Following the recognition of von Willebrand disease (VWD) in 1926 and the cloning of the gene for von Willebrand factor (VWF) in 1985, significant advances have been made in our fundamental knowledge of both the disease and the protein. Some of this new knowledge has also begun to impact the clinical management of VWD. First, the progressive increase in our understanding of the molecular genetic basis of VWD has resulted in rational applications of molecular testing to complement the current range of phenotypic tests for VWD. These molecular genetic strategies are most effectively directed at the prenatal diagnosis of type 3 VWD and confirmatory testing for types 2B and 2N disease. In contrast, the use of molecular testing to clarify the diagnosis of type 1 VWD is of marginal benefit, at best. In terms of VWD therapies, a new recombinant VWF concentrate has recently completed successful clinical trials and is now awaiting more widespread application. There have even been some preclinical successes with VWF gene transfer although the clinical rationale for this therapeutic strategy needs careful consideration. Much more remains to be learnt about the biology of VWF and further translational advances for the enhancement of VWD care will inevitably be realized.
It is now 87 years since Erik von Willebrand described the occurrence of a severe mucocutaneous bleeding disease in several families living in the Aland archipelago . At the time, it was clear that there was a hereditary component to this condition and that its inheritance pattern was distinct from hemophilia. With these features in mind, the disease was given the name hereditary pseudohemophilia, but the mechanistic basis of the bleeding did not begin to be elucidated until 30 years later. In 1957, Professor Inge Marie Nilsson and Margareta Blombäck revisited the investigation of the Aland Island families and documented prolonged bleeding times and a reduction in the level of plasma antihemophilic globulin (now recognized as factor VIII—FVIII) in affected patients. They also showed that the infusion of plasma fraction 1–0 (containing several large proteins including VWF, FVIII, and fibrinogen) corrected the hemostatic defect in these patients .
The next major advance in understanding the basis of this disease occurred in 1971, with the development of an immunological test to differentiate VWF and FVIII . This test now made it possible to diagnose patients with VWD and to identify those with hemophilia A in whom isolated FVIII deficiency existed . This new assay was also used to demonstrate that variant forms of defective VWF could be demonstrated in some patients , and by 1987, the subclassification of VWD had reached a peak through the application of phenotypic analysis, with the description of 11 type 1 variants and 13 variants of type 2 VWD .
However, as with a large number of monogenic traits, it was the cloning of the VWF cDNA, simultaneously by four laboratories in 1985, that initiated the advances in knowledge pertaining to the biology of VWF and the clinical management of VWD that we have witnessed over the past 27 years [7-10].
The purpose of this review article is to highlight how the new molecular knowledge of VWD has been translated into advances in the clinical care for this condition.
Brief summary of VWF genetics and protein structure
The VWF gene on chromosome 12p comprises 178 kb of genomic DNA arranged into 52 exons ranging in size from 41 bp to 1.3 kb (exon 28) . The genetic analysis of VWF is complicated by the fact that a partial VWF pseudogene is located on chromosome 22 that replicates the chromosome 12 sequence between exons 23 and 34, with a 3% variance .
The VWF gene encodes a 9 kb transcript that is translated into a 2813 amino acid prepropolypeptide sequence . The 22 amino acid presequence enables entry into the endoplasmic reticulum, and the 741 residue propeptide mediates alignment of VWF dimers into N-linked multimeric forms. The VWF propeptide is cleaved from the mature subunit in the golgi and is secreted into the plasma as disulfide-linked dimers . The 2050 residue mature VWF subunit possesses binding domains for FVIII (D'D3 domains), platelet glycoprotein Ib (A1 domain), collagen (A1 and A3 domains), and platelet glycoprotein IIb/IIIa (C1 domain). These multiple binding functions enable VWF to mediate platelet adhesion and aggregation  and to protect FVIII from premature proteolysis by activated protein C . A recent study utilizing electron microscopy images of VWF to establish relationships between protein sequence and structure has presented a re-annotated arrangement of the protein .
Following secretion from its cell of synthesis (either endothelial cells , ~85% of plasma VWF, or platelets—derived from megakaryocytic expression ), VWF is subjected to limited proteolysis by the metalloprotease ADAMTS-13, through cleavage between Tyr1605–Met1606 in the VWF A2 domain . Access to this cleavage site requires partial unfolding of the A2 domain through a shear-mediated mechanism .
Background to the application of molecular science in von Willebrand disease
During the past two decades, there has been a growing appreciation of the clinical impact of VWD around the world. Ironically, while the first described index case of VWD bled to death during her 4th menstrual period, the adverse effect on the quality of life of women with VWD during their reproductive years has only recently been translated into more focused care for this population [22, 23]. This increase in awareness of the disease, along with laboratory tests of the phenotype that are temporally variable  and challenging to perform for many less experienced laboratories, has spurred the exploration of alternative, molecular strategies for disease identification.
Translational advances for von Willebrand disease diagnosis
In any population, the ratio of types 1, 2, and 3 VWD will be approximately 65–75:25–30:1–6 , although with enhanced phenotypic testing (especially, with higher quality multimer analysis) more type 2 variants become apparent . In 2013, the optimal diagnostic strategies will vary for each of these subtypes of the disease (Fig. 1).
Type 3 von Willebrand disease diagnosis
Type 3 disease should be readily apparent from VWF:Ag and VWF:RCo levels < 0.05 IU mL−1 and a FVIII:C of < 0.10 IU mL−1 . Many type 3 cases will have undetectable levels of VWF, while there are rare atypical cases in which FVIII:C levels may be > 0.10 IU mL−1. Thus, for this VWD subtype, there are only two legitimate rationales for utilizing molecular approaches to diagnosis: first, in the context of prenatal testing where mutation-specific diagnosis can be accomplished as early as 10 weeks into a pregnancy on chorionic villus material [28, 29] and second, to obtain information that might facilitate prediction of alloantibody development. This latter, rare but critical, complication of VWF concentrate use has been associated in a small number of reports with VWF gene deletions [30, 31]. Thus, while no systematic genotype/phenotype analysis has been performed to assess this association, it is reasonable to assume that any null (as opposed to missense) mutations would be more likely to be associated with alloantibody generation .
All other molecular studies of type 3 VWD are aimed at a better understanding of the biosynthetic pathogenesis of this severe quantitative trait [33-36], and while this knowledge is biologically interesting, the potential for translational diagnostic advances is currently limited.
Type 2 von Willebrand disease diagnosis
In contrast to type 3 VWD, the diagnosis of type 2 forms of the disease may often cause problems using routine tests of the VWF phenotype. While some of these tests are relatively reproducible, others such as the VWF:RCo and VWF multimer analysis are challenging to standardize and interpret [37-39] and can produce data that result in diagnostic uncertainty. In some of these instances, the addition of molecular analysis will be of diagnostic assistance [29, 40].
Molecular diagnosis of type 2A von Willebrand disease
In type 2A VWD, where the loss of high molecular weight VWF multimers (HMWM) and markedly reduced ristocetin-induced platelet agglutination should be diagnostic, ~80% of the causative missense mutations are located in exon 28 of the VWF gene and encode amino acid substitutions in the VWF A2 domain. These substitutions result in both group I (lack of production of HMWM) and group II (enhanced ADAMTS-13-mediated proteolysis) pathogenic mechanisms  but unfortunately, only a few of the > 70 type 2A VWD mutations have been investigated in sufficient detail to predict which mechanism is associated with specific mutations. One exception to this statement is the mutations resulting in the 2A(IIE) subtype in which D3 domain missense mutants predominate. Nevertheless, a recent in depth study of type 2A pathogenesis suggests that many of these mutations are associated with a complex combination of mechanisms affecting VWF multimer formation, ADAMTS-13-mediated proteolysis, secretion, and storage .
Thus, overall, the potential for molecular testing in type 2A VWD to provide diagnostic clarity and predictive information concerning desmopressin responsiveness is limited to those few previously reported mutations for which detailed pathogenic studies have been completed. Aside from these cases, the diagnosis of type 2A VWD is still more efficiently made using standard phenotypic tests.
Molecular diagnosis of type 2B von Willebrand disease
The characteristic phenotype of type 2B VWD involves low VWF:Ag and VWF:RCo levels and usually a low VWF:RCo/VWF:Ag ratio (< 0.6); absent VWF HMWMs; mild thrombocytopenia associated with platelet clumping on blood smears, and most importantly, evidence of increased sensitivity to ristocetin-induced platelet agglutination. While this constellation of features is often apparent, the potential for molecular confirmation of the diagnosis is significant and practically feasible with almost all of the > 20 causative missense mutations being located in exon 28 (encoding residues in the A1 domain of VWF). Not only does the molecular analysis of type 2B VWD provide diagnostic confirmation, but it also rules out the complicating matter of the pathological genocopy, platelet-type VWD, in which mutations are located in the GPIBA gene .
In addition to providing diagnostic benefit, the molecular definition of type 2B mutations has also been shown to predict the phenotypic pattern and severity of the disease for at least some of the common, recurrent mutations. Thus, patients carrying the P1266L/Q or R1308L variants rarely develop thrombocytopenia (even after desmopressin infusion) and have low bleeding tendencies, while patients that have inherited the V1316M, R1308C, and S1310F mutations have a more severe phenotype .
Molecular diagnosis of type 2M von Willebrand disease
Type 2M VWD represents the loss of function equivalent of type 2B disease, in which missense mutations, most of which are located in the VWF A1 domain, result in significantly reduced binding to the platelet GpIb alpha receptor. In this VWD subtype, a reduced VWF:RCo/VWF:Ag ratio (< 0.6) is associated with a normal multimer pattern.
The main differential diagnosis of type 2M VWD is type 1 disease, where again multimer patterns are normal and sometimes VWF:RCo/VWF:Ag ratios may approach or be < 0.6. The small amount of evidence that exists suggests that the lower the VWF:RCo/VWF:Ag ratio is, the more likely it is that a type 2M missense mutation will be found in the VWF A1 domain and the less likely that desmopressin will produce an adequate therapeutic response (i.e. more likely to be type 2M as opposed to type 1 VWD) .
Overall, without additional knowledge of specific genotype–phenotype correlations, the current molecular definition of ‘classical’ type 2M VWD is not sufficiently beneficial to merit routine analysis for this disease subtype.
Molecular diagnosis of collagen-binding variants of type 2M von Willebrand disease
While the original designation of type 2M VWD applied to variants with loss of platelet dependent function, several families have now been described in whom collagen-binding defects have been characterized. There are two collagen-binding domains in VWF: the A3 domain that binds to collagen types I and III  and the A1 domain that binds collagen types I, III, and VI [47, 48]. The patients who have been diagnosed with collagen-binding mutations have usually been heterozygous for missense substitutions in either the A3 or A1 domain. Their bleeding phenotypes have been mild, and the only abnormal phenotypic test has been a reduced collagen-binding assay result.
Four collagen-binding missense substitutions have been identified in the VWF A3 domain: S1731T, W1745C, S1783A, and H1786D [49-51]. These mutants show a pattern of dominant inheritance with reduced binding in heterozygotes to collagens I and III.
Very recently, the first mutations adversely affecting binding to collagen VI have been described: R1399H, S1387I, and Q1402P . These missense variants are all located in the A1 domain. However, to complicate matters, these variants have also been found in some healthy individuals without bleeding symptoms. More studies are needed to determine the impact of abnormal collagen binding on the clinical bleeding phenotype . In some instances, it appears that these variants are sufficient alone to result in bleeding, whereas in other instances, these changes may act as secondary genetic modifiers of the phenotype.
The recent increased interest in collagen-binding defects in VWF raises important questions about the role of this investigation in patients with mild bleeding disorders [53, 54]. While phenotypic tests of collagen binding have been employed in diagnostic laboratories for a considerable period of time, their uptake for routine diagnostic purposes has shown major geographical variances. The recent determination of small numbers of families with collagen-binding defects suggests that this phenotypic test should probably be adopted as a routine component of the diagnostic algorithm for patients with mild bleeding symptoms and that any functional deficit should also be complemented by sequence analysis of the A1 and A3 regions of VWF.
Molecular diagnosis of type 2N von Willebrand disease
Type 2N VWD is now a well-recognized genocopy of mild/moderate hemophilia A in which the molecular defect in VWF results in poor binding to FVIII . The > 25 different mutant type 2N alleles are clustered between exons 17–25 of the VWF gene and result in missense substitutions in the FVIII-binding D'/D3 region of the protein .
Type 2N VWD is a recessive trait, and thus, affected subjects acquire either two type 2N mutant alleles (homozygosity or compound heterozygosity) or one type 2N mutation and a VWF null allele. This latter group of 2N patients may also exhibit lower plasma levels of VWF:Ag in addition to low FVIII levels.
There is certainly a role for the molecular diagnosis of type 2N VWD. First, the alternative, functional FVIII-binding assay is not readily available in many locations and can produce results that are indeterminate. However, most importantly, the genetic test will provide definitive information that can be used to evaluate genotype/phenotype associations in type 2N patients. Thus, the most common mutation, R854Q, is associated with FVIII levels of ~25% and a reasonable likelihood that desmopressin administration will provide a satisfactory therapeutic benefit, for at least minor procedures. In contrast, the recurrent missense substitutions, R816W and T791M, result in levels of FVIII < 10% and poor desmopressin responsiveness.
Molecular diagnosis of type 1 von Willebrand disease
Type 1 VWD represents the most common form of VWD in all populations, accounting for approximately 65–75% of cases. The phenotypic diagnosis is made on the basis of a clinical bleeding history, laboratory evidence of mild/moderate reductions in normally functional plasma VWF and often, the presence of a family history. Most cases of type 1 VWD show a dominant inheritance pattern but the condition exhibits significant levels of incomplete penetrance and variable expressivity of the phenotype . The plasma levels of VWF are also influenced by a range of environmental factors such as physical stress, hormones, intercurrent illness, and aging, all of which can significantly complicate the phenotypic diagnosis of type 1 disease.
During the past decade, several studies have begun to reveal the genetic background to this quantitative trait [58-61] and while there may have been early hopes that a straightforward genetic answer might be forthcoming to facilitate type 1 VWD diagnosis, exactly the opposite has occurred.
All four of the initial type 1 VWD genetic studies (comprising > 500 index patients) have documented candidate mutations in the promoter, exons, and splice junctions of the VWF gene in ~65% of index cases. While the absence of putative mutations in these regions of the gene in 35% of affected subjects is noteworthy, equally concerning is the fact that the pathogenic mechanism associated with many of the documented candidate variants (the majority of which are missense substitutions) has yet to be characterized. The importance of this latter shortcoming has been highlighted by a recent report in which several of these type 1 VWD candidate mutations were found in subjects with no bleeding history . Many of these subjects were of African American origin, further emphasizing the additional complexity that ethnicity will bring to the interpretation of molecular analysis in type 1 disease. This theme has recently been further highlighted by results of VWF gene analysis from the 1000 Genomes Project where again multiple VWF variants have been found in healthy subjects including several of the previously designated type 1 VWD candidate mutations .
Although there is no consensus, the rationale to undertake molecular genetic diagnosis for type 1 VWD currently appears tenuous, at best [63, 64]. Even for what might be the most therapeutically justifiable variant of type 1 disease to evaluate, type 1C VWD, in which accelerated clearance of the mutant VWF will result in an exaggerated but short-lived response to desmopressin, the performance of genetic analysis will add nothing more to a standard 4-h desmopressin trial of therapy. The baseline VWF:Ag levels in these type IC patients are usually unequivocally low (< 0.20 IU mL−1), and their abnormal response to desmopressin is readily apparent [65, 66]. Similarly, evidence that genetic analysis can provide information that would obviate the need to perform a trial of desmopressin for other type 1 patients is marginal .
Translational advances for the treatment of von Willebrand disease
At a time when there is significant excitement in the hemophilic community concerning the imminent introduction of a variety of new treatments into the clinic [68, 69], it is notable that a similar phenomenon has not occurred for the future management of VWD.
There are several reasons why this may be the case. First, the natural history of bleeding in VWD, especially for type 1 disease, is, of course, associated with far less morbidity, and even in patients with type 3 disease, many patients do not bleed frequently enough to merit long-term prophylactic therapy. Secondly, the current therapies for VWD including desmopressin , plasma-derived VWF/FVIII concentrates, and adjunctive therapies such as antifibrinolytic agents, and for women with VWD-associated menorrhagia, the use of oral contraceptives and the progestagen-eluting intrauterine device  are all highly effective and safe. Thus, neither the clinical rationale nor commercial incentive for significant innovation in the treatment of VWD is obvious.
Despite the relatively satisfactory state of therapy for VWD, there are still a number of challenges that require further investigation to optimize management approaches. These situations include the resistant nature of gastrointestinal bleeding in VWD, particularly when associated with angiodysplasia; the development of improved schedules of VWF dosing for major surgery and secondary long-term prophylaxis and finally, the development of strategies for treating patients with anti-VWF alloantibodies.
Against this background, there are several molecular innovations that justify further discussion (Fig. 2).
The most significant change to the VWD treatment landscape is the very recent arrival of a recombinant VWF (rVWF) concentrate that has just completed its phase III clinical trial [72, 73]. This product is manufactured from an existing Chinese hamster ovary cell line, where the VWF is initially expressed in combination with recombinant FVIII (rFVIII). In the preparation of the recombinant concentrate, the two proteins are initially separated and purified in isolation, before being reconstituted at a ratio of 1.3:1 (VWF:RCo to FVIII:C) prior to infusion. In vitro analysis of this material shows that the rVWF displays a full array of large and ultra-large MW multimers as would be expected from a product that has not been exposed to ADAMTS-13 proteolysis during production.
The clinical trials of this new concentrate have demonstrated excellent hemostatic efficacy and safety. Excellent control of bleeding has been documented and a detailed evaluation of immunogenicity has shown no evidence of either functionally neutralizing or non-neutralizing antibodies that result in accelerated VWF clearance. Of note, two of the clinical trial participants were found to possess VWF-binding non-neutralizing antibodies (and had reduced recoveries and shorter VWF half-lives), but these antibodies were present at trial enrollment and were presumably related to prior therapy with a plasma-derived VWF-FVIII concentrate. Other key features of these studies have documented the rapid disappearance of the ultra high molecular weight VWF multimers and the coincident demonstration of ADAMTS-13-mediated proteolytic fragments of VWF within 15 minutes of product infusion. Furthermore, there were no clinical or laboratory signs of a microangiopathic state after administration of the concentrate.
We now await the licensing and more widespread clinical use of this new recombinant product. Interesting questions remain as to the clinical niche for this concentrate and whether the final product will be distributed as an rVWF/rFVIII mixture or as a stand-alone VWF concentrate.
Additional translational considerations for future von Willebrand factor therapies
While not directly aimed at the treatment of VWD, there is growing interest in developing strategies that would substantially prolong the half-life of FVIII, which currently appears to be limited by FVIII's interaction with VWF . Although the upcoming generation of novel FVIII products will extend the protein's half-life by 1.5- to 2-fold, it remains to be seen whether this will substantially alter clinical practise. With this background, future strategies to extend the FVIII half-life more significantly will need to employ approaches that either modify FVIII in such a way that its dependence on VWF is eliminated or that generate VWF molecules that possess an extended survival. Clearly, this latter group of molecules, if attainable, would be of interest for the treatment of VWD as well as hemophilia A.
von Willebrand factor gene therapy
If the rationale for translational advances in protein replacement for VWD is limited, it is hardly surprising that strategies for VWF gene transfer have not attracted much attention. Realistically, the only type of VWD where one would even consider the application of gene transfer would be type 3 disease, and even in this very small population, there are variable bleeding patterns, with many patients not requiring prophylactic treatment.
Despite, the lack of a clear clinical rationale, there have been preclinical investigations of the feasibility of VWF gene transfer that have shown some promise . From the point of view of a gene therapist, VWF represents a significant practical challenge: its 9 kb cDNA is too big to package in most gene therapy vectors, and its complex post-translational modification may not be achievable in all cell types.
The two VWF gene transfer approaches that have been reported to date utilize different transgene delivery strategies. The first approach involves the ex vivo delivery of the VWF transgene via lentivector transduction of syngeneic endothelial progenitor cells . In contrast, the second report describes an in vivo transgene delivery approach using hydrodynamic injection . It should be emphasized that while both of these studies document therapeutically relevant levels of VWF in the mouse model of type 3 VWD, the translation of these findings to human VWD is not likely in the foreseeable future.
Future considerations for translational medicine advances in von Willebrand disease
It is now 27 years since the cloning of the VWF gene, and our knowledge of VWF biology and genetics has advanced dramatically. In parallel with these basic advances, there have also been some translational benefits that have been highlighted in this review. However, if one is to be honest, and in comparison with the translational advances associated with its partner protein FVIII, the pragmatic benefits for the diagnosis and clinical management of VWD derived from 27 years of molecular science are modest. Why is this, and does this apparent lack of translational progress suggest that we should abandon further basic studies of this protein?
Some of the clinical and practical reasons for the relative dearth of new VWD diagnostics and therapies have already been detailed in the section above, addressing new treatments. In addition, and not previously addressed, this is a very large and biologically complex protein that undergoes a highly structured biosynthetic process and whose hemostatic function is regulated by a unique, shear-mediated proteolytic processing step. Even if there were a high clinical demand for innovation, delivering on such a promise would be challenging.
Nevertheless, and to emphatically answer the second question posed above, of course we must continue to investigate the physiology and pathology of this protein, because it is inevitable that we will learn things along the way that will be of clear translational value. The best current and ongoing example of this phenomenon relates to the FVIII half-life developments alluded to above. This clinically driven objective highlights the key importance of understanding more about the VWF–FVIII interaction and the natural life cycles of both of these proteins. There is also increasing evidence that VWF is involved in a range of additional biological functions relating to inflammation, vascular and cancer biology. Only through the attainment of new fundamental information will we be able to pursue translational advances that focus on the multiple roles of VWF .
Finally, while most people would argue that the application of molecular testing to the diagnosis of type 1 VWD is not currently of value, our ongoing investigation of this complex quantitative trait will certainly yield knowledge that will be of direct or indirect benefit to the future clinical management of VWD and other similarly complex genetic conditions.
von Willebrand disease is the most prevalent inherited bleeding disorder in humans, and studies over the past three decades have substantially enhanced our knowledge of the biology of VWF. Some of these advances have begun to impact the diagnosis and clinical treatment of the disease, and we now await further translational progress as we continue to characterize the basic biology of this multifaceted protein.
DL is the recipient of a Canada Research Chair in Molecular Hemostasis. The author's VWD research is supported by the Canadian Institutes of Health Research, the Heart and Stroke Foundation of Canada, the US National Institutes of Health, and the Canadian Hemophilia Society.
Disclosure of Conflict of Interest
The author states that he has no conflict of interest.