Clinical Chemistry 48:6
891– 899 (2002)
Evaluation of the Anticoagulants EDTA and
Citrate, Theophylline, Adenosine, and
Dipyridamole (CTAD) for Assessing Platelet
Activation on the ADVIA 120 Hematology System
Marion Macey,1* Urooj Azam,1 Desmond McCarthy,2 Lee Webb,1
E. Sabrinah Chapman,3 David Okrongly,3 David Zelmanovic,3 and Adrian Newland1
Background: Monitoring of platelet activation by the
ADVIA® 120 Hematology System requires an anticoag-
ulant and protocol that ensures that platelets are
sphered and their activation status is not altered artifac-
tually in vitro.
Methods: Blood from healthy controls was collected
into tripotassium EDTA; citrate, theophylline, adeno-
sine, and dipyridamole (CTAD); or a combination of
both (E/C) and stored at ambient temperature or at 4 °C
(E/C only) and then analyzed between 0 and 180 min
later on the ADVIA 120. In addition, immunofluores-
cent flow cytometry was used to identify activated
platelets and platelet-leukocyte aggregates.
Results: In blood stored with all three anticoagulants,
the platelet count changed little, but the mean platelet
volume (MPV) at first decreased and then increased,
whereas the mean platelet component (MPC; an indica-
tor of activation) changed in a reciprocal manner. The
changes in MPV and MPC, which reflect platelet spher-
ing and swelling, were greatest between 30 and 60 min
in blood stored at ambient temperature, irrespective of
which anticoagulant was used, and between 60 and 180
min when blood anticoagulated with E/C was stored at
4 °C. In all anticoagulants, the percentages of platelets
expressing CD62P and of leukocytes in platelet-leuko-
cyte aggregates increased significantly (P <0.01) over
1
Department of Haematology, The Royal London Hospital, Whitechapel,
London E1 1BB, United Kingdom.
2
School of Biological Sciences, Queen Mary (University of London),
London E1 4NS, United Kingdom.
3 4
Laboratory Testing Segment, Diagnostics Division, Bayer Corporation,
Tarrytown, NY 10591.
*Author for correspondence. Fax 44-0207-377-7016; e-mail M.G.Macey@
mds.qmw.ac.uk.
Received January 21, 2002; accepted March 22, 2002.
891
Hematology
180 min at ambient temperature. Only minimal (<2%)
increases occurred when blood with E/C was stored at
4 °C.
Conclusions: When determining platelet activation ex
vivo on the ADVIA 120, blood should be collected
into E/C, stored at 4 °C, and analyzed between 60
and 180 min later; these conditions ensure maximum
platelet sphering without concurrent artifactual platelet
activation.
© 2002 American Association for Clinical Chemistry
It has been suggested by multiple studies [for original
references see Refs. (1, 2 )] that the measurement of indi-
cators of platelet activation during routine hematologic
investigations might offer advantages in the clinical eval-
uation and management of patients at risk from throm-
botic and other diseases. The ADVIA® 120 Hematology
System, recently introduced by the Bayer Corporation, is
an automated analyzer that in addition to measuring the
conventional hematologic indices, also provides some
activation-related information about platelets. It measures
the intensity of light scattered by platelets at two different
cone angles (2–3° and 5–15°) and from the paired values
computes platelet volume (PV)4 and platelet component
(PC) concentration on a cell by cell basis. The latter values
are then averaged to provide the mean platelet compo-
nent (MPC) concentration (expressed in g/dL), a measure
of platelet density that is correlated with the platelet
activation state (1 ). Mean platelet mass (in pg) is com-
Nonstandard abbreviations: PV, platelet volume; PC, platelet component;
MPC, mean PC; MPV, mean PV; ACD, acid-citrate-dextrose; CTAD, citrate,
theophylline, adenosine, and dipyridamole; E/C, tripotassium EDTA plus
CTAD; TS; Tyrode’s salt solution; FITC, fluorescein isothiocyanate; and PE,
phycoerythrin.
892 Macey et al.: Assessing Platelet Activation on the ADVIA 120
puted from the mean PV (MPV) and MPC. Data for
individual platelets are presented in cytograms, whereas
the mean value for each index is tabulated (3 ). Several
studies in which the ADVIA 120 has been used to monitor
platelet activation have already been published, but little
information is available at present concerning the effects
of preanalytical variables and, in particular, the correct
choice of anticoagulant and optimum conditions for sam-
ple storage. For reasons of simplicity and economy, the
ideal anticoagulant would be one that enabled informa-
tion on platelet activation to be obtained as part of the full
blood profile, but this might not be practicable.
EDTA is the anticoagulant recommended for full blood
cell counts and white blood cell differential analysis by
the NCCLS (4 ), principally for its cell preservation prop-
erties. However, an attempt to select and standardize on a
particular salt of EDTA (5 ) has been abandoned and
replaced by the H1-A standard. The International Council
for Standardization in Hematology currently recom-
mends the dipotassium salt of EDTA as the anticoagulant
for full blood counts (6, 7 ). In Europe and Japan, it is the
preferred anticoagulant for this purpose, whereas in the
US and the United Kingdom the tripotassium salt of
EDTA is more commonly used (8 ). Under optimal condi-
tions (appropriate anticoagulant concentration and anal-
ysis within 1– 4 h after phlebotomy), the choice of dipo-
tassium EDTA or tripotassium EDTA makes little
difference to the results of full blood cell counts and white
blood cell differential analyses (9 ). Heparin is not used
because it is considered too expensive, it activates plate-
lets (10 ), and it affects the staining properties of cells.
Citrate is used as an anticoagulant primarily for coagula-
tion studies. However, a pilot study using the CELL-DYN
4000 hematology system (Abbott) indicated that it can be
used instead of EDTA for routine full blood cell counts,
provided that corrections are made to take account of the
different dilution factor. Furthermore, peripheral blood
smears stained with Wright’s Giemsa prepared from
citrated blood are indistinguishable from those prepared
from EDTA-anticoagulated blood, thus allowing blood
collected into just one tube to be used for both hemato-
logic and coagulation analyses (11 ).
Two main requirements must be met when monitoring
platelet activation ex vivo: (a) a venipuncture procedure
must be used that minimizes spontaneous platelet activa-
tion and (b) blood must be collected into a medium that
will not only prevent coagulation, but will also preserve
the activation status of platelets until the samples can be
analyzed (12–15 ). None of the above-mentioned com-
monly used anticoagulants is able to prevent platelet
activation, and the extent to which preanalytical activa-
tion occurs is markedly dependent on the anticoagulant
into which blood is collected (10 ). If blood is collected into
EDTA, platelets quickly change shape from discs with a
2– 4 ␮m diameter and a thickness of 0.5 ␮m to spheres
covered by long thin filopodia (8 ). The sphering of
platelets in EDTA is initially isovolumetric (16, 17 ), but
almost immediately, their apparent size changes over
⬃1–2 h until a state of semiequilibrium is reached
(8, 17, 18 ). During this time, there may be apparent in-
creases (⬎20%) or decreases (⬃10%) in the MPV, depend-
ing on whether measurements are made by impedance or
light scattering, respectively [see Refs. (19, 20 ) for original
references]. This has led to the suggestion that EDTA is
not a suitable anticoagulant for PV analysis (21–25 ).
Furthermore, if blood from certain individuals is antico-
agulated with EDTA, the platelets aggregate, causing an
apparent thrombocytopenia to be recorded (8, 26 ) that in
some, but not all, cases may result from the presence of
agglutinating anti-platelet antibodies (24 ). EDTA was
used as an anticoagulant when a prototype of the ADVIA
120 was being evaluated (3 ) and has also been used more
recently in several studies of platelet status (1, 27, 28 ).
Provided that the effects of EDTA on platelets are under-
stood, the data generated could still be clinically useful
(20 ).
For many years, citrate was the anticoagulant preferred
by most investigators undertaking platelet studies [see
Ref. (2 ) for original references], mainly because sodium
citrate causes less spontaneous activation of platelets in
vitro than does EDTA (10 ). When blood is collected into
citrate, there is initially little or no change in platelet shape
and volume. However, in citrate, platelets slowly adopt a
spherical shape (1 ) and, as in EDTA, swell progressively
over a period of 1–2 h [3–10% increase in volume by
impedance procedures, depending on the concentration
of the sodium citrate used (17, 18, 20, 29 )]. For these
reasons, citrate was originally considered unreliable for
the measurement of platelet volume (29 ), but acid-citrate-
dextrose (ACD) (25 ) and a combination of ACD and
sodium EDTA (29 ) have been recommended instead
because the MPV values obtained are stable over time.
Citrate-based anticoagulants have been used for the de-
termination of platelet indices in the ADVIA 120 (1, 3 ),
suggesting that platelet sphering may not be essential for
the analysis. However, the standard deviation of the MPC
(recorded as platelet component distribution width) is
initially greater in citrate than in EDTA because the light
scattering characteristics of disc-shaped platelets, unlike
those of spheres, are dependent on their orientation (3 ).
Specific inhibitors of platelet function have been added
to anticoagulants in attempts to minimize preanalytical
activation in vitro. An example of this strategy, which is
available commercially in Diatube-H Vacutainers (BD
Biosciences), comprises citrate, theophylline, adenosine,
and dipyridamole (commonly known as CTAD) (30 ).
Theophylline and dipyridamole inhibit cAMP phospho-
diesterase activity (31 ), and adenosine stimulates mem-
brane adenylyl cyclase (32 ). The consequent increase in
platelet cAMP and the inhibition of Ca2⫹-mediated re-
sponses lead to a reduction in platelet activation (33, 34 ).
Dipyridamole has the disadvantage of being light sensi-
tive, and CTAD anticoagulant tubes should be stored
appropriately. Flow cytometric studies of platelet surface
 Clinical Chemistry 48, No. 6, 2002
antigens have shown that whole-blood samples anticoag-
ulated with CTAD are stable at 20 °C under standard
laboratory conditions for up to 4 h after venesection
(2, 35 ).
Although platelets are neither perfectly sphered nor
homogeneous in EDTA, the platelet sphering that occurs
in EDTA is considered advantageous for obtaining accu-
rate PV and PC (and hence MPV and MPC) values in the
ADVIA 120 because the computations are based on the
assumption that scattering is by spherical particles (3 ).
However, the platelet-activating property of EDTA makes
its use as an anticoagulant for measuring patient platelet
activation status questionable. CTAD was recently used
with the ADVIA 120 system in a study (36 ) in which it
was shown that patients with chest pain had lower MPC
values than controls, but being citrate-based, CTAD is
unlikely to sphere platelets. We therefore investigated, in
a comparative manner, the consequences of using CTAD,
EDTA, and the two anticoagulants together (E/C) for
assessing platelet activation ex vivo. Samples were ana-
lyzed in the ADVIA 120 and by immunofluorescent flow
cytometry for up to 180 min postvenesection to assess the
relative stability of the platelets in each anticoagulant. The
first aim of the study was to determine which anticoagu-
lant would be optimal in minimizing platelet activation,
whereas the second was to assess whether the optimal
anticoagulant for platelet studies would also be suitable
for making full blood cell counts and white blood cell
differentials. We present data showing that the two anti-
coagulants together (E/C; UK patent application number
0027309.4) ensure platelet sphering with minimal activa-
tion and that blood in the combined anticoagulant is even
more stable over time when stored at 4 °C rather than at
ambient temperature.
Materials and Methods
materials
Tyrode’s salt solution (TS; 0.265 g/L CaCl2 䡠 2 H2O, 0.214
g/L MgCl2 䡠 6 H2O, 0.2 g/L KCl, 1.0 g/L NaHCO3, 8.0
g/L NaCl, 0.05 g/L Na2HPO4, 1.0 g/L glucose) and
human thrombin (10 units) were from Sigma. Tripotas-
sium EDTA and CTAD in Vacutainers were from BD
Biosciences; the latter were stored in light-protective
boxes and removed just before use.
antisera
Fluorescein isothiocyanate (FITC)-conjugated mouse
IgG1, FITC-CD62P, and phycoerythrin (PE)-conjugated
CD45 were from Immunotech. Mouse FITC-IgG2a and
FITC-CD42a were from BD Biosciences.
blood samples
Blood was collected from the antecubital vein of seven
healthy individuals (median age, 35 years) who had not
taken any medication, including aspirin or aspirin-con-
taining products, in the previous 48 h and who had
previously given informed consent. All procedures were
893
in accordance with the current (October 2000) revision of
the Helsinki Declaration.
assessment of platelet activation on the
advia 120
Whole-blood samples were collected into Vacutainers that
contained tripotassium EDTA, CTAD, or E/C. For the
last, blood was collected first into tripotassium EDTA and
then immediately transferred to a Vacutainer containing
CTAD. Samples were stored at ambient temperature and
analyzed immediately and at 30, 60, 120, and 180 min after
venesection. The platelet count, MPV, and MPC were
determined using the ADVIA 120 hematology system
(Bayer Corporation). Platelet counts made in blood anti-
coagulated with CTAD and E/C were corrected for dilu-
tion (dilution factors were 1.11 and 1.125, respectively).
The system was calibrated and standardized before use
with ADVIA SETpoint Hematology Control and ADVIA
OPTIpoint, respectively (Bayer). In one series of experi-
ments (n ⫽ 4), these analyses were performed on blood
samples anticoagulated with E/C but stored at 4 °C to
investigate the effect of cooling on platelet activation.
measurement of percentages of platelets
expressing cd62p and of leukocytes with
platelets attached (platelet-leukocyte
aggregates)
Anticoagulated blood (5 ␮L) was labeled at ambient
temperature with FITC-isotype control (5 ␮L); FITC-
CD62P (5 ␮L); PE-CD45 (5 ␮L) and FITC-isotype control
(5 ␮L); or PE-CD45 (5 ␮L) and FITC-CD42a (5 ␮L) in 90 ␮L
of TS for 5 min. Previous studies have shown that
antibody binding is complete within this time (37 ). Sam-
ples were diluted to 1 mL with TS and analyzed imme-
diately by flow cytometry.
flow cytometry
Blood cells were analyzed on a FACScan (BD Biosciences)
equipped with CellQuest® software. The flow cytometer
was calibrated and standardized before use with fluoro-
chrome-labeled beads (Fluorospheres; Dako).
For the analysis of CD62P expression, data were ac-
quired in real time with a primary gate set on a dual-
parameter histogram of forward light scatter logarithmic
scale (abscissa) and side light scatter logarithmic scale
(ordinate). This facilitated identification of the platelets
within the blood and was confirmed by the analysis of
CD42a expression. Background fluorescence was assessed
with platelets labeled with the FITC-conjugated isotype
control antibody. Cursors were set in a single-parameter
dot plot of frequency (ordinate) and green fluorescence
intensity (abscissa) so that ⬍1% of the platelets stained
positively with the control antibody. Changes in CD62P
expression (green fluorescence logarithmic scale), to-
gether with those of forward and side light scatter, were
then recorded on the gated platelets.
For the analysis of platelet-leukocyte aggregates, cells
894 Macey et al.: Assessing Platelet Activation on the ADVIA 120

were analyzed first in histograms (Fig. 1, A and D) of side
light scatter (logarithmic scale; ordinate) and orange flu-
orescence (logarithmic scale; abscissa). Leukocytes identi-
fied by their positive staining with PE CD45 were gated
(R1) to dot plots (Fig. 1, B and E) of green fluorescence
(logarithmic scale; abscissa) and orange fluorescence (log-
arithmic scale; ordinate). Events that were both green and
orange (R2) were considered platelet-leukocyte aggre-
gates and recorded as a percentage of a total of 10 000
gated leukocytes. Platelet-leukocyte aggregates could
then be gated to histograms (Fig. 1, C and F) of side light
scatter (logarithmic scale; ordinate) and orange fluores-
cence (logarithmic scale; abscissa) to identify, by their
characteristic side light scatter, which leukocytes were
forming platelet-leukocyte aggregates.
by titration) of human thrombin. A concentration of
thrombin was chosen that would stimulate low-level
platelet activation as determined by CD62P expression.
Thrombin (15 ␮L) was added to blood (210 ␮L) that had
been anticoagulated with tripotassium EDTA to give a
final concentration of 1.2 units/L and incubated at ambi-
ent temperature with FITC-CD62P (25 ␮L). After 10 min,
a sample (40 ␮L) was removed, added to CTAD (5 ␮L),
and incubated for an additional 20 min. Control blood
samples anticoagulated with tripotassium EDTA or
CTAD to which no thrombin had been added were also
incubated with FITC-CD62P for 30 min. Aliquots (5 ␮L) of
blood were removed from each reaction tube at 0, 10, 20,
40, and 60 min; diluted with TS (995 ␮L); and analyzed
immediately by flow cytometry.

investigation of inhibitory effect of ctad on statistical analysis

thrombin-activated platelets
In one series of experiments (n ⫽ 3), the effect of CTAD on
platelet activation was investigated in blood anticoagu-
lated with tripotassium EDTA that had been incubated
with a suboptimal concentration (determined previously
Results from the flow cytometer and the ADVIA 120
Hematology System were compared using the paired
t-test to test for significant differences between the same
sample analyzed at different times. To take into account
multiple comparisons, P ⬍0.01 was considered signifi-

Fig. 1. Analysis of platelet-leukocyte aggregates in whole blood.

Blood was stained with PE-conjugated CD45 and FITC-conjugated CD42a. Leukocytes were identified (region R1) by their positive staining with PE-CD45 in a plot of side
scatter (logarithmic scale; ordinate) vs orange fluorescence (logarithmic scale; abscissa; dot plots A and D) and were displayed in a plot of green fluorescence
(logarithmic scale; abscissa) and orange fluorescence (logarithmic scale; ordinate; dot plots B and E). Events that were both green (CD42a) and orange (CD45; region
R2) were considered platelet-leukocyte aggregates. Back-gating of these events to a plot of side scatter (logarithmic scale; ordinate) vs orange fluorescence (logarithmic
scale; abscissa; dot plots C and F) showed that the majority of aggregates had the side light scatter characteristics of granulocytes and monocytes. An example of the
analysis performed on blood from a healthy control is illustrated in plots A–C and that from a patient with inflammatory bowel disease in plots D–F.
 Clinical Chemistry 48, No. 6, 2002 895

cant. Results were also compared by ANOVA to test for

significant differences between means, and the post hoc
Scheffé test was applied for multiple comparisons.

Results
platelet count
Platelet counts in all three anticoagulants immediately
after venesection did not differ significantly and were
within the reference interval. No significant change in
platelet count occurred over 180 min in blood stored with
the different anticoagulants at ambient temperature (Ta-
ble 1) or when blood anticoagulated with E/C was stored
at 4 °C.

mpv
MPV values decreased initially in all anticoagulants and
then increased again. When blood was stored at ambient
Fig. 2. Changes in MPV with time in blood anticoagulated with
temperature, the nadir was at 30 min in all anticoagulants,
tripotassium EDTA (⽧), CTAD (Œ), and E/C (f) at ambient temperature
but when blood anticoagulated with E/C was stored at and with E/C at 4 °C (䡺).
4 °C, MPV values remained low between 60 and 180 min. During storage at ambient temperature, the MPV values obtained at all times
When blood was stored at ambient temperature, the MPV were significantly lower (P ⬍0.04) in blood that had been anticoagulated with
tripotassium EDTA than in blood anticoagulated with CTAD or E/C. In blood
values obtained at all times were significantly lower (P anticoagulated with E/C, the MPV values after storage at 4 °C and at ambient
⬍0.04) when it had been anticoagulated with tripotassium temperature differed significantly (P ⬍0.01) at all times between 30 and 180
EDTA than when it had been anticoagulated with CTAD min inclusive. Symbols represent the mean values of four (E/C at 4 °C) or seven
(CTAD, E/C, and tripotassium EDTA) experiments. Bars, SE.
or E/C. Moreover, in blood anticoagulated with E/C, the
MPV values after storage at 4 °C and at ambient temper-
ature differed significantly (P ⬍0.01) at all times between expression of cd62p on platelets
30 and 180 min inclusive (Fig. 2). Only low percentages of platelets expressed CD62P
shortly after venesection (mean ⫾ SE, 1.10% ⫾ 0.61%,
mpc 1.22% ⫾ 0.62%, and 1.28% ⫾ 0.85% in tripotassium EDTA,
In direct contrast to the results for MPV values (above), CTAD, and E/C, respectively), but the percentages in-
the MPC values increased initially and then remained creased when blood was stored at ambient temperature.
approximately constant or decreased. Maximal values Values at 180 min were significantly higher in blood
were reached at 30 min in CTAD and E/C and between 30
and 60 min in tripotassium EDTA when blood was stored
at ambient temperature and at 60 to 180 min when it was
stored in E/C at 4 °C. MPC values were significantly
higher at all times in blood stored at ambient temperature
with tripotassium EDTA than with CTAD (P ⬍0.04) or
E/C (P ⬍0.02). In blood stored with CTAD or tripotas-
sium EDTA at ambient temperature, the MPC values at 30
and 180 min differed significantly (P ⬍0.01). In blood
anticoagulated with E/C and stored at 4 °C or at ambient
temperature, the values differed significantly (P ⬍0.01) at
all times between 30 and 180 min inclusive (Fig. 3).

Table 1. PCs in blood stored with different anticoagulants.a

Anticoagulant Fig. 3. Changes in MPC with time in blood anticoagulated with
tripotassium EDTA (⽧), CTAD (Œ), and E/C (f) at ambient temperature
Time, min K3EDTAb CTAD E/C E/C at 4 °C and with E/C at 4 °C (䡺).
0 235 ⫾ 13 229 ⫾ 14 230 ⫾ 14 231 ⫾ 14 MPC values were significantly higher at all times in blood stored at ambient
30 246 ⫾ 14 229 ⫾ 14 238 ⫾ 14 233 ⫾ 19 temperature with tripotassium EDTA than with CTAD (P ⬍0.04) or E/C (P ⬍0.02).
180 246 ⫾ 5 224 ⫾ 7 225 ⫾ 10 237 ⫾ 18 In blood stored with CTAD or tripotassium EDTA at ambient temperature, the
MPC values at 30 and 180 min differed significantly (P ⬍0.01), whereas in blood
a
Mean (⫾SE) of seven experiments at ambient temperature except where anticoagulated with E/C and stored at 4 °C or at ambient temperature, the
otherwise indicated. values differed significantly (P ⬍0.01) at all times between 30 and 180 min
b
Tripotassium EDTA. inclusive. Symbols represent the mean values of four (E/C at 4 °C) or seven
(CTAD, E/C, and tripotassium EDTA) experiments. Bars, SE.
896 Macey et al.: Assessing Platelet Activation on the ADVIA 120

anticoagulated with tripotassium EDTA (23.05% ⫾ 1.54%)

than with E/C (8.45% ⫾ 0.79%; P ⬍0.01), and were lowest
in blood anticoagulated with CTAD (4.14% ⫾ 0.79%; P
⬍0.01). When blood samples anticoagulated with E/C
were stored at 4 °C, there were only minimal increases in
the number of CD62P-positive platelets, from 0.53% ⫾
0.12% at 0 min to 1.07% ⫾ 0.57% at 180 min (Fig. 4).

platelet-leukocyte aggregate formation

Immediately after venesection, a small percentage of
leukocytes that were associated with platelets could be
found in blood from all donors irrespective of the antico-
agulant into which it had been collected (3.50% ⫾ 0.71%,
3.95% ⫾ 0.97% and 2.82% ⫾ 1.05% in tripotassium EDTA,
CTAD, and E/C, respectively). In all anticoagulants, the
percentage of platelet-leukocyte aggregates increased
markedly when blood was stored at ambient temperature.
The increases at 180 min were greater in blood anticoag-
ulated with CTAD (18.88% ⫾ 2.06%) than with tripotas-
sium EDTA (13.50% ⫾ 1.74%) and were lowest in blood
that had been anticoagulated with E/C (7.81% ⫾ 1.43%).
However, when blood samples anticoagulated with E/C
were incubated at 4 °C, there were only minimal increases
in the percentage of platelet-leukocyte aggregates over
180 min (Fig. 5).
effect of ctad on thrombin-activated platelets
To ascertain whether CTAD could effectively inhibit
further responses by platelets that had already encoun-
tered an agonist, blood that had been collected into
tripotassium EDTA was incubated alone or with a subop-
timal concentration of thrombin. After 10 min, an aliquot
of the thrombin-stimulated blood was added to CTAD.
Platelet activation, as monitored by CD62P expression,
Fig. 5. Changes in the percentage of leukocytes forming platelet-
leukocyte aggregates (monitored by CD45/CD42a expression) with
time in blood anticoagulated with tripotassium EDTA (⽧), CTAD (Œ),
and E/C (f) at ambient temperature and with E/C at 4 °C (䡺).
Symbols represent the mean values of four (E/C at 4 °C) or seven (CTAD, E/C,
and tripotassium EDTA at ambient temperature) experiments; bars, SE. There
were significant differences (P ⬍0.01) between the percentages of leukocytes
forming platelet-leukocyte aggregates at 0 and 180 min in blood stored with all
anticoagulants except E/C at 4 °C.
was completely inhibited by the addition of CTAD,
whereas progressive activation occurred in blood that had
been anticoagulated only with tripotassium EDTA and, as
expected, was greater in these samples when thrombin
had been added than when it had been omitted (Fig. 6).
stability of edta and ctad when mixed before
use
Because the above-mentioned results suggested that E/C
might be a better anticoagulant for platelet studies than
either tripotassium EDTA or CTAD alone, the effect of

Fig. 4. Changes in the percentage of CD62P-positive platelets with

time in blood anticoagulated with tripotassium EDTA (⽧), CTAD (Œ),
and E/C (f) at ambient temperature and with E/C at 4 °C (䡺).
Symbols represent the mean values of four (E/C at 4 °C) or seven (CTAD, E/C,
and tripotassium EDTA at ambient temperature) experiments; bars, SE. There
were significant differences (P ⬍0.001) between the percentages of CD62P-
positive platelets at 0 and 180 min in samples stored with all anticoagulants
except E/C at 4 °C.
Fig. 6. Changes in the percentage of CD62P-positive platelets with
time in blood anticoagulated with tripotassium EDTA and stimulated
with (〫) and without thrombin (⽧), blood anticoagulated with tripotas-
sium EDTA and stimulated with thrombin for 10 min when CTAD was
added (f), and blood anticoagulated with CTAD without thrombin (Œ).
Symbols represent the mean values of three experiments; bars, SE.
 Clinical Chemistry 48, No. 6, 2002 897

premixing the two components was investigated. Blood
was collected into mixtures of tripotassium EDTA and
CTAD that had been prepared either 14 days or immedi-
ately before use and into tripotassium EDTA that was
then mixed with CTAD (as had been done previously
throughout the study). All samples were subsequently
stored at 4 °C. Values for routine hematologic and platelet
activation indices measured on the ADVIA 120 soon after
venesection were similar irrespective of whether the two
anticoagulants were mixed before or after blood collec-
tion. Moreover, the values remained essentially un-
changed when also analyzed at 3, 6, and 24 h. Immuno-
fluorescence assays showed that the percentage of
leukocytes involved in aggregates with platelets increased
slightly (⬍5% increases at 24 h) in all samples over 24 h
(results not shown).
advia 120 hematology results in blood
anticoagulated with edta, ctad, or e/c
To investigate whether blood anticoagulated with E/C
could be used for routine analysis of hematologic indices,
the results obtained on the ADVIA 120 from seven control
samples anticoagulated with tripotassium EDTA, CTAD,
or E/C were compared. There were no significant differ-
ences in the values obtained for the white blood cell
counts, the red cell counts, and the hemoglobin concen-
tration in blood samples anticoagulated with the three
anticoagulants. However, the hematocrit and several
platelet indices were significantly different (P ⬍0.01) in
blood anticoagulated with CTAD and E/C compared
with that anticoagulated with tripotassium EDTA (Table
2).
Discussion
The data presented here clearly show that in blood stored
with EDTA, CTAD, or E/C at ambient temperature, the
MPV decreased between 0 and 30 min but then increased
again. The results for EDTA are largely in agreement with
those of previous studies in which MPV was determined
by optical procedures [e.g., Ref. (23 )] and have been
interpreted as representing the initial sphering and sub-
sequent swelling of platelets (19, 20 ).
Consistent with recent reports, CTAD largely inhibited
the increases in CD62P expression that occurred on plate-
lets stored for 180 min at ambient temperature in blood
anticoagulated with tripotassium EDTA (1, 2, 35 ). Fur-
thermore, when blood that had been anticoagulated with
tripotassium EDTA was stimulated with thrombin, the
subsequent addition of CTAD inhibited platelet degran-
ulation and the increases in CD62P expression that other-
wise occurred in blood stored with just tripotassium
EDTA. We have previously demonstrated that after in
vitro stimulation of tripotassium EDTA-anticoagulated
whole blood, increases in CD62P expression are accom-

Table 2. Comparison of ADVIA 120 hematology resultsa in blood anticoagulated with tripotassium EDTA, CTAD, and
E/C measured at 30 min after venesection.
Index K3EDTAb CTADc E/Cd E/C at 4 °Cd
9
PLTs, 10 /L 246 ⫾ 14 229 ⫾ 14 238 ⫾ 14 242 ⫾ 14
MPV, fL 7.7 ⫾ 0.27 8.4 ⫾ 0.24e 8.6 ⫾ 0.32e 7.6 ⫾ 0.27
PDW, % 58.9 ⫾ 3.24 53.8 ⫾ 2.40e 55.6 ⫾ 2.24 62.6 ⫾ 3.24
PCT, % 0.19 ⫾ 0.01 0.17 ⫾ 0.01e 0.18 ⫾ 0.01 0.20 ⫾ 0.14
MPC, g/dL 28.6 ⫾ 0.52 27.1 ⫾ 0.40e 26.6 ⫾ 0.66e 26.2 ⫾ 0.46e
PCDW, g/dL 5.00 ⫾ 0.09 7.2 ⫾ 0.13e 6.9 ⫾ 0.38e 4.8 ⫾ 0.13
WBCs, 109/L 5.35 ⫾ 0.40 5.27 ⫾ 0.42 5.36 ⫾ 0.40 5.90 ⫾ 0.46
Neutrophils, 109/L 3.13 ⫾ 0.24 3.07 ⫾ 0.24 3.10 ⫾ 0.22 3.20 ⫾ 0.12
Lymphocytes, 109/L 1.61 ⫾ 0.20 1.65 ⫾ 0.23 1.69 ⫾ 0.24 1.66 ⫾ 0.14
Monocytes, 109/L 0.32 ⫾ 0.04 0.28 ⫾ 0.03 0.30 ⫾ 0.04 0.28 ⫾ 0.01
Eosinophils, 109/L 0.11 ⫾ 0.04 0.12 ⫾ 0.04 0.11 ⫾ 0.03 0.13 ⫾ 0.03
Basophils, 109/L 0.06 ⫾ 0.01 0.03 ⫾ 0.01 0.04 ⫾ 0.01 0.04 ⫾ 0.01
RBCs, 1012/L 4.77 ⫾ 0.20 4.82 ⫾ 0.21 4.87 ⫾ 0.21 4.71 ⫾ 0.16
HGB, g/dL 14.7 ⫾ 0.40 14.8 ⫾ 0.40 15.1 ⫾ 0.44 14.2 ⫾ 0.37
HCT, L/L 0.421 ⫾ 0.01 0.434 ⫾ 0.01e 0.438 ⫾ 0.01e 0.426 ⫾ 0.01
MCV, fL 88.7 ⫾ 1.82 90.5 ⫾ 1.89e 90.4 ⫾ 1.93e 90.9 ⫾ 1.82e
MCH, pg 30.9 ⫾ 0.56 30.9 ⫾ 0.63 31.1 ⫾ 0.56e 30.3 ⫾ 0.66
MCHC, g/dL 34.9 ⫾ 0.20 34.2 ⫾ 0.20e 34.4 ⫾ 0.23e 33.3 ⫾ 0.27e
RDW, % 12.5 ⫾ 0.10 12.6 ⫾ 0.11 12.6 ⫾ 0.12 12.9 ⫾ 0.13
a
Mean (⫾SE).
b
K3EDTA, tripotassium EDTA; PLT, platelet; PDW, platelet distribution width; PCT, platelet crit; PCDW, platelet component distribution width; WBC, white blood cell;
RBC, red blood cell; HGB, hemoglobin; HCT, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin
concentration; RDW, red blood cell distribution width.
c
Values corrected for increased dilution (dilution factor, 1.11).
d
Values corrected for increased dilution (dilution factor, 1.125).
e
Significantly different (P ⬍0.01) from result obtained with tripotassium EDTA.
898 Macey et al.: Assessing Platelet Activation on the ADVIA 120
panied by a concurrent decrease in MPC (1 ). We now
show that at ambient temperature, the MPC decreases
spontaneously more in tripotassium EDTA than in CTAD
or in E/C, consistent with the notion that EDTA causes
platelet activation. In blood anticoagulated with CTAD or
E/C and stored at ambient temperature, MPC values
were largely similar over 180 min and were significantly
lower than those in blood stored in tripotassium EDTA.
Taken together, the changes in MPV and MPC values at
ambient temperature suggest that in blood anticoagulated
with E/C, the platelets become sphered by 30 min with-
out degranulating. This was confirmed by the low per-
centage of platelets expressing CD62P in E/C-anticoagu-
lated blood, which was comparable to that found in blood
anticoagulated with CTAD alone.
Blood anticoagulated with CTAD contained higher
numbers of platelet-leukocyte aggregates than did blood
that had been anticoagulated with tripotassium EDTA or
E/C (Fig. 5). The reasons for this are not immediately
apparent, but seem dependent on the presence of EDTA.
Somewhat paradoxically, EDTA has been shown to affect
platelet membrane-bound receptors in a way that en-
hances, rather than diminishes, granule secretion and
aggregation (10 ). However, external Ca2⫹ is required for
aggregation, and because EDTA is a better chelator of
Ca2⫹ than citrate, it may have a greater inhibitory effect in
these respects. If this explanation is true, then it is possible
that CTAD might not have completely inhibited increases
in CD62P expression and that activated platelets were
present but remained uncounted because they were at-
tached to leukocytes. These results suggest that platelet-
leukocyte aggregate formation should be monitored dur-
ing studies of platelet activation in whole blood and also
highlight the fact that the formation of platelet-leukocyte
aggregates ex vivo is anticoagulant dependent. It has
previously been shown that when blood anticoagulated
with either sodium citrate or ACD is cooled to ⱕ20 °C, the
platelets undergo granule secretion and reversibly de-
velop a spherical morphology with pseudopodia (38, 39 ).
This low-temperature-induced activation was shown to
be dependent on release of intracellular free calcium. We
have shown (unpublished data) in blood anticoagulated
with sodium citrate incubated at 4 °C that there is a
time-dependent increase in CD62P expression on platelets
and a significant increase in platelet-leukocyte aggregate
formation. Presumably the inhibitory effects of CTAD on
intracellular calcium mobilization prevent granule release
and subsequent platelet-leukocyte aggregate formation.
The crucial finding of this study is that E/C effectively
spheres platelets in 60 min at 4 °C without simultaneously
causing degranulation, or in 30 min at ambient tempera-
ture with only minimal activation, thereby allowing the
accurate measurement of MPC on the ADVIA 120. More-
over, if blood anticoagulated with E/C is stored at 4 °C,
there are only minimal changes in all indicators of platelet
activation for at least 180 min. This finding will be
important for clinical studies because 30 min to 3 h is
usually sufficient time for a sample collected in a ward or
clinic to reach the laboratory for analysis. In fact, prelim-
inary studies (unpublished results) indicate that the time
period at 4 °C could probably be safely extended to 6 h. It
is known that the inhibitory effects of the platelet antag-
onists in CTAD begin to dissipate at 3– 4 h when blood is
stored at ambient temperature and that, if necessary,
storage times could probably be prolonged by increasing
their concentrations (2 ). Before MPC values can be used
diagnostically, it seems essential to (a) have a set of
reference reagents for calibrating the MPC scale on the
ADVIA 120 so that interlaboratory comparisons can be
made and (b) to determine reference intervals for this
marker in healthy control populations. To this end, refer-
ence reagents are currently being developed and refer-
ence intervals are being established by the Bayer Corpo-
ration. It is also worth noting that the temperature (4 °C or
ambient) at which blood that had been anticoagulated
with E/C was stored markedly affected the MPV and
MPC values that were subsequently obtained. These
effects may be caused by the actin-mediated shape change
(from smooth discs into spiny spheres with irregular
projections) that platelets undergo at temperatures
⬍15 °C (40 ).
The ability of E/C to inhibit platelet-leukocyte aggre-
gate formation ex vivo indicates that this combined anti-
coagulant is also suitable for the investigation of these
interactions in clinical studies. Indeed, with this anticoag-
ulant, significantly greater numbers of platelet-leukocyte
aggregates (5.16 ⫾ 1.48) were found in the blood of
patients (n ⫽ 62) with inflammatory bowel disease than in
healthy controls (n ⫽ 20; mean ⫾ SE, 3.43 ⫾ 0.82; P ⫽ 0.03;
unpublished data, an example of which is illustrated in
Fig. 1). A recent preliminary study (41 ) has shown that
CTAD is suitable for the analysis of routine hematology
indices on the CELL DYN 4000 (Abbott) and SE-9000
(Sysmex Corporation), although the results for platelet
count and MPV were ⬃10% lower than in dipotassium
EDTA irrespective of which analyzer was used. Our
preliminary results (Table 2) suggest that E/C is probably
a satisfactory anticoagulant for routine analysis on the
ADVIA 120, although certain algorithms may need to be
altered. Additional studies are needed to confirm this
point. From a clinical viewpoint, it is not practical to
collect blood into one Vacutainer and then pour it into a
second or to mix the contents of two Vacutainers. For this
reason, we are currently undertaking studies, in associa-
tion with BD Biosciences and the Bayer Corporation, to
develop tubes containing both anticoagulants.
References
1. Macey MG, Carty E, Webb L, Chapman ES, Zelmanovic D, Ok-
rongly, et al. Use of mean platelet component to measure platelet
activation on the ADVIA® 120 Haematology System. Cytometry
1999;38:250 –5.
2. Mody M, Lazarus AH, Semple JW, Freedman J. Preanalytical
 Clinical Chemistry 48, No. 6, 2002
requirements for flow cytometric evaluation of platelet activation:
choice of anticoagulant. Transfus Med 1999;9:147–54.
3. Zelmanovic D, Colella GM, Hetherington, EJ, Chapman ES, Pa-
seltiner L, inventors. Automated method and device for identifying
and quantifying platelets and for determining platelet activation
state using whole blood samples. US Patent 5,817,519, 1998.
4. National Committee for Clinical Laboratory Standards. Evacuated
tubes and additives for blood specimen collection, 4th ed.;
approved standard. NCCLS Document H1-A4. Villanova, PA: NC-
CLS, 1996.
5. National Committee for Clinical Laboratory Standards. Additives to
blood collection devices: EDTA; tentative standard H35-T. Wayne,
PA: NCCLS, 1992.
6. International Council for Standardization in Haematology. Recom-
mendations of the International Council for Standardization in
Haematology for Ethylenediaminetetraacetic Acid Anticoagulation
of Blood for Blood Cell Counting and Sizing. International Council
for Standardization in Haematology: Expert Panel on Cytometry.
Am J Clin Pathol 1993;100:371–2.
7. Lewis SM. Harmonization and standardization of the blood count:
the role of the ICSH cytometry panel. Sysmex J Int 1999;9:1–7.
8. Okada T. Development and problems of automated hematology
analyzers. Sysmex J Int 1999;9:52–7.
9. Goossens W, Van Duppen V, Verwilghen RL. K2- or K3-EDTA: the
anticoagulant of choice in routine haematology. Clin Lab Haematol
1992;13:291–5.
10. Golanski J, Pietrucha T, Baj Z, Greger J, Watala C. Molecular
insights into the anticoagulant-induced spontaneous activation of
platelets in whole blood—various anticoagulants are not equal.
Thromb Res 1996;83:199 –216.
11. Perrotta G, Roberts L, Glazier J, Schumacher HR. Use of sodium
citrate anticoagulant for routine hematology analysis on the
CELL-DYN® 4000: an opportunity to enhance efficiency in the
clinical laboratory. Lab Hematol 1998;4:156 – 62.
12. George JN, Thio LL, Morgan RK. Quantitative analysis of platelet
membrane glycoproteins; effect of platelet washing and isolation
of platelet density subpopulations. Thromb Res 1981;23:69 –77.
13. Hawiger J. Platelet secretory pathways: an overview. Methods
Enzymol 1989;169:191–5.
14. Michelson AD. Flow cytometry: a clinical test of platelet func-
tion—a review. Blood 1996;87:4925–36.
15. Wu KK. Platelet activation and arterial thrombosis. Lancet 1994;
344:991–5.
16. Rowan RM. Platelet size distribution analysis: principles, tech-
niques and potential clinical utility. Hematol Rev 1986;1:109 –
44.
17. Threatte GA, Adrados C, Ebbe S, Breecher G. Mean platelet
volume. The need for a reference method. Am J Clin Pathol
1984;81:769 –72.
18. Bath PWM. The routine measurement of platelet size using
sodium citrate alone as the anticoagulant. Thromb Haemost
1993;70:687–90.
19. Dow RB. The clinical and laboratory utility of platelet volume
parameters. Aust J Med Sci 1994;15:118 –25.
20. Jackson SR, Carter JM. Platelet volume: laboratory measurement
and clinical application. Blood Rev 1993;7:104 –13.
21. Lippi U, Cappelletti P, Schinell M, Signori D. Mean platelet
volumes: facts or artifacts? Am J Clin Pathol 1985;84:111–3.
22. Lippi U, Schinella M, Moena N, Nicoli M. Unpredictable effects of
K3EDTA on mean platelet volume. Am J Clin Pathol 1987;87:391–3.
23. Trowbridge EA, Reardon DM, Hutchinson D, Pickering C. The
routine measurement of platelet volume. A comparison of light-
scattering and aperture-impedance technologies. Clin Phys
Physiol Meas 1985;6:221–38.
24. Wehmeier A, Schneider W. Platelet volume parameters as a
899
diagnostic tool: the influence of anticoagulation and storage
conditions on platelet impedance volume. Klin Wochenschr 1989;
67:980 – 4.
25. O’Malley T, Ludlam CA, Fox KA, Elton RA. Measurement of platelet
volume using a variety of different anticoagulants and antiplatelet
mixtures. Blood Coagul Fibrinolysis 1996;7:431– 6.
26. Berkman N, Michaeli Y, Or R, Eldor A. EDTA-dependent pseudo-
thrombocytopenia: a clinical study of 18 patients and a review of
the literature. Am J Hematol 1991;36:195–201.
27. Brumitt DR, Barker HF. The determination of a reference range for
new platelet parameters produced by the Bayer ADVIA120 full
blood count analyser. Clin Lab Haematol 2000;22:103–7.
28. Stanworth SJ, Denton K, Monteath J, Patton WN. Automated
counting of platelets on the Bayer ADVIA®120 analyser. Clin Lab
Haematol 1999;21:113–7.
29. Thompson CB, Diaz DD, Quinn PG, Lapins M, Kurtz SR, Valeri CR.
The role of anticoagulation in the measurement of platelet vol-
umes. Am J Clin Pathol 1983;l80:327–32.
30. Constant G, Gouault-Heilmann M, Martinoli JL. Heparin inactiva-
tion during blood storage: its prevention by blood collection in
citric acid, theophylline, adenosine, dipyridamole-CTAD mixture.
Thromb Res 1983;31:365–74.
31. Salzman EW, Kensler PC, Levine L. Cyclic 3⬘5⬘-adenosine mono-
phosphate in human blood platelets: regulatory role of cyclic AMP
in platelet function. Ann N Y Acad Sci 1972;201:61–71.
32. Mills DCB, Smith JB. The influence on platelet aggregation of
drugs that affect the accumulation of adenosine 3⬘5⬘-cyclic mono-
phosphate in platelets. Biochem J 1971;121:185–96.
33. Broekman MJ, Biroa AM, Marcus AJ. Inhibition of human platelet
reactivity by endothelium-derived relaxing factor from human um-
bilical vein endothelial cells in suspension: blockade of aggrega-
tion and secretion by an aspirin-insensitive mechanism. Blood
1991;78:1033– 40.
34. Kroll MH, Schafer AI. The analysis of ligand-receptor interaction in
platelet activation. In: Joseph M, ed. Immunopharmacology of
platelets. The handbook of immunopharmacology. Orlando, FL:
Academic Press, 1995:31– 63.
35. Kuhne T, Hornstein A, Semple J, Chang W, Blanchette V, Freed-
man J. Flow cytometric evaluation of platelet activation in blood
collected into EDTA vs. Diatube-H, a sodium citrate solution
supplemented with theophylline, adenosine, and dipyridamole.
Am J Hematol 1995;50:40 –5.
36. Ahnadi CE, Chapman ES, Okrongly D, Hoang T, Lucena-Fernandes
F, Grant AM. Evaluation of platelet activation in patients with chest
pain by the ADVIA®120 hematology system [Abstract]. XIIIth
International Symposium on Technical Innovations in Laboratory
Hematology, Banff, Canada, 2000; 41. http://mmserver.cjp.
com/gems/blood/PlateletActivation.pdf (Accessed July 2001).
37. Carty E, Macey M, Rampton DS. Inhibition of platelet activation by
5-aminosalicylic acid in inflammatory bowel disease. Aliment
Pharmacol Ther 2000;14:1169 –79.
38. Oliver AE, Tablin F, Walker NJ, Crowe JH. The internal calcium
concentration of human platelets increases during chilling. Bio-
chim Biophys Acta 1999;1416:349 – 60.
39. Maurer-Spurej E, Pfieler G, Maurer N, Linder H, Glatter O, Devine
D. Room temperature activates human blood platelets. Lab Invest
2001;81:581–92.
40. Hoffmeister KM, Falet H, Toker A, Barkalow KL, Stossel TP,
Hartwig JH. Mechanisms of cold-induced platelet actin assembly.
J Biol Chem 2001;276:24751–9.
41. Tsuda I, Kastami T, Kondo H, Tatsumi N. CTAD (citric acid,
theophylline, adenosine and dipyrimadole) as an anticoagulant for
hematological analysis [Abstract]. XIIIth International Symposium
on Technical Innovations in Laboratory Hematology, Banff, Can-
ada, 2000:37.