Practical diagnostic considerations
Pre-analytical considerations
Clear preanalytical patient guidance regarding infection, time after potential surgery, and smoking, fasting or heavy exertion (should not be less than 24 hours) should be provided.
Recommendations
Blood samples should preferably be collected in the coagulation laboratory to avoid preanalytical challenges.
Samples should ideally be collected from fasting and resting subjects who have refrained from smoking and caffeine ingestion on the day of testing. Exercise, stress, infections, and pregnancy all elevate VWF and FVIII levels and may obscure the diagnosis of mild VWD type 1. Blood can be collected without consideration of the menstrual cycle [1,2].
Use a standardized, atraumatic blood collection protocol with minimal stasis.
In children and teenagers, application of anesthetic plaster prior to blood sampling should be used.
Use needles between 19 and 21 gauge to prevent vein trauma or reduced blood flow, leading to platelet activation.
The first 3-5 mL of blood should not be used for VWD testing.
VWD tests require the use of 105-109 mmol/L (3.2 %) buffered trisodium citrate tubes for blood collection.
Tubes must be filled completely to ensure the proper 9:1 ratio of blood to anticoagulant.
The collection tubes should be mixed gently immediately after filling, avoiding excessive agitation or mixing.
Care should be taken to remove platelets from plasma by centrifuging to prevent platelet VWF from contaminating plasma on freezing and thawing the sample. Some laboratories use double centrifugation. If not tested immediately, plasma samples should be frozen immediately at -70°C.
If transportation of plasma samples is needed, samples should be kept frozen. If not frozen there is a risk of cold activation if cooling of the plasma sample occurs during transportation. This results in decreased VWF and FVIII activities and, therefore, may lead to misdiagnosis of VWD [3].
Screening tests should include complete blood count with differential leukocyte analysis and platelet count, Owren and/or Quick PT, APTT, VWF activity (we recommend to use both platelet-dependent and collagen binding activity assays), VWF:Ag, FVIII level, and ABO blood type. In urgent situations, the use of PFA-100/200 may help to exclude moderate and severe forms of VWF. Prolongation of the closure times is sensitive but not specific for VWD and may indicate anemia, thrombocytopenia, platelet dysfunction or antiplatelet agents.
Defining the subtype of VWD: A ratio of VWF:activity/VWF:Ag <0.7 defines type 2 VWD and RIPA and VWF:CB and/or VWF:multimers, and VWF:FVIIIB further defines the type 2 subtype. Mutational analysis may also help to define the VWD subtype. The use of VWF:CB and VWF:multimer assays will reduce potential errors in VWD misidentification and may add information in the differentiation between 2A and 2M subtypes [4,5].
Genotyping VWD
Genotyping of VWD has its main impact when the VWD phenotype is difficult to discern, and it helps to find the differential diagnosis, such as hemophilia A or platelet defects. Genetic analysis is also useful in prenatal diagnostics and in genetic counseling. With new and powerful genotyping techniques, the knowledge about new variants and modifying genetic factors is steadily increasing. However, pathogenic variants are not always easy to be identified in all VWD subtypes, especially in VWD with mild clinical phenotype (i.e., VWD type 1). Pathogenic variants are most often found in severe quantitative VWF deficiency (VWD type 3) and in qualitative deficiencies (VWD type 2), in contrast, this is not the case in <65% of VWD type 1 alleles [6–8].
VWD type 1 is the most prevalent subtype, and it is in majority of cases dominantly inherited but some patients may display recessive inheritance with more than one mutation. The mutations are located throughout the VWF gene. Genetic analysis in case of suspicion of VWD type 1 is not recommended for other purposes than exclusion of VWD.
The locations of the VWD type 2 mutations depend on the VWD subtype, 2A, 2B, 2 M or 2N, but are usually found in a limited number of exons. Most often, they are caused by missense mutations that change specific amino acids, resulting in either loss- or gain-of-function phenotypes. In type 2A, the mutations are most often caused by variants in exon 28 that encodes the A2 domain of the VWF molecule or by mutations in exons 22 and 25-27 (D3 domain). Variants causing type 2A have also been found in exons 11-16 (D2 domain) and in exons 51-52 (CK domain). In type 2B, the pathogenic mutations are almost invariably gain-of function variants found in exon 28, in the region encoding the A1 domain, leading to enhanced interaction with platelet GPIb molecule, and leaving the in vivo hemostasis deficient. In VWD type 2M the mutations are found in exon 28 (A1 domain) or exons 29-32 (A3 domain). The inheritance of VWD type 2 N is recessive and the mutations, found in exons 17-20 or 24-25, impair the binding of FVIII to VWF and result in shortened FVIII half-life and thus, lower levels of FVIII in plasma.
VWD type 3 is inherent in a recessive way and the pathogenic variants are also found throughout the entire VWF gene, but null mutations, resulting in complete loss of VWF synthesis, are found in majority of described cases. An investigation from 2013 concluded that VWD type 3 in Finland is caused mainly by two founder mutations, unlike reports from other populations [9].
Mutations causing VWD and polymorphisms of the VWF gene can be found in an online VWF database on Internet (dbs.eahad.org) that is supported by the European Association for Haemophilia and Allied Disorders (EAHAD).
Bleeding symptoms
VWD is characterized by prolonged and reoccurring mucocutanous bleeds, e.g., epistaxis and menorrhagia, bleeding after tooth extraction or surgery, and bleeding after minor wounds. Typically, the bleeding symptoms originate from several sites. Joint bleeds and frequent gastrointestinal bleeds occur in the most severe cases [10]. Suggestive mucocutaneous bleeding symptoms are defined as:
Nose bleeding, ≥2 episodes without a history of trauma not stopped by short compression of <10 min, or ≥1 episode requiring blood transfusion.
Cutaneous hemorrhage and bruising with minimal or no apparent trauma, as a presenting symptom or requiring medical treatment.
Prolonged bleeding from trivial wounds, lasting ≥15 min or recurring spontaneously during the 7 days after wounding.
Oral cavity bleeding that requires medical attention, such as gingival bleeding, or bleeding with tooth eruption or bites to lips and tongue.
Spontaneous gastrointestinal bleeding requiring medical attention or resulting in acute or chronic anemia, unexplained by ulceration or portal hypertension.
Heavy, prolonged, or recurrent bleeding after tooth extraction or other oral surgery, such as tonsillectomy and adenoidectomy, requiring medical attention.
Menorrhagia resulting in acute or chronic anemia, or requiring medical treatment, not associated with known structural lesions of the uterus, e.g. congenital defects, myomas.
Bleeding from other skin or mucous membrane surfaces requiring medical treatment (e.g., eye, ear, respiratory tract, genitourinary tract, other than uterus).
Newborns can show cephalohematoma, umbilical bleeding or bleeding after circumcision.
Bleeding score
A bleeding score (BS) has been developed in a large European cohort of patients with type 1 VWD with an aim to quantitatively evaluate the severity of bleeding symptoms and correlation with clinical and laboratory features [11]. The BS showed a strong significant inverse relation with VWF:RCo, VWF:Ag or FVIII:C. Higher BS was related with increasing likelihood of VWD, and a mucocutaneous BS was strongly associated with bleeding after surgery or tooth extraction. The relative importance of different bleeding symptoms was also described (Figure 1) [11]. ISTH has developed a standardized bleeding questionnaire and a defined interpretation grid for computation of a final BS, also referred to as the ISTH/SSC Bleeding Assessment Tool (BAT) [10]. This kind of BS is a useful screening tool for VWD, and its use is encouraged at the Nordic Hemophilia Centers, see appendix 1. A version, better suited for pediatric patients, was published in 2009 [12].
Criteria for family history
A positive family history compatible with VWD (except for types 2N and 3) usually is associated at least with one first-degree relative, or two second-degree relatives, who have a personal history of objective, significant mucocutaneous bleeding events and laboratory tests compatible with VWD. When available, the use of VWF mutations or genetic markers linked to the VWF locus may allow examination of more distant relatives and may allow asymptomatic relatives with low VWF levels to provide evidence of inheritance.
Clinical criteria for VWD type 1
Type 1 VWD: is a hereditary bleeding disorder due to quantitative deficiency of VWF. In most cases type 1 is inherited as an autosomal dominant trait. The diagnosis therefore is based upon criteria for symptoms, VWF deficiency, and inheritance, all of which must be satisfied. These include significant mucocutaneous bleeding tendency, laboratory tests with VWF levels below 35 IU/dL at least in two occasions and either a positive family history for VWD type 1 or an appropriate VWF mutation. Asymptomatic individuals are typically children who have not yet been challenged with trauma or invasive procedures that could cause bleeding.
Special considerations on the diagnosis of type 1 VWD
In the investigation of type 1 VWD, the bleeding history is particularly important. The relative impact of bleeding manifestations observed in the European study (MCMDC-VWD1) is presented in Figure 1, giving the odds ratio of various bleeding symptoms for the risk of a VWD based on 154 families studied [11].
However, the biochemical diagnosis of type 1 VWD in persons with a mild deficiency in VWF represents a diagnostic dilemma. In three major cohorts, genetic analysis failed to detect mutations in approximately 35 % of patients with a type 1 VWD diagnosis [14]. Clinically, patients with a very low concentration of circulating VWF most often present with a distinct bleeding tendency, hereby qualifying for a bleeding disorder diagnosis, while other patients with less suppressed and marginally low levels of VWF and equivalence between VWF activity and VWF:Ag constitute a grey zone between a healthy state and overt VWD. Linkage studies have further revealed that linkage between members in VWD type 1 families and the VWF gene locus is weak, when VWF:RCo levels are >45 IU/dL9 [6]. Recent work has further demonstrated an inverse relationship between VWF mutations and levels of VWF in these patients. “Genetically confirmed” VWD type 1 has a high diagnostic sensitivity only at VWF:Ag and VWF:RCo levels below 35 IU/dL (Figure 1), which seems a reasonable diagnostic cut-off level. However, it should be noted that marginally low VWF levels not qualifying for VWD criteria, may be associated with increased mucocutaneous bleeds as shown in two case-control studies [2,15] and that such individuals may carry VWD associated mutations [16].
Low VWF: subjects with borderline VWF levels of 35-50 IU/dL may carry a risk for bleeding events rather than a disease. However, these individuals may be completely asymptomatic, unless a coinciding hemostatic risk factor would appear. Patients with borderline VWF levels and bleeding problems should be referred to a hematologist with knowledge of coagulation disorders. The distinction between low VWF trait and definite VWD type 1 will be useful for certain clinical and genetic studies. Alternative or additional diagnoses should be re-considered for patients with possible VWD, especially including the co-existence of platelet disorders.
The presence of blood group O generally predicts lower levels of VWF compared with the non-O state and blood group O itself represents a risk factor for lowered VWF and may show increased bleeding tendency. The influence of the blood group on circulating VWF is not caused by differences in expression rates but is rather due to a blood group-dependent shift in clearance.
Based on recent progress in understanding of type 1 VWD, it is advisable not to assign a diagnosis of VWD type 1 to persons with intermediately lowered plasma VWF (i.e. 35 – 50 IU/dL), but rather to denote symptomatic persons with VWF in that range as having a mild bleeding disorder or a risk for bleeding under situations which compromise hemostasis. This tendency and VWF levels may improve with aging.
Criteria for VWD type 2
VWD type 2 is defined by low levels of functional VWF activity (platelet-dependent activity and/or collagen binding activity), and a low VWF activity /VWF:Ag ratio (<0.7) (except type 2N). Type 2 is further subdivided into subtypes 2A, 2B, 2M and 2N, depending on the type of functional defect (Table 1, Figure 1).
Type 2A and the classical form of type 2B lack the high molecular weight multimers, whereas the rare subtype of 2B called the type Malmö/New York has all multimers. Also, type 2M have a full set of multimers, albeit the separate bands may be aberrant. RIPA is decreased in type 2A and 2M but increased in type 2B.
Type 2A is inherited as a dominant trait, but a recessive inheritance has been described [18]. Two groups of mutations in the A2 domain of the VWF subunit cause the lack of HMWM in type 2A [19]. Group 1 mutations affect intracellular transport, assembly, storage, and secretion of VWF multimers, and group 2 mutations cause increased susceptibility to proteolysis in plasma [20]. Type 2B is caused by mutations in the A1 domain and is characterized by an increased sensitivity to ristocetin in the RIPA test. In addition, mutations in the D1 and D2 domain of the propeptide or in the D3 domain of the mature protein may cause a multimerization defect by affecting intramolecular disulfide bonding within the D3 domain [21,22].
Type 2M is similar to type 2A and characterized by a decreased binding of the VWF to platelets, but in contrast to 2A, all multimers are present. Mutations have been found in the A1 domain of the VWF. The type 2M-Vicenza subtype is characterized by the presence of multimers that are larger than normal and mutations are found in the D3 domain [23].
FVIII deficiency is the typical feature of type 2N VWD due to FVIII binding defect caused by specific mutations in the VWF gene. The levels of FVIII ranges from 1 – 40 IU/dL but is usually above 5 IU/dL [24]. The phenotypic diagnosis of type 2N is based on measuring the ability of VWF to bind FVIII (VWF:FVIIIB assay). The FVIII/VWF ratio is typically decreased (<0.7), but it may be only slightly decreased in compound heterozygotes for a type 2N mutation and a mutation causing a quantitative VWF deficiency. Very low VWF:FVIII binding capacity (<20 %) is indicative of VWD type 2N in homozygous or compound heterozygous state.
Criteria for VWD type 3
VWD type 3 is inherited as a recessive trait and is defined by virtual absence of VWF and very low levels of FVIII (<10 IU/dL). It is very rare, with a prevalence of about 2-3 cases per million in the Nordic area. Higher figures have been diagnosed in countries with a high degree of consanguinity. Bleeding symptoms are usually moderate to severe, and many type 3 patients require replacement therapy with VWF-containing concentrates. The patients are usually homozygotes or compound heterozygotes for mutations in the VWF gene. Missense mutations are found in approx. 15 – 20%, which can result in a less severe phenotype compared to patients having 2 null alleles [25]. Bleeding symptoms are more prevalent in obligatory carriers than in the normal population, but not as frequent as in patients with type 1 VWD [26].