2006 Schattauer GmbH, Stuttgart
Platelets and Blood Cells
Dose- and time-dependent antiplatelet effects of aspirin
Christina Perneby1, N. Hkan Walln1,2, Cathy Rooney3, Desmond Fitzgerald4, Paul Hjemdahl1
1
Department of Medicine, Clinical Pharmacology Unit, Karolinska University Hospital (Solna), Stockholm, Sweden; 2Department of Medicine,
Karolinska Institutet at Danderyd's Hospital, Danderyd, Danderyd, Sweden; 3Department of Clinical Pharmacology, Royal College of
Surgeons of Ireland, Dublin, Ireland; 4Molecular Medicine Laboratory, Conway Institute, University College Dublin, Ireland
Summary
Aspirin is widely used, but dosages in different clinical situations
and the possible importance of aspirin resistance are debated.
We performed an open cross-over study comparing no treat-
ment (baseline) with three aspirin dosage regimens 37.5 mg/
day for 10 days,320 mg/day for 7 days,and,finally,a single 640 mg
dose (cumulative dose 960 mg) in 15 healthy male volunteers.
Platelet aggregability was assessed in whole blood (WB) and pla-
telet rich plasma (PRP).The urinary excretions of stable thromb-
oxane (TxM) and prostacyclin (PGI-M) metabolites, and bleed-
ing time were also measured.Platelet COX inhibition was nearly
complete already at 37.5 mg aspirin daily, as evidenced by >98 %
suppression of serum thromboxane B2 and almost abolished ar-
achidonic acid (AA) induced aggregation in PRP 26 h after dos-
ing. Bleeding time was similarly prolonged by all dosages of as-
Keywords
Platelet function, thromboxane, prostacyclin, acetylsalicylic acid,
dosage
Introduction
Acetylsalicylic acid, commonly referred to as aspirin, is a widely
used and very well documented antiplatelet drug that lowers car-
diovascular morbidity and mortality (1, 2). The antiplatelet ef-
fects of aspirin are related to irreversible inhibition of platelet
cyclooxygenase (COX), and reduced production of the potent
vasoconstrictor and platelet activator thromboxane A2 (TxA2)
(2). However, aspirin has limitations as an antithrombotic agent,
and there is still debate about the optimal dosages in different
clinical situations. Aspirin has variable antiplatelet activity when
assessed by different methods, and the poorly defined phenom-
enon of aspirin resistance is much discussed. Using various
test methods, 557 % of patients in different studies were judged
not to have optimal antiplatelet effects during aspirin treatment
(3), and several authors have reported an increased risk of suffer-
ing major cardiovascular events in post hoc analyses of studies
Correspondence to:
Prof. Paul Hjemdahl, MD, PhD
Department of Medicine
Clinical Pharmacology Unit
Karolinska University Hospital (Solna)
SE-171 76 Stockholm, Sweden
Tel.: +46 8 5177 5293 or 92, Fax +46 8 30 85 29
E-mail: Paul.Hjemdahl@ki.se
652
pirin. Once daily dosing was associated with considerable re-
covery of AA induced platelet aggregation inWB after 24 hours,
even after 960 mg aspirin. Collagen induced aggregation in WB
with normal extracellular calcium levels (hirudin anticoagulated)
was inhibited <40% at all dosages. TxM excretion was incom-
pletely suppressed, and increased <24 hours after the cumu-
lative 960 mg dose. Aspirin treatment reduced PGI-M already at
the lowest dosage (by 25%), but PGI-M excretion and platelet
aggregability were not correlated. Antiplatelet effects of aspirin
are limited in WB with normal calcium levels. Since recovery of
COX-dependent platelet aggregation occurred within 24 hours,
once daily dosing of aspirin might be insufficient in patients with
increased platelet turnover.
Thromb Haemost 2006; 95: 6528
among such patients (24). However, prospective studies vali-
dating the clinical utility of identifying biochemical or func-
tional aspirin resistance are lacking.
TxA2 generated by activated platelets forms a positive feed-
back loop which accelerates the local platelet activation (5). Less
than 5% residual capacity to generate TxA2 is enough to fully
sustain thromboxane dependent platelet aggregation (6, 7), and
full aggregation was achieved in vitro with only 2.5 % of the pla-
telet population being aspirin free (7). Thus, nearly complete ir-
reversible inhibition of platelet COX is required for effective
antithrombotic aspirin therapy. Interestingly, recovery of TxB2
was found already 4 hours after ingestion of 650 mg aspirin (7).
Chronic daily administration of as little as 0.45 mg/kg aspirin
provides sufficient platelet COX inhibition, defined as a vir-
tually complete inhibition of serum TxB2, in healthy volunteers
(8). In clinical trials, however, daily aspirin dosages of 75150 mg
appear to be most effective in patients with atherosclerotic vas-
Financial support:
The study was supported by grants from the Swedish Science Council (5930), the Swed-
ish Heart-Lung Foundation, the Karolinska Institute, the Stockholm County Council, and
the Irish Heart Foundation.
Received October 5, 2005
Accepted after revision February 2, 2006
Prepublished online March 17, 2006 DOI 10.1160/TH05100653
 Perneby, et al.: Dose- and time-dependent aspirin effects
cular disease, and the clinical efficacy of aspirin is reduced or
lost at doses below 75 mg/day (1, 2). A recent retrospective
analysis of two GPIIb/IIIa inhibitor trials suggested that a higher
dose of aspirin may afford better protection against myocardial
infarction (9), but a similar analysis from the CURE study indi-
cated that 100 mg/day was more effective than 200 mg/day
(10). Prospective, placebo-controlled trials have shown excellent
protection with 75 mg aspirin daily in patients with stable (11) or
unstable (12) angina pectoris.
The effect of repeated daily low doses of aspirin is cumu-
lative, and steady state with regard to platelet COX inhibition is
reached within 7 days (8).After a single full dose of aspirin, COX
activity recovers by some 10% per day, as a function of platelet
turnover (2), which may vary both within and between patients.
Phasic increases in thromboxane synthesis were observed in pa-
tients with unstable angina, indicating that platelet activation
may occur during spontaneous ischemia (13). Furthermore, dia-
betes, preeclampsia and advanced coronary artery disease are as-
sociated with increased platelet turnover (1416). Thus, the
requirements for dosages and dosing intervals with aspirin treat-
ment may differ in different clinical situations.
At low dosages, first pass hepatic metabolism of aspirin li-
mits the systemic exposure of the patients to COX inhibition,
thus preserving the vasodilating and antiplatelet effects of en-
dothelial prostacyclin (17). Higher doses of aspirin may, how-
ever, attenuate this potentially antithrombotic mechanism (2,
18). Experiments in knock-out mice support an antithrombotic
role for endogenous prostacyclin (19). The importance of prosta-
cyclin in thrombotic processes in humans is not completely
understood, but several studies have recently demonstrated unfa-
vourable cardiovascular outcomes in patients treated with
COX-2 inhibitors (20), and this may be related to suppression of
prostacyclin production (21).
In the present study we compared the effects of different as-
pirin dosages, from 37.5 mg daily to a cumulative single dose of
960 mg, on various aspects of platelet function in healthy volun-
teers. We studied platelet aggregation in platelet rich plasma and
in whole blood, and we measured the urinary excretion of
thromboxane and prostacyclin metabolites, as well as TxB2 gen-
eration in clotting whole blood. We used hirudinized blood to
avoid the artefactual enhancement of the thromboxane depend-
ence of platelet aggregation (and thus the efficacy of aspirin)
which is seen when citrate is used as anticoagulant and extracel-
lular calcium levels are low (2224). We also investigated the
time dependence of the effects of different aspirin dosages to
clarify if there is significant recovery of COX-dependent platelet
function during the normal dosing interval of 24 hours.
Material and methods
Subjects
Fifteen healthy male volunteers with easily accessible antecubital
veins for blood sampling with minimal artefacts participated in
the study. They were non-smokers (3 used snuff), 2239 years old
and had a body mass index of 22.9 (range 2026). They were in-
structed to avoid antiplatelet drugs 2 weeks before their first visit.
Informed consent was obtained from all subjects, and the Ethics
Committee of the Karolinska Hospital had approved the study.
653
Study design
The study was an open cross-over study with measurements at
baseline (untreated) and during treatment with three different
dosages of aspirin. Measurements were performed at baseline,
and after aspirin treatment with 37.5 mg/day for 10 days (half a
tablet Trombyl 75 mg, Pharmacia Sverige AB), and 320 mg/
day for 7 days (one tablet Alka Seltzer, Bayer, Germany). Fin-
ally, they took a single dose of 640 mg aspirin (two tablets of
Alka Seltzer) 12h after the last 320 mg dose (cumulative dose
thus 960 mg on the last day). There was a 1014 day wash-out
period between the low dose and the intermediate/high dose in-
vestigations. The volunteers were instructed to refrain from caf-
feine, tobacco or food intake >12 hours before each visit.
Blood sampling was performed after 30 min rest in the supine
position on all occasions. Between samplings the volunteers
were free to move about, but were instructed not to perform rig-
orous physical activity. At baseline, the sampling was performed
three times during 24 hours (9 AM, 1 PM, and again at 9 AM) be-
fore the treatment was initiated. Thereafter, the subjects received
37.5 mg aspirin daily (low dose) for 10 days, and on the 10th day
blood sampling was performed 1.52 hours (9 AM), 6 hours
(1 PM), and 24 hours after the last dose of aspirin. The same pro-
cedure was performed after the last ingested dose of 320 mg as-
pirin; i.e. on the 7th day of treatment. However, after the 9 AM
sampling (1.52h after drug ingestion) the volunteers took an ad-
ditional single 640 mg aspirin dose to ensure complete platelet
COX-1 inhibition. Sampling was then repeated 4 hours and
24 hours after this high bolus dose.
Each subject brought his morning urine to the laboratory be-
fore the first blood sampling and when he returned the next day
for 24 h sampling. Urine samples were also collected at 1 PM on
days of experiment. These 1 PM samples reflect urinary excre-
tions of analytes <4h (960 mg) or <6h (37.5 mg) after drug in-
take; diurnal variability was examined by 1 PM measurements
also at baseline. Baseline values were the means of the two noc-
turnal samples before treatment.
Blood sampling
Venous blood was sampled without stasis from an antecubital
vein through a 19 G butterfly needle (Abott Ltd, Sligo, Ireland)
and was anticoagulated with recombinant desulphatohirudin
(CGP 39393; 11 700 ATU/mg; Ciba-Geigy, Basle, Switzerland;
diluted in physiological saline, final concentration 20 g/ml).
Two centrifugations at 190 x g and 1,400 x g for 10 minutes were
used to prepare platelet rich plasma (PRP) and platelet poor plas-
ma (PPP) immediately after blood sampling. For analyses using
the PFA-100 system (see below), blood was anticoagulated with
sodium citrate (final concentration 0.38 % w/v) according to in-
structions from the manufacturer.
Whole blood aggregometry
Platelet aggregation in whole blood was studied using an imped-
ance aggregometer (Chrono-log model 570-VS Four Sample;
Chrono-log Corp, Haverton, PA, USA). The blood was anti-
coagulated with hirudin and diluted 1:1 with physiological sa-
line. Samples were preincubated at 37oC for 5 minutes, after
which agonists were added. Agonists used were collagen type 1
(Horm, Nycomed Arzneimittel, Munich, Germany) at final con-
 Perneby, et al.: Dose- and time-dependent aspirin effects
centrations of 1, 3 or 5 M and arachidonic acid (AA) dissolved
in ethanol (Sigma Chemical Co., St. Louis, MO, USA) at final
concentrations of 0.2, 0.5 or 1 mM. Each experiment included a
test to verify that the final ethanol concentration in the sample
had no effect per se. The amplitude of aggregation was measured
after 8 minutes.
Aggregometry in platelet rich plasma (PRP)
A four-channel platelet aggregation profiler (PAP-4, Bio-Data
Corporation, Hatboro, PA, USA) was used to study platelet ag-
gregation in PRP. Adenosine diphosphate (ADP; Sigma Chemi-
cal Co.) was diluted in Tris buffer and the EC50 for ADP (i.e. the
concentration required for half-maximal aggregation) was deter-
mined by a dose-response procedure in which the extent of ag-
gregation after 4 minutes was measured; for details see (24). Pla-
telet COX inhibiting effects of aspirin treatment were deter-
mined as the ability of AA (final concentrations 0.5 and 1 mM)
to induce aggregation in PRP.
Bleeding time
Standardized transverse incisions were made on the lateral volar
side of the forearm, at a constant venous pressure of 40 mmHg,
using a disposable device (Surgicutt, Ortho Diagnostics, Raritan,
NJ, USA). Each incision was made 10 mm distally to the pre-
vious one. Total bleeding time was measured by collecting blood
on filter papers at 15 s intervals, according to a technique pre-
viously shown to reveal aspirin effects in our laboratory (25).
Thromboxane related measurements
Thromboxane B2 (TxB2) is the chemically stable and inactive hy-
dration product of TxA2, and measurements of TxB2 in serum re-
flect platelet COX inhibition. 11-dehydro-TxB2 is the major
metabolite of TxB2 which is excreted in man (26).
Urinary 11-dehydro-TxB2 (TxM) was determined by enzyme
immunoassay (Cayman Chemicals, Ann Arbor, MI, USA) using
a sample work-up procedure previously described and validated
by us (27). Two ml of centrifuged urine was diluted 1:2 with
63 mM ammonium bicarbonate buffer pH 8.6 and was incubated
for 3 hours to convert 11-dehydro-TxB2 to its open ring form be-
fore extraction with Bond-elute Certify-II columns (Varian). The
analyte was eluted with 2% formic acid in methanol. The eluate
was evaporated in a vacuum centrifuge, resuspended in buffer
(pH 8.6), and incubated 6 hours before analysis. Urinary TxM is
expressed in relation to urinary creatinine. Data obtained with
this modified assay correlate well with results obtained with GC-
MS analysis of 11-dehydro-TxB2 (27).
Serum TxB2 was determined using commercially available
kits for measurements of TxB2 (Cayman Chemicals). Sample
preparation and analysis were according to instructions from the
manufacturer. Serum samples were only obtained from 9 indi-
viduals, as the assay was established and included in the study at
a later stage.
Urinary 2,3 dinor 6-keto prostaglandin F1 (PGI-M)
Analysis of the 2,3-dinor-6-keto-prostaglandin F1 metabolite of
prostacyclin in urine seems to be the most accurate method for
assessment of systemic (extrarenal) prostacyclin biosynthesis
(28). However, influences of physical activity, day to day vari-
654
ation and large interindividual variation have to be taken into
consideration.
Urinary 2,3-dinor-6-keto prostaglandin F1 (PGI-M) was de-
termined by LC/MS-MS (Rooney C, Harhen B, Fitzgerald D,
Treumann A, manuscript in preparation). Briefly, 2 ml of urine
spiked with 4 ng of the internal standard 2,3-dinor-6-keto prostag-
landin F1-D3 (Biomol, PA, USA) was adjusted to pH 3 and centri-
fuged at 950 x g for 5 minutes before being applied to Empore
solid phase extraction disc cartridges (3M, USA) from which the
PGI-M was eluted with 1 ml of 1% methanol in ethyl acetate. Fol-
lowing liquid-liquid extraction with 50 mM sodium borate (pH
8.0) and derivatization with methoxyamine (Sigma, UK), the puri-
fied and derivatized organic extract was loaded onto an omni-
sphere C18 reversed phase column (3 m x 2.1 x 100 mm, Varian,
UK) and eluted with a linear gradient from 5% to 45% acetonitrile
in 10 mM ammonium acetate. The effluent was analyzed by a
triple quadrupole tandem mass spectrometer (API 3000, Applied
Biosystems). The transition of the ion at m/z 370 to 150.2 was
monitored for the analyte and that of the ion at m/z 373 to 150.2
was monitored for the internal standard, using an optimized colli-
sion energy of 28 V. Concentrations of PGI-M were determined by
calculating the peak height ratios for sample and internal standard.
Data from two out of nine samples in each of two individuals
are missing due to technical problems with the analysis. PGI-M
data from one individual were completely excluded due to varia-
bility related to poor compliance with the restrictions of physical
activity during the experiment. Thus, the PGI-M data presented
are from 1214 individuals.
Platelet function test (PFA-100) and platelet counts
The PFA-100 analyzer (Dade Behring, Germany) was used for
measurements of platelet-related primary hemostasis capacity of
citrate-anticoagulated whole blood under high shear conditions
(29). Closure time (CT) is the time needed to form a platelet plug
occluding the aperture cut in a collagen/epinephrine coated
membrane. Measurements could only be performed in 6 individ-
uals. Platelets were counted by a semiautomatic cell counter
(Cellanalyzer 460, Medonic Solna).
Statistics
Data are presented as mean values SEM. Normally distributed
variables were compared with paired t-tests, whereas bleeding
time data were compared with the Wilcoxon test and Spearmans
rank correlation. Effects of aspirin treatment were analyzed by
2-factor repeated measures ANOVA:s for overall effects, fol-
lowed by appropriate after-testing. The software used was Statis-
tica, version 5.5 for PC. A p value <0.05 was considered signifi-
cant.
Results
Aspirin treatment did not change platelet or leukocyte counts or
mean platelet volume.
Bleeding time
There was no diurnal variation of the bleeding time without treat-
ment (6.30.4 to 6.10.5 min). Ingestion of 37.5 mg aspirin for
10 days increased the bleeding time to 10.41.0 min (p<0.001)
 Perneby, et al.: Dose- and time-dependent aspirin effects

Figure 1: Comparison of AA-induced aggregation (final concentrations 0.2, 0.5 and 1.0 mM) in whole blood anticoagulated with hi-
rudin after different doses of aspirin. Blood samples were taken 1.52, 46 or 24 hours after the latest ingested dose. Low dose is 37.5 mg/day
for 10 days (filled symbols). High-dose is 320 mg/day for 7 days and a single dose of 640 mg taken immediately after the 1.52 hour sample (open
symbols). Thus, 6 h is 4 h after a cumulative dose of 960 mg aspirin in the high dose condition.

1.52 h after the last dose with no significant variation 24 h
thereafter. The bleeding time was 11.71.3 min (p<0.001) 1.52
h after 320 mg aspirin. It was 12.22.3 min 4 h after 960 mg as-
pirin (p<0.001), and similar after 24 hours (12.31.4 min). Eight
subjects had <60% increments of bleeding time after 37.5 mg as-
pirin, but similarly low increments after 320 mg. Bleeding time
responses to 37.5 and 320 mg aspirin were correlated (r=0.53;
p=0.041; Spearman test).
Whole blood aggregometry
There was no dose-response relationship for platelet aggregation
induced by 0.21 mM AA in whole blood, and no significant
diurnal variability without aspirin treatment. However, the re-
sponse became time- and dose-dependent during aspirin treat-
ment (Fig. 1). Thus, 37.5 mg aspirin suppressed aggregation
mainly at low concentrations of AA; the inhibitory effects of
low-dose aspirin were significantly attenuated after 24 h com-
pared to 6 h (p<0.01) (Fig. 1). When using the lowest concen-
tration of AA (0.2 mM), platelet aggregation was almost com-
pletely inhibited after 320 and 960 mg aspirin (p<0.001 for
both), but there was less complete inhibition with higher concen-
trations of AA, and significant recovery 24 h after high-dose as-
pirin (Fig. 1). In fact, the aggregatory response to 1 mM AA was
almost normalized 24 h after 960 mg aspirin (p<0.001 compared
to 4 h value).
Treatment with 37.5 mg aspirin slightly attenuated platelet ag-
gregation in hirudinized whole blood induced by 1 M collagen, as
the impedance decreased by 12.5 5.1% (from 19.80.9 to
17.11.2 Ohm; p<0.05). The inhibition was more pronounced
after 320 mg aspirin (36.45.4%; p<0.001), but increasing the
dose to 960 mg caused no further inhibition. Higher collagen con-
centrations (3 and 5M) yielded nearly maximal platelet aggre-
gation in hirudinized whole blood with all aspirin dosages.
Aggregometry in platelet rich plasma (PRP)
AA (1.0 mM) yielded 891.3% aggregation in hirudinized PRP
before treatment. All doses of aspirin (37.5 mg, 320 mg and
960 mg) inhibited AA-induced platelet aggregation almost com-
pletely (to less than 1%), indicating good compliance and effec-
tive platelet COX inhibition. Three individuals had small aggre-
gatory responses (35 %) to AA after 37.5 mg aspirin, which
were abolished after 320 and 960 mg.
The platelet sensitivity to ADP decreased after 37.5 mg as-
pirin, as the EC50 for ADP increased by 6511% (from
1.260.21 to 1.970.31 M; p=0.01). Higher dosages decreased
the responsiveness to ADP less (EC50:s increased by 239% and
2613% after 320 and 960 mg, respectively). One individual
showed extreme increases in the EC50 for ADP (e.g. from 1.3 to
275 M after 960 mg), and was excluded from the analysis.
There was no significant time dependence of aspirin effects on
ADP induced aggregation in PRP.
Serum thromboxane B2 and urinary 11-dehydro-
thromboxane B2 (TxM)
There was no significant diurnal variability of the urinary TxM
excretion. The nocturnal TxM excretion was reduced by
74.02.0% (p<0.001) after 37.5 mg aspirin, and there was was
no difference at <6 compared to <24 h after the last dose. After
320 mg aspirin the degree of inhibition was 82.01.8% (p<0.001
vs. baseline, and 37.5 mg). Four hours after 960 mg aspirin the
TxM excretion was reduced by 83.92.8% (p<0.001 vs. base-
line, and 37.5 mg), but there was significant recovery after <24
hours to 78.51.4% inhibition (TxM excretion increased from
9.41.1 to 14.81.5 ng/mmol creatinine; p<0.001) (Fig. 2A).
The TxB2 production in serum was inhibited by 98.50.7%
(n=8) after 37.5 mg aspirin, and by 98.90.5% (n=9) after 320 mg
aspirin.

655
 Perneby, et al.: Dose- and time-dependent aspirin effects
Figure 2: Urinary excretion of TxM (A) and PGI-M (B) express-
ed as ng/mmol creatinine after different doses of aspirin. Night
urine (open columns) was sampled the night following the last ingested
dose at each dose level; the untreated baseline data are the means of
two consecutive nights (before and after no treatment). Day urine
(dashed columns ) was sampled 46 hours after the ingestion of 37.5 mg
and the cumulative 960 mg dose.
Figure 3: Comparison of platelet inhibition afforded by aspirin
at different dosages (37.5 960 mg) using different platelet
function assays. Illustrated in the graph are: platelet aggregation in
whole blood (WB) using arachidonic acid (AA) or collagen (Coll.);
platelet aggregation in platelet rich plasma (PRP) using AA as agonist;
serum (S) TxB2; and thromboxane (TxM) metabolite excretion in urine.
Urinary 2,3-dinor-6-keto prostaglandin F1 (PGI-M)
The urinary PGI-M excretion was 16.51.6 and 14.61.0 ng/
mmol creatinine in nocturnal and 1 PM samples, respectively,
without treatment. The nocturnal PGI-M excretion decreased by
257% (p<0.05) after 37.5 mg aspirin, and by 2814%
(p=0.066) after 320 mg (Fig. 2B). PGI-M excretion during the
day tended to decrease (by 188%; p=0.055) with 37.5 mg as-
pirin, and was reduced by 2610% (p<0.05) 4 h after 960 mg as-
pirin, but recovered (from 10.61.5 to 13.92.0 ng/mmol creati-
nine; p<0.05) in the ensuing nocturnal sample (Fig. 2B).
Closure time, PFA-100
The closure time with the collagen/epinephrine cartridge tended
to increase after 10 days of treatment with 37.5 mg aspirin (from
99.310.4 to 153.632 s; ns), and was significantly prolonged
(to 179.332.2 s; p<0.05) after 320 mg aspirin. There was a sig-
nificant correlation between closure time and bleeding time after
aspirin (r=0.78; p<0.01).
Comparison of methods and correlations
Figure 3 compares the inhibitory effects of aspirin treatment at
various dosages on platelet related variables in the study. It may
be seen that the parameter most sensitive to inhibition by aspirin
was AA induced aggregation in PRP, followed by serum TxB2.
TxM excretion was less markedly, but dose-dependently reduced
by aspirin treatment. A dose related but weak inhibition of col-
lagen induced aggreation in hirudinized whole blood was also
found with the lowest collagen concentration, and some in-
hibition of PGI-M excretion.
Changes in PGI-M excretion were not correlated to changes
in the platelet related variables tested in this study (bleeding
time, platelet aggregation in PRP or whole blood, or TxM excre-
tion) during aspirin treatment at any dose level. Platelet aggre-
gation in whole blood stimulated by low AA concentrations was
correlated to serum TxB2 levels after 37.5 mg aspirin (e.g.
r=0.73, p<0.05, at 0.2 mM AA); correlations disappeared with
higher AA concentrations and aspirin dosages.
Discussion
The present study shows that once daily dosing of aspirin may be
insufficient even though there is nearly complete inhibition of
platelet COX activity early after dosing. Aspirin treatment sup-
pressed TxB2 in serum by more than 98% and abolished AA-
induced platelet aggregation in PRP already at a very low dose
level (37.5 mg daily), but there was significant recovery of pla-
telet function 24 hours after dosing even with high dosages of as-
pirin. Thus there was incomplete suppression of TxM excretion
in urine, and AA-induced platelet aggregation could still occur in
whole blood. Furthermore, we found that treatment with 37.5 mg
aspirin decreased the sensitivity to stimulation by ADP in pla-
telet rich plasma, but that this effect was less pronounced (not
significant) after higher doses of aspirin, in accordance with our
earlier results (30). Bleeding time increased similarly after all
dosages of aspirin, and was not related to changes in the urinary
excretion of PGI-M. Thus, the effectiveness of aspirin is highly
dependent on the method used to evaluate its antiplatelet effect.
656
 Perneby, et al.: Dose- and time-dependent aspirin effects
When platelet function is investigated in vitro, the anticoagu-
lant used may artifactually modify platelet responses. Sodium ci-
trate is the most commonly used anticoagulant, which lowers
extracellular calcium and enhances the antiplatelet effects of as-
pirin (2224). We used hirudin (a highly selective thrombin in-
hibitor) to preserve normal ionized calcium levels, and found li-
mited inhibition by aspirin of collagen- and AA-induced platelet
aggregation. Thus, AA-induced platelet aggregation occurred
even after 960 mg aspirin if the AA concentration was high
enough. During low-dose aspirin treatment (37.5 mg/day) pla-
telet aggregation in hirudinized whole blood was virtually unaf-
fected when stimulated by 1.0 mM AA. Collagen induced pla-
telet aggregation in hirudinized whole blood was minimally af-
fected by aspirin treatment.
A problem with in vitro studies of platelet function in PRP is
that centrifugation of the blood may remove the largest and most
active platelets along with the red and white blood cells, which
are known to influence platelet behaviour (31, 32). Responses to
antiplatelet agents like aspirin may also be influenced by the re-
moval of other blood cells. For example, the antiplatelet efficacy
of low-dose aspirin (50 mg/day) is attenuated by the presence of
erythrocytes (31). Leukocytes can, via transcellular prostanoid
metabolism, influence platelets even if the platelet COX is in-
hibited (33, 34), and leukocyte-platelet cross-talk may contribute
to platelet responsiveness in whole blood (32). The less efficient
antiaggregatory effect of aspirin treatment in whole blood com-
pared to PRP may thus reflect erythrocyte enhancement, extra-
platelet sources of PGH2, and/or transcellular formation of TxA2
despite efficient inhibition of COX in platelets.
This study illustrates the importance of distinguishing be-
tween inhibition of thromboxane synthesis and inhibition of pla-
telet activation, as reflected by, e.g., platelet aggregation, which
may be more clinically relevant. Thus, we found that the very low
37.5 mg dose of aspirin, which appears to be subtherapeutic (1,
2), effectively inhibited platelet-dependent thromboxane
formation with very modest effects on aggregation in whole
blood. Nucleated cells are capable of regenerating the COX-1
enzyme after a couple of hours and provide PGH2 to platelets to
bypass the inhibition of platelet COX-1 (34, 35). There was less
pronounced inhibition of TxM excretion compared to serum
TxB2 generation, and partial recovery of TxM excretion at the
end of the dosing interval, indicating that aspirin-insensitive
(presumably non-platelet) thromboxane synthesis had occurred
in vivo.
Several studies have shown that aspirin effects may be li-
mited, and that some patients at risk may not benefit from aspirin
protection against cardiovascular events. There are large inter-
individual differences in various responses to aspirin treatment
(3, 4). However, the phenomenon which is often named aspirin
resistance should not be confused with treatment failure,
which may be related to aspirin insensitive disease mechanisms
and poor compliance rather than true or biochemical aspirin
resistance (24). Eikelboom, et al. found that incomplete sup-
pression of thromboxane generation (measured as urinary TxM)
was associated with an increased risk of suffering MI, stroke or
cardiovascular death (36). The aspirin dosages prescribed were
80325 mg/day, but the timing of sampling in relation to dosing,
657
and patient compliance are not known. We found that 960 mg as-
pirin resulted in more pronounced inhibition of urinary TxM
than 37.5 mg some 5 hours after ingestion, but that this differ-
ence had decreased 24 hours after ingestion. Chamorro, et al.
showed that non-responders to 300 mg aspirin (defined as pa-
tients with recurrent stroke within 12 months despite aspirin
treatment) had less effectively inhibited AA dependent platelet
aggregation than patients without events (37). However, a higher
dose of aspirin (600 mg/day) resulted in similar inhibition of re-
sponses to AA in the two groups (37). This is in accordance with
our results showing more effective inhibition of AA-induced pla-
telet aggregation in whole blood with higher doses of aspirin,
and that aggregation was correlated to serum TxB2 levels at a low
aspirin dosage. Thus, the ability of aspirin to block platelet
COX-1 (true aspirin resistance) should be clearly distin-
guished from treatment failure when discussing aspirin resis-
tance (4), and the time-dependence of inhibitory effects should
be considered in this context.
We found slight reductions of PGI-M excretion after aspirin
treatment at all dose levels. After 960 mg aspirin there was a
further reduction of PGI-M excretion <4 h after dosing, but a sig-
nificant recovery after <24 hours. Vesterqvist, et al. showed that
the PGI-M excretion in urine was reduced during only three
hours after ingestion of high-dose aspirin (38). Others have
found different degrees of inhibition of PGI-M excretion after
aspirin treatment (17, 39). Thus, findings differ between studies,
but it appears that aspirin treatment may reduce prostacyclin pro-
duction even at low dosages. We found no relationship between
changes in PGI-M excretion and any platelet function variable.
The relationship between surrogate markers for platelet func-
tion and clinical events is a critical issue determining dosing
regimens. There is no single or accepted platelet function test
that is optimal for the monitoring of aspirin treatment. Using uri-
nary TxM and whole blood aggregometry in combination may
be the most suitable approach for the evaluation of in vivo effects
of aspirin treatment, but AA-induced aggregation in PRP may
provide the best assessment of platelet COX inhibition. How-
ever, these approaches have not been evaluated in relation to
clinical outcomes during aspirin treatment. Furthermore, the
presently observed recovery of COX-dependent platelet aggre-
gation between doses might need consideration regarding the do-
sage interval; perhaps twice daily treatment might be more effi-
cient in patients with increased platelet turnover. Our findings of
dose- and time-dependent recovery of platelet function within 24
h after dosing of aspirin are, e.g., of relevance for the inability of
100 mg aspirin every other day to reduce cardiac events in the
Womens Health Study (40), as such a dose regimen may provide
incomplete COX-1 inhibition during at least part of the dosing
interval.
In conclusion, this study illustrates the complexity involved
in evaluating the effectiveness of aspirin therapy, as different ef-
fect parameters yield very different results. Importantly, aspirin
provides only weak inhibition of platelet aggregation in whole
blood with normal calcium levels, despite effective platelet COX
inhibition. Furthermore, there is recovery of platelet function
within a normal 24 h dosing interval. The optimal dose of aspirin
has been discussed for several years. Defining the optimal as-
 Perneby, et al.: Dose- and time-dependent aspirin effects
pirin dose (and the dosing interval) for each patient group, and
establishing well documented means of monitoring individual
patients may further improve the secondary prevention of athero-
thrombotic disease by aspirin treatment.
References
1. Antithrombotic Trialists Collaboration. Collabor-
ative meta-analysis of randomised trials of antiplatelet
therapy for prevention of death, myocardial infarction,
and stroke in high risk patients. BMJ 2002; 324: 718.
2. Patrono C, Coller B, Fitzgerald GA, et al: Platelet-
active drugs: the relationships among dose, effective-
ness, and side-effects. Chest 2004; 126: 234S-264S.
3. Sanderson S, Emery J, Baglin T, et al. Narrative re-
view: aspirin resistance and its clinical implications.
Ann Intern Med 2005; 142: 37080.
4. Cattaneo M. Aspirin and clopidogrel efficacy,
safety, and the issue of drug resistance. Arterioscl
Thromb Vasc Biol 2004; 24: 19807.
5. Rihal CS, Flather M, Hirsch J, et al. Advances in
antithrombotic drug therapy for coronary artery dis-
ease. Eur Heart J 1995; 16 (Suppl D): 1021.
6. Reilly IAG, Fitzgerald GA. Inhibition of thrombox-
ane formation in vivo and ex vivo: implications for ther-
apy with platelet inhibitory drugs. Blood 1987; 69:
1806.
7. Di Minno G, Silver MJ, Murphy S. Monitoring the
entry of new platelets into the circulation after inges-
tion of aspirin. Blood 1983; 61: 10815.
8. Patrignani P, Filabozzi P, Patrono C. Selective
cumulative inhibition of platelet thromboxane produc-
tion by low-dose aspirin in healthy subjects. J Clin In-
vest 1982; 69: 136672.
9. Quinn MJ, Aronow HD, Califf RM, et al. Aspirin
dose and six-month outcome after an acute coronary
syndrome. J Am Coll Cardiol 2004; 43: 9728.
10. Peters RJG, Mehta SR, Fox KAA, et al. Effects of
aspirin dose when used alone or in combination with
clopidogrel in patients with acute coronary syndromes;
observations from the Clopidogrel in Unstable angina
to prevent Recurrent Events (CURE) study. Circulation
2003; 108: 16827.
11. Juul-Mller S, Edvardsson N, Jahnmatz B, et al.
Double-blind trial of aspirin in primary prevention of
myocardial infarction in patients with stable chronic
angina pectoris. Lancet 1992; 340: 14215.
12. The RISC group. Risk of myocardial infraction and
death during treatment with low dose aspirin and intra-
venous heparin in men with unstable coronary artery
disease. Lancet 1990; 336: 82730.
13. Fitzgerald DJ, Roy L, Catella F, et al. Platelet acti-
vation in unstable coronary disease. N Engl J Med
1986; 315: 9839.
14. Hamsten A, Svensson J, Waldius G, et al. Shortened
megakaryocyte-platelet regeneration time in young
survivors of myocardial infarction. Am Heart J 1985;
110: 115460.
15. Wallenburg HCS, van Kessel PH. Platelet life span
in pregnancies resulting in small-for gestational age in-
fants. Am J Obstet Gynecol 1979; 134: 73942.
Acknowledgements
The authors are grateful for the expert technical assistance with platelet re-
lated parameters by Maud Daleskog, Maj-Christina Johansson and Pia Hil-
lelsson (Stockholm), and for discussions and advice about prostacyclin
metabolite measurements toAchimTreumann and Brendan Harhen (Dublin).
16. Tschoepe D. The activated megacaryocyte-platelet-
system in vascular disease: focus on diabetes. Sem
Thromb Hemost 1995; 21: 15260.
17. Clarke RJ, Mayo G, Price P, et al. Supression of
thromboxane A2 but not of systemic prostacyclin by
controlled-release aspirin. N Engl J Med 1991; 325:
113741.
18. Fitzgerald GA, Oates JA, Hawiger J, et al. Endogen-
ous biosynthesis of prostacyclin and thromboxane and
platelet function during chronic administration of as-
pirin in man. J Clin Invest 1983; 71: 67688.
19. Kobayashi T, Tahara Y, Matsumoto M, et al. Roles
of thromboxane A2 and prostacyclin in the devel-
opment of atherosclerosis in apoE-deficient mice. J
Clin Invest 2004; 114: 78494.
20. Antman EM, DeMets D, Loscalzo J. Cyclooxyge-
nase inhibition and cardiovascular risk. Circulation
2005; 112: 75970.
21. Fitzgerald GA. Coxibs and cardiovascular disease.
N Engl J Med 2004; 351: 170911.
22. Packham MA, Kinlough-Rathbone RL, Mustard
JF. Thromboxane A2 causes feedback amplification in-
volving extensive thromboxane A2 formation on close
contact of human platelets in media with a low concen-
tration of ionized calcium. Blood 1987; 70: 64751.
23. Bretschneider E, Glusa E, Schrr K. ADP-, PAF-
and adrenaline-induced platelet aggregation and
thromboxane formation are not affected by a thromb-
oxane receptor antagonist at physiological external
Ca++ concentrations. Thromb Res 1994; 75: 23342.
24. Walln N, Ladjevardi M, Albert J, et al. Influence of
different anticoagulants on platelet aggregation in
whole blood; a comparison between citrate, low mol-
ecular mass heparin and hirudin. Thromb Res 1997; 87:
1517.
25. Albert J, Norman M, Walln NH, et al. Inhaled ni-
tric oxide does not influence bleeding time or platelet
function in healthy volunteers. Eur J Clin Invest 1999;
29: 9539.
26. Catella F, Healy D, Lawson J A, et al. 11-dehydro-
thromboxane B2: a quantitave index of thromboxane A2
formation in the human circulation. Proc Natl Acad Sci
1986; 83: 58615.
27. Perneby C, Granstrm E, Beck O, et al. Optimiz-
ation of an enzyme immunoassay for 11-dehydro-
thromboxane B2 in urine: comparison with GC-MS.
Thromb Res 1999; 96: 42736.
28. Fitzgerald GA, Pedersen AK, Patrono C. Analysis
of prostacyclin and thromboxane biosynthesis in car-
diovascular disease. Circulation 1983; 67: 11747.
29. Harrison P, Robinson MSC, Mackie IJ, et al. Per-
formance of the platelet function analyser PFA-100 in
testing abnormalities of primary haemostasis. Blood
Coag Fibrinol 1999; 10: 2531.
658
30. Walln NH, Held C, Rehnqvist N, et al. Impact of
treatment with acetylsalicylic acid on the proaggrega-
tory effects of adrenaline in vitro in patients with stable
angina pectoris: influence of the anticoagulant. Clin
Sci (Lond) 1993; 85: 57783.
31. Valles J, Santos T, Aznar J, et al. Erythrocyte pro-
motion of platelet reactivity decreases the effectiveness
of aspirin as an antithrombotic therapeutic modality.
Circulation 1998; 97: 3505.
32. Li N, Hu H, Lindqvist M, et al. Platelet-leukocyte
cross talk in whole blood. Arterioscler Thromb Vasc
Biol 2000; 20: 27028.
33. Camacho M, Vila L. Transcellular formation of
thromboxane A2 in mixed incubations of endothelial
cells and aspirin treated platelets strongly depends on
the prostaglandin I-synthase activity. Thromb Res
2000; 99: 15564.
34. Karim S, Habib A, Lvy-Toledano S, et al. Cyclo-
oxygenase-1 and 2 of endothelial cells utilize exogen-
ous or endogenous arachidonic acid for transcellular
production of thromboxane. J Biol Chem 1996; 271:
120428.
35. Cipollone F, Patrignani P, Greco A, et al. Differen-
tial suppression of thromboxane biosynthesis by indo-
bufen and aspirin in patients with unstable angina. Cir-
culation 2001; 96: 110919.
36. Eikelboom JW, Hirsh J, Weitz JI, et al. Aspirin-re-
sistant thromboxane biosynthesis and the risk of myoc-
ardial infarction, stroke or cardiovascular death in pa-
tients at high risk for cardiovascular events. Circulation
2002; 105: 16505.
37. Chamorro A, Escolar G, Revilla M, et al. Ex vivo re-
sponse to aspirin differs in stroke patient with single or
recurrent events: a pilot study. J Neurol Sci 1999; 171:
1104.
38. Vesterqvist O. Rapid recovery of in vivo prostacyc-
lin formation after inhibition by aspirin. Eur J Clin
Pharmacol 1986; 30: 6973.
39. Toghi H, Konno S, Tamura K, et al. Effects of low-
to-high doses of aspirin on platelet aggregability and
metabolites of thromboxane A2 and prostacyclin.
Stroke 1992; 23: 14003.
40. Ridker PM, Cook N, I-Min L, et al. A randomised
trial of low-dose aspirin in the primary prevention of
cardiovascular disease in women. N Engl J Med 2005;
352: 1293304.