CLIN. CHEM. 41/4, 51 0-514 (1995)
#{149}
Coagulation
Four Agglutination Assays Evaluated for Measurement of von Willebrand Factor
(Ristocetin Cofactor Activity)
Anthony
A. M. Ermens,’
Piet
J. de Wild, Huib L. Vader, and Fedde van der Graaf
The concentration of von Willebrand factor (vWf) in patients’
plasma can be determined by measuring the nstocetin
cofactor activity (vWf R:Co). However, this vWf R:Co assay
is time consuming, which limits its routine use. Several
commercial
vWf R:Co tests, based on agglutination of
lyophilized fixed platelets, are available. We evaluated the
slide tests and aggregometer assays from Behring and
Organon Teknika and compared them with the classic vWf
R:Co aggregometer method. The within-run and betweenrun precisions of the two slide tests were better than those
of the aggregometer methods. The correlation studies between the four commercial assays and the classic aggregation method were based on 23 plasma samples (range:
15-450% vWf R:Co). The correlation coefficients, which
ranged from 0.923 to 0.950, did not differ significantly (P
>0.1). All four commercial assays gave significantly lower
vWf R:Co values than the classic aggregation method (P
<0.01). We conclude that commercially available fixed
platelets can be used for the rapid measurement of vWf
R:Co with a slide test. The use of the aggregometer is time
consuming and may result in a lower precision.
IndexIng Terms: coagulation Is/ide test/platelets
factor (vWf) plays an essential role in
platelet
adhesion and coagulation.2 Reduced concentrations or a modified structure
of vWf may cause coagulopathy (1,2). The quantitative
determination
of vWf is
Von Wifiebrand
both
therefore of great importance
in the diagnosis
of von
Wifiebrand disease, which is the most common congenital
hemorrhagic
disorder.
The concentration
of vWf in
plasma can be assessed by its ability to promote agglutination of platelets in the presence of the antibiotic ristocetin A (3). In the ristocetin cofactor assay, a source of
normal platelets, a standard
amount of ristocetin A, and
the patient’s plasma are combined and the velocity of
agglutination
of the platelets
is determined
with an
aggregometer.
The amount of ristocetin
cofactor activity
(vWf R:Co) is then interpolated
from a previously
propared calibration
curve (3). However, this classic vWf
R:Co assay
is time
as fresh
consuming,
platelet
suspen-
sions have to be prepared from donor blood.
The measurement
ofvWfR:Co
can also be performed
Clinical
Laboratories, Sint Joseph Hospital,
5500 MB, Veldhoven, The Netherlands.
‘Author for correspondence.
2Nonsdard
abbreviations:
P.O. Box 7777,
Fax (31)40588929.
510
29, 1994; accepted
January
Materials and Methods
Samples
Citrated plasma was prepared by mixing nine volumes of freshly drawn blood obtained by venipuncture
with one volume of sodium citrate (106 mmol/L) in
polypropylene
tubes. After centriftigation
of the blood
(10 mm, 1500g), the plasma
was stored in polypropylene
vials at -70#{176}C
before use. All samples were
analyzed within 6 months of blood collection.
“Normal
plasma”
was pooled citrated
plasma
collected from 44 apparently
healthy donors who had not
taken any medication
(including aspirin and oral contraceptives).
The plasma
was pooled and frozen at
-70#{176}C
in 0.5-mL aliquots in polypropylene
vials.
For the precision studies, two pooled samples with a
low (-30% of the activity level of normal plasma) and a
high (-80% of the activity level of normal plasma) vWf
R:Co activity, respectively,
were prepared from two
patients’
samples.
Plasma samples from 23 patients
were used for the
correlation study. Among them were six patients with
von
Willebrand
disease
factor; vWf
5, 1995.
CLINICAL CHEMISTRY, Vol. 41, No. 4, 1995
plasma.
type
I, one
patient
with
von
Willebrand disease type lib, and three von Willebrand
disease patients receiving 1-desamino-8-D-arginine
Vasopressin
therapy.
The remaining
13 patients
were
selected
from the hospital
inpatient
population
and
had no known coagulation
disorders
according to their
histories
and normal values for prothrombin
time and
activated partial thromboplastin
time.
Classic Aggregometer
vWf, von Willebrand
R:Co, ristocetin cofactor activity; and PRP, platelet-rich
Received August
formaldehyde-fixed
(4-6)
and (or) lyophilized
(7)
platelets.
These treatments
allow long-term
storage of
platelets
for vWf R:Co determination,
as they retain
the receptor for vWf. This can reduce the vWf R:Co
assay time considerably.
Several vWf R:Co tests, all
based on ristocetin-induced
agglutination
of lyophilized, fixed platelets,
are now commercially
available.
Monitoring
of platelet agglutination
is done either by
an aggregometer
or by visual inspection
on a special
slide. However, little is known about the analytical and
practical performance of these commercial tests. In this
study we evaluate
two commercially
available
vWf
R:Co reagents (Behring and Organon Teknika) used in
both an aggregometer
and a slide method.
with
This
Weiss
Method
is carried out essentially
as described
by
et al. (3), with some minor modifications.
Briefly,
test
mL of donor blood
is collected
in a plastic tube
containing
5 mL of sodium citrate (106 mmolIL). Platelet-rich plasma (PR?) is prepared by centriftigation
for
15 mm at 175g. The PR? is centrifuged
for 15 miii at
360g. The supernate is discarded and the platelets
are
resuspended
in 5 mL Tris-EDTA buffer (EDTA, 2 gfL,
Tris 0.25 mol/L
in saline, pH 7.3). The platelets
are
washed three times and finally resuspended
in TrisEDTA buffer at a concentration
of 250 X i09/L. In an
aggregometer
cuvette,
425 pi of this platelet suspension is pipetted together with 50 pL of (diluted) plasma
and 25 L of ristocetin
A (20 g/L; Diamed, Cressiersur-Morat, Switzerland).
Agglutination
is monitored in
an aggregometer
(Lumi aggregation
module
series
1000B; Payton, Scarborough,
Canada) by recording the
increase
of transmission
over several minutes.
The
slope of this transmission
curve is proportional
to the
vWfR:Co. The amount ofvWfR:Co
activity in a plasma
sample
(expressed
as the percentage
of the normal
plasma
activity level) was read from a calibration
curve, constructed
from serial dilutions
of normal
plasma (100%) with isotonic saline to final vWf R:Co
activities of 100%, 50%, 25%, 12.5%, and 6.3%.
45
Organon vWf R:Co Reagent
The vWf R:Co test kit of Organon (Durham,
NC)
contains
two vials of lyophilized
platelets
with buffer
and stabilizers.
Ristocetin A, a normal
reference
plasma (not used in this study), and two control plasmas are also provided in lyophilized
form by Organon.
The ristocetin A is reconstituted
with distilled water.
The lyophilized
plasmas are reconstituted
in a buffer
(0.05 mol/L Tris, 0.1 mol/L NaCl, pH 7.35) provided in
the test kit. According to the manufacturer,
the stability of the reconstituted
platelets is -8 h at 2-8#{176}C.
For the aggregometer
method the manufacturer’s
directions
in the test kit recommend
the use of a
suspension
of -100
X i0 platelets/L.
However,
we
found that concentration
too low to obtain reliable
changes in transmittance.
We therefore increased the
platelet concentration
to -250 X i0fi.,, in accordance
with the concentration
used in the classic method and
in the Behring aggregometer
method.
A vial of lyophilized
platelets is reconstituted
with 2
mL of distilled
water containing
1 g/L ristocetin
A.
After standing
for 30 mm at room temperature
the
platelet suspension
is ready for use. Into an aggregometer cuvette 450 giL of platelet suspension
is pipetBehring vWf R:Co Reagent
ted; after incubation
for 2 miii, 50 giL of plasma is
The vWf R:Co reagent
of the Behring
(Marburg,
added. The velocity of agglutination
is determined
as
Germany)
test kit consists
of five vials containing
described above. The calibration
curve, prepared with
stabilized human platelets,
ristocetin
A, and EDTA in
normal plasma, had a detection
limit of 12.5% vWf
lyophilized
form. According to the manufacturer,
the
R:Co; no reliable changes in transmittance
could be
stability of reconstituted
platelets is -5 days at 2-8#{176}C. recorded below this vWf R:Co activity.
For the Behring
aggregometer
method one vial of
For the Organon
slide test a vial of lyophilized
lyophilized
platelets
is reconstituted
with 5 mL of
platelets is reconstituted
with 2 mL of distilled water.
isotonic saline containing
1 g/L ristocetin A. The reFrom this suspension
0.1 mL is pipetted into a test
agent is then ready for use and contains -260 X iO
tube together with 25 giL of ristocetin
A (10 g/L). After
platelets/L.
In an aggregometer
cuvette
450 L of
preincubation
for 10 mm at room temperature,
5 giL of
platelet
suspension
is pipetted
together with 50 p.L of
undiluted
plasma is added. After swirling
for 5 s the
(diluted)
plasma. The agglutination
is monitored
as
mixture
is pipetted onto a slide and is gently agitated
described above. The calibration curve is prepared
with
until agglutination
occurs. The time required to obtain
serial dilutions of the normal plasma mentioned before.
4+ agglutination
is recorded. The percentage
of vWf
For the Behring slide test a vial of lyophilized platelets
R:Co activity in the plasma sample can be determined
is reconstituted
with 1 mL of distilled water to a concenby interpolation
on a calibration
curve prepared from
tration of 1300 X 109/L. According to the manufacturer
serial dilutions of normal plasma.
the concentration
of ristocetin A in the vial is sufficient
for agglutination
in the slide test. Diluted plasma (50 L)
Evaluation of vWf R:Co Assays
is pipetted onto a slide and mixed with 50 giL of platelet
The precision of the vWf R:Co tests is determined
suspension. The slide is slowly agitated for 1 miii and left
with the low-level
(-30%)
and high-level
(-‘-80%)
to stand for another minute. With use of an indirect light
plasma samples.
The within-run
CV is based on 10
source and a dark background,
the slide is inspected for
subsequent
vWf R:Co measurements
made on 1 day.
the presence of agglutination.
Serial stepwise dilutions of
The between-run
CV is based on 10 vWfR:Co measurea plasma sample are prepared. The dilution factor of the
most diluted plasma
sample
that still gives distinct agments made on 10 days over a 2-week period.
The correlation
between
the four commercial
vWf
glutination
and the dilution factor of the next plasma
dilution (which does not give distinct agglutination)
are
R:Co assays and the classic aggregometer
method is
based on 23 plasma samples containing
various vWf
determined.
The vWfR:Co of the sample is obtained as a
range by multiplying
these two dilution factors with the
R:Co activities (range: 15-450% vWf R:Co according to
detection limit of the vWf R:Co reagent. This detection
the classic aggregometer
method). The regression lines
limit is provided by Behring and is based on serial
between the five tests are calculated
according to the
method of Passing
and Bablok (8). The correlation
dilutions
of a standard plasma. Reduction of the intervals
between dilutions allows the determination
of vWf R:Co
coefficients
are determined
with the Spearman
rank
within narrower limits.
test.
The results
of the four commercial
vWf R:Co
CLINICAL CHEMISTRY, Vol. 41, No. 4, 1995
511
assays
tically
and the classic aggregometer
method are statiscompared
by use of the signed rank test.
Resufts
Table 1 shows the within-run
and between-run
CVs
of the studied
vWf R:Co assays.
The results
of the
Behring
slide test are presented
as the mean of the
reported
ranges,
which varied from 5% to 10%. At both
the low and high level of vWf R:Co, the slide tests
showed
ods.
better precisions
There
is
than the aggregometer
a substantial
difference
meth-
between
the
mean within-run
and between-run
values of the low
level in the classic aggregometer
method.
The correlations
between the four commercial
vWf
R:Co assays and the classic aggregometer
method are
presented in Fig. 1. The differences between the correlation coefficients
were not statistically
significant
(P
>0.1). All four commercial
assays gave significantly
Table 1. Precision of the studied vWf R:Co assays.
High-level vWt R:Co(80% normal)
Low-level vWf R:Co (30% normal)
Within-run
Behring slide
Organon slide
Behring aggregometer
Organon aggregometer
Classic aggregometer
0
(n
=
10)
Between-run (n
=
10)
Mean, %
CV %
Mean, %
CV, %
34
31
37
34
42
9
6
11
12
12
36
31
30
34
30
9
6
11
15
12
100
200
300
400
Aggr.gomster method with fresh platslsts
WithIn-run
Mean, %
85
89
80
79
87
(n
=
10)
CV, %
5
3
15
8
9
Aggr.gomstsr method with fresh
Between-run (n
Mean, %
88
85
83
78
86
=
10)
CV, %
5
5
19
14
14
platelets
I
I.
0
100
200
300
400
Aggr.gometer
method with fresh platelets
0
100
Aggrsgom.ter
200
300
400
method with fresh platelets
Fig. 1. Correlation between vWf R:Co measured by the classic aggregometer method and (A) the Behring aggregometer method, (B) the
Behnng slide test, (C) the Organon aggregometer method, and (D) the Organon slide test (n = 23 for all, r = 0.942, 0.944, 0.923, and 0.950,
respectively).
The dotted line indicates the line of identity;the solid line represents the line of best fit. 4) Intercept -0.043(95% confidence interval -8.061-6.667), slope 0.766
(0.667-0.837); (B) Intercept 7.312(-2.667-19.706),
slope 0.706(0.588-0.827);
(C)Intercept -1.057(-11.571-6.667), slope 0.562(0.467-0.714); (D)intercept
13.449 (3.667-23.227), slope 0.709 (0.547-0.870).
512
CLINICAL CHEMISTRY, Vol. 41, No. 4, 1995
lower vWf R:Co values than the classic aggregation
method (Organon
slide test P <0.01, other tests P
<0.001). The sample from the patient with type lib von
Willebrand
disease did not behave differently from the
samples of patients with type I von Willebrand disease
10
in any of the performed
assays.
The correlation
between the aggregometer
and slide methods
of the two
vWf R:Co reagents
is presented
in Fig. 2. In both cases
the slide method
yielded
higher vWf R:Co results
than
6
the aggregometer
method (Behring P <0.05, Organon
P <0.001).
However,
the differences
were
less pronounced
than
those between the slide tests and the
classic aggregometer
method.
Typical calibration
curves of the three aggregometer
methods are shown in Fig. 3. The Behring aggregometer method gave the highest agglutination
response to
vWfR:Co.
Discussion
vWf is well recognized
platelet
adhesion
and
as an important
factor in both
coagulation
(1, 9). Quantifica-
/
B
/
/
‘U
/
U,
..
DI
4.
U.
2
p.
..
0
#{149}li’-i10
15
20
100
40
vWf R:Co (%)
Fig. 3. Calibration curves of the classic (-), Behring (--) and
Organon (.. . .) aggregometer methods.
Tg a is defined as the slope of the steepest part of the agglutination-induced
absorbance
curve.
tion of vWf in a plasma sample can be done either by an
immunochemical
assay (vWf antigen)
or by measurement of the ristocetin
cofactor activity of vWf, that is,
its ability to promote agglutination
of platelets in the
presence
of the antibiotic
ristocetin
(2). vWf R:Co
determination
with the venom coagglutinin
botrocetin
has also been developed
(10). vWfR:Co determination
e
has become
U
C
essential
in the assessment
with suspected
von Willebrand
In 1973 a quantitative
assay
gating property of vWf was
suspension
this test
of freshly
washed
has not been routinely
of any
patient
disease.
for the ristocetin-aggredeveloped,
involving
a
platelets
(3). However,
used
because
it re-
quires
method
Behrlng aggregometer
an aggregometer
and laborious preparation.
The
test was later modified to include
lyophilized
and (or)
paraformaldehyde-fixed
platelets.
Further
modifications include the use of a macroscopic platelet agglutination test (slide test) (11, 12). In the few reports that
have been published on the usefulness
of these assays
commercially
available
vWf R:Co slide tests
were
shown to be easy to perform
and correlated
well with
each other (11) or with a vWf R:Co aggregometer
method
involving
fixed platelets
(12). However,
until
now
0
50
100
150
200
Organon aggregometer
250
method
Fig. 2. Correlation between vWf R:Co measured by (/1) the Behring
aggregometer method and the Behnng slide test and by (B) the
Organon aggregometer method and the Organon slide test (n = 23,
r = 0.955 and 0.935, respectively).
The dotted line indicates the line of identity: the solid line represents the line
of best fit. (A) Intercept 5.000(95% confidence interval -0.455-9.594), slope
1.000(0.894-1.091); (B) Intercept 17.653(1.556-28.667), slope 1.027(0.8671.347).
no data
on the
analytical
performance
of these
commercial
assays had been presented
nor were they
compared
with the classic aggregometer
method
of
Weiss et al. (3). We therefore
evaluated
two commercially available
vWf R:Co reagents that can be used in
both a slide test and in an aggregometer
method and
compared
them with the classic aggregometer
method.
The analytical
evaluation
indicates
that both slide
tests have a better precision
than the aggregometer
methods,
possibly because the aggregometer
methods
require more manual
steps. In addition,
the determination of the recorded slope of transmittance
is always
arbitrary,
being done visually. The observed difference
between
the low-level
within-run
and between-run
mean
caused
of the
classic
aggregometer
by the high imprecision
method
may
be
of the assay.
CLINICAL CHEMISTRY. Vol. 41, No. 4, 1995
513
Comparison
of the
calibration
curves
of the
three
aggregometer
methods
shows that the agglutination
response of fixed platelets
of the Organon kit, even at a
high platelet concentration,
is low. Consequently,
the
precision
of this vWfR:Co
assay is low, especially in the
lower vWf R:Co range.
Analysis
of patients’
correlation
between
the classic
aggregometer
mercial
samples
the
four
revealed
a comparable
commercial
method.
Also,
the
assays
and
four
com-
gave significantly
lower vWf R:Co reclassic aggregometer
method. However,
comparison
of the slide test and aggregometer
method
of the vWf R:Co reagents showed only small differences
in the recorded vWf R:Co activities. We are not able to
sults
assays
than
explain
the
tion method
the
discrepancy
and the
between
other vWf
the classic
aggregaR:Co assays,
as all
methods
but one (the Behring
slide method) were
calibrated against the same normal plasma. The clinical importance
of this phenomenon
is limited if the
reference values used are based on the vWfR:Co assay
applied.
Compared with the classic aggregometer
method, all
four commercial vWf R:Co assays require less time, as
the preparation
of a fresh platelet
suspension
is not
necessary.
Among the commercial
vWf R:Co assays the
slide tests can be performed the most rapidly. However,
the simplicity
of the procedures
of the two slide tests
does not imply that they are easy to perform. To obtain
reproducible
results,
the visual monitoring
of the agglutination
requires
an experienced
laboratory
technician.
The number of vWf R:Co determinations
that can be
performed per aliquot of reconstituted
platelets is difficult to compare, as it is dependent
on the vWf R:Co
content of the samples and the number of patients’
samples that are analyzed
at one time. However, both
aggregometer
methods
require substantially
more reagent than the slide tests.
514
CLINICAL CHEMISTRY, Vol. 41, No. 4, 1995
In conclusion,
our results
show that both slide methods merit use as tests for the rapid determination
of
vWf R:Co in plasma.
Since
no special
apparatus
is
needed
to perform
a slide test, the determination
of
vWf R:Co may become
feasible
for many laboratories.
References
1. Ruggeri ZM, Ware J. von Wiflebrand factor. FASEB J 1993;7:
308-16.
2. Triplett AT. Laboratory diagnosis of von Willebrand’s disease.
Mayo Clin Proc 1991;66:832-40.
3. Weiss HJ, Hoyer LW, Rickles FR, Varma A, Rogers J. Quasititative assay of a plasma factor deficient in von Wilebrand’s
disease that is necessary for platelet aggregation. J Clin Invest
1973;52:2708-17.
4. MacFarlane DE, Stibbe J, Kirby EP, Zucker MB, Grant HA,
McPherson
J. A method for assaying von Willebrand’s
factor
(ristocetin cofactor). Thromb Diath Haemorrh
1975;34:306-8.
5. Allain JP, Cooper HA, Wagner RH, Brinkhous KM. Platelets
fixed with paraformaldehyde:
a new reagent for assay of von
Willebrand factor and platelet aggregation factor. J Lab Clin Med
1975;85:318-28.
6. Brinkhous
KM, Read MS. Fixation platelets
and platelet
agglutination/aggregation
tests. Methods Enzymol, Part A 1989;
169:149-63.
7. Brinkhous KM, Read MS. Preservation
of platelet receptors
for platelet aggregating factor/von Willebrand factor by air drying, freezing, or lyophilization:
new stable platelet preparations
for von Willebrand factor assays. Thromb Res 1978;13:591-7.
8. Passing H, Bablok W. A new biometrical procedure for testing
the equality of measurements from two different analytical methods. Application of linear regression procedures for method comparison studied in clinical chemistry, part I. J Clin Chem Clin
Biochem 1983;21:709-20.
9. Ruggeri ZM. Structure and function of von Willebrand factor:
relationship
to von Willebrand’s
disease. Mayo Clin Proc 1991;
66:847-61.
10. Brinkhous
1CM, Read MS. Use of venom coagglutinin
and
lyophylized platelets in testing for platelet-aggregating
von Wilebrand factor. Blood 1980;55:517-20.
11. Wright RD, Krauss JS. A comparison
of two macroscopic
platelet agglutination
assays for von Willebrand factor. Ann Clin
Lab Sci 1990;20:73-8.
12. Casonato A, Fabris F, Vicariotto M, Girolanii A. The evaluation of factor VIII antigen by means of a simple slide test. Am J
Clin Pathol 1985;84:107-11.