor (VWF). FVIII is then activated by thrombin at the site of vascular injury. This leads to B-domain loss, release from VWF, and formation of an active heterotrimer. Two mechanisms of FVIII inactivation have been proposed: spontaneous A2 domain dissociation and APC cleavage, with the former being thought to be the main mechanism.

This study elegantly combines biochemical and in vivo studies to characterize FVIII inactivation by APC cleavage and A2 domain dissociation and how these are affected by FIXa and FX. The study uses wild-type FVIII (WT-FVIII) and multiple FVIII variants: APC resistant (single: R336Q or R562Q; dual: FVIII-QQ), A2 stabilized (D519V/E665V: FVIII-VV), and APC resistant/A2 stabilized (FVIII-VVQQ). After FVIII activation, activity was lost for these variants as expected, that is, APC-resistant variants rapidly lost function due to A2 loss and A2-stabilized variants retained activity as APC is not generated in the plasma assays. After WT-FVIII activation, rapid loss of activity occurred due to A2 dissociation which was unaffected by APC. APC differentially recognized activated FVIII and inactivates this faster than WT-FVIII. The role of FIXa was studied to find whether this affects APC inactivation in addition to its known role in A2 domain stabilization. In the absence of FIXa, rapid loss of FVIII activity was found due to A2 loss and APC minimally contributed to inactivation. Increasing FIXa saturation resulted in APC inactivation becoming the more prominent mechanism through stabilization of the A2 domain, although this protection was incomplete. Additional studies demonstrated that FIXa provided some protection from APC cleavage through disruption at Arg562. FX was then found to protect against APC inactivation both biochemically and in vivo by disrupting APC cleavage at Arg336, although this was incomplete at normal FX concentrations. The key aspect of this study was finding that both inactivation mechanisms occur in vivo. Using a tail-clip bleed model in hemophilia A mice, increased potency was found for APC-resistant and/or A2-stabilized variants. This supports both A2 dissociation and APC cleavage occurring in vivo. A limitation to these findings is that the contribution of these pathways is dependent on FIXa, FX, and APC concentrations present at the injury site which is unknown.

The FVIII variants described in this study will generate interest as these may provide new options to address some of the limitations of current gene therapy approaches for treatment of hemophilia A.³ Supporting this, a recent study demonstrated stable FVIII expression and improved efficacy in hemophilia A mice treated with an adeno-associated virus (AAV) vector containing the APC-resistant FVIII-QQ variant compared with WT-FVIII.⁴ There are several areas that require consideration before translation of these variants into the clinic. In these studies, potency was assessed in vivo as this is not directly measurable using conventional plasma assays.^1,4 This poses questions on how to measure and compare outcomes from clinical trials for these variants. A second consideration is whether these approaches could result in an increased risk of thrombosis. Two gene therapy studies have recently reported transient supraphysiological expression of FVIII or FIX with associated thrombotic events.^5,6 Studies in immunodeficient hemophilia A mice treated using an AAV vector containing the APC-resistant FVIII-QQ variant revealed normal survival even with high FVIII expression.⁴ In mice with an induced prothrombotic variant, although normal survival was found when FVIII-QQ levels were below the normal range, decreased survival was found when FVIII levels were normal or elevated. Taken together, this highlights a need for greater understanding of interindividual variation in response to gene therapy, assessment of thrombotic risk, and a need to maintain some degree of physiological regulation. Finally, immunogenicity should be considered as all 4 residues mutated in these variants demonstrate a high degree of conservation across species.⁷ Although murine studies have not demonstrated increased immunogenicity to date, ongoing studies will be required if these approaches are taken forward into clinical trials.^4,8 

In summary, this study provides evidence that both A2 domain dissociation and APC cleavage occur in vivo and interactions with FIXa and FX modulate FVIII inactivation.

Conflict-of-interest disclosure: P.B. has received research funding from BioMarin and Octapharma; consulting fees or honoraria from BioMarin, Octapharma, Pfizer, Institute for Medical and Nursing Education, Novo Nordisk, and CSL Behring; and travel funding (conference attendance) from Octapharma, CSL Behring, and Pfizer.