Screening for von Willebrand Disease With a New Analyzer

Using High Shear Stress: A Study of 60 Cases

By Edith Fressinaud, Agne`s Veyradier, Florence Truchaud, Isabelle Martin,
Catherine Boyer-Neumann, Marc Trossaert, and Dominique Meyer
We have evaluated the performance of a new analyzer using
high shear stress, the PFA-100 (Platelet Function Analyzer,
Dade International, Massy, France), for screening of patients
with von Willebrand disease (vWD). Whole citrated blood is
aspirated through a capillary to the central aperture of a
membrane coated with collagen and with a platelet agonist
(either epinephrine or adenosine diphosphate [ADP]). The
time required to obtain occlusion of the aperture by a
platelet plug is defined as the closure time (CT). We studied
60 patients with different types of vWD and 96 normal
subjects. Fourteen subjects with hemophilia and 15 patients
with a platelet disorder were also analyzed. When omitting
results from two patients with type 2N, the 58 other patients
with type 1, type 2A, type 2B, type 3, or acquired vWD all

exhibited an abnormal occlusion with collagen-ADP (sensitivity, 100%) and 56 of 58 had an abnormal CT with collagenepinephrine (sensitivity, 96.5%). Only two patients with mild
type 1 were not detected with collagen-epinephrine. In
comparison, the bleeding time (BT) was normal in 20 patients: 17 with type 1, two with type 2A, and one with
acquired vWD (sensitivity, 65.5%). The specificity of the
PFA-100 was over 95% with both types of cartridges. Thus,
the analyzer is well adapted to routine testing, as it has the
advantages of simplicity and ease of execution, and demonstrates a high sensitivity, clearly superior to that of BT, for
the screening of patients with vWD.
r 1998 by The American Society of Hematology.

that allows a rapid and simple determination of platelet function

in primary hemostasis. We used the PFA-100 to screen 60
patients with different types of vWD. The present study shows
that the PFA-100 is strikingly more sensitive than the BT in
screening patients with vWD type 1, type 2A, type 2B or type 3,
and acquired forms of the disease.

ON WILLEBRAND DISEASE (vWD), which is the most

common inherited bleeding disorder with an incidence as
high as 1%,1 results from quantitative or qualitative defects of
von Willebrand factor (vWF).2 vWF is a large multimeric
glycoprotein and its degree of polymerization and integrity of
specific domains are essential for function of the protein.3,4
vWF is involved in platelet adhesion to the injured vessel wall,
platelet spreading, and platelet-platelet interactions under conditions of high shear.3,4 These functions require interaction of
vWF with two platelet receptors, glycoprotein (GP) Ib/IX, and
GPIIb/IIIa.5-8 vWF also serves to transport factor VIII (F.VIII)
and to protect it from proteolytic degradation.3,4
Recently, a pathophysiologic classification for vWD was
introduced by Sadler.9 Quantitative defects of vWF may either
be classified as partial (type 1) or total (type 3). vWD type 2
(subtypes 2A, 2M, 2B, and 2N) corresponds to a qualitative
defect, and distinct mutations have been found in the vWF gene
corresponding to different functional domains of vWF.10,11 Type
2A is characterized by both a decreased interaction of vWF with
platelets and lack of the highest sized multimers. In contrast,
type 2M corresponds to a decreased interaction of vWF with
platelets, which is not associated with a loss of high molecular
weight multimers. Type 2B is characterized by an increased
affinity of vWF for GPIb/IX. Type 2N is due to a qualitative
abnormality of vWF causing defective interaction with F.VIII,
which is not anymore protected from rapid proteolysis.
At present, the diagnosis of vWD relies on the determination
of the bleeding time (BT) and assessment of vWF antigen
(vWFAg), vWF ristocetin cofactor activity (vWFRCo), and F.
VIII activity. Ristocetin-induced platelet agglutination (RIPA)
and distribution of vWF multimers allow for further diagnosis
and classification.
Historically, the BT has been considered as an essential
screening test for primary hemostasis disorders and particularly
for vWD, except in type 2N, probably because of the lack of a
better alternative. However, the accuracy, validity, predictibility, and reproducibility of the BT have frequently been questioned.12 As a result, the BT is usually replaced by assays such
as vWFRCo and/or vWFAg that require more time and skill to
perform. In this study, we tested a new instrument, the PFA-100
(Platelet Function Analyzer, Dade International, Massy, France),
Blood, Vol 91, No 4 (February 15), 1998: pp 1325-1331

MATERIALS AND METHODS

Patients. Sixty patients with vWD were included in the study after
obtaining appropriate consent. Twenty-six men and thirty-four women,
with a mean age of 32.4 years (range, 2 to 77) were investigated. These
patients had not received vWF concentrates or DDAVP for at least 1
month. vWD had been previously diagnosed by a personal or familial
bleeding history and results of laboratory tests such as vWFAg,
vWFRCo, F.VIII, RIPA, and multimeric composition of vWF. For this
study, all parameters were retested and patients were classified according to the revised classification of vWD.9 The following vWD types
were represented in the 60 patients: type 1, n 5 36; type 2A, n 5 10;
type 2B, n 5 3; type 3, n 5 4; and type 2N, n 5 2. Among the 36
patients with type 1, 32 of them exhibited all criteria for definite type
1 vWD, ie, a personal history of significant mucocutaneous bleeding,
results of vWFRCo and vWFAg ,2 standard deviation (SD) below the
mean of the normal population when adjusted to the blood type and a
positive family history of vWD. The four other patients had possible
type 1 vWD, ie, positive personal history and laboratory criteria, but no
evidence of familial vWD. We also studied five patients with acquired
vWD associated with benign (n 5 3) or malignant (n 5 2) B-cell
disorders.
Fifteen patients with a previously well-defined platelet disorder were

From the Laboratoire dHematologie, CHU Hotel Dieu, Nantes; and

the Laboratoire dHematologie, Hopital Antoine-Becle`re, Clamart, and
INSERM U.143, Hopital de Bicetre, Kremlin-Bicetre, France.
Submitted January 29, 1997; accepted October 8, 1997.
Address reprint requests to Prof Dominique Meyer, MD, INSERM
U143, Hopital Bicetre, Secteur Violet, Porte 19, 94275 Le KremlinBicetre Cedex, France.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. section 1734 solely to indicate
this fact.
r 1998 by The American Society of Hematology.
0006-4971/98/9104-0003$3.00/0
1325

1326

also investigated: two with Glanzmann thrombasthenia, three with

pseudo- or platelet-type vWD, four with congenital storage pool
disease, and six with aspirin-like defect.
Control group. One hundred and ten subjects devoid of any primary
hemostasis disorder were used as controls. They had not taken any
medication for at least 2 weeks. Ninety-six healthy volunteers (31 men
and 65 women) served as the first control group. Their mean age was
34.5 years (range, 15 to 65). These individuals had no history of
bleeding. Fourteen subjects with hemophilia A (n 5 12) or B (n 5 2)
were used as the second control group. Informed consent was obtained
from each of them.
Laboratory tests. Hematocrit and platelet counts were performed
with a Coulter STKS cell counter (Coulter Corporation, Coultronics,
France). Blood samples were also collected for blood group. Citrated
whole blood (1 vol of 3.8% sodium citrate mixed with 9 vol of blood)
was obtained by clean venipuncture. The activated partial thromboplastin time (APTT) was measured with PTT-LT (Diagnostica Stago,
Asnieres, France) using an STA analyzer (Diagnostica Stago). F.VIII
activity was performed by a one-stage clotting assay based on the APTT
using F.VIII-deficient plasma (Diagnostica Stago) on the STA. vWFAg
was measured by enzyme-linked immunosorbent assay (ELISA) with a
commercial kit (Asserachrom vWF, Diagnostica Stago). vWFRCo was
assayed by aggregometry using a commercially available kit from
Behring (Marburg, Germany), which consists of lyophilized platelets
and ristocetin A. All results are the mean of two determinations and
expressed as IU/dL of plasma. The Second International Reference
Preparation for Factor VIII-related activities (National Institute for
Biological Standards and Control, London, UK) was used as a standard.
Multimeric composition of plasma vWF was estimated by sodium
dodecyl sulfate gel electrophoresis as previously described.13 RIPA was
performed in platelet-rich plasma on an aggregometer (ThromboAgregametre, Regulest, France) with ristocetin (Diagnostica Stago) at
three final concentrations: 0.5, 1.0, and 1.5 mg/mL. Platelet aggregation
was performed at 37C in platelet-rich plasma on the same aggregometer using 2.5 and 5 mol/L of adenosine-58-diphosphate (ADP), 5
mol/L of epinephrine from Diagnostica Stago, 1.0 and 4.0 g/mL of
equine collagen from Hormon Chemie, Munich, Germany, or 1.5
mmol/L of arachidonic acid from Helena Laboratories, Beaumont, TX.
BT was determined by a modified Ivy technique using the Simplate
sterile disposable device (Organon Teknika Corp, Durham, NC) according to the instructions of the manufacturer.
PFA-100 system. The PFA-100 is a high shear-inducing device that
simulates primary hemostasis after injury to a small vessel.14-16 The
system consists of a microprocessor-controlled instrument and a
disposable test cartridge. The test cartridge contains a reservoir for
citrated whole blood and a capillary (200 m internal diameter)
surmounted by a cup containing a biologically active membrane with a
central aperture (approximately 150 m diameter). The membrane is
coated with fibrillar type I equine tendon collagen. Additionally, either
epinephrine (10 g) or ADP (50 g) is present on the membrane. These
agents provide a controlled stimulation of platelets as the blood sample
passes through the aperture. The analyzer provides a constant negative
pressure that aspirates whole blood (800 L) through the capillary into
the cup where it comes into contact with the membrane and then passes
through the aperture. In response to stimulation by collagen, in
conjunction with either epinephrine or ADP, as well as by high shear
rates (5,000 to 6,000 second-1), platelets adhere and aggregate on the
membrane surface at the area surrounding the aperture. During the
course of measurement, a platelet plug forms that ultimately occludes
the aperture and blood flow is stopped. The time required to obtain
occlusion of the aperture is defined as the closure time (CT).
All normal subjects and patients were tested with both types of
cartridges (collagen/epinephrine or collagen/ADP). For each cartridge
type, results were based on the mean of duplicate testing. If results of
duplicate tests deviated by more than 20%, a third test was performed.

FRESSINAUD ET AL

Citrated whole blood was stored at room temperature less than 4 hours
before testing.
Statistical analysis. Mean, SD, and reference ranges were determined for all parameters. The normal range for the PFA-100 was
calculated as mean 6 2 SD of the healthy volunteers group. Sensitivity
was calculated as the percentage of correctly detected vWD patients,
using the formula: sensitivity (%) 5 true positive 3 100/true positive 1
false negative. Specificity was calculated as the percentage of values
within the normal range among the total control group (healthy
volunteers and hemophiliacs), using the formula: specificity (%) 5 true
negative 3 100/true negative 1 false positive. Positive predictive value
(PPV) was estimated as the percentage of patients with vWD among the
subjects whose CT is prolonged, using the formula: PPV 5 true
positive 3 100/true positive 1 false positive. Negative predictive value
(NPV) was estimated as the percentage of subjects devoid of any
primary hemostasis disorder among the subjects whose CT is in the
normal range, using the formula: NPV 5 true negative 3 100/true
negative 1 false negative. The analysis of correlations was performed
with statgraphics (Statview; Abacus Concepts, Berkeley, CA) as simple
regression.
RESULTS

Blood group, hematocrit, and platelet count. Blood group

type O was found in 32 healthy volunteers (33.3%), non-O in 64
(66.7%). Blood group O was found in seven subjects with
hemophilia (50%), non-O in three (21.4%), and was unknown
in four (28.6%). In vWD patients, 27 (45%) were of group O, 30
(50%) of group non-O, the blood group being unknown for
three (5%). In all controls and patients, the hematocrit was .
35% and platelet count was . 150 3 109/L, except in one
patient with type 2B vWD (90 3 109/L).
vWF assays. Table 1 shows the extreme values, mean, and
SD of vWF levels in the 96 healthy volunteers. Table 2 indicates
the results in the 14 hemophiliacs. Table 3 shows vWFAg and
vWFRCo levels in the 36 patients with type 1 vWD. Among
these patients, 18 had vWFRCo levels between 26 and 39
UI/dL, 10 between 11 and 25 IU/dL, and eight between 1 and 10
IU/dL. Results in patients with type 2 (2A, 2B, 2N), type 3, and
acquired vWD are reported in Table 4. One patient with type 2B
was tested during pregnancy, with a normal vWFRCo level (88
IU/dL) and a high vWFAg level (208 IU/dL).
BT. For evident ethical reasons, BT was not performed in
the control group. The values in vWD patients are indicated in
Tables 3 and 4. The BT was normal in the two patients with type
2N vWD. Only 38 of 58 patients with other types of vWD
showed a prolonged BT (. 8.5 minutes) (Fig 1). Twenty
patients had a normal BT: two with type 2A, one with an
acquired form, and 17 with type 1 (Fig 1). When excluding
results in type 2N vWD, the sensitivity of the BT, calculated
from these results, was 65.5%.
Table 1. Mean and SD of Laboratory Parameters
in 96 Healthy Volunteers

Extreme values
Mean
SD
Mean 6 2 SD

CT (sec)

vWFAg
(IU/dL)

vWFRCo
(IU/dL)

ADP

Epi

56-207
102.8
31.5
40-166

58-209
101.7
32
38-140

66-126
89
15.2
59-120

77-186
120
19.9
80-160

Abbreviation: CT, closure time.

SCREENING OF VWD WITH THE PFA-100 ANALYZER

1327

Table 2. Laboratory Parameters in 14 Patients With Hemophilia

F.VIII
(IU/dL)

Patients

Hemophilia A
1
2
3
4
5
6
7
8
9
10
11
12
Hemophilia B
1
2

,1
,1
,1
,1
3
3
5
11
14
16
25
27
1.5
3

CT (sec)

vWFAg
(IU/dL)

vWFRCo
(IU/dL)

ADP

Epi

66
200
60
73
157
57
68
147
110
76
148
87

54
164
60
68
157
54
56
125
101
71
137
84

97
63
119
119
91
103
99
101
89
92
101
87

108
87
138
117
98
123
133
108
115
126
120
132

74
190

58
186

109
60

128
77

Table 3. Laboratory Parameters in 36 Patients With Type 1 vWD

CT (sec)

Patients

vWFAg
(IU/dL)

vWFRCo
(IU/dL)

BT
(min)

ADP

Epi

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

48
40
34
62
39
45
37
35
38
45
42
38
35
29
59
48
31
38
28
23
39
30
38
24
24
44
26
15
23
12
15
27
12
29
9
9

39
39
38
38
36
35
34
34
34
34
31
30
28
28
27
27
26
26
22
22
22
22
16
15
14
13
12
11
10
10
10
10
9
9
9
5

10
8
6
8.5
18
6
6
16
5
6
5.5
8
18.5
6
.20
3.5
7
17
15
12
17
15
8.5
6.5
.20
.20
.20
.20
11.5
9
.20
11
4
8.5
.20
7.5

136
147
154
148
193
141
144
210
145
138
127
129
197
170
.250
141
.250
.250
.250
.250
181
181
197
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250

137
173
202
164
159
184
183
.250
218
180
174
191
.250
.250
.250
219
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250

CT with collagen-epinephrine test cartridges. The extreme

values of CT obtained in 96 normal subjects (Table 1) were 77
to 186 seconds, with a mean value of 120 seconds and a mean
CV of 5.4%. This control group was used to calculate the
normal range, indicating an upper limit of 160 seconds (mean 1
2 SD). Four healthy volunteers had a slightly prolonged CT
(167, 169, 174, and 186 seconds, respectively). The 14 hemophiliacs had a normal CT (Table 2). Among the 110 controls
without any primary hemostasis disorder, the frequency of
prolonged CT was 3.6%, indicating a specificity of 96.4%. Type
2N patients exhibited normal occlusion ( Fig 2). Fifty-six of the
remaining 58 patients showed abnormal occlusion (Tables 3 and
4). All patients with type 3 and type 2A vWD had infinite CT,
that is greater than 250 seconds (Fig 2). All patients with type
2B had a prolonged CT, even the patient with a normal
vWFRCo level. Among the 41 patients with acquired or type 1
vWD, 28 showed an infinite CT (Fig 2). Only two patients with
type 1 had a normal CT (137 and 159 seconds) and their level of
vWFRCo was 39 and 36 IU/dL, respectively (Table 3). When
excluding results from type 2N vWD, the sensitivity of the test
with collagen-epinephrine cartridges was 96.5%. The CT with
collagen-epinephrine exhibited a PPV of 93.3% and an NPV of
98.2%.
CT with collagen-ADP test cartridges. In the normal subjects (Table 1), the extreme values of CT were 66 to 126 seconds
Table 4. Laboratory Parameters in 24 Patients
With Types 2A, 2B, 2N, Type 3, and Acquired vWD

Patients

Type 2A
1
2
3
4
5
6
7
8
9
10
Type 2B
1*
2
3
Type 2N
1
2
Type 3
1
2
3
4
Acquired form
1
2
3
4
5

vWFAg
IU/dL

CT (sec)

vWFRCo
IU/dL

BT
(min)

ADP

Epi

75
92
85
79
69
30
72
45
92
50

43
24
23
22
22
14
14
,3
,3
,3

8.5
8.5
18
.20
.20
.20
.20
.20
.20
.20

.250
.250
.250
.250
.250
.250
.250
.250
.250
.250

.250
.250
.250
.250
.250
.250
.250
.250
.250
.250

208
51
61

88
17
15

9
.20
14

.250
.250

184
.250
.250

83
67

78
54

81
94

135
141

,1
,1
,1
,1

,3
,3
,3
,3

.250
.250
.250
.250

.250
.250
.250
.250

49
42
48
27
14

40
36
33
21
3

140
179
.250
.250
.250

166
.250
.250
.250
.250

5.5
4
.20
.20
.20
.20
12
7.5
9
18
10

*During pregnancy (5 months).

Obstruction of the capillary by platelet aggregates.

1328

FRESSINAUD ET AL

Fig 1. Comparison of BT (Simplate) and measurement of CT with collagen-ADP and collagen-epinephrine cartridges in the PFA-100 in patients
with von Willebrand disease (n 5 60). Each type of vWD is identified by the following symbol: (X) type 1, (W) type 2A, (M) type 2B, (Q) type 2N,
(N) type 3, and (S) acquired vWD. The interrupted lines represent the upper limit of the normal range (8.5 minutes) for the BT and the mean value
1 2 SD of the control group for CT (120 seconds with collagen-ADP and 160 seconds with collagen-epinephrine).

with a mean value of 89 seconds (mean CV of 4.5%), indicating

an upper limit of 120 seconds (mean 1 2 SD). Only one healthy
volunteer had a slightly prolonged CT (126 seconds). All of the
hemophiliacs had a normal CT (Table 2). Thus, less than 1% of
the control group showed a prolonged CT, indicating a specificity of 99%. The CT was normal in type 2N vWD. All other

patients exhibited abnormal occlusion (Fig 3). The CT was

infinite (. 250 seconds) in all patients with type 3 and type 2A
vWD, in two with type 2B, in 18 with type 1, and in three with
acquired vWD (Fig 3). In the pregnant patient with type 2B, the
CT could not be estimated because of rapid flow obstruction of
the capillary. When omitting results from type 2N patients, the

Fig 2. Measurement of CT with collagen-epinephrine cartridges in the PFA-100 in normal subjects and in patients with vWD disease. Each
closed circle is the mean of duplicate testing in an individual. The solid line represents the mean value in the normal subjects (120 seconds); the
interrupted line across the figure represents the upper limit of normal (160 seconds), which was calculated as the mean (m) 1 2 SD.

SCREENING OF VWD WITH THE PFA-100 ANALYZER

1329

Fig 3. Measurement of CT with collagen-ADP cartridges in the PFA-100 in normal subjects and in patients with vWD. Each closed circle is the
mean of duplicate testing in an individual. The solid line represents the mean value in the normal subjects (89 seconds); the interrupted line
across the figure represents the upper limit of normal (120 seconds), which was calculated as the mean (m) 1 2 SD.

sensitivity of the test with collagen-ADP cartridges was 100%.

The CT collagen-ADP exhibited a PPV of 98.3% and an NPV of
100%.
Correlations. Figure 4 shows the correlation between CT
and vWFRCo levels in the 96 healthy volunteers: R 5 0.62 with
ADP and R 5 0.56 with epinephrine. The correlation between
CT and vWFRCo levels in patients with vWD could not be
calculated because a large number of patients had an infinite CT
(. 250 seconds): n 5 37 with collagen-ADP cartridges and n 5
44 with collagen-epinephrine cartridges (Tables 3 and 4).
However, results in type 1 vWD indicated that on the whole the

Fig 4. Correlation between

vWFRCo levels and CT with either collagen-epinephrine or collagen-ADP in 96 normal subjects.

prolongation of the CT was inversely proportional to the level

of vWFRCo (Table 3).
DISCUSSION

The detection of vWD is still a challenge, especially in the

mild forms. Because there is an abnormal interaction of
platelets with the subendothelium in all types of vWD except
type 2N,17 the BT has been considered in the past as a useful
screening tool to detect these patients. Due to poor sensitivity,
lack of reproducibility and large variability, the BT is however

1330

not a suitable method for screening of vWD. The poor

sensitivity of the BT has been illustrated by the Italian working
group,12 showing that about 20% of patients with moderate
vWD have a normal BT. Our study confirms these findings and
indicates that 47% of patients with type 1 vWD, 20% with type
2A, and 20% with acquired vWD, have a normal BT.
An optimal screening procedure to detect patients with a
primary hemostasis disorder needs to be reproducible, sensitive,
and specific. The PFA-100 shows excellent reproducibility as
results of duplicate tests deviated by more than 20% in only 2%
of controls and 0% of patients.
The sensitivity of the PFA-100 is also highly satisfactory.
When omitting results in type 2N vWD (showing as expected a
normal CT), the PFA-100 detected all patients with vWD using
the collagen-ADP cartridges and all but two patients using the
collagen-epinephrine cartridges. The latter two patients had a
positive history of mild bleeding and laboratory criteria (vWFRCo levels of 39 and 36 IU/dL) in favor of vWD, but no
evidence for an inherited form. According to the recent recommendations of the Standardization and Scientific Committee
(SSC) of the International Society on Thrombosis and Haemostasis (ISTH), both patients had possible and not definite
type 1 vWD. In addition, it is well known that individuals with
group O have lower levels of vWF18 than those of other groups
and may have a very mild bleeding tendency or none at all.
Knowing that these two subjects were of group O, they may also
be considered as borderline normal subjects and do not really
limit the excellent sensitivity of the PFA-100 in detecting
patients with vWD.
The specificity of the PFA-100 was assessed by studying a
large control group devoid of any primary hemostasis disorder
(healthy volunteers and hemophiliacs); we detected four of 96
normal subjects whose CT with epinephrine was either slightly
prolonged or prolonged, and one whose CT was borderline with
ADP. A forgotten medication with any drug interacting with
primary hemostasis may have explained the four prolonged CT
with epinephrine, although the platelet aggregation tests were
found to be normal. As expected, results in hemophilia showed
that the CT was not affected by coagulation abnormalities. The
good specificity of CT was confirmed by the estimation of
predictive values. Positive and negative predictive values were
excellent, the latter being essential for a screening test used to
exclude the diagnosis of vWD.
The relationship between BT and plasma vWFRCo levels has
been analyzed by several investigators and Weiss,19 in particular, reported a good correlation between both tests. In the
present study, we analyzed the correlation between the CT
measured in the PFA-100 and the plasma vWFRCo levels. The
correlation was fair in normal subjects, but could not be
calculated in vWD patients because a large number exhibited an
infinite CT due to the high sensitivity of the method.
A recent study16 has shown the involvement of vWF in
platelet adhesion and aggregation under the high shear stress
conditions of the PFA-100 by demonstrating that monoclonal
antibodies to vWF, specific for the collagen-, GPIb/IX-, or
GPIIb/IIIa-binding sites of vWF, caused a dose-dependent
prolongation of the CT. In contrast, a polyclonal antibody to
fibrinogen did not affect the CT.16 In fact, we tested a patient
with afibrinogenemia and found a normal CT with both types of

FRESSINAUD ET AL

cartridges (results not shown). These results are in agreement

with those of perfusion studies6 indicating that under high shear
stress flow conditions, vWF is the only protein to play a role in
platelet adhesion and aggregation. This has been corroborated
by experiments of shear-induced platelet aggregation.20
Interestingly, the PFA-100 was capable of detecting patients
with type 2B vWD. This is in agreement with our results of
perfusion studies21 and with experiments using the blood
filtration test showing defective platelet retention and aggregation in all vWD patients tested.22 On the contrary, 2B variants
are not detected by shear-induced platelet aggregation measured
in the cone-and-plate viscometer23 where there is an increased
aggregation, paradoxically masking the defective platelet adhesion and thrombus formation in these patients. In vivo, binding
of abnormal vWF to platelet GPIb results in occupation of the
receptor so that it is no longer available for mediating platelet
adhesion. The PFA-100 system appears to detect a decreased
ability of platelets to adhere rather than an increased tendency
of platelets to aggregate. Therefore, this system may mirror the
defective hemostatic function of vWF in type 2B as well as in
all other types of vWD, except type 2N. Thus, the PFA-100
reflects vWF-dependent adhesive interactions as they occur in
vivo and is a global predictor of vWF-dependent platelet
function under high shear stress.
The PFA-100 may be used in a more general setting to predict
the bleeding tendency resulting from functional platelet alterations in patients with defects other than vWD and to monitor
vWD patients treated by deamino-8-D-arginine vasopressin
(DDAVP). Studies using monoclonal antibodies to GPIb or to
GPIIb/IIIa, and aspirin14-16 have shown the usefulness of the the
PFA-100 for detecting patients with inherited or acquired
platelet dysfunction. Indeed, we tested patients with various
platelet disorders and found that the PFA-100 is an excellent
analyzer for their screening. All patients showed prolonged CT
using epinephrine and ADP, except three with an aspirin-like
defect whose CT was normal with ADP (Table 5). However,
ADP is known not to be discriminating for the diagnosis of
aspirin-like defects.15 In regard to the therapeutic monitoring of
vWD patients, we found that DDAVP completely corrected the
Table 5. CT in 15 Patients With Platelet Disorders
CT (sec)
Platelet Disorder

Glanzmann thrombasthenia
(n 5 2)
Pseudo (platelet)-vWD
(n 5 3)
Storage pool disease (n 5 4)

Aspirin-like defect (n 5 6)

Normal range (mean 6 2 SD)

Patients

ADP

Epi

1
2
1
2
3
1
2
3
4
1
2
3
4
5
6

.250
.250
.250
.250
.250
.250
.250
.250
.250
150
130
121
115
104
81
59-120

.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
.250
228
216
210
80-160

SCREENING OF VWD WITH THE PFA-100 ANALYZER

CT with both types of cartridges in the 11 cases with type 1

vWD that we tested (results not shown), emphasizing the
usefulness of the PFA-100 as compared with the other tests (BT,
vWFRCo, and vWFAg).
In conclusion, the PFA-100 is well adapted to routine testing,
as it has the advantage of simplicity and ease of execution. It
provides fast results and uses the same citrated blood that is
routinely drawn for other coagulation testing; the latter is
particularly useful in emergency situations, especially before
surgery. The test can be safely delayed for up to 4 hours from
the time of blood sampling, provided that the blood is kept at
room temperature. When expressed by its sensitivity, specificity, and predictive values, the clinical performance of the test is
excellent in vWD with both types of cartridges, possibly
slightly better in the presence of ADP. The high sensitivity
demonstrated by the PFA-100 analyzer is clearly superior to that
of BT for the detection of patients with vWD. This leads to the
proposal of the following strategy: if the CT is within the
normal range, one may exclude the diagnosis of vWD (except
type 2N) and probably of a hereditary or acquired platelet
disorder; if the CT is abnormal, a medical questioning should
first search for any medication capable of interacting with
platelet aggregation. Secondly, the CT should be retested before
measuring vWFRCo and/or vWFAg levels and investigating for
platelet function defects.
ACKNOWLEDGMENT
We are most grateful to Marlies Ledford for her helpful comments.
We also thank Christine Euzen and Anne-Lise Marville-Gigot for expert
secretarial assistance.
REFERENCES
1. Rodeghiero F, Castaman G, Dini E: Epidemiological investigation
of the prevalence of von Willebrands disease. Blood 69:454, 1987
2. Gralnick HR, Ginsburg D: von Willebrand disease, in Beutler A,
Lichtman MA, Coller BS, Kipps TJ (eds): Williams Hematology, (ed 5).
New York, NY, McGraw-Hill, 1995, p 1458
3. Ruggeri ZM, Ware J: The structure and function of von Willebrand factor. Thromb Haemost 67:594, 1992
4. Meyer D, Girma JP: von Willebrand factor: Structure and
function. Thromb Haemost 70:99, 1993
5. Weiss HJ, Turitto VT, Baumgartner HR: Effect of shear rate on
platelet interaction with subendothelium in citrated and native blood. I.
Shear rate-dependant decrease of adhesion in von Willebrands disease
and the Bernard-Soulier syndrome. J Lab Clin Med 92:750, 1978
6. Weiss HJ, Hawiger J, Ruggeri ZM, Turitto VT, Thiagarajan P,
Hoffmann T: Fibrinogen-independent platelet adhesion and thrombus
formation on subendothelium mediated by glycoprotein IIb-IIIa complex at high shear rate. J Clin Invest 83:288, 1989
7. Savage B, Shattil SJ, Ruggeri ZM: Modulation of platelet function
through adhesion receptors: A dual role for glycoprotein IIb-IIIa
(integrin aIIb b3) mediated by fibrinogen and glycoprotein Ib-von
Willebrand factor. J Biol Chem 267:11300, 1992

1331

8. Goto S, Salomon Dr, Ikeda Y, Ruggeri ZM: Characterization of

the unique mechanism mediating the shear-dependent binding of
soluble von Willebrand factor to platelets. J Biol Chem 270:23352,
1995
9. Sadler JE: A revised classification of von Willebrand disease. For
the Subcommittee on von Willebrand factor of the Scientific and
Standardization Committee of the International Society on Thrombosis
and Haemostasis. Thromb Haemost 71:520, 1994
10. Ginsburg D, Sadler JE: von Willebrand disease: A database of
point mutations, insertions and deletions. Thromb Haemost 69:177,
1993
11. Mazurier C, Meyer D: Molecular basis of von Willebrand
disease, in Haemophilia, Baillie`res Clinical Haematology (vol 9).
London, UK, 1996, p 229
12. Italian Working Group: Spectrum of von Willebrands disease: A
study of 100 cases. Br J Haematol 35:101, 1977
13. Meyer D, Obert B, Pietu G, Lavergne JM, Zimmerman TS:
Multimeric structure of factor VIII/von Willebrand factor in von
Willebrands disease. J Lab Clin Med 95:590, 1980
14. Kundu SK, Heilmann EJ, Sio R, Garcia C, Davidson RM,
Ostgaard RA: Description of an in vitro platelet function analyzer
PFA-100TM. Semin Thromb Hemost 21:106, 1995 (suppl 2)
15. Mammen EF, Alshameeri RS, Comp PC: Preliminary data from a
field trial of the PFA-100TM system. Semin Thromb Hemost 21:113,
1995 (suppl 2)
16. Kundu SK, Heilmann E, Sio R, Garcia C, Ostgaart R: Characterization of an in vitro platelet function analyzer, PFA-100TM. Clin Appl
Thromb Hemost 2:241, 1996
17. Meyer D, Fressinaud E, Gaucher C, Lavergne JM, Hilbert L,
Ribba AS, Jorieux S, Mazurier C and the INSERM Network on
molecular abnormalities in von Willebrand disease: Gene defects in 150
unrelated French cases with type 2 von Willebrand disease: From the
patient to the gene. Thromb Haemost 78:451, 1997
18. Gill JC, Endres-Brooks J, Bauer PJ, Marks WJ, Montgomery
RR: The effect of ABO blood group on the diagnosis of von Willebrand
disease. Blood 69:1691, 1987
19. Weiss HJ: Relation of von Willebrand factor to bleeding time. N
Engl J Med 291:420, 1974 (letter)
20. Ikeda Y, Handa M, Kawano K, Kamata T, Murata M, Araki Y,
Anbo H, Kawai Y, Watanabe K, Itagaki I, Sakai K, Ruggeri ZM: The
role of von Willebrand factor and fibrinogen in platelet aggregation
under varying shear stress. J Clin Invest 87:1234, 1991
21. Fressinaud E, Federici AB, Castaman G, Mannucci PM, Meyer
D: Ex vivo experimental thrombosis in variants of von Willebrand
disease. C R Acad Sci Paris, Sciences de la vie/Life Sci 316:1260, 1993
22. OBrien JR, Salmon GP: Shear stress activation of platelet
glycoprotein IIb-IIIa plus von Willebrand factor causes aggregation,
filter blockage and the long bleeding time in von Willebrands disease.
Blood 70:1354, 1987
23. Pareti IF, Cattaneo M, Carpinelli L, Zighetti ML, Bressi C,
Mannucci PM, Ruggeri ZM: Evaluation of the abnormal platelet
function in von Willebrand disease by the blood filtration test. Thromb
Haemost 75:460, 1996