876

Activation of Human Platelets by Cocaine

Aaron D. Kugelmass, MD; Atsushi Oda, MD; Kevin Monahan, MD;
Claudia Cabral, MS; J. Anthony Ware, MD

Background. Cocaine ingestion has been associated with thrombosis of coronary as well as peripheral
arteries, but the mechanism by which cocaine promotes thrombus formation is unknown. Accordingly, we
determined whether cocaine activates human platelets by flow cytometric analysis of whole blood to which
cocaine was added.
Methods and Results. Activated platelets were detected by "two-color" flow cytometric analysis of the
binding of fluorescently labeled antibodies directed against either platelet-associated fibrinogen or
P-selectin, which are found on the surface of platelets only after stimulation. Platelets were distinguished
from other constituents of whole blood by their ability to bind an anti-glycoprotein Ib antibody bound to
both activated and resting platelets. Incubation of whole blood with cocaine, in concentrations of 10 ,uM
to 13 mM, induced significant increases in both platelet-associated fibrinogen (range of increase, 45±12%
to 125±40%o) and P-selectin expression (36±15% to 112±24%). In platelets suspended in either buffer or
plasma, however, P-selectin expression was detected only at the highest cocaine concentration (85±13%
increase in plasma and 59±7% in buffer). Neither aspirin nor the ADP scavenger apyrase inhibited
cocaine-induced P-selectin expression. Cocaine inhibited the uptake of '4C-radiolabeled serotonin by
platelets (IC., 8.7 1tM). P-selectin expression and fibrinogen binding were found after the addition of
cocaine alone to blood taken from some but not all donors; however, platelet activation in response to
submaximal concentrations of the agonists ADP or epinephrine was enhanced by a low concentration of
cocaine added to blood from every donor.
Conclusions. Cocaine, in concentrations similar to those found clinically, induces activation of
individual platelets studied in whole blood from some but not all donors, and platelet response to
physiological agonists is enhanced by cocaine. Thus, cocaine-induced platelet activation may contribute to
Downloaded from http://ahajournals.org by on August 5, 2020

thrombosis following cocaine ingestion. (Circulation. 1993;88:876-883.)

KEY WORDs * cocaine * thrombosis * platelets * P-selectin * fibrinogen

O f the many and varied cardiovascular disorders
that have been associated with cocaine inges-
tion, most attention has been generated by the
cases of myocardial infarction temporally related to
cocaine consumption, because they occur most often in
patients much younger than those in whom coronary
artery disease usually is manifested clinically. More-
over, more than one third of the patients who have
sustained a cocaine-related myocardial infarction and
subsequently have undergone either angiographic or
pathological examination did not have significant coro-
nary artery stenosis.' The pathophysiological basis for
cocaine-induced myocardial infarction is not yet clear;
multiple contributing mechanisms have been postu-
lated, including adrenergic-mediated increases in myo-
cardial oxygen demand, coronary vasospasm, endothe-
lial dysfunction, and accelerated atherosclerosis.12 It
also is possible that cocaine promotes the formation of
coronary thrombi; such a mechanism could account not
only for those infarctions in which a thrombus is dem-
onstrated but also for those patients with angiographi-
Received August 5, 1992; revision accepted May 3, 1993.
From the Charles A. Dana Research Institute and the Harvard-
Thorndike Laboratory of the Departments of Medicine (Cardio-
vascular Division) and Surgery, Beth Israel Hospital and Harvard
Medical School, Boston, Mass.
Correspondence to Dr Ware, Cardiovascular Division, Beth
Israel Hospital, 330 Brookline Ave, Boston, MA 02215.
cally "normal" coronary arteries, as the absence of
discernible thrombus may reflect the endogenous
thrombolytic capacity of an otherwise normal artery,
particularly in younger patients. The possibility that
cocaine exerts a prothrombotic effect is bolstered by
reports of cocaine-induced pulmonary infarction and
thrombosis of peripheral arteries and veins.3-6
An attractive hypothesis to explain the association of
thrombosis with cocaine ingestion is that cocaine either
directly or indirectly induces activation of platelets. The
demonstration of thrombi comprising platelet aggre-
gates in the coronary arteries of patients with fatal
cocaine-associated myocardial infarction7 supports this
theory. Although other local anesthetics, including
lidocaine8 and dibucaine,9"10 have been shown to inhibit
platelet aggregation to standard agonists in vitro, no
effect of cocaine on aggregation of human platelets has
been reported. The purpose of this study, therefore, was
to determine if cocaine activates human platelets. To
test this hypothesis, we used a sensitive assay modified
from that described by Shattil and colleagues" to detect
activation of individual platelets in whole blood, thus
allowing study of platelet activation while preserving
cellular communication with erythrocytes and leuko-
cytes, both of which can affect platelet function.'213 This
flow cytometric assay detects fluorochrome conjugated
antibodies directed against epitopes expressed on plate-
lets only after activation, thus permitting the assessment
 Kugelmass et al Activation of Human Platelets by Cocaine
of activation of individual platelets in either platelet-
rich plasma, gel-filtered platelets, or whole blood. In
this study, platelet activation was determined by the
detection of P-selectin (previously known as GMP-140
or PADGEM14), which is a protein contained in the
platelet a-granule membrane that becomes expressed
on the surface when this membrane fuses with the
platelet surface during the release reaction,15 and asso-
ciation of fibrinogen with the platelet surface, an event
that is an activation-dependent precursor of secondary
platelet aggregation.16
Methods
Preparation of Whole Blood and
Platelet-Rich Plasma
Blood was obtained from healthy, nonfasting volun-
teers who denied cocaine ingestion and who had not
taken aspirin within 2 weeks of donation. Blood was
collected from a clean antecubital vein via a 19-gauge
needle with the use of a light tourniquet. The initial 5
mL of sample was discarded, and the remainder was
anticoagulated with 3.8% trisodium citrate (Sigma
Chemical, St Louis, Mo) (1:9 dilution). Platelet-rich
plasma and platelets washed by gel filtration and sus-
pended in buffer were prepared using standard tech-
niques described previously.17,18
Platelet Stimulation
Aliquots of whole blood, gel-filtered platelets, or
platelet-rich plasma (500 gL) were immediately trans-
ferred to polystyrene tubes. Standard agonists ADP (10
,uM) (Sigma) and phorbol 12,13-dibutyrate (PdBu) (0.2
Downloaded from http://ahajournals.org by on August 5, 2020
,uM) (Sigma) and equal volumes of appropriate solvents
0.9% NaCl and dimethyl sulfoxide (DMSO) (Sigma)
were mixed without stirring in aliquots of blood, gel-
filtered platelets, or platelet-rich plasma so that the
effect of cocaine might be compared with that of known
platelet stimulators. Cocaine solution (provided by the
Beth Israel Hospital Pharmacy as a 4% solution with
0.5% chlorobutanol as preservative) was mixed without
stirring in aliquots of blood or platelet-rich plasma to
provide the following final cocaine concentrations: 13
mM, 1.3 mM, 130 uM, and 10 jIM. Stock 0.5% chloro-
butanol solution was mixed with aliquots of control
blood, gel-filtered platelets, or platelet-rich plasma.
Samples were incubated for 2 minutes at room temper-
ature for ADP, PdBu, DMSO, or NaCl or for 5 minutes
for cocaine and chlorobutanol solutions. Times were
chosen by preliminary experiments that demonstrated
the time course at which peak effect was achieved. After
incubation, samples were fixed with a 1:1 dilution of
1.5% paraformaldehyde in phosphate-buffer solution
(PBS) at 4'C19 and processed that day.
Preparation of Antibodies and Controls
The murine monoclonal antibody 6D1 (a kind gift of
Barry Coller, MD), which binds to the platelet-specific
glycoprotein Ib (GpIb),20 was used to distinguish plate-
lets from erythrocytes and leukocytes, to which there is
minimal binding of this antibody. Goat anti-mouse
immunoglobulin (Ig) G conjugated to phycoerythrin
(GAM-PE) (Sigma) was used to detect 6D1 binding to
platelets. S12, a murine monoclonal antibody that binds
to human P-selectin, was kindly provided by Dr Rodger
877
McEver, Oklahoma Medical Research Foundation.21
S12 was conjugated to fluorescein isothiocyanate
(FITC) by standard methods.22 Rabbit anti-human fi-
brinogen antibody (RAHFg) was raised in rabbits and
purified, as previously described.23 This antibody was
then conjugated to FITC (RAHFg-FITC). Because the
limited supplies of this antibody were exhausted, a
second IgG anti-fibrinogen antibody, raised in goats
against human fibrinogen (GAHFg), was used in later
experiments to assess platelet fibrinogen binding.
GAHFg was also conjugated to fluorescein (GAHFg-
FITC) (Atlantic Antibody, Stillwater, Minn). Although
both anti-fibrinogen antibodies bound to platelets only
after their activation, as described below, their binding
characteristics were not identical; therefore, they were
used individually in separate sections of this study. No
attempt was made to combine results on fibrinogen
binding obtained using these different antibodies.
To confirm that antibody binding was specific to
platelet activation and did not reflect an artifact of
sample processing or microaggregate formation, fluo-
rescence from paired fluorophore-conjugated isotype
antibodies that did not bind blood cell antigens were
compared with that associated with the activation-
dependent antibodies. Samples treated with standard
agonists and cocaine were labeled, as below, with FITC
conjugated anti-fibrinogen antibodies and S12-FITC in
parallel with FITC conjugated control antibodies with
equal or greater fluorescence-to-protein ratios. Mouse
IgG-FITC (Sigma) served as a control for S12-FITC,
and rabbit IgG-FITC (Sigma) was the control for
RAHFg-FITC. Goat anti-human haptoglobin-FITC
(IgG) (Atlantic Antibody) was used to control for
GAHFg-FITC.
Flow Cytometric Evaluation of Platelet Activation
Blood cells were labeled with the appropriate anti-
body using a technique similar to that previously report-
ed."1 After cell fixation with paraformaldehyde, which
was used to prevent time-dependent expression of P-se-
lectin and fibrinogen,"1 5 ,L aliquots of blood or
platelet-rich plasma were diluted in 50 ,uL aliquots of
PBS in duplicate. Although fixation of cells before the
addition of antibody has been shown to decrease the
amount of epitope detected using this assay,11 prelimi-
nary experiments revealed improved consistency of an-
tibody binding to platelets if fixation preceded the
addition of antibodies, and thus this sequence was
adopted for these experiments. These samples were
then incubated with 6D1 at a final concentration of 10
,g/mL for 20 minutes at room temperature. Next,
samples were centrifuged at 800g for 1 minute and
washed with 50 ,uL PBS. The duplicate samples were
similarly incubated with GAM-PE (24 ,ug/mL), spun,
and washed with PBS. Individual samples were then
incubated with S12-FITC (30 ,g/mL) or an anti-fibrin-
ogen antibody, RAHFg-FITC (30 gg/mL), or GAHFg-
FITC (30 ,g/mL) for 20 minutes at room temperature.
Samples then were diluted with PBS to a final volume of
600 ,uL and immediately examined by flow cytometry.
Processed samples of blood, gel-filtered platelets, and
platelet-rich plasma were analyzed in a Becton Dickin-
son FacStar Plus flow cytometer (Becton Dickinson,
Braintree, Mass). Fluorescent chromophores were ex-
cited with a 5-W argon laser at 200 mW power with a
 878 Circulation Vol 88, No 3 September 1993
wavelength of 488 nm. FITC and PE fluorescence were
detected using 530±11- and 575±11-nm band-pass fil-
ters, respectively. To distinguish platelets from other
cells and debris, two complementary techniques were
used, as previously described.1" A fluorescent threshold
was set to analyze only those cells that bound 6D1-
GAM-PE, effectively excluding erythrocytes and leuko-
cytes from platelets. The platelet population was further
defined by their forward and right-angle light scatter
characteristics, and a gate was set around these parti-
cles, as has been previously described.11 S12 and
RAHFg binding were then determined by analyzing
5000 platelets per sample at a rate of less than 1,000
cells per second. Light scatter and fluorescence data
were obtained with gain settings in a logarithmic mode,
and data were analyzed with a Hewlett-Packard Consort
30 software package (Palo Alto, Calif). Fluorescence
was expressed as mean log fluorescence per cell. Mean
log fluorescence was converted to a linear scale and
calibrated with fluorescent standard beads (Flow Cy-
tometry Standards Corporation, Research Triangle
Park, NC), as previously described.24
Enhancement Studies
To assess whether cocaine could enhance platelet
activation induced by standard agonists, whole blood
was incubated with cocaine, ADP, epinephrine (Abbott
Laboratories, North Chicago, Ill), or a combination of
cocaine and either ADP or epinephrine. Aliquots of
citrated whole blood (500 lL) were incubated for 5
minutes at room temperature without stirring with ADP
(0.2 to 20 t&M), epinephrine (1 MM), cocaine (10 or 130
Downloaded from http://ahajournals.org by on August 5, 2020
MM), or combinations of epinephrine and cocaine or
ADP and cocaine. Platelets were then fixed with para-
formaldehyde, as described above. P-selectin expression
and fibrinogen binding to the platelet were assayed
using S12-FITC and GAHFg-FITC, respectively.
Inhibition of Thromboxane A2 Formation and ADP
To investigate if cocaine-induced release of a-granule
products required cyclooxygenase metabolites of plate-
let arachidonic acid, citrated whole blood was pre-
treated with aspirin (1 mM) (Sigma) for 30 minutes at
room temperature. Donor-paired, citrated whole blood
that had been standing at room temperature for the
same period of time served as a control for time-
dependent platelet activation. Thereafter, blood was
incubated with standard agonists or cocaine, as de-
scribed above. S12 and GAHFg binding in the aspirin-
treated samples were compared with paired controls.
Similarly, to investigate if ADP, derived from either
platelets themselves or erythrocytes, was necessary for
cocaine to induce platelet activation, citrated whole
blood was pretreated with the ADP scavenger apyrase
(Sigma) (2 units/mL) for 5 minutes. Preliminary exper-
iments using flow cytometry demonstrated that apyrase
completely eliminated the ability of ADP to increase
P-selectin expression of platelets in whole blood, thus
confirming and extending findings with apyrase in whole
blood aggregometry studies reported previously.25-27
Subsequently, incubation with cocaine and standard
agonists followed by flow cytometric determination of
S12 binding was performed.
(0)
-W_i
hi
0
a
z
log FLUORESCENCE UNITS
U)
W
J
W
4
-I
_.
CL
z
log FLUORESCENCE UNITS
FIG 1. Profile histogram offlow cytometry-detected fluores-
cence associated with P-selectin expression (top) and fibrin-
ogen binding (bottom) in whole blood. Top, P-selectin expres-
sion as determined by S12 binding: mean platelet log
fluorescence after incubation with control solvent, phorbol
12,13-dibutyrate (PDBU) (0.2 gM) as a positive control, and
cocaine (13 mM). Bottom, Fibrinogen binding as determined
by rabbit anti-human fibrinogen antibody binding: mean
platelet logfluorescence after incubation with control solvent,
ADP (10 ,M) as a positive control; and cocaine (13 mM).
5-Hydroxytryptamine Uptake
Studies of GC-radiolabeled serotonin (`4C-5HT)
were performed in plasma to assess whether cocaine
prevented uptake of serotonin. Aliquots of platelet-
rich plasma, which were first incubated with control
(chlorobutanol solution), cocaine (13 mM), and imip-
ramine (CIBA-GEIGY) (2 ,uM) for 20 minutes at
room temperature, and 14C-5HT (Amersham) were
then added for 30 minutes at room temperature. Total
available '4C-5HT was determined by scintigraphy of a
100 gL sample. The treated platelet-rich plasma then
was centrifuged, and scintigraphy was performed on
the platelet-free supernatant. 14C-SHT uptake was
then calculated as 1-(free cpm/total cpm). Samples
were performed in duplicate.
Statistical Analysis
All data are given as the mean±SEM. Statistical
significance was assessed by Student's t test.
Results
Addition of cocaine to whole blood resulted in a
rightward shift of the platelet population fluorescence
curve with the appearance of a subpopulation of acti-
vated platelets, as shown in representative histograms
for RAHFg and S12 binding, which reflect detection of
 Kugelmass et al Activation of Human Platelets by Cocaine 879

11

o
120-

100
I
i 80-

160- 1 :
d' .5
*S 40-
: /
1

* whole blood
I 20- :/~ o PdBu
W
jf * GFP
0
L., A PRP
1
-2u 1 2 3 4 5 6 7 8 9 10
0.01 0.1 1 10 100 Time (minutes)
lcocainel mM
FIG 3. Plot of time course of cocaine-induced platelet P-se-

1
180-

160-

140-
120-
{ lectin expression in response to cocaine (13 mM). Each point
represents the mean ±SEM value ofpercent maximal increase
above the fluorescence observed in unstimulated platelets in
five separate determinations.

100- ated fibrinogen binding also increased in a typical

.5 80- concentration-response fashion on exposure to cocaine
(Fig 2, bottom). Addition of cocaine (10 MtM) resulted in
0 cocaine a 45±12% (P<.01) increase in RAHFg binding. Treat-
60-
40- ment with higher concentrations of cocaine (130 uM
O ADP
20- and 13 mM) induced 61±19% (P<.01) and 125±40%
U 1 | l
(P<.03) increases in RAHFg binding, respectively.
0.01
9
0.1 ; 10
-
100 Treatment with ADP (10 MiM), a potent stimulus of
[cocainCe mM platelet fibrinogen binding, induced a 166±28% in-
FIG 2. Plots of effect of cocaine on P-selectin expression crease in RAHFg binding. Thus, incubation of whole
Downloaded from http://ahajournals.org by on August 5, 2020

(top) and fibrinogen binding (bottom). The increase in fluo- blood with cocaine induced significant increases in
rescence is expressed as a percentage increase in fluorescence platelet fibrinogen binding as well as release of the
per platelet versus control and standard agonists. Each point
contents of a-granules.
represents the mean ±SEM value for 5 to 13 separate deter- To determine if cocaine-induced platelet activation
minations. Top, P-selectin expression, as determined by S12 occurred with a similar frequency in our population of
binding, as a function of the cocaine concentration in whole
normal donors tested as described, we designated a
blood, platelet-rich plasma (PRP), and gel-filtered platelets
positive response, or "platelet activation," as mean
fluorescence per platelet of more than 3 SDs above the
(GFP). The increase in fluorescence is expressed as a percent- mean fluorescence of unstimulated platelets to which
age increase in fluorescence per platelet above control (un- saline and 0.5% chlorobutanol in water had been added.
stimulated). Each point represents the mean ±SEM value of By this criterion, samples from three of six donors
percent maximal change in 5 to 13 separate determinations. demonstrated platelet activation at cocaine concentra-
The increase in fluorescence induced by phorbol 12,13- tions of 10 and 130 MtM, as determined by fibrinogen
dibutyrate (PdBu) (0.2 uM) is provided for comparison. binding, whereas a cocaine concentration of 13 mM
Bottom, Platelet-associated fibrinogen, as determined by rab- resulted in an increased fibrinogen binding response in
bit anti-human fibrinogen antibody binding, as a function of four of five donors. Similar results are seen for S12
cocaine concentration; the increase in fluorescence induced by binding, as platelets from 89% of donors responded to a
ADP (10 MM) is provided for comparison. cocaine concentration of 13 mM. For cocaine concen-
trations of 10 and 130 MiM, fewer donors were positive
platelet-associated fibrinogen and surface expression of (36% and 27%, respectively). Thus, there was notable
P-selectin (Fig 1). Binding of S12, expressed as fluores- heterogeneity of response to cocaine among donors, in
cence per platelet, increased by 112±24% (P<.001)
both quantitative change in S12 or RAHFg binding and
the threshold drug concentration to which platelets
when blood was exposed to cocaine at 13 mM, 77±12% responded.
(P<.001) at a cocaine concentration of 1.3 mM, and To determine whether constituents of whole blood
16±16% (P=NS) at a cocaine concentration of 130 ,M other than plasma contributed to the increased expres-
(Fig 2, top). A cocaine concentration of 10 ,M induced sion of P-selectin on the surface of cocaine-treated
a 36±15% (P<.04) increase in S12 binding. For com- platelets, the above results were compared with the
parison, incubation with the phorbol ester PdBu (0.2 stimulation by cocaine of platelets suspended in plate-
MM), which promotes marked release of the contents of let-rich plasma. Addition of the highest concentration of
a-granules, resulted in a 62±22% increase. Cocaine- cocaine (13 mM) resulted in a significant increase in S12
induced S12 binding achieved a maximum after 5 min- binding, 85±13% (P<.002); at the lower cocaine con-
utes incubation with the drug (Fig 3). Platelet-associ- centrations, however, no significant difference from
 880 Circulation Vol 88, No 3 September 1993
control was observed. Likewise, treatment of gel-fil-
tered platelets with cocaine resulted in a significant rise
in S12 binding, 59±7% (P<.001), only at high concen-
trations (13 mM cocaine). When these results are
directly compared with the results in whole blood (Fig 2,
top) it is apparent that P-selectin expression of platelets
suspended in buffer (gel-filtered platelets) or plasma
(platelet-rich plasma) that results from exposure to
cocaine parallels that of platelets in whole blood but
that the latter is enhanced. Thus, although very high
concentrations of cocaine can directly stimulate P-selec-
tin expression in platelets, other constituents of whole
blood enhance the stimulatory effect of submaximal
concentrations of cocaine.
As platelets are activated in vivo under circumstances
of multiple stimuli, we sought to determine whether
cocaine could enhance the stimulatory effects of ago-
nists that mimic physiological mediators likely to be
present after cocaine ingestion. Whole blood was incu-
bated with ADP, epinephrine, submaximal concentra-
tions of cocaine, or the combination of cocaine and
ADP or epinephrine, as described in "Methods." En-
hancement was defined as an increase in mean fluores-
cence per antibody-labeled platelet induced by the
combination of cocaine and agonist that was more than
the sum of the increases in mean fluorescence per
platelet separately induced by cocaine and the agonist.
In all samples (n=4), enhancement in S12 and/or
GAHFg binding could be demonstrated. We found
variability among donors regarding specific agonists and
concentrations; however, the effect of a standard ago-
nist could be reproducibly enhanced by cocaine at
submaximal concentrations (10 and 130 MiM). The mean
Downloaded from http://ahajournals.org by on August 5, 2020
values of these determinations are shown in Fig 4, top
and bottom, for P-selectin expression and platelet-
bound fibrinogen, respectively.
To determine if cocaine-induced a-granule secretion
was dependent on metabolites of cyclooxygenase or
ADP, whole blood was pretreated with aspirin or the
ADP scavenger apyrase and then incubated with co-
caine, followed by determination of S12 binding. In no
sample (n=4 for each agent) was there a significant
difference in S12 binding between aspirin or apyrase-
treated platelets and paired controls. Thus, cocaine-
induced platelet activation does not depend on ADP or
arachidonic acid metabolites.
Cocaine is known to inhibit presynaptic uptake of the
amine serotonin in neural tissue. As platelets also
imbibe, store, and release significant amounts of sero-
tonin, we sought to establish if cocaine also inhibited its
uptake by platelets. Cocaine inhibited `4C-5HT uptake
in a concentration-dependent manner (Fig 5) with an
IC50 of 8.7 uM. By comparison, imipramine completely
inhibited 14C-5HT uptake, and the chlorobutanol con-
trol demonstrated no significant effect (84+6% maxi-
mum uptake).
Discussion
The results of this study indicate that cocaine induces
release of platelet a-granule contents and the binding of
fibrinogen to the platelet surface and thereby implicate
cocaine as an agonist capable of activating platelets in
whole blood. Control experiments using matched anti-
bodies that do not bind to platelets confirm that the
increased fluorescence of platelets is specific for the
30-
U =ocaine
25- ocaIne & agonist-sum
0
0 ocaine & agonist-conbinatn
X 20-
8 15-
1 10-
0s
5-
* cocaine
25- Q cocaine & agonist-sum
conet cocaine&
of aponh1,combonat)on
x 20-
LI 15-
~10-
5-
FiG 4. Bar graphs of enhancement of epinephrine- and
ADP-induced platelet activation in whole blood by cocaine.
Bars represent the mean-±SEM response for near-threshold
concentrations of cocaine only (10 MM or 130 MiM), the sum
of the activation induced by these concentrations of cocaine
and that induced by either ADP or epinephrine added to a
separate aliquot of blood, and the observed combination of
these concentrations of cocaine added together with epineph-
rine or ADP. Top, Percent increase x100 in P-selectin
expression above that induced by submaximal concentrations
of epinephrine or ADP alone, as determined by S12 binding.
Bottom, Percent increase x 100 in platelet fibrinogen binding,
as determined by goat anti-human fibrinogen antibody bind-
ing, above that induced by submaximal concentrations of
epinephrine orADP alone. Bars represent mean ±SEM values
of four separate determinations.
activated state and is not the result of microaggregates.
Significant increases in platelet-associated fibrinogen
were demonstrated at a cocaine concentration of 10 ,uM
and increased further with higher drug concentrations.
Similarly, release of platelet a-granule contents, as
measured by S12 binding, revealed a concentration-
dependent response to cocaine added to whole blood.
Neither platelet cyclooxygenase metabolites nor ADP,
whether released by proximate platelets or by erythro-
cytes, are essential for cocaine-induced platelet activa-
tion. Cocaine, however, does inhibit serotonin (5HT)
reuptake, as shown by these and previous studies.28
Thus, cocaine may increase the local concentration of
this endogenous agonist at the surface of the platelet, an
effect that might also enhance the degree of activation
induced by other agonists, such as epinephrine, when
present in subthreshold concentrations.
 Kugelmass et al Activation of Human Platelets by Cocaine
11
Q might be established by selective addition of isolated
.0001 0.001 0.01 0.1 1 10
[cocainel mM
FIG 5. Plot of inhibition of platelet uptake of 14C-labeled
serotonin (SHT) by cocaine. Each point represents the
mean +SEM value of three separate determinations, each in
duplicate.
Cocaine, at lower concentrations, enhances the effect
of ADP and epinephrine, as do other standard platelet
agonists; similar results have been reported in rabbit
platelets suspended in plasma and stimulated by low
concentrations of cocaine and arachidonic acid.29 Our
laboratory previously has demonstrated that enhance-
ment of platelet activation by pairs of agonists, each in
subthreshold concentration, has characteristic features
regardless of the identity of the agonists.30 These fea-
tures include an initial rise in intracellular Ca'+ and
phosphorylation of intracellular protein substrates that
Downloaded from http://ahajournals.org by on August 5, 2020
correlate closely to fibrinogen binding to the platelet
surface.31 Our observation that the binding of fibrino-
gen can be enhanced by combinations of cocaine and
either ADP or epinephrine suggests that the effect of
cocaine on platelets may be mediated through mecha-
nisms of activation common to other platelet agonists.
Comparisons of the effects of cocaine on platelets in
whole blood, gel-filtered platelets, and platelet-rich
plasma provide further insight into the interaction of
platelets with other blood cells. The extent of P-selectin
expression induced by cocaine in our studies is similar to
that detected by George and colleagues19 using radiola-
beled S12 antibody to detect platelet activation by
standard agonists added to human whole blood. Flow
cytometric determination of P-selectin expression in
washed platelets following maximal stimulation with
thrombin as reported by this group has, however, dem-
onstrated a sevenfold increase in epitope expression,32
which is more than that induced by either cocaine or
PdBu in our studies. Differences in the platelet medium,
as well as variance in the flow cytometric method,
including differences in fixation techniques, may ac-
count for the lesser degree of P-selectin expression in
our experiments. In the direct comparisons reported in
this study, the response of platelets to cocaine in whole
blood was greater than that of platelets washed by gel
filtration or of those suspended in plasma, suggesting
that erythrocytes, leukocytes, or other blood cells may
amplify platelet activation, a finding that is consistent
with platelet activation induced by other stimulants.
The presence of erythrocytes can enhance platelet
response to standard agonists,12 and neutrophils can
also promote platelet activation, an effect that appar-
881
ently is mediated by leukotriene B4.'3 The association of
activated platelets with monocytes both in vitro and in
vivo has also been described, although whether either
cell type enhances the other has not been estab-
lished.33-35 Which, if any, of the cells existing in whole
blood enhance the effect of cocaine on blood platelets
blood components.
This potential of blood constituents to amplify the
effect of an agonist emphasizes the usefulness of the
method used in these experiments to investigate the
effect of cocaine on platelet activation. Previous studies
by our laboratory36 and others29 found that cocaine
could not directly cause aggregation of platelets sus-
pended in platelet-rich plasma, as assessed by turbido-
metric methods. The platelet stimulation by cocaine as
detected by flow cytometric analysis of individual plate-
lets in whole blood may reflect both greater sensitivity
and preservation of intercellular interaction.
The concentration range of cocaine that induces
fibrinogen binding and the release of the contents of
a-granules and prevents platelet uptake of serotonin is
likely to be relevant to the clinical problem of cocaine
abuse. Although the most pronounced effect of cocaine
on platelet function in these studies is with millimolar
concentrations of the drug, our data indicate that
significant increases in a-granule release, fibrinogen
binding, and inhibition of serotonin uptake are evident
at a cocaine concentration of 10 gM.
In addition, this submaximal concentration of cocaine
enhances epinephrine- and ADP-induced platelet acti-
vation. Cocaine has been shown to increase the systemic
concentrations of catecholamines, particularly epineph-
rine, in animal models.37'38 Thus, platelet activation that
may result from this increase in catecholamine activity
is likely to be enhanced by cocaine, and together with
the direct platelet activation induced by cocaine may
represent a potent prothrombotic effect. Serum cocaine
levels of 80 1AM have been reported in autopsy studies of
patients suffering cocaine-associated sudden death.39 As
the drug is rapidly metabolized by serum cholinester-
ases in vivo,4041 it is probable that premortem serum
concentrations in these patients were manyfold higher.
Experimental studies of cocaine ingestion in volunteers
have revealed serum cocaine levels as high as 50
gM.42-44 Moreover, it also is likely that the local blood
concentrations of cocaine in local capillary networks at
the site of ingestion, through which platelets circulate,
may be manyfold higher than those found in the periph-
eral circulation.
Because this study demonstrates that cocaine induces
platelet activation, one may question why apparently
only a minority of those who ingest the drug develop
thrombotic consequences. Our data indicate that there
is marked individual variability in cocaine's effect on
platelets, as samples from fewer than 50% of donors
met our criteria for significant release of a-granule
products at a cocaine concentration of 10 gM. Thus,
one might advance the hypothesis that only a subset of
those who ingest cocaine in "lower" amounts will de-
velop significant platelet activation and sustain clinically
apparent thrombosis, perhaps facilitated by concomi-
tant states that predispose to activation of platelets or
soluble factors of clotting.
 882 Circulation Vol 88, No 3 September 1993
Cocaine-induced platelet activation may contribute to
cardiovascular disorders other than acute thrombosis.
The release of platelet-derived mitogens, including
platelet-derived growth factor, epidermal growth factor,
and transforming growth factor-p, from platelet a-gran-
ules may explain the premature atherosclerosis associ-
ated with cocaine abuse,4546 particularly if the endothe-
lium is altered by such use. Similarly, the elevation of
local concentrations of vasoactive substances such as
serotonin, as a result of diminished uptake, might
contribute to the coronary vasospasm described with
cocaine administration.47-49 Thus, it is likely that acute
vascular disorders associated with cocaine represent the
interplay of several pathophysiological pathways rather
than resulting from a single cause. Cocaine-induced
platelet activation potentially can contribute to many of
these processes and may represent a final event leading
to the clinical presentation of acute thrombotic
occlusion.
Acknowledgments
This work was supported by US Public Health Service grants
HL-38820 and DA-06306. Dr Ware is the recipient of a
Research Career Award in Thrombosis (HL-02271). Dr
Kugelmass is the recipient of an Individual National Research
Service Award (HL-08816). We wish to thank Drs Barry
Coller and Rodger McEver for their generous donations of
antibodies and Ms Marianne Smith for her technical assis-
tance. Flow cytometry was performed at the Beth Israel
Hospital Research Flow Cytometry facility.
References
1. Minor RL, Scott DS, Brown DD, Winniford MD. Cocaine-induced
Downloaded from http://ahajournals.org by on August 5, 2020
myocardial infarction in patients with normal coronary arteries.
Ann Intern Med. 1991;115:797-806.
2. Kloner RA, Hale S, Alker K, Rezkalla S. The effects of acute and
chronic cocaine use on the heart. Circulation. 1992;85:407-419.
3. Delaney K, Hoffman RS. Pulmonary infarction associated with
crack cocaine use in a previously healthy 23-year old woman. Am
J Med. 1991;91:92-94.
4. Zamora-Quezada JC, Dinerman H, Stadecker MJ, Kelly JJ.
Muscle and skin infarction after free-basing cocaine (crack). Ann
Intern Med. 1988;108:564-565.
5. Sharff JA. Renal infarction associated with intravenous cocaine
use. Ann Emerg Med. 1984;13:1145-1147.
6. Lisse JR, Davis CP, Thurmond-Anderle M. Cocaine abuse and
deep venous thrombosis. Ann Intern Med. 1989;110:571-572.
7. Simpson RW, Edwards RD. Pathogenesis of cocaine-induced
ischemic heart disease. Arch Pathol Lab Med. 1986;110:479-484.
8. Grant GJ, Ramanathan S, Patel N, Turndorf H. The effects of
local anesthetics on maternal and fetal platelet function. Acta
Anaesthesiol Scand. 1989;33:409-412.
9. Peerschke EI. Platelet membrane alterations induced by the local
anesthetic dibucaine. Blood. 1986;68:463-471.
10. Feinstein MB, Fiekers J, Frazer C. An analysis of the mechanism
of local anesthetic inhibition of platelet aggregation and secretion.
J Pharmacol Exp Ther. 1976;197:215-228.
11. Shattil SJ, Cunningham M, Hoxie JA. Detection of activated
platelets in whole blood using activation-dependent monoclonal
antibodies and flow cytometry. Blood. 1987;70:307-315.
12. Santos MT, Valles J, Marcus AJ, Safler LB, Broekman MJ, Islam
N, Ullman HL, Eiroa AM, Aznar J. Enhancement of platelet
reactivity and modulation of eicosanoid production by intact eryth-
rocytes. J Clin Invest. 1991;87:571-580.
13. Maschio AD, Evangelista V, Rajtar G, Chen ZM, Cerletti C, de
Gaetano G. Activated polymorphonuclear leukocytes stimulate
platelet function. Am J Physiol. 1990;258:H870-H879.
14. Bevilacqua M, Butcher E, Furie B, Furie B, Gallatin M, Gimbrone
M, Harlan J, Kishimoto K, Lasky L, McEver R, Paulson J, Rosen
S, Seed B, Siegelman M, Springer T, Stoolman L, Tedder T, Varki
A, Wagner D, Weissman I, Zimmerman G. Selectins: a family of
adhesion receptors. Cell. 1991;67:223.
15. Steinberg PE, McEver RP, Shuman MA, Jacques YV, Bainton
DF. A platelet alpha-granule membrane protein (GMP-140) is
expressed on the plasma membrane after activation. J Cell Biol.
1985;101:880-886.
16. Warkentin TE, Powling MJ, Hardisty RM. Measurement of
fibrinogen binding to platelets in whole blood by flow cytometry: a
micromethod for the detection of platelet activation. Br J
Haematol. 1990;76:387-394.
17. Oda A, Daley JF, Kang JH, Smith M, Salzman EW. Heterogeneity
in filamentous actin content among individual human blood
platelets. Blood. 1992;79:920-927.
18. Oda A, Druker BJ, Smith M, Salzman EW. Association of pp60src
with Triton X-100 insoluble residue in human blood platelets
requires platelet aggregation and actin polymerization. J Biol
Chem. 1992;267:20075-20081.
19. George JN, Pickett EB, Saucerman S, McEver RP, Kunicki TJ,
Kieffer N, Newman PJ. Platelet surface glycoproteins: studies on
resting and activated platelets in normal subjects, and observations
in patients during adult respiratory distress syndrome and cardiac
surgery. J Clin Invest. 1986;78:340-348.
20. Coller BS, Peerschke EI, Scudder LE, Sullivan CA. Studies with a
murine monoclonal antibody that abolishes ristocetin-induced
binding of von Willebrand factor to platelets: additional evidence
in support of GPIb as a platelet receptor for von Willebrand factor.
Blood. 1983;61:99-110.
21. McEver RP, Martin MN. A monoclonal antibody to a membrane
glycoprotein binds only to activated platelets. J Biol Chem. 1984;
259:9799-9804.
22. Rinderknecht H. Ultra-rapid fluorescent labelling of proteins.
Nature Lond. 1963;193:167-168.
23. Lindon JL, McManama G, Kushner L, Merrill EW, Salzman EW.
Does the conformation of adsorbed fibrinogen dictate platelet
interactions with artificial surfaces? Blood. 1986;68:355-362.
24. Muirhead KA, Schmitt TC, Muirhead AR. Determination of
linear fluorescence intensities from flow cytometric data accu-
mulated with logarithmic amplifiers. Cytometry. 1983;3:251-256.
25. Kato T, Wokalek H, Schopf E, Hovestadt I, Ernst M, Rietschel E,
Fischer H. Zymosan-induced chemiluminescence of granulocytes
in incubated human whole blood: the role of platelets. Tohoku J
Exp Med. 1983;141:423-431.
26. Abi-Younes SA, Ayers ML, Myers AK. Mechanism of ethanol-
induced aggregation in whole blood. Thromb Res. 1991;63:481-489.
27. Petty AC, Scrutton MC. Platelet aggregation in whole blood: is the
response to adrenaline, 5-hydroxytryptamine and PAF a direct
consequence of stimulation by these agonists? Thromb Res. 1989;
54:151-159.
28. Bevan J, Heptinstall S. Effects of ketanserin and mepyramine on
platelet aggregation and on the uptake of 5-hydroxytryptamine
into platelets. Thromb Res. 1983;30:415-423.
29. Togna G, Tempesta E, Togna AR, Dolci N, Cebo B, Caprino L.
Platelet responsiveness and biosynthesis of thromboxane and pros-
tacyclin in response to in vitro cocaine treatment. Haemostasis.
1985;15:100-107.
30. Ware JA, Smith M, Salzman EW. Synergism of platelet-
aggregating agents: role of elevation of cytoplasmic calcium. J Clin
Invest. 1987;80:267-271.
31. Saitoh M, Salzman EW, Smith M, Ware JA. Activation of protein
kinase C in platelets by epinephrine and A23187: correlation with
fibrinogen binding. Blood. 1989;74:2001-2006.
32. Johnston GI, Pickett EB, McEver RP, George JN. Heterogeneity
of platelet secretion in response to thrombin demonstrated by
fluorescence flow cytometry. Blood. 1987;69:1401-1403.
33. Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Dynamics
of leukocyte-platelet adhesion in whole blood. Blood. 1991;78:
1730-1737.
34. Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Activated
and unactivated platelet adhesion to monocytes and neutrophils.
Blood. 1991;78:1760-1769.
35. Rinder CS, Bonan JL, Rinder HM, Mathew J, Hines R, Smith BR.
Cardiopulmonary bypass induces leukocyte-platelet adhesion.
Blood. 1992;79:1201-1205.
36. Kugelmass AD, Ware JA. Cocaine and coronary artery
thrombosis. Ann Intern Med. 1992;116:776-777.
37. Chiueh CC, Kopin IJ. Centrally mediated release by cocaine of
endogenous epinephrine and norepinephrine from the sympatho-
adrenal medullary system of unanesthetized rats. J Pharmacol Exp
Ther. 1978;205:148-154.
38. Schwartz AB, Boyle W, Janzen D, Jones RT. Acute effects of
cocaine on catecholamines and cardiac electrophysiology in the
conscious dog. Can J Cardiol. 1988;4:188-192.
 Kugelmass et al Activation of Human Platelets by Cocaine
39. Virmani R, Robinowitz M, Smialek JE, Smyth DF. Cardiovascular
effects of cocaine: an autopsy study of 40 patients. Am Heart J.
1988;115:1068-1076.
40. Stewart DJ, Inaba T, Tang BK, Kalow W. Hydrolysis of cocaine in
human plasma by cholinesterase. Life Sci. 1977;20:1557-1564.
41. Inaba T, Stewart DJ, Kalow W. Metabolism of cocaine in man. Clin
Pharmacol Ther. 1978;23:547-552.
42. Paly D, Jatlow P, Van Dyke C, Jeri FR, Byck R. Plasma cocaine
concentrations during cocaine paste smoking. Life Sci. 1982;30:
731-738.
43. Van Dyke C, Barash PG, Jatlow P, Byck R. Cocaine: plasma
concentrations after intranasal application in man. Science. 1976;
191:859-861.
44. Perreault CL, Hague NL, Ransil BJ, Morgan JP. The effects of
cocaine on intracellular Ca2l handling and myofilament Ca 2+
responsiveness of ferret ventricular myocardium. Br J Pharmacol.
1990;101:679-685.
Downloaded from http://ahajournals.org by on August 5, 2020
883
45. Dressler FA, Malekzadeh S, Roberts WC. Quantitative analysis of
amounts of coronary artery narrowing in cocaine addicts. Am J
CardioL 1990;65:303-308.
46. Kolodgie FD, Virmani R, Cornhill JF, Herderick EE, Smialek J.
Increase in atherosclerosis and adventitial mast cells in cocaine
abusers: an alternative mechanism of cocaine-associated
coronary vasospasm and thrombosis. JAm Coil Cardiol. 1991;17:
1553-1560.
47. Schachne JS, Roberts BH, Thompson PD. Coronary-artery spasm
and myocardial infarction associated with cocaine use. N EngI J
Med. 1984;310:1665-1666.
48. Lange RA, Cigarroa RG, Yancy CW, Willard JE, Popma JJ, Sills
MN, McBride W, Kim AS, Hillis LD. Cocaine-induced coronary-
artery vasoconstriction. N EngI J Med. 1989;321:1557-1562.
49. Isner JM, Chokshi SK. Cocaine and vasospasm. N EngI J Med.
1989;321:1604-1606. Editorial.