Inhibitor formation is among the most serious complications of hemophilia treatment, being estimated to occur in 25% or more of those with hemophilia A . Inhibitor formation is a T-cell-dependent immune response to foreign infused factor VIII that renders lifesaving FVIII concentrate treatment ineffective and leads to poorly controlled bleeding, severe morbidity, frequent hospitalization, and high healthcare costs . Treatment of inhibitors is difficult: the use of bypass agents, FVIIa or FIX complex, is unpredictable and and the results are suboptimal. Immune tolerance induction, a program of regular FVIII infusions over a prolonged period of time to eradicate inhibitors, is inconvenient, costly, and ineffective in up to 20% of cases . The lack of effective hemostasis results in uncontrolled bleeding and significant morbidity. Inhibitors also complicate the current standard of care, three times weekly FVIII prophylaxis (preventive FVIII), to prevent joint bleeds and arthropathy , recommended by the Medical and Scientific Advisory Committee of the National Hemophilia Foundation . Inhibitors complicate the placement of central lines required to infuse standard prophylaxis, and may complicate gene transfer, if directed at FVIII expressed by the transgene. Thus, a major goal of hemophilia care is to prevent the formation of inhibitors.
Among the approaches to prevent inhibitor formation in patients with hemophilia A is the avoidance of risk factors associated with inhibitor formation. Thus, whereas patient-related risk factors, such as race, family history, and FVIII genotype, cannot be changed, those that are treatment-related, such as dose, intensity, and purity of factor, can be modified . Some approaches to reducing inhibitor formation include limiting or avoiding high-dose, high-intensity or high purity-factor product exposure. Regarding the latter, a number of small, uncontrolled clinical studies have confirmed that von Willebrand factor (VWF)-containing FVIII concentrates result in lower immunogenicity (inhibitor formation) than pure FVIII products in patients with hemophilia A [7–9]. Furthermore, VWF-containing products have shown more success in immune tolerance induction in hemophilia A patients with inhibitors than have pure FVIII concentrates . Moreover, despite the presence of ultralarge VWF multimers in these products, with their potential to cause platelet aggregation or microvascular thrombosis, VWF-containing FVIII concentrates have shown similar safety and efficacy to FVIII concentrates, with no reports of thrombosis [7–9].
What is the mechanism by which the simple presence of VWF in clotting factor concentrates protects FVIII against inhibitor formation in clinical patients? It is known that FVIII circulates as a complex with VWF, which is required for FVIII stability and survival in the circulation, as in cell culture [11,12]. When vessel damage occurs, FVIII is activated through thrombin cleavage, binds to phospholipid surfaces on damaged cells and platelets, and serves as a cofactor for FIXa to activate FX through the intrinsic pathway, with subsequent clot formation. VWF serves as the glue enabling platelets to bind to damaged vascular endothelium and form platelet plugs, allowing subsequent fibrin clot formation. Yet, without VWF binding to FVIII, FVIII would be inactivated by proteinases in the circulation, preventing normal clot formation. In the presence of anti-FVIII inhibitor antibodies, normal FVIII function is blocked and normal clot formation is prevented. Anti-FVIII antibodies are directed against epitopes on the heavy and light chains of FVIII, primarily the A2 domain of the heavy chain and the C2 domain of the light chain . The latter site, the C2 domain, is the phospholipid-binding site  where VWF binds to FVIII, and also where anti-FVIII antibodies interfere with or block FVIII phospholipid binding.
How does VWF reduce the immunogenicity of FVIII? First, by complexing with and protecting FVIII from degradation by plasma proteinases, VWF may prolong antigen presentation. Second, VWF is known to mask FVIII epitopes within the A2, A3 and C2 domains , primarily within the light chain C2 domain (Fig. 1). Thus, by masking the FVIII C2 epitope, VWF may partially block the binding of inhibitor to FVIII and reduce the potential for inhibitor formation. In one of the first published studies evaluating the effects of a VWF-containing concentrate on the treatment of inhibitor patients, Berntorp  showed that FVIII recovery was significantly improved in an inhibitor patient with a high-responding inhibitor against the FVIII C2 domain, following treatment with a VWF-containing concentrate. Corroborating this findings, Suzuki demonstrated that, in a panel of 14 inhibitor patient plasma samples, those with C2 domain specificity were less active against VWF-bound FVIII than purified or recombinant FVIII . Recently, Delignat et al. reported that VWF may also reduce the immune response to FVIII by interfering with antigen recognition. In in vitro and in vivo hemophilia A models, they found that VWF prevented internalization of FVIII by bone marrow dendritic cells (antigen-presenting cells), which hindered antigen recognition of FVIII . VWF has also been shown to interfere with FVIII binding to lectin receptors specific for mannose residues, which is important in antigen presentation of human therapeutic self-proteins to CD4+ T lymphocytes . In addition, VWF-containing FVIII concentrates show reduced FVIII inactivation  and increased thrombin generation  and FVIII recovery . Finally, the binding of VWF to FVIII in the circulation to protect it from degradation increases the half-life of FVIII in the circulation, which may allow increased contact with tolerogenic marginal zone B cells in the spleen .
Despite these findings, there has been controversy regarding whether VWF reduces FVIII immunogenicity. This has arisen, in part, because findings in some studies in the hemophilia A mouse model have differed from those in clinical studies. What is the basis for this difference between the clinical and mouse model data? First, the clinical findings, although impressive, are based on a small number of uncontrolled studies. Second, the mouse studies have methodologic differences, in particular the sequence in which VWF is added to the reaction mixture to simulate VWF-containing FVIII products.
The study by Shi et al.  takes us one step closer to the answer. In a series of elegant in vitro and in vivo studies (Table 1), Shi et al. demonstrate that VWF exerts its protective effect as a complex with FVIII, rather than by competitive binding of VWF with inhibitor antibodies to FVIII. In a series of in vitro mixing studies, combining FVIII or FVIII plus VWF with inhibitors from one of three sources (plasma from immunized VWFnullFVIIInull mice; purified plasma IgG from an inhibitor patient; or human mAb from B-cell clones from an inhibitor patient), Shi et al. observed that residual FVIII activity was greater when VWF was in the mixture than when FVIII alone was present. The VWF protective effect was attributed to formation of aVWF–FVIII complex, resulting in less FVIII being available to bind inhibitor. The elegant work of Shi et al. also explains why previous studies , in which inhibitor antibody was already present in test mixtures, i.e. the inhibitor has a chance to bind with FVIII present, unprotected by VWF, found no reduction in immunogenicity.
|1. Plasma from FVIII-immunized VWFnullFVIIInull mouse|
|2. Purified IgG plasma from hemophilia A inhibitor patient|
|3. Human mAb from B-cell clone from inhibitor patient|
|In vitro studies||Outcome: residual FVIII (chromogenic assay) and inhibitor titer (Bethesda)|
|Inhibitor (1) + rhFVIII||Findings: there is higher residual FVIII and a lower inhibitor titer when VWF is in the admixture|
|Inhibitor (2) + rhFVIII||Mechanism: there is greater inhibitor binding by VWF, so inhibitor is unable to bind FVIII|
|Inhibitor (3) + rhFVIII||Less bound FVIII results in higher residual FVIII in chromogenic assay|
|Inhibitor (1) + rhFVIII + rhVWF||Less bound FVIII results in lower immunogenicity and lower anti-FVIII titer|
|Inhibitor (2) + rhFVIII + rhVWF|
|Inhibitor (3) + rhFVIII + rhVWF|
|In vivo studies||Outcome: tail clip bleeding time|
|Inhibitor (1) + rhFVIII||Findings: there is shorter bleeding time following tail clip when VWF is in the admixture|
|Inhibitor (2) + rhFVIII||Mechanism: there is greater inhibitor binding by VWF, so less FVIII is bound by inhibitor|
|Inhibitor (3) + rhFVIII||Less bound FVIII results in a higher FVIII level and better hemostasis after tail clip|
|Inhibitor (1) + rhFVIII + rhVWF|
|Inhibitor (2) + rhFVIII + rhVWF|
|Inhibitor (3) + rhFVIII + rhVWF|
In in vivo studies, Shi et al. demonstrated that, when FVIIInull mice were infused first with FVIII and then with inhibitors, allowing FVIII to form a protective complex with VWF before encountering inhibitors, the mice survived tail clip. By contrast, when inhibitors were infused before FVIII, and FVIII was exposed to VWF and inhibitor antibody (preformed) at the same time (before a protective complex could be formed with VWF), most animals, and all at high titer inhibitor, did not survive tail clip.
Thus, the difference between results from clinical studies and mouse model studies appears to be explained by the critical formation of VWF–FVIII complex, i.e. the protection afforded by VWF when mixing experiments allowed preassociation of infused FVIII with VWF. This is analogous to the clinical situation in which VWF-containing concentrate is infused into an inhibitor patient who has circulating pre-existing FVIII antibody.
In summary, the studies by Shi et al. provide bench evidence for the clinical observations that VWF-containing FVIII concentrate products may reduce or prevent inhibitor formation in hemophilia A patients. These findings raise the question of whether VWF-containing products should be used to treat hemophilia A to reduce or prevent inhibitor formation. The answer to this question will require confirmation in randomized clinical trials. If confirmed, these findings could have significant clinical and economic impacts, alleviating the most difficult management issue of hemophilia, inhibitor formation. Moreover, if inhibitor formation could be prevented rather than treated, quality of life could be significantly improved by the avoidance of onerous daily tolerizing factor infusions, complicating port infections, and poorly controlled bleeding . Finally, if clinical trials prove that VWF-containing FVIII concentrates prevent inhibitor formation, this could impact on the single greatest unmet health need in hemophilia, i.e. inhibitor prevention, and would challenge the current paradigm that FVIII concentrate is the treatment of choice for hemophilia A.
VWF binds to FVIII A3 and C2 domains, thereby preventing inhibitor binding to FVIII.