Skip left side navigation and go to content

The Diagnosis, Evaluation and Management of von Willebrand Disease

Diagnosis and Evaluation


The evaluation of a person for possible VWD or other bleeding disorders may be initiated because of a variety of clinical indications (see Figure 3). These indications and situations may include evaluation of: (1) an asymptomatic person who will undergo a surgical or interventional procedure; (2) persons who present with current symptoms of or a history of increased bleeding, abnormal laboratory studies, and/or a positive family history of a bleeding disorder; or (3) persons who present with a prior diagnosis of VWD but do not have supporting laboratory documentation. In all cases, the initial step in assessment should focus on key aspects of the person’s clinical history to determine whether the person may benefit from further diagnostic evaluation. This section is divided into two parts. The first part uses a summary of the medical literature to provide suggested questions for an initial assessment of persons presenting for concerns about bleeding issues or for evaluation prior to procedures that may increase their risk of bleeding. Using the answers to the initial assessment, the second part focuses on a strategy for optimal laboratory assessment of those persons who potentially have bleeding disorders and suggests guidelines for interpretation of laboratory results.

Evaluation of the Patient

History, Signs, and Symptoms

The initial clinical assessment of a person who is being evaluated for VWD should focus on a personal history of excessive bleeding throughout the person’s life and any family history of a bleeding disorder. The history of bleeding should identify the spontaneity and severity, sites of bleeding, duration of bleeding, type of insult or injury associated with bleeding, ease with which bleeding can be stopped, and concurrent medications—such as aspirin, other nonsteroidal antiinflammatory drugs (NSAIDs), clopidogrel (Plavix™), warfarin, or heparin—at the onset of bleeding. Particularly when an invasive procedure is anticipated, the person should be asked whether he or she is currently taking any of these medications and also whether he or she has any history of liver or kidney disease, blood or bone marrow disease, or high or low platelet counts. If a history of any of these illnesses is present, further appropriate evaluation or referral should be undertaken.

Clinical manifestations. The most common presenting symptoms in persons subsequently diagnosed with VWD are summarized in Table 7. Symptoms usually involve mucous membranes and skin sites, and bleeding is of mild to moderate severity (bleeding that does not require blood transfusions and usually does not require visits to the physician) for most persons who have VWD, reflecting the predominance of type 1 VWD. However, life—threatening bleeding (CNS, gastrointestinal) can occur in persons who have type 3 VWD, in some persons who have type 2 VWD, and rarely in persons who have type 1 VWD. Uncommon bleeding manifestations, such as hemarthrosis, are more common in persons who have a more severe deficiency, especially those who have type 3 VWD.85,136 Clinical symptoms may also be modified by coexisting illnesses or other medications. For example, use of aspirin or other NSAIDs can exacerbate the bleeding tendency, whereas use of oral contraceptives can decrease bleeding in women who have VWD.

The clinical evaluation of bleeding symptoms is a challenge, because mild bleeding symptoms are also very common in healthy populations (Table 7, shaded column). Responses to questionnaires used to survey healthy controls indicate that they identify themselves as having specific bleeding manifestations as frequently as persons who have VWD, particularly type 1 VWD (Table 7).137,138,140,143 In addition, a family history of bleeding was reported by 44 percent of healthy children undergoing tonsillectomy143 and by 35 percent138 or 60 percent144 of persons referred because of bleeding. Because bleeding symptoms are so prevalent, it may be impossible to establish a causal relationship between bleeding and low VWF.

Figure 3. Initial Evaluation For VWD or Other Bleeding Disorders

Figure 3. Initial Evaluation For VWD or Other Bleeding Disorders. This figure is explained in detail in the section, 'Diagnostic Recommendations. I. Evaluation of Bleeding Symptoms and Bleeding Risk by History and Physical Examination'

Initial evaluation strategy to determine which patients would most benefit from further diagnostic evaluation for von Willebrand disease (VWD) Left Upper Box: Individuals would be asked three questions about their personal or family bleeding history which, if any are positive, would lead to a second set of questions selected for their sensitivity and specificity for VWD (Box 1, below). Those patients answering positively to one or more of the second set of questions would benefit from laboratory evaluation. Right Boxes: Patients presenting with specific information or a concern about bleeding would be asked the Box 1 questions and the initial 3 questions if not already asked, and would also undergo laboratory evaluation.

Some of the most important clinical issues in VWD apply specifically to women, particularly menorrhagia. Studies of women who have VWD report a high prevalence of menorrhagia (Table 7), although the definition of menorrhagia is not clearly specified in most of these studies and the diagnostic criteria for VWD are not uniform. The sensitivity of menorrhagia as a predictor of VWD may be estimated as 32–100 percent. However, menorrhagia is a common symptom, occurring with a similar frequency in healthy controls and women who have VWD; therefore, it is not a specific marker for VWD (Table 7). In a survey of 102 women who had VWD and were registered at hemophilia treatment centers in the United States, 95 percent reported a history of menorrhagia, but 61 percent of controls also reported a history of menorrhagia.145 Studies have reported a prevalence of VWD of between 5–20 percent among women who have menorrhagia.146–152 Therefore, the specificity of menorrhagia as a predictor of VWD can be estimated as 5–20 percent. Three findings that predict abnormal menstrual blood loss of >80 mL include:

  • Clots greater than approximately 1 inch in diameter
  • Low serum ferritin
  • Changing a pad or tampon more than hourly153

Box 1. Suggested Questions for Screening Persons for a Bleeding Disorder
  1. Do you have a blood relative who has a bleeding disorder, such as von Willebrand disease or hemophilia?
  2. Have you ever had prolonged bleeding from trivial wounds, lasting more than 15 minutes or recurring spontaneously during the 7 days after the wound?
  3. Have you ever had heavy, prolonged, or recurrent bleeding after surgical procedures, such as tonsillectomy?
  4. Have you ever had bruising, with minimal or no apparent trauma, especially if you could feel a lump under the bruise?
  5. Have you ever had a spontaneous nosebleed that required more than 10 minutes to stop or needed medical attention?
  6. Have you ever had heavy, prolonged, or recurrent bleeding after dental extractions that required medical attention?
  7. Have you ever had blood in your stool, unexplained by a specific anatomic lesion (such as an ulcer in the stomach, or a polyp in the colon), that required medical attention?
  8. Have you ever had anemia requiring treatment or received blood transfusion?
  9. For women, have you ever had heavy menses, characterized by the presence of clots greater than an inch in diameter and/or changing a pad or tampon more than hourly, or resulting in anemia or low iron level?

Sources: Dean JA, Blanchette VS, Carcao MD, Stain AM, Sparling CR, Siekmann J, Turecek PL, Lillicrap D, Rand ML. von Willebrand disease in a pediatric—based population—comparison of type 1 diagnostic criteria and use of the PFA—100® and a von Willebrand factor/collagenbinding assay. Thromb Haemost 2000 Sep;(3):401–409; Drews CD, Dilley AB, Lally C, Beckman MG, Evatt B. Screening questions to identify women with von Willebrand disease. J Am Med Womens Assoc 2002;57(4):217–218; and Laffan M, Brown SA, Collins PW, Cumming AM, Hill FG, Keeling D, Peake IR, Pasi KJ. The diagnosis of von Willebrand disease: a guideline from the UK Haemophilia Centre Doctors’ Organization. Haemophilia 2004 May;10(3):199–217.

Table 7. Common Bleeding Symptoms of Healthy Individuals and Patients Who Have VWD

Symptoms Normals
(n = 500;137
n= 341;‡138
n = 88;‡‡139
n= 60‡‡140)
All types VWD
(n = 264;137
n = 1,885141)
Type 1 VWD
(n = 42;†142
n = 671136)
Type 2 VWD
(n = 497136)
Type 3 VWD
(n = 66;136
n = 38585)
Epistaxis 4.6–22.7 38.1–62.5 53–61 63 66–77
Menorrhagia* 23–68.4 47–60 32 32 56–69
Bleeding after dental extraction 4.8–41.9 28.6–51.5 17–31 39 53–70
Ecchymoses 11.8–50 49.2–50.4 50 N.R. N.R.
Bleeding from minor cuts or abrasions 0.2–33.3 36 36 40 50
Gingival bleeding 7.4–47.1 26.1–34.8 29–31 35 56
Postoperative bleeding 1.4–28.2 19.5–28 20–47 23 41
Hemarthrosis 0–14.9 6.3–8.3 2–3 4 37–45
Gastrointestinal bleeding 0.6–27.7 14 5 8 20

* Calculated for females above 13 to 15 years of age.
‡341 individuals were sent a questionnaire, but the precise number of patients responding was not provided.
‡‡ Study included women only.
† Study included males only.
N.R., Not reported.

Identification of people who may require further evaluation for inherited bleeding disorders. Since other "bleeding symptoms" besides menorrhagia are reported frequently by persons who have apparently normal hemostasis, it is important to use questions that can best identify persons who have a true bleeding disorder. Sramek and colleagues138 used a written questionnaire with patients who had a proven bleeding disorder. When the responses were compared to those of a group of healthy volunteers, the most informative questions were related to: (1) prolonged bleeding after surgery, including after dental extractions, and (2) identification of family members who have an established bleeding disorder (Table 8, columns 2–5). A history of muscle or joint bleeding may also be helpful when associated with the above symptoms.

General questions that relate to isolated bleeding symptoms—such as frequent gingival bleeding, profuse menstrual blood loss, bleeding after delivery, and epistaxis in the absence of other bleeding symptoms—were not informative.138 The study also found that an elaborate interview after referral to a hematologist was not particularly helpful when attempting to distinguish persons who have a true bleeding disorder from persons who have a "suspected" bleeding disorder, implying that the selection of those with bleeding disorders had already been made by the referring physician.138

Drews et al.139 attempted to develop a questionnairebased screening tool to identify women who might benefit from a diagnostic workup for VWD. They conducted a telephone survey of 102 women who had a diagnosis of type 1 VWD and were treated at a hemophilia treatment center compared with 88 friends who were controls. With the exception of postpartum transfusions, all study variables were reported more frequently by women who had VWD than by their friends (Table 8, columns 6 and 7). In addition, positive responses to multiple questions were more likely to be obtained from patients who have an inherited bleeding disorder.139 An important limitation of this study is that these women were more symptomatic than most women diagnosed as having type 1 VWD, indicating a more severe phenotype of the disease; this fact might decrease the sensitivity of the questions in the setting of persons who have milder type 1 VWD and fewer symptoms.

More recently, Rodeghiero and colleagues155 compared responses to a standardized questionnaire obtained from 42 obligatory carriers of VWD (from well—characterized families) to responses from 215 controls. The questionnaire covered 10 common bleeding symptoms (including all symptoms in Table 7, and postpartum hemorrhage), with assigned scores for each ranging from 0 (no symptoms) to 3 (severe symptoms, usually including hospitalization and/or transfusion support). With this instrument, the researchers found that having a cumulative total bleeding score of 3 in men, or 5 in women, was very specific (98.6 percent) but not as sensitive (69.1 percent) for type 1 VWD. Limitations of this study include that it was retrospective and that the person administering the questionnaire was aware of the respondent’s diagnosis. This questionnaire is available online.155

A similar retrospective case–control study154 used a standardized questionnaire like that of Rodegherio et al.155 to assess bleeding symptoms of 144 index cases who had type 1 VWD, compared to 273 affected relatives, 295 unaffected relatives, and 195 healthy controls. The interviewers were not blinded to subject’s status. At least one bleeding symptom was reported by approximately 98 percent of index cases, 89 percent of affected relatives, 32 percent of unaffected relatives, and 12 percent of healthy controls. The major symptoms of affected persons (excluding index cases) included bleeding after tooth extraction, nosebleeds, menorrhagia, bleeding into the skin, postoperative bleeding, and bleeding from minor wounds. Using a bleeding score calculated from the data for comparison, the severity of bleeding diminished with increasing plasma VWF, not only for subjects who had low VWF levels but throughout the normal range as well. Although the mean bleeding score was significantly different between several groups, the distribution was sufficiently broad that the bleeding score could not predict the affected or unaffected status of individuals.

Table 8. Prevalences of Characteristics in Patients Who Have Diagnosed Bleeding Disorders Versus Healthy Controls

Symptom Univariate analysis* Multivariate analysis* Women who have VWD† Type 1 VWD families‡
Odds ratio 95% CI Odds ratio 95% CI Sensitivity 95% CI Odds ratio 95% CI
Family members have an established bleeding disorder 97.5 38.3–248 50.5 12.5–202.9        
Profuse bleeding from small wounds 67.2 28.4–159 30.0 8.1–111.1     16.7 2.0–137.7
Profuse bleeding at site of tonsillectomy/adenoidectomy 27.7 8.0–96.1 11.5 1.2–111.9        
Easy bruising 12.7 8.0–20.2 9.9 3.0–32.3 9.8 4.8–17.3 8.1 2.1–30.5
Profuse bleeding after surgery 23.0 10.6–50.1 5.8 1.3–26.4 52.9 42.8–62.9 8.9 3.6–21.8
Muscle bleeding (ever) 13.3 6.4–27.7 4.8 0.7–31.4 9.8 4.8—17.3    
Frequent nosebleeds 3.5 2.0–6.2 3.8 0.9–15.7 61.8 51.6–71.2 4.9 2.4–10.0
Profuse bleeding at site of dental extraction 39.4 20.6–75.5 3.2 0.9–11.3 54.9 44.7–64.8 4.6 2.5–8.4
Blood in stool (ever) 2.8 1.7–4.6 2.8 0.7–11.7 13.7 7.7–22.0 1.6 0.6–4.3
Family members with bleeding symptoms 28.6 15.0–54.6 2.5 0.7–9.4        
Joint bleeding (ever) 8.6 4.8–15.2 2.5 0.6–10.2 20.6 13.2–29.7    
Menorrhagia 5.4 3.0–9.8 2.5 0.6–9.9     5.1 2.6–10.1
Hemorrhage at time of delivery 5.3 2.3–12.0 2.1 0.3–13.5 50.0 39.9–60.1 0.9 0.3–3.2
Frequent gingival bleeding 2.8 1.9–4.2 0.7 0.3–2.0 76.5 67.0–84.3 1.3 0.3–6.7
Hematuria (ever) 3.2 1.8–5.6 0.5 0.1–2.3        

Sources: Sramek A, Eikenboom JC, Briet E, Vandenbroucke JP, Rosendaal FR. Usefulness of patient interview in bleeding disorders. Arch Intern Med 1995 Jul;155(13):1409–1415; Drews CD, Dilley AB, Lally C, Beckman MG, Evatt B. Screening questions to identify women with von Willebrand disease. J Am Med Womens Assoc 2002;57(4):217–218; and Tosetto A, Rodeghiero F, Castaman G, Goodeve A, Federici AB, Batlle J, Meyer D, Fressinaud E, Mazurier C, Goudemand J, et al. A quantitative analysis of bleeding symptoms in type 1 von Willebrand disease: results from a multicenter European study (MCMDM—1 VWD). J Thrombos Haemostas 2006;4:766–773.
* Univariate and multivariate analyses from reference comparing 222 patients who had a known bleeding disorder (43 percent mild VWD) to 341 healthy volunteers.138
† Compiled from responses to a questionnaire sent to 102 women, who had type 1 VWD, in a hemophilia treatment center.139
‡ Compiled from interviews comparing affected vs. unaffected family members of patients who have type 1 VWD. The index cases (patients who have VWD) were not included in the analysis (Tosetto et al. 2006, and personal communication from Dr. Francesco Rodeghiero on behalf of coauthors).154
CI, confidence interval

In a related study, bleeding symptoms were assessed with the same questionnaire in 70 persons who were obligatory carriers of type 3 VWD, 42 persons who were obligate carriers of type 1 VWD (meaning affected family members of index cases who had type 1 VWD), and 215 persons who were healthy controls.156 Carriers of type 3 VWD were compared with carriers of type 1 VWD to address the question of whether the distinct types of VWF mutations associated with these conditions predisposed to the same or different severity of bleeding. Approximately 40 percent of carriers of type 3 VWD, 82 percent of carriers of type 1 VWD, and 23 percent of healthy controls had at least one bleeding symptom. The major bleeding symptoms in carriers of type 3 VWD were bleeding into skin and postsurgical bleeding. The results suggest that carriers of type 3 VWD are somewhat distinct, as they have bleeding symptoms more frequently than healthy controls but less frequently than persons who have or are carriers of type 1 VWD. Usually, carriers of type 1 VWD have lower VWF levels than carriers of type 3 VWD.

Family history. Although a family history that is positive for an established bleeding disorder is useful in identifying persons who are likely to have VWD, such a history is frequently not present. This is most commonly the case for persons who have milder forms of VWD and whose family members may have minimal, if any, symptoms. As shown in Table 8, the presence of a documented bleeding disorder in a family member is extremely helpful in deciding which persons to evaluate further, whereas a family history of bleeding symptoms is less helpful.

Box 1 summarizes suggested questions that can be used to identify persons who should be considered for further evaluation for VWD with laboratory studies.

Physical examination. The physical examination should be directed to confirm evidence for a bleeding disorder, including size, location, and distribution of ecchymoses (e.g., truncal), hematomas, petechiae, and other evidence of recent bleeding. The examination should also focus on findings that may suggest other causes of increased bleeding, such as evidence of liver disease (e.g., jaundice), splenomegaly, arthropathy, joint and skin laxity (e.g., Ehlers—Danlos Syndrome), telangiectasia (e.g., hereditary hemorraghic telangiectasia), signs of anemia, or anatomic lesions on gynecologic examination.

Acquired von Willebrand Syndrome (AVWS). Persons who have AVWS present with bleeding symptoms similar to those described, except that the past personal and family history are negative for bleeding symptoms. AVWS may occur spontaneously or in association with other diseases, such as monoclonal gammopathies, other plasma cell dyscrasias, lymphoproliferative diseases, myeloproliferative disorders (e.g., essential thrombocythemia), autoimmune disorders, valvular and congenital heart disease, certain tumors, and hypothyroidism.117,157 The evaluation should be tailored to finding conditions associated with AVWS.

Laboratory Diagnosis and Monitoring

An algorithm for using clinical laboratory studies to make the diagnosis of VWD is summarized in Figure 4.

Ideally, a simple, single laboratory test could screen for the presence of VWD. Such a screening test would need to be sensitive to the presence of most types of VWD and would have a low false—positive rate. Unfortunately, no such test is available. In the past, the activated partial thromboplastin time (PTT) and bleeding time (BT) were recommended as diagnostic tests. These tests were probably satisfactory for detecting severe type 3 VWD, but as variant VWD and milder forms of VWD were characterized, it became apparent that many of the persons who have these conditions had normal PTT and normal BT results.

An initial hemostasis laboratory evaluation (see Box 2) usually includes a platelet count and complete blood count (CBC), PTT, prothrombin time (PT), and optionally either a fibrinogen level or a thrombin time (TT). This testing neither "rules in" nor "rules out" VWD, but it can suggest whether coagulation factor deficiency or thrombocytopenia might be the potential cause of clinical bleeding. If the mucocutaneous bleeding history is strong, consider performing initial VWD assays (VWF:Ag, VWF:RCo, and FVIII) at the first visit.

Box 2. Initial Laboratory Evaluation of Hemostasis
  • CBC and platelet count
  • PTT
  • PT
  • Fibrinogen or TT (optional)

Figure 4. Laboratory Assessment For VWD or Other Bleeding Disorders

Figure 4. Laboratory Assessment For VWD or Other Bleeding Disorders. This figure is explained in detail in the section, 'Diagnostic Recommendations. II.  Evaluation by Laboratory Testing'

* Isolated decreased platelets may occur in VWD type 2B.
† Correction in the PTT mixing study immediately and after 2—hour incubation removes a factor VIII (FVIII) inhibitor from consideration.
Investigation of other intrinsic factors and lupus anticoagulant also may be indicated.
CBC, complete blood count; PT prothrombin time; PTT partial thromboplastin time; RIPA, ristocetin—induced platelet aggregation; TT, thrombin time; VWF:Ag, VWF antigen; VWF:RCo, VWF ristocetin cofactor activity.

If the initial clinical evaluation suggests a bleeding disorder, the "initial hemostasis tests" should be ordered, followed by or along with the next tests ("initial VWD assays") indicated in the algorithm. Referral to a hemostasis specialist is appropriate for help in interpretation, repeat testing, and specialized tests.

Some centers add a BT or a platelet function analyzer (PFA—100®) assay to their initial laboratory tests. The BT test is a nonspecific test and is fraught with operational variation. It has been argued that it was a population—based test that was never developed to test individuals.158 Variables that may affect results include a crying or wiggling child, differences in the application of the blood pressure cuff, and the location, direction, and depth of the cut made by the device.

This test also has a potential for causing keloid formation and scarring, particularly in non—Caucasian individuals. The PFA—100® result has been demonstrated to be abnormal in the majority of persons who have VWD, other than those who have type 2N, but its use for population screening for VWD has not been established.159–162 Persons who have severe type 1 VWD or who have type 3 VWD usually have abnormal PFA—100® values, whereas persons who have mild or moderate type 1 VWD and some who have type 2 VWD may not have abnormal results.163–165 When persons are studied by using both the BT and PFA—100®, the results are not always concordant.162,164,166

When using the PTT in the diagnosis of VWD, results of this test are abnormal only if the FVIII is sufficiently reduced. Because the FVIII gene is normal in VWD, the FVIII deficiency is secondary to the deficiency of VWF, its carrier protein. In normal individuals, the levels of FVIII and VWF:RCo are approximately equal, with both averaging 100 IU/dL. In type 3 VWD, the plasma FVIII level is usually less than 10 IU/dL and represents the steady state of FVIII in the absence of its carrier protein. In persons who have type 1 VWD, the FVIII level is often slightly higher than the VWF level and may fall within the normal range. In persons who have type 2 VWD (except for type 2N VWD in which it is decreased), the FVIII is often 2–3 times higher than the VWF activity (VWF:RCo).167,168 Therefore, the PTT is often within the normal range. If VWF clearance is the cause of low VWF, the FVIII reduction parallels that of VWF, probably because both proteins are cleared together as a complex.

Initial Tests for VWD

Box 3 lists the initial tests commonly used to detect VWD or low VWF. These three tests, readily available in most larger hospitals, measure the amount of VWF protein present in plasma (VWF:Ag), the function of the VWF protein that is present as ristocetin cofactor activity (VWF:RCo), and the ability of the VWF to serve as the carrier protein to maintain normal FVIII survival, respectively. If any of the above tests is abnormally low, the next steps should be discussed with a coagulation specialist, who may recommend referral to a specialized center, and/or repeating the laboratory tests plus performing additional tests.

VWF:Ag is an immunoassay that measures the concentration of VWF protein in plasma. Commonly used methods are based on enzyme—linked immunosorbent assay (ELISA) or automated latex immunoassay (LIA). As discussed below, the standard reference plasma is critical and should be referenced to the World Health Organization (WHO) standard. The person’s test results should be reported in international units (IU), either as international units per deciliter (IU/dL) or as international units per milliliter (IU/mL). Most laboratories choose IU/dL, because it is similar to the conventional manner of reporting clotting factor assays as a percentage of normal.

VWF:RCo is a functional assay of VWF that measures its ability to interact with normal platelets. The antibiotic, ristocetin, causes VWF to bind to platelets, resulting in platelet clumps and their removal from the circulation. Ristocetin was removed from clinical trials because it caused thrombocytopenia. This interaction was developed into a laboratory test that is still the most widely accepted functional test for VWF. (In vivo, however, it is the high shear in the microcirculation, and not a ristocetin—like molecule, that causes the structural changes in VWF that lead to VWF binding to platelets.)

Box 3. Initial Tests for VWD
  • VWF:Ag
  • VWF:RCo

Several methods are used to assess the platelet agglutination and aggregation that result from the binding of VWF to platelet GPIb induced by ristocetin (ristocetin cofactor activity, or VWF:RCo). The methods include: (1) time to visible platelet clumping using ristocetin, washed normal platelets (fresh or formalinized), and dilutions of patient plasma; (2) slope of aggregation during platelet aggregometry using ristocetin, washed normal platelets, and dilutions of the person’s plasma; (3) automated turbidometric tests that detect platelet clumping, using the same reagents noted above; (4) ELISA assays that assess direct binding of the person’s plasma VWF to platelet GPIb (the GPIb may be derived from plasma glycocalicin) in the presence of ristocetin;169–171 and (5) the binding of a monoclonal antibody to a conformation epitope of the VWF A1 loop.172 Method 5 can be performed in an ELISA format or in an automated latex immunoassay. It is not based on ristocetin binding. The first three assays (above) may use platelet membrane fragments containing GPIb rather than whole platelets. The sensitivity varies for each laboratory and each assay; in general, however, Methods 1 and 2, which measure platelet clumping by using several dilutions of the person’s plasma, are quantitative to approximately 6–12 IU/dL levels. Method 3 is quantitative to about 10–20 IU/dL. Method 4 can measure VWF:RCo to <1 IU/dL, and a variation of it can detect the increased VWF binding to GPIb seen in type 2B VWD.173 Some automated methods are less sensitive and require modification of the assay to detect <10 IU/dL. Each laboratory should define the linearity and limits of its assay. Several monoclonal ELISAs (Method 5) that use antibodies directed to the VWF epitope containing the GPIb binding site have been debated because the increased function of the largest VWF multimers is not directly assessed.174

The ristocetin cofactor activity (VWF:RCo) assay has high intra– and interlaboratory variation, and it does not actually measure physiologic function. The coefficient of variation (CV) has been measured in laboratory surveys at 30 percent or greater, and the CV is still higher when the VWF:RCo is lower than 12–15 IU/dL.175–179 This becomes important not only for the initial diagnosis of VWD, but also for determining whether the patient has type 1 versus type 2 VWD (see discussion on VWF:RCo to VWF:Ag ratio, below). Despite these limitations, it is still the most widely accepted laboratory measure of VWF function. Results for VWF:RCo should be expressed in international units per deciliter (IU/dL) based on the WHO plasma standard.

FVIII coagulant assay is a measure of the cofactor function of the clotting factor, FVIII, in plasma. In the context of VWD, FVIII activity measures the ability of VWF to bind and maintain the level of FVIII in the circulation. In the United States, the assay is usually performed as a one—stage clotting assay based on the PTT, although some laboratories use a chromogenic assay. The clotting assay, commonly done using an automated or semiautomated instrument, measures the ability of plasma FVIII to shorten the clotting time of FVIII—deficient plasma. Because this test is important in the diagnosis of hemophilia, the efforts to standardize this assay have been greater than for other hemostasis assays. FVIII activity is labile, with the potential for spuriously low assay results if blood specimen collection, transport, or processing is suboptimal. Like those tests discussed above, it should be expressed in international units per deciliter (IU/dL) based on the WHO plasma standard.

Expected patterns of laboratory results in different subtypes of VWD, depicted in Figure 5, include results of the three initial VWD tests (VWF:Ag, VWF:RCo, FVIII) and results of other assays for defining and classifying VWD subtypes. The three initial tests (or at least the VWF:RCo and FVIII assays) are also used for monitoring therapy.

Other Assays To Measure VWF, Define/Diagnose VWD, and Classify Subtypes

The VWF multimer test, an assay that is available in some larger centers and in commercial laboratories, is usually performed after the initial VWD testing indicates an abnormality, preferably using a previously unthawed portion of the same sample or in association with a repeated VWD test panel (VWF:Ag, VWF:RCo, FVIII) using a fresh plasma sample. VWF multimer analysis is a qualitative assay that depicts the variable concentrations of the different—sized VWF multimers by using sodium dodecyl sulfate (SDS)—protein electrophoresis followed by detection of the VWF multimers in the gel, using a radiolabeled polyclonal antibody or a combination of monoclonal antibodies. Alternatively, the protein is transferred to a membrane (Western blot), and the multimers are identified by immunofluorescence or other staining techniques.99,180,181

Figure 5. Expected Laboratory Values in VWD

Figure 5. Expected Laboratory Values in VWD. This figure shows sample expected laboratory patterns, with relative increases and decreases, for VWF:Ag, VWF:RCo, FVIII, RIPA, LDRIPA, PFA100 CT, BT, Platelet count, and VWF multimer pattern. Expected sample results are shown for normal subjects, and for patients with VWD (Types 1, 2A, 2B, 2M, 2N, and 3), and PLTVWD. This figure is explained in the sections 'Expected patterns of laboratory results in different subtypes of VWD' and 'Other Assays to Measure VWF, Define/Diagnose VWD, and Classify Subtypes' in VWD.

The symbols and values represent prototypical cases. In practice, laboratory studies in certain patients may deviate slightly from these expectations.

L, 30—50 IU/dL; ↓, ↓↓, ↓↓↓, relative decrease; ↑, ↑↑, ↑↑↑, relative increase; BT, bleeding time; FVIII, factor VIII activity; LD—RIPA, low—dose ristocetin—induced platelet aggregation (concentration of ristocetin ≤ 0.6 mg/mL); N, normal; PFA—100® CT, platelet function analyzer closure time; RIPA, ristocetin—induced platelet aggregation; VWF, von Willebrand factor; VWF:Ag, VWF antigen; VWF:RCo, VWF ristocetin cofactor activity.

*Note: persons who have platelet—type VWD (PLT—VWD) have a defect in their platelet GPIb. Laboratory test results resemble type 2B VWD, and both have a defect in their LD—RIPA. In the VWF:platelet binding assay (see text), persons who have type 2B VWD have abnormally increased platelet binding. Normal persons and those who have PLT—VWD have no binding of their VWF to normal platelets at low ristocetin concentrations.

Note: this figure is adapted from and used by permission of R.R. Montgomery.

Multimer assays are designated as "low resolution" (which differentiate the largest multimers from the intermediate and small multimers) or "high resolution" (which differentiate each multimer band of the smaller multimers into three to eight satellite bands). For diagnostic purposes, the low—resolution gel systems are used primarily; these systems help to differentiate the type 2 VWD variants from types 1 or 3 VWD. Figure 6 illustrates the differences between these two techniques with regard to the resolution of high— and low—molecular—weight multimers. It should be noted that multimer appearance alone does not define the variant subtype and that only types 2A, 2B, and platelet—type VWD (PLT—VWD) have abnormal multimer distributions with relative deficiency of the largest multimers. An exception is Vicenza variant VWD with ultralarge VWF multimers and low VWF. For more information about VWF multimer findings in type 2 VWD variants, see descriptions above (type 2 VWD) and associated references.

Low—Dose RIPA. RIPA and VWF platelet—binding assay (VWF:PB assay) are two tests that are performed to aid in diagnosing type 2B VWD. RIPA may be done as part of routine platelet aggregation testing. RIPA is carried out in platelet—rich plasma, using a low concentration of ristocetin (usually <0.6 mg/mL, although ristocetin lots vary, resulting in the use of slightly different ristocetin concentrations). This low concentration of ristocetin does not cause VWF binding and aggregation of platelets in samples from normal persons, but it does cause VWF binding and aggregation of platelets in samples from patients who have either type 2B VWD or mutations in the platelet VWF receptor. The latter defects have been termed platelet—type (PLT—VWD) or pseudo VWD, and they can be differentiated from type 2B VWD by VWF:PB assay. At higher concentrations of ristocetin (1.1–1.3 mg/mL), RIPA will be reduced in persons who have type 3 VWD. However, the test is not sufficiently sensitive to reliably diagnose other types of VWD.

Figure 6. Analysis of VWF Multimers

Figure 6. Analysis of VWF Multimers. This figure is explained in detail in the legend below.

The distribution of VWF multimers can be analyzed using sodium dodecyl sulfate (SDS)—agarose electrophoresis followed by immunostaining. Low—resolution gels (0.65% agarose, left side) can demonstrate the change in multimer distribution of the larger multimers (top of the gel), while high—resolution gels (2–3% agarose, right side) can separate each multimer into several bands that may be distinctive. For example, the lowest band in the 0.65% gel (1) can be resolved into 5 bands in the 3% agarose gel, but the 3% gel fails to demonstrate the loss of high molecular weight multimers seen at the top in the 0.65% gel. The dotted lines (1) indicate the resolution of the smallest band into several bands in the 3% agarose gel. In each gel, normal plasma (NP) is run as a control. Type 1 VWD plasma has all sizes of multimers, but they are reduced in concentration. Type 2A VWD plasma is missing the largest and intermediate multimers, while type 2B VWD plasma is usually missing just the largest VWF multimers. No multimers are identified in type 3 VWD plasma. Patients who have thrombotic thrombocytopenic purpura (TTP) may have larger than normal multimers when studied with low—resolution gels.
Note: Used by permission of R.R. Montgomery.

VWF: platelet—binding (VWF:PB) assay measures the binding of VWF to normal paraformaldehyde—fixed platelets using low concentrations of ristocetin (usually 0.3–0.6 mg/mL).182 The amount of VWF bound to the fixed platelets is determined by using a labeled antibody. Normal individuals, or those who have types 1, 2A, 2M, 2N, and 3 VWD, exhibit minimal or no binding to platelets at the concentration of ristocetin used, but patients who have type 2B VWD exhibit significant binding that causes their variant phenotype (a loss of high—molecular—weight multimers, decreased ristocetin cofactor activity, and thrombocytopenia). Both type 2B VWD and platelet—type VWD have agglutination of platelet—rich plasma (PRP) to low—dose ristocetin, but the VWF:PB assay can differentiate type 2B VWD from platelettype VWD. Only VWF from persons who have type 2B VWD has increased VWF:PB, while VWF from persons who have platelet—type VWD has normal VWF:PB with low doses of ristocetin.

VWF collagen—binding (VWF:CB) assay measures binding of VWF to collagen. The primary site of fibrillar collagen binding is in the A3 domain of VWF. Like the ristocetin cofactor assay, the collagen binding assay is dependent on VWF multimeric size, with the largest multimers binding more avidly than the smaller forms. The VWF:CB assay performance and sensitivity to VWD detection or discrimination among VWD subtypes is highly dependent on the source of collagen, as well as on whether type 1 collagen or a mixture of type 1/3 collagen is used.183,184 Only a few patients have been identified who have specific collagen—binding defects that are independent of multimer size, and the defects have been associated with a mutation of VWF in the A3 domain.73 The prevalence of such defects is unknown. The place of VWF:CB in the evaluation of VWD has not been established. In principle, however, patients who have defects in collagen binding may have a normal VWF:RCo and thus escape clinical diagnosis unless a VWF:CB assay is performed. Limited studies suggest that supplementary VWF:CB testing, complementing assays of VWF:RCo and VWF:Ag, can improve the differentiation of type 1 VWD from types 2A, 2B, or 2M VWD.175,185,186

VWF:FVIII binding (VWF:FVIIIB) assay measures the ability of a person’s VWF to bind added exogenous FVIII and is used to diagnose type 2N VWD.75,77,78,187,188 The assay is performed by capturing the person’s VWF on an ELISA plate, removing the bound endogenous FVIII, and then adding back a defined concentration of exogenous recombinant FVIII. The amount of FVIII bound is determined by a chromogenic FVIII assay. The level of this bound FVIII is then related to the amount of the person’s VWF initially bound in the same well. In clinical experience, Type 2N VWD is usually recessive; the person is either homozygous or compound heterozygous (one allele is type 2N, and the other is a type 1 or "null" allele). In either case, the VWF in the circulation does not bind FVIII normally, and the concentration of FVIII is thus decreased.

The VWF:RCo to VWF:Ag ratio can aid in the diagnosis of types 2A, 2B, and 2M VWD and help differentiate them from type 1 VWD. For example, VWF:RCo/VWF:Ag <0.6189 or <0.7 has been used as a criterion for dysfunctional VWF.8,190 A similar approach has been proposed for the use of the VWF:CB/VWF:Ag ratio.8,190 In type 2A VWD, the ratio is usually low; and in type 2B VWD, the VWF:RCo/VWF:Ag ratio is usually low but may be normal. In type 2M VWD, the VWF:Ag concentration may be reduced or normal, but the VWF:RCo/VWF:Ag ratio will be <0.7. One study70 determined the VWF:RCo/VWF:Ag ratio in nearly 600 individuals with VWF levels <55 IU/dL who had normal VWF multimers. The study used this ratio to identify families who had type 2 VWD, but most centers do not have the ability to establish normal ranges for patients who have low VWF. Additionally, the VWF:RCo assay has a coefficient of variation (CV) as high as 30 percent or more, depending on methodology, whereas the CV for the VWF:Ag assay is somewhat lower. The high intrinsic variability of the VWF:RCo assay, especially at low levels of VWF, can make the VWF:RCo/VWF:Ag ratio an unreliable criterion for the diagnosis of type 2 VWD.175,177–179 (See Recommendations II.C.1.a. and III.B.1, below). It is important that the same plasma standard be used in both the VWF:RCo and VWF:Ag assays and that the normal range for the VWF:RCo/VWF:Ag ratio and its sensitivity to types 2A and 2M VWD be determined in each laboratory. Because no large multicenter studies have evaluated the precise ratio that should be considered abnormal, a ratio in the range of less than 0.5–0.7 should raise the suspicion of types 2A, 2B, or 2M VWD. Further confirmation should be sought by additional testing (e.g., repeat VWD test panel and VWF multimer study or sequencing of the A1 region of the VWF gene).191

ABO blood types have a significant effect on plasma VWF (and FVIII) concentrations.43,192 Individuals who have blood type O have concentrations approximately 25 percent lower compared to persons who have other ABO blood types. The diagnosis of type 1 VWD occurs more frequently in individuals who have blood group type O.43 Table 9 illustrates the significant effect of blood type on VWF:Ag level.

Table 9. Influence of ABO Blood Groups on VWF:Ag

ABO Type N VWF:Ag mean Range
O 456 74.8* 35.6–157.0* (41–179)
A 340 105.9* 48.0–233.9* (55–267)
B 196 116.9* 56.8–241.0* (65–275)
AB 109 123.3* 63.8–238.2* (73–271)

*U/dL; IU/dL
Source: Gill JC, Endres—Brooks J, Bauer PJ, Marks WJ, Jr., Montgomery RR. The effect of ABO blood group on the diagnosis of von Willebrand disease. Blood 1987 Jun;69(6):1691–1695. Copyright American Society of Hematology, used with permission. In this publication, VWF:Ag was expressed as U/dL, but the range in IU/dL (WHO) is higher for all blood groups, as noted in the values in parentheses (personal communication from R.R. Montgomery, J.L. Endres, and K.D. Friedman).

Although it has been recommended to stratify reference ranges for VWF:Ag and VWF:RCo with respect to blood group O and nongroup O,193,194 evolving limited information supports the concept that, despite the ABO blood grouping and associated VWF reference ranges, the major determinant of bleeding symptoms or risk is low VWF.189,195,196 Therefore, referencing VWF testing results to the population reference range, rather than to ABO—stratified reference ranges, may be more useful clinically.

Platelet VWF studies are performed by some laboratories, including VWF:RCo, VWF:Ag, and VWF multimers, using VWF extracted from washed platelets. The methods and interpretations of these studies, however, are not well standardized.

DNA sequencing of patient DNA has been used to make a molecular diagnosis of variants of type 2 VWD,197–199 but DNA sequencing is not widely available. Most of the mutations found in types 2B, 2M, and 2N VWD cluster in the cDNA that directs the synthesis of specific regions of VWF (see Figure 2).200 In the common forms of type 2A VWD, in which the VWF is spontaneously cleaved by ADAMTS13, mutations cluster in the A2 domain (which contains the cleavage site). In the less common type 2A variants of VWD, in which multimer formation is inhibited, the mutations may be scattered throughout the gene. In most persons who have type 1 VWD, the genetic mutations have not been established, although several studies are being conducted at present to characterize these mutations.

Assays for Detecting VWF Antibody

Assays for detecting anti—VWF antibodies are not as well established as the assays for detecting antibodies to FVIII in patients who have hemophilia A. Some patients who have AVWS do appear to have anti—VWF antibodies that decrease the half—life of infused VWF. Although a few antibodies do inhibit VWF function and can be demonstrated in 1:1 mixing studies with normal plasma using the VWF:RCo assay, most anti—VWF antibodies are not "inhibitors" of VWF function. The presence of these antibodies, however, promotes rapid clearance of VWF. The plasma level of VWF propeptide (VWFpp) is normally proportionate to the level of VWF:Ag, and the VWFpp level can be measured to aid in the detection of the rapid clearance of VWF. Accelerated plasma clearance of VWF:Ag—as occurs in some patients who have AVWS, in those who have certain type 1 VWD variants, or in those who have type 3 VWD and have alloantibodies to VWF—is associated with an increase in the ratio of VWFpp to VWF:Ag.201,202 Persons who have type 3 VWD, with large deletions of the VWF gene, are prone to develop alloantibodies to transfused VWF.203 Patients who have AVWS, VWF antibodies, or mutations that affect VWF clearance can be studied using VWF—survival testing after administration of DDAVP or VWF concentrate.

Making the Diagnosis of VWD

Scoring systems and criteria for assessing the bleeding history and the probability of having VWD, especially type 1 VWD, are in evolution but have not yet been subjected to prospective studies outside of defined populations.155,194 Establishing the diagnosis of VWD in persons who have type 2 VWD variants and type 3 VWD is usually straightforward, based on the initial VWD tests (described in Box 3 above, Initial Tests for VWD). Treatment depends on the specific subtype (e.g., type 2A, 2B, 2M, or 2N), which is determined by additional tests including VWF multimer analysis. In contrast, the diagnosis of type 1 VWD is often more difficult,21,44,93,114,204 partly because not all persons who have decreased levels of VWF have a molecular defect in the VWF gene. Whether individuals who do not have an abnormality in the VWF gene should be diagnosed as having VWD or should be given another designation is currently under consideration (see section on "Type 1 VWD Versus Low VWF"). The reasons for reduced VWF levels in many of these persons who have a normal VWF gene sequence are not understood. A "low" VWF level is believed to confer some bleeding risk, despite having a normal VWF gene, and those persons who have clinical bleeding and low VWF may benefit from treatment to raise the VWF level. Most clinicians would agree that persons having VWF levels below 30 IU/dL probably have VWD. It is likely that most of these persons have a mutation in the VWF gene. Currently, several large European Union, Canadian, and U.S. studies are trying to define that frequency. Persons whose plasma VWF levels are below the lower limit of the laboratory reference range, but >30 IU/dL, may have VWD but are sometimes referred to as having "possible type 1 VWD" or "low VWF." There is no generally accepted designation for these persons. Although type 3 VWD is usually the result of inheriting two "null" alleles, the heterozygous "carriers" in these families do not universally have a significant bleeding history; therefore, type 3 VWD has been called a recessive disorder.21,44,101

Special Considerations for Laboratory Diagnosis of VWD

Repeated testing for VWD is sometimes needed to identify low levels of VWF. Stress—including surgery, exercise, anxiety, crying in a frightened child, as well as systemic inflammation, pregnancy, or administration of estrogen/oral contraceptives— can cause an increase in plasma levels of VWF and mask lower baseline values. VWF levels vary with the menstrual cycle, and lowest values are detected on days 1–4 of the menstrual cycle. However, the importance of timing of the testing with respect to the menstrual cycle is not clear. Family studies may be helpful to diagnose hereditary decreases in VWF levels.

Problems may occur in preparing samples for testing. As noted, anxiety may falsely elevate the VWF and FVIII levels, and the setting for phlebotomy should be as calm as possible. It is important that the sample be obtained by atraumatic collection of blood, drawn into the appropriate amount of citrate anticoagulant. The College of American Pathologists (CAP), as well as the Clinical Laboratory Standards Institute (CLSI, formerly NCCLS), recommend collecting blood into 3.2 percent citrate, although some laboratories still use 3.8 percent citrate. Fasting or nonlipemic samples should be used for testing, and icteric or hemolyzed samples may also compromise the quality of testing results.193,205 If a person has polycythemia or profound anemia, the amount of anticoagulant should be adjusted on the basis of nomograms for this purpose. Blood should be centrifuged promptly to obtain plasma, and the plasma should remain at room temperature if assays are to be completed within 2 hours. Whole blood should not be transported on wet ice (or frozen).206,207 If plasma samples are frozen, they should be thawed at 37°C to avoid formation of a cryoprecipitate. Plasma assays should be performed on "platelet—poor" or "platelet—free" plasma.193 Although a small number of platelets may not significantly affect studies done on fresh plasma, freezing these samples may result in the release of proteases or platelet membrane particles that affect plasma assays for VWF. Thus, plasmas should be centrifuged carefully. Some laboratories perform double centrifugation to ensure platelet removal. The integrity of samples may suffer during transport to an outside laboratory, and steps should be taken that can best ensure prompt delivery of frozen samples. (See Table 10, below.)

The VWF reference standard is critical to the laboratory diagnosis of VWD. When possible, all laboratory assays of VWF should use the same standard to avoid artifactual discrepancies. Results of VWF assays can be reported in international units (IUs) only if they have been referenced to the WHO standard for that analyte. If a reference plasma pool is used, it is usually reported as a percentage of normal, as it cannot be called an IU. To assist the comparison, IUs are usually expressed as IU/dL so that the reported values have the same range as "percentage of normal plasma" values.

Table 10. Collection and Handling of Plasma Samples for Laboratory Testing

Phlebotomy conditions—An atraumatic blood draw limits the exposure of tissue factor from the site and the activation of clotting factors, minimizing falsely high or low values.

Patient stress level—Undue stress, such as struggling or crying in children or anxiety in adults, may falsely elevate VWF and FVIII levels. Very recent exercise can also elevate VWF levels.

Additional conditions in the person—The presence of an acute or chronic inflammatory illness may elevate VWF and FVIII levels, as may pregnancy or administration of estrogen/oral contraceptives.

Sample processing—To prevent cryoprecipitation of VWF and other proteins, blood samples for VWF assays should be transported to the laboratory at room temperature. Plasma should be separated from blood cells promptly at room temperature, and the plasma should be centrifuged thoroughly to remove platelets. If plasma samples will be assayed within 2 hours, they should be kept at room temperature. Frozen plasma samples should be carefully thawed at 37°C and kept at room temperature for <2 hours before assay.

Sample storage—Plasma samples that will be stored or transported to a reference laboratory must be frozen promptly at or below –40°C and remain frozen until assayed. A control sample that is drawn, processed, stored, and transported under the same conditions as the tested person’s sample may be helpful in indicating problems in the handling of important test samples.

Laboratory variables also occur. The variability (CV) of the VWF:RCo assay is high (20–30 percent or greater) and the CV of the VWF:Ag assay is also relatively high (10–20 percent or greater), as is the CV for the FVIII assay.175,177–179,183,208–210 The quality of laboratory testing also varies considerably among laboratories (high interlaboratory CV). Coupled with variability of VWF and FVIII contributed by conditions of the patient and the blood sample, the high variability of these three diagnostic tests can contribute to difficulty in diagnosing VWD or classifying the VWD subtype (e.g., type 1 vs. type 2 variant, using the VWF:RCo to VWF:Ag ratio). Some of the more specialized tests, such as VWF multimer analysis likely also have high variability of test performance and interpretation,180,181 and they are often not available at local testing laboratories.

Summary of the Laboratory Diagnosis of VWD

The diagnosis of VWD can be complex, and no single diagnostic approach is suitable for all patients. Improvements in laboratory testing and quality, along with further research into the frequency of mutations of the VWF gene, alterations of other proteins that result in reduced VWF levels, and the correlation of clinical symptoms with laboratory test levels will be necessary to place the diagnosis of VWD on a more secure foundation. (See Table 10, above.)

The following recommendations include specific clinical history, physical findings, laboratory assays, and diagnostic criteria that this Panel suggests will allow the most definitive diagnosis of VWD.

  • Tests such as the bleeding time, PFA—100®, or other automated functional platelet assays have been used but there are conflicting data with regard to sensitivity and specificity for VWD.162,164,166 Therefore, the Panel believes current evidence does not support their routine use as screening tests for VWD.
  • The Panel believes that platelet—based assays should be used for the ristocetin cofactor method.
  • The Panel emphasizes the importance of the timing of the phlebotomy for assays, with the patient at his/her optimal baseline as far as possible. (For example, VWF levels may be elevated above baseline during the second and third trimesters of pregnancy or during estrogen replacement, during acute inflammation such as the perioperative period, during infections, and during acute stress.) The careful handling and processing of the sample is also critical, particularly if the sample will be sent out for testing at a distant location.

Diagnostic Recommendations

The recommendations are graded according to criteria described in "Clinical Recommendations—Grading and Levels of Evidence" and Table 1. Evidence tables are provided for recommendations given a grade of B and having two or more references.

  1. Evaluation of Bleeding Symptoms and Bleeding Risk by History and Physical Examination Summarized in Figure 3, and Box 1.
    1. Ask the following broad questions:
      1. Have you or a blood relative ever needed medical attention for a bleeding problem, or have you been told you had a bleeding problem? Grade B, level IIb138
        If the answer is "Yes" to either of the broad questions above, ask the additional probes:
        1. Have you needed medical attention for bleeding? After surgery? After dental work? With trauma?
        2. Have you ever had bruises so large they had lumps?
          Grade B, level IIb138
      2. Do you have or have you ever had:
        1. Liver or kidney disease?
        2. A blood or bone marrow disorder?
        3. A high or low platelet count?
          If the answer is "Yes" to any of these questions, obtain relevant details. Grade C, level IV
      3. Are you currently taking, or have you recently taken anticoagulation or antiplatelet medications (warfarin, heparin, aspirin, NSAIDs, clopidogrel)?
        If the answer is "Yes", obtain relevant details. Grade C, level IV
    2. If answers to questions I.A.1 are positive, ask if the patient or any blood relatives have had:
      1. A bleeding disorder, such as von Willebrand disease or hemophilia?
      2. Prolonged bleeding, heavy, or recurrent from:
        1. Trivial wounds, lasting more than 15 minutes or recurring spontaneously during the 7 days after the wound?
        2. Surgical procedures, such as tonsillectomy?
      3. Bruising with minimal or no apparent trauma, especially if you could feel a lump?
      4. Spontaneous nosebleeds that required more than 10 minutes to stop or needed medical attention?
      5. Dental extractions leading to heavy, prolonged, or recurrent bleeding?
      6. Blood in your stool, unexplained by a specific anatomic lesion (such as an ulcer in the stomach, or a polyp in the colon), that required medical attention?
      7. Anemia requiring treatment or received a blood transfusion?
      8. For women, heavy menses, characterized by the presence of clots greater than an inch in size and/or changing a pad or tampon more than hourly, or resulting in anemia or low iron level?
        If answers to above questions I.B.1–8 are positive, obtain relevant specific information. Grade B, level IIb138,139 See Evidence Table 1.
    3. Perform a physical examination to include evaluation for:
      1. Evidence for a bleeding disorder, including size, location, and distribution of ecchymoses (e.g., truncal), hematomas, petechiae, and other evidence of recent bleeding and/or anemia. Grade C, level IV
      2. Evidence that suggests other causes or risks of increased bleeding, such as jaundice or spider angiomata (liver disease), splenomegaly, arthropathy, joint and skin laxity (e.g., Ehlers—Danlos Syndrome), telangiectasia (e.g., hereditary hemorraghic telangiectasia), or evidence of anatomic lesions on gynecologic examination. Grade C, level IV

return to text, after Figure 3

Laboratory testing should be guided by the history and physical findings (section I.) and the initial laboratory evaluation (see II.A., below). For example, findings of liver disease may lead to a different or additional laboratory evaluation rather than an evaluation for VWD (see II.B., below).

  1. Evaluation by Laboratory Testing
    1. Initial laboratory evaluation for the etiology of a bleeding disorder should include:
      1. A complete blood count (CBC including platelet count), prothrombin time (PT), activated partial thromboplastin time (PTT), and optionally either thrombin time or fibrinogen level.
      2. If laboratory abnormalities besides the PTT are present (the platelet count may also be decreased in type 2B VWD), in conjunction with the history and physical examination findings, consider bleeding disorders other than VWD or additional underlying diseases.
      3. If the mucocutaneous bleeding history is strong, consider performing initial VWD assays at the first visit (see II.B., below).
      4. If there are no abnormalities on initial blood testing, or if there is an isolated prolonged PTT that corrects on the 1:1 mixing study, the following three tests for VWD should be performed (II.B., below), unless another cause for bleeding has been identified and VWD is not likely (see Figure 4). For further laboratory evaluation, physicians may consider referral to a hemostasis center because of the special sample handling and testing requirements (see Table 10).
        Grade C, level IV
    2. Initial tests for diagnosing or excluding VWD include the following three tests:
      1. VWF:RCo
      2. VWF:Ag
      3. Factor VIII activity
        Grade B, level III3,43,175,176 See Evidence Table 2.
    3. If any one of the above test results is abnormally low, a discussion with or a referral to a hemostasis expert is appropriate. In addition to repeating the initial three tests (in most cases), the specialist may recommend appropriate studies from the following:
      1. The first set of additional tests may include:
        1. Evaluation of the ratio of VWF activity (VWF:RCo and/or VWF:CB) to VWF antigen (only in laboratories that have defined reference ranges for the ratio[s]) Grade B, level III70,71,91,163,175,190 See Evidence Table 3.
        2. VWF multimer study
          Grade B, level III181
        3. Ristocetin—induced platelet aggregation
          Grade B, level III46
        4. VWF collagen binding activity (VWF:CB)
          Grade B, level IIb175,185,186 See Evidence Table 4.
      2. Studies in selected patients, especially those who have discordantly low FVIII activity compared to VWF levels and who are suspected of having type 2N VWD, should include a FVIII binding assay (VWF:FVIIIB) Grade B, level IIb77,78,188 See Evidence Table 5.
      3. Additional studies in selected persons may include:
        1. Gene sequencing
          Grade C, level IV
        2. Assays for antibodies to VWF
          Grade C, level IV
        3. Platelet—binding studies
          Grade B, level III182
  2. Making the Diagnosis
    1. Clinical criteria.Clinical criteria. These criteria include personal and/or family history and/or physical evidence of mucocutaneous bleeding. Until further validation of scoring systems and criteria for assessing bleeding history and the probability of VWD, especially type 1 VWD, the Expert Panel suggests that an increasing number of positive responses to the questions about bleeding (Figure 3, and Box 1) and abnormal findings on physical examination increase the likelihood that an individual has a bleeding disorder, including possible VWD.
    2. Laboratory criteria. The values in the following table represent prototypical cases without additional VWF (or other disease) abnormalities in the patient. In practice, exceptions occur, and repeat testing and clinical experience are important and may be necessary for interpretation of laboratory results.
      1. Although published evidence is limited, for defining the ratio of VWF:RCo/VWF:Ag to use for distinguishing type 1 VWD versus type 2 VWD variants (A, B, or M), the Expert Panel recommends a ratio of <.5–0.7 until more laboratories clearly define a reference range using large numbers of normal subjects and persons who have type 1 VWD and type 2 VWD variants.
        Grade C, level IV70,71,91,163,189,190
      2. The panel currently recommends that 30 IU/dL be used as the "cut—off" level for supporting the definite diagnosis of VWD for the following reasons:
        • There is a high frequency of blood type O in the United States, and it is associated with "low" VWF levels;43
        • Bleeding symptoms are reported by a significant proportion of normal individuals;137–140 and
        • No abnormality in the VWF gene has been identified in many individuals who have mildly to moderately low VWF:RCo levels
          Grade C, level IV100–102
    3. This recommendation does not preclude the diagnosis of VWD in individuals with VWF:RCo of 30–50 IU/dL if there is supporting clinical and/or family evidence for VWD. This recommendation also does not preclude the use of agents to increase VWF levels in those who have VWF:RCo of 30–50 IU/dL and may be at risk for bleeding.
Condition VWF:RCo (IU/dL) VWF:Ag (IU/dL) FVIII Ratio of VWF:RCo/ VWF:Ag
Type 1 <30* <30* ↓ or Normal >0.5–0.7
Type 2A <30* <30–200*† ↓ or Normal <0.5–0.7
Type 2B <30* <30–200*† ↓ or Normal Usually <0.5–0.7
Type 2M <30* <30–200*† ↓ or Normal <0.5–0.7
Type 2N 30–200 30–200 ↓↓ >0.5–0.7
Type 3 <3 <3 ↓↓↓(<10 IU/dL) Not applicable
"Low VWF" 30–50 30–50 Normal >0.5–0.7
Normal 50–200 50–200 Normal >0.5–0.7

↓ refers to a decrease in the test result compared to the laboratory reference range.
* <IU/dL is designated as the level for a definitive diagnosis of VWD; there are some patients with type 1 or type 2 VWD who have levels of VWF:RCo and/or VWF:Ag of 30–50 IU/dL.
The VWF:Ag in the majority of individuals with type 2A, 2B, or 2M VWD is <50 IU/dL.

return to text, after Figure 4

Back to Top
Back to Table of Contents

Skip footer links and go to content
Twitter iconTwitterExternal link Disclaimer         Facebook iconFacebookimage of external link icon         YouTube iconYouTubeimage of external link icon         Google+ iconGoogle+image of external link icon