Journal of Thrombosis and Haemostasis, 10: 521–528
ORIGINAL ARTICLE
Reversal of the anti-platelet effects of aspirin and clopidogrel
C . L I , * J . H I R S H ,   C . X I E , à M . A . J O H N S T O N § and J . W . E I K E L B O O M –
*Department of Cardiology, First Affiliated Hospital of Nanjing Medical University, Nanjing, China;  Department of Medicine, MacMaster
University, Hamilton, ON; àPopulation Health Research Institute, MacMaster University, Hamilton, ON; §Hemostasis Reference Laboratory,
Henderson Hospital, Hamilton, ON; and –Thrombosis service, Hamilton General Hospital, Hamilton, ON, Canada
To cite this article: Li C, Hirsh J, Xie C, Johnston MA, Eikelboom JW. Reversal of the anti-platelet effects of aspirin and clopidogrel. J Thromb
Haemost 2012; 10: 521–28.
Summary. Background: Guidelines recommend stopping aspi-
rin and clopidogrel 7 to 10 days before surgery to allow time for
replacement of permanently inhibited platelets by newly
released uninhibited platelets. Objectives: The purpose of the
present study was to determine the rate of offset of the anti-
platelet effects of aspirin and clopidogrel after stopping
treatment and the proportion of untreated donor platelets that
are required to reverse their anti-platelet effects. Meth-
ods: Cohort 1 consisted of 15 healthy subjects who received
aspirin 81 mg day)1 or clopidogrel 75 mg day)1 for 7 days and
underwent serial blood sampling until platelet function testing
results normalized. Cohort 2 consisted of 36 healthy subjects
who received aspirin 325 mg day)1, clopidogrel 75 mg day)1,
aspirin 81 mg day)1 plus clopidogrel 75 mg day)1 or no
treatment for 7 days and underwent a single blood sam-
pling. Results: In cohort 1, arachidonic acid (AA)-induced
light transmission aggregation (LTA) returned to baseline levels
in all subjects within 4 days of stopping aspirin, coinciding with
the partial recovery of plasma thromboxane B2 concentrations.
ADP-induced LTA did not return to baseline levels until
10 days after stopping clopidogrel. In cohort 2, AA-induced
LTA in patient treated with aspirin reached control levels after
mixing with 30% untreated donor platelets whereas ADP-
induced LTA in patients treated with clopidogrel reached
control levels only after the addition of 90% or more donor
platelets. Conclusions: Platelet aggregation recovers within
4 days of stopping aspirin but clopidogrel must be stopped
for 10 days to achieve a normal aggregatory response.
Keywords: aspirin, clopidogrel, peri-operative treatment, plate-
let aggregation.
Correspondence: Chunjian Li, Department of Cardiology, First
Affiliated Hospital of Nanjing Medical University, 300, Guangzhou
Road, Nanjing 210029, China.
Tel.: +86 25 83718836 6018; fax: +86 25 83674380.
E-mail: lijay@njmu.edu.cn
Received 13 November 2011, accepted 14 January 2012
 2012 International Society on Thrombosis and Haemostasis
DOI: 10.1111/j.1538-7836.2012.04641.
Introduction
Aspirin and clopidogrel are effective anti-platelet drugs for the
prevention of cardiovascular events but are often stopped
before surgery because they can cause bleeding [1,2]. The
timing of stopping anti-platelet treatment before procedures is
important because prolonged discontinuation can increase the
risk of ischemic events whereas delayed discontinuation can
increase the risk of bleeding [3]. The American College of Chest
Physicians and the American College of Cardiology/American
Heart Association guidelines recommend stopping aspirin and
clopidogrel 7 to 10 days before surgery [4,5] based on their
irreversible platelet-inhibitory effects and the premise that
recovery of hemostatic function after stopping treatment
requires the replacement of irreversibly inhibited platelets by
newly synthesized uninhibited platelets released from the bone
marrow. An alternative approach to stopping anti-platelet
therapy before surgery might be to transfuse patients with
uninhibited donor platelets [6,7]. However, it is uncertain
whether platelet transfusion can fully reverse the anti-platelet
effects of aspirin and clopidogrel and, if so, the proportion of
non-treated platelets required to reverse their effects.
The first objective of the present study was to compare the
time course of recovery of platelet function after stopping
aspirin and clopidogrel treatment. The second objective was to
determine the proportion of untreated donor platelets required
to fully reverse the platelet inhibitory effects of aspirin and
clopidogrel.
Methods
Study population
We designed two cohort studies, and enrolled 51 healthy
subjects aged at least 18 years who did not have a history of
cardiovascular disease and were not taking anti-platelet
therapy from January 2009 to July 2010 in Hamilton General
Hospital, Hamilton, Canada. Exclusion criteria were (i) allergy
or intolerance to aspirin or clopidogrel; (ii) subjects at a high
risk of bleeding (e.g. thrombocytopenia [platelet count
< 150 · 109 L)1]), known bleeding diathesis, active peptic
522 C. Li et al

Bleeding time
Aspirin: 81mg qd×7days Withdrawal of aspirin Days 1,3,5,8
(n = 7)
Platelet function testing
RPFA
Cohort 1
Days 1,2,3,4,5

Withdrawal of clopidogrel
Clopidogrel: 75mg qd×7days LTA (PLAA, PLADP, PLCOLL)
(n = 8) Platelet function testing Days 1,2,3,4,5,6,8,10,12,14

Plasma TXB2
Days 1,2,3,4,5,6,8,10,12,14
Healthy
Volunteers 325mg qd×7days Mixing
Aspirin:
(n = 51) Serum TXB2
(n = 5) studies
Days 1,2,3,4,5,6,8,10,12,14
Platelet
Control: no anti-platelet drug ×7days
function
(n = 5) Urinary 11-dH-TXB2
testing
Days 1,2,3,4,5,6,8,10,12,14

Clopidogrel: 75mg qd×7days

(n = 7) Mixing
Cohort 2 studies LTA (PLAA, PLADP, PLCOLL)
Control: no anti-platelet drug ×7days Platelet
(n = 7) function
testing
Aspirin: 81mg qd +
Clopidogrel: 75mg qd×7days Mixing Plasma TXB2
(n = 6) studies
Platelet
Control: no anti-platelet drug ×7days
function
(n = 6)
testing

Fig. 1. Study design.

ulcer disease; (iii) cigarette smoking; (iv) diabetes; (v) preg-
nancy; and (vi) consumption of drugs within the preceding
week that potentially could interfere with the anti-platelet
effects of aspirin (e.g. non-steroidal anti-inflammatory drugs)
or clopidogrel (e.g. proton pump inhibitors). The study design
is shown in Fig. 1.
The present study complies with the Declaration of Helsinki,
and approved by the Research Ethics Board of Hamilton
Health Science Corporation. All study subjects provided
written informed consent.
Cohort 1: recovery of platelet function after stopping aspirin
and clopidogrel
Study treatments Fifteen subjects were treated with either
aspirin (PharmaScience Inc., Montreal, QC, Canada) 81 mg
(n = 7) or clopidogrel (Sanofi-aventis, Paris, France) 75 mg
(n = 8) taken once daily at 08.00 hours for 7 days.
into a sterile tube simultaneously with the blood sampling, then
transferred into a 2-mL tube and frozen at )80 C until further
analysis.
Bleeding time For subjects treated with aspirin, bleeding
time was measured using the Surgicutt device (International
Technidyne Corporation, Edison, NJ, USA) by a single
operator (C. Li) on days 1, 3, 5, and 8 after the last dose of
aspirin. Briefly, the volunteerÕs arm was placed on a table with
the volar surface exposed. A blood pressure cuff was placed
around the upper arm and inflated to 40 mmHg for the duration
of the test. An area on the volar surface of the forearm was
swabbed with alcohol and allowed to air dry. A single incision
was made parallel to and 5 cm below the antecubital crease
using the Surgicutt device and the blood was blotted with filter
paper every 30 s until bleeding stopped. Care was taken during
the blotting not to touch the edge of the clot and wound. The
time from the incision to the end of bleeding was measured.

Blood and urine collection Venous blood samples were
collected by venipuncture into two 4.5-mL draw vacutainer
tubes (Becton, Dickinson and Company, Mississauga,
Canada) containing 0.105 M buffered sodium citrate (3.2%),
one 2-mL draw Greiner Bio-One vacuette (Accumetrics, San
Diego, CA, USA) containing 0.105 M buffered sodium citrate
(3.2%) and one 4-mL draw BD vacutainer tube without
additive immediately before starting the study drug and on
days 1 to 6, 8, 10, 12 and 14 after the last dose of the study drug.
In patients treated with aspirin, 10 mL of urine was collected
Laboratory analyzes Light transmittance aggregation
(LTA) was performed in platelet-rich plasma using
arachidonic acid (AA), ADP and collagen as agonists. All
agonists were obtained from Chronolog (Havertown, PA,
USA). In aspirin-treated subjects, we also assessed platelet
function using the rapid platelet function assay (RPFA) aspirin
cartridge (Accumetrics, San Diego, CA, USA) and we
measured serum thromboxane B2 (TXB2), plasma TXB2 in
platelet-rich plasma after AA aggregation and urinary 11-
dehydro thromboxane B2 (11-dH-TXB2).

 2012 International Society on Thrombosis and Haemostasis


Cohort 2: reversal of the anti-platelet effects of aspirin and
clopidogrel using donor platelets
Study treatments Eighteen healthy subjects received 7 days
of treatment with aspirin (NovoPharm Limited, Scarborough,
ON, Canada) 325 mg day)1 (n = 5); clopidogrel 75 mg day)1
(n = 7); or the combination of aspirin 81 mg day)1 and
clopidogrel 75 mg day)1 (n = 6) taken at 20.00 hours each
day. Another 18 healthy subjects received no anti-platelet
therapy and served as controls.
Blood collection Venous blood samples were collected by
venipuncture into two 4.5-mL draw BD vacutainer tubes
containing 0.105 M buffered sodium citrate (3.2%) at
08.00 hours in the morning after the last dose of the study drug.
Laboratory analyzes Platelet-rich plasma (PRP) from
subjects who received anti-platelet therapy was mixed with
increasing proportions of PRP from untreated controls, with
one untreated subject serving as the control for one treated
subject. The proportion of control platelets mixed with
inhibited platelets was calculated based on platelet numbers,
starting at 10% and increasing by 10% increments. AA, ADP
and collagen-induced platelet aggregation were measured
before and after mixing. Plasma TXB2 concentrations were
measured in PRP from subjects treated with aspirin and in
those treated with the combination of aspirin and clopidogrel.
Laboratory procedures
Sample preparation PRP and platelet-poor plasma (PPP)
were prepared shortly after blood collection by spinning the
sample at 200 g for 8 min. The PRP was carefully removed and
the remaining blood centrifuged at 2465 g for 10 min to obtain
PPP. The centrifuge temperature was maintained at 22 C.
Platelet counts were adjusted by the addition of PPP to the
PRP to achieve a count of 250 · 109 L)1. Platelet aggregation
studies were completed within 3 h of preparation of PRP.
Venous blood was allowed to clot in a warmed additive free
BD vacutainer and then incubated at 37 C for 1 h. The blood
was spun at 2000 g for 15 min and the serum was removed from
thecellularelementsandstoredat)80 CuntilassayedforTXB2.
Platelet aggregation Platelet aggregation studies were
performed using a Chronolog 560VS/490-2D aggregometer
(Chronolog Corporation, Havertown, PA, USA).
Immediately after preparation of PRP, 500 lL was
transferred into each of the three test tubes, with 500 lL
PPP set as a control. After 2 min of warming, PRP and PPP
were put in testing places and were warmed for a further
2 min. Final concentrations of agonists were: AA 1 mM, ADP
5 lM and collagen 1 lg mL)1, respectively. The aggregated
plasma samples were stored at )80 C until assayed for TXB2.
Rapid platelet function assay The rapid platelet function
assay (RPFA) is a semi-quantitative platelet function test which
measures turbidimetric platelet aggregation as an increase of
 2012 International Society on Thrombosis and Haemostasis
Reversal of aspirin and clopidogrel effects 523
light transmittance as a result of platelet agglutination. The test
was performed using the VerifyNow AspirinWorks assay
(Accumetrics) according to the manufacturerÕs instructions.
Briefly, at the instrument prompt, a cartridge was inserted into
the instrument with the needleÕs protection sheath removed.
The Greiner Bio-one vacuette containing 2 mL of citrated
venous blood was gently mixed and inserted onto the needle of
the cartridge. Results were reported in aspirin reaction units
(ARU), a function of the rate at which platelets aggregate. A
result of < 550 ARU indicates an ÔadequateÕ response to
aspirin, whereas ‡ 550 ARU indicates an ÔinadequateÕ or low
response to aspirin.
Serum and plasma TXB2 Serums and plasma TXB2 was
analyzed using the Amersham Biotrak Thromboxane B2
ELISA kit (GE Healthcare, Montreal, Canada) according to
the manufacturerÕs instructions. In-house controls were tested
with each batch.
Urinary 11-dH-TXB2 Urinary 11-dH-TXB2 was measured
using a commercially available AspirinWorks enzyme
immunoassay kit (Corgenix Inc., Broomfield, CO, USA)
according to the manufacturerÕs instructions. This assay has
an inter-assay and an intra-assay coefficient of variation of
5.8% and 3.7%, respectively. Kit urine controls with assigned
values for 11-dH-TXB2 were tested with each batch.
Statistical analyzes
Results of categorical variables are presented as number and
percentage and results of continuous variables as
means ± standard deviations or median (interquartile range
[IQR]) where appropriate. The statistical significance of
differences in means and medians was analyzed using two-
way analysis of variance or the Kruskal–Wallis rank test, and
DunnettÕs multiple comparison tests. The three-parameter
logistic function was adopted to explore the relation between
AA-induced platelet aggregation (PLAA) and plasma TXB2
concentration after stopping aspirin treatment. Recovery ratios
of AA- and ADP-induced aggregation ([Xn-X1/baseline-
X1] · 100%; Xn represent PLAA, PLADP of day 2–6, 8, 10
and 12 after aspirin and clopidogrel withdrawal; X1 represent
AA- or ADP-induced aggregation on day 1 after aspirin and
clopidogrel withdrawal; baseline represent AA- or ADP-
induced platelet aggregation before taking aspirin and
clopidogrel) after aspirin and clopidogrel withdrawal were
compared using two-factor repeated measurements ANOVA.
Similarly, the recovery curves of plasma TXB2, serum TXB2
and urinary 11-dH-TXB2 after withdrawal of aspirin were
compared with the recovery of AA-induced platelet aggrega-
tion (PLAA) after withdrawal of aspirin and with the recovery
of ADP-induced platelet aggregation (PLADP) after the with-
drawal of clopidogrel. All analyzes were performed using SAS
9.1.3 (SAS Institute Inc., Cary, NC, USA) and STATA11
(StataCorp LP, College Station, TX, USA). A two-sided
P-value < 0.05 was considered statistically significant.
524 C. Li et al

Table 1 Subject basic characteristics

Cohort 1 Cohort 2

Aspirin Clopidogrel Aspirin Control Clopidogrel Control Combination Control

n 7 8 5 5 7 7 6 6
Males (%) 3 (43) 6 (75) 3 (60) 2 (40) 5 (71) 3 (43) 2 (33) 2 (33)
Age (years) 39 ± 7 25 ± 1 36 ± 10 36 ± 9 38 ± 9 35 ± 9 43 ± 8 34 ± 8
Weight (kg) 77.4 ± 13.5 63.1 ± 6.1 73.9 ± 9.4 61.6 ± 5.9 64.1 ± 7.9 61.1 ± 9.0 77.4 ± 11.9 64.1 ± 6.1

Combination: aspirin and clopidogrel; kg, kilogram; n, number.

Results A

2500
100
PLAA (%) Plasma TXB2 (ng ml–1)

AA-induced platelet aggregation (%)

Key baseline characteristics of study subjects are presented in
Table 1.

500 1000 1500 2000

80

Plasma TXB2 (ng mL–1)

60
Cohort 1: recovery of platelet function after stopping aspirin
and clopidogrel

40
Recovery of platelet function after stopping
aspirin Compared with baseline levels, PLAA was

20
maximally suppressed on day 1 after completion of aspirin
treatment (2.4 ± 1.6% vs. 83.6 ± 4.6%, P < 0.0001) and

0
remained fully inhibited on day 2. Platelet function began to Baseline 1 2 3 4 5 6 8 10 12 14
Days after stopping aspirin
recover on day 3 and returned to baseline levels by day 4 B
(84.6 ± 4.0% vs. 83.6 ± 4.6%, P = 1.0) (Fig. 2A). ADP and
ADP-induced platelet aggregation (%)
10 20 30 40 50 60 70 80

collagen-induced platelet aggregation also recovered to

baseline levels within 4 days of the last dose of aspirin
(Appendix 1).
AA-induced platelet aggregation in whole blood as mea-
sured by RPFA was maximally suppressed on day 1 after the
completion of aspirin treatment (503.7 ± 81.3 vs. 657.6 ± 8.7
ARU, P < 0.0001) and recovered to baseline level by day 4
(635.6 ± 25.2 vs. 657.6 ± 8.7 ARU, P = 0.80).
Plasma TXB2 concentrations were maximally suppressed on
0

day 1 after the completion of aspirin treatment (median: 6.5,

Baseline 1 2 3 4 5 6 8 10 12
IQR 4.5–19.4 vs. 1732.5, IQR 1343.7–1954.5 ng mL)1, Days after stopping clopidogrel
P < 0.0001). Unlike the recovery of PLAA, plasma TXB2 levels
Fig. 2. Offset of the anti-platelet effects of aspirin (A) and clopidogrel (B).
did not recover to baseline levels until 10 days after the last dose
(A) Demonstrates recovery of AA-induced platelet aggregation (PLAA)
of aspirin (1423.5, IQR 1310.9–1555.0 vs. 1732.5, IQR 1343.7– and plasma thromboxane B2 (TXB2) (n = 7). (B) Demonstrates recovery
1954.5 ng mL)1, P = 0.56) (Fig. 2A). The pattern of recovery of ADP-induced platelet aggregation (PLADP) (n = 8). Error bars repre-
of serum TXB2 and urinary 11-dH-TXB2 was similar to that of sent two standard errors above and below the mean. Baseline indicates
plasma TXB2 with complete recovery not occurring until 8– immediately before ingestion of the first dose of study drug.
10 days after completion of aspirin therapy (Appendix 2).
The mean bleeding time increased from 352 ± 126 to achieved at a plasma TXB2 level of 387 ng mL)1, which is
472 ± 142 s (P = 0.03) on day 1 after completion of aspirin equivalent to the EC50 value (Appendix 3).
therapy. Bleeding time was not measured on day 4 but by day 5
Recovery of platelet function after stopping
had returned to baseline levels (318 ± 101 vs. 352 ± 126 s,
clopidogrel PLADP was maximally inhibited on day 1 after
P = 0.82).
completion of clopidogrel treatment (24.6 ± 11.1% vs.
A plot of PLAA and plasma TXB2 after stopping aspirin
73.8 ± 7.6%, P < 0.0001) and full recovery did not occur
treatment revealed a symmetric sigmoid relation
  until day 10 (72.0 ± 9.6% vs. 73.8 ± 7.6%, P = 1.0)
84:7184 2 (Fig. 2B).
PLAA ¼ ; R ¼ 0:9961 ;
1 þ e0:2979ðPlasma TXB2 385:7650Þ
Comparison of the recovery curves between PLAA and
with abrupt recovery of PLAA in all subjects when plasma plasma TXB2, serum TXB2, urinary 11-dH-TXB2 after
TXB2 levels exceeded 400 ng mL)1. Median PLAA was stopping aspirin, and PLADP after stopping

 2012 International Society on Thrombosis and Haemostasis

 Reversal of aspirin and clopidogrel effects 525

clopidogrel The recovery of PLAA after stopping aspirin was donor platelets (75.7 ± 9.6% vs. 76.0 ± 15.5%, P = 1.00)
significantly more rapid than that of plasma TXB2 (Fig. 3B).
(P < 0.0001), serum TXB2 (P < 0.0001) and urinary 11-
dH-TXB2 (P = 0.0023), and the recovery of PLADP after Reversal of the anti-platelet effects of the combination of
stopping clopidogrel (P = 0.0086). In contrast, the recovery aspirin and clopidogrel Among subjects treated with the
curves of PLADP after stopping clopidogrel were similar to that combination of aspirin and clopidogrel, suppression of PLAA
of plasma TXB2, serum TXB2 and urinary 11-dH-TXB2 after was fully reversed compared with controls after adding 20%
stopping aspirin (P = 0.64, 0.18 and 0.86, respectively). donor platelets (73.7 ± 13.9% vs. 84.3 ± 4.8%, P = 0.57)
(Fig. 3C). Plasma TXB2 was fully inhibited to 15.84 ng mL)1
and did not return to control levels (1545.50 ng mL)1) until all
Cohort 2: reversal of the anti-platelet effects of aspirin and
the platelets were replaced by the donor platelets (Fig. 3C).
clopidogrel using donor platelets
Similarly, PLADP was fully reversed only after the addition of
Reversal of the anti-platelet effects of aspirin Among 90% or more donor platelets (Fig. 3D).
subjects treated with aspirin, suppression of PLAA was fully
reversed compared with controls after mixing with 30%
Discussion
uninhibited control platelets (84.4 ± 9.1% vs. 78.4 ± 13.7%,
P = 0.99) (Fig. 3A). A similar reversal pattern was evident for The present results indicate that the patterns of recovery of
collagen and ADP-induced platelet aggregation (Appendix 4). thromboxane production after stopping aspirin and of PLADP
In contrast, plasma TXB2 concentrations did not approximate after stopping clopidogrel are similar, with full recovery taking
control levels until all the platelets were replaced by the approximately 10 days. In contrast, the recovery of PLAA after
uninhibited donor platelets (Fig. 3A). stopping aspirin is much more rapid, with complete recovery
occurring within 4 days. Differences between aspirin and
Reversal of the anti-platelet effects of clopidogrel Among clopidogrel in the rate of recovery of platelet aggregation after
subjects treated with clopidogrel, suppression of PLADP was stopping treatment are paralleled by differences in the propor-
fully reversed only after mixing with 90% or more uninhibited tion of untreated control platelets required in mixing studies to

A B
2500
100

Plasma TXB2 (ng mL–1)

0 10 20 30 40 50 60 70 80

PLAA (%)
ADP-induced platelet aggregation (%)
AA-induced platelet aggregation (%)

500 1000 1500 2000

80

Plasma TXB2 (ng mL–1)

20 40 60 0

0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 80 100 Proportion of non-clopidogrelized platelets (%)
Proportion of non-aspirinated platelets (%)
C D
2500
100

0 10 20 30 40 50 60 70 80

PLAA (%) Plasma TXB2 (ng mL–1)

ADP-induced platelet aggregation (%)
AA-induced platelet aggregation (%)

500 1000 1500 2000

80

Plasma TXB2 (ng mL–1)

20 40 60 0

0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 80 100 Proportion of non-aspirinated and non-clopidogrelized platelets (%)
Proportion of non-aspirinated and non-clopidogrelized platelets (%)

Fig. 3. Reversal of the anti-platelet effects of aspirin (A), clopidogrel (B) and the combination of aspirin and clopidogrel (C and D) compared with
controls. (A) Demonstrates reversal of AA-induced platelet aggregation (PLAA) and plasma thromboxane B2 (TXB2) (n = 5 pairs). (B) Demonstrates
reversal of ADP-induced platelet aggregation (PLADP) (n = 7 pairs). (C) Demonstrates reversal of PLAA and plasma TXB2 (n = 6 pairs). (D) Dem-
onstrates reversal of PLADP (n = 6 pairs). Error bars represent two standard errors above and below the mean.

 2012 International Society on Thrombosis and Haemostasis

526 C. Li et al
reverse their anti-platelet effects, where reversal of inhibition of
AA-induced aggregation by aspirin is achieved when the
mixture contains 30% non-aspirinated platelets whereas inhi-
bition of PLADP is achieved when the mixture contains at least
90% uninhibited platelets.
The similar patterns of recovery of thromboxane production
after stopping aspirin and of PLADP after stopping clopidogrel
are explained by the irreversible platelet-inhibitory effects of the
drugs. Aspirin permanently inhibits the cyclooxygenase activity
of prostaglandin H-synthase-1 (also referred to as COX-1),
thereby blocking platelet thromboxane production [8], and the
active metabolite of clopidogrel permanently inhibits the
platelet P2Y12 receptor, thereby blocking ADP-induced plate-
let aggregation [9]. Full recovery of platelet thromboxane
production and PLADP does not occur until permanently
inhibited platelets are replaced by de novo newly synthesized
platelets, usually after 10 days (When platelet turnover is
normal, approximately 10% of platelets are replaced each day
[10]). The recovery of PLAA 4 days after stopping aspirin
coincides with the replacement of approximately 40% of
aspirinated platelets with newly released non-aspirinated
platelets. This finding is consistent with in vitro mixing studies
that demonstrated reversal of PLAA when the mixture
contained only 30% non-aspirinated platelets. Both lines of
evidence indicate that in aspirin-treated subjects, only 30% to
40% uninhibited platelets are required to produce sufficient
thromboxane A2 to achieve a normal aggregation response, by
activating the thromboxane A2 receptor of their own and the
adjacent aspirinated platelets. Thromboxane from non-platelet
sources (e.g. monocytes, endothelial cells) can also stimulate
platelets in the presence of aspirin because aspirin selectively
blocks COX-1 and does not inhibit the platelet thromboxane
receptor [8].
Our results demonstrating rapid recovery of platelet function
after stopping aspirin are consistent with those reported by
Zisman et al. [11] who found that PLAA normalized within
3 days of stopping aspirin and by Feldman et al. [12] who
reported that serum TXB2 levels remained suppressed for at
least 7 days after stopping aspirin. Likewise, delayed recovery
of PLADP after stopping clopidogrel in the present study is
similar to the rate of recovery reported in other studies as
measured using the VerifyNow P2Y12 assay [13] and bleeding
time [14]. However, none of these previous studies explored the
effects of mixing with uninhibited platelets on the platelet
inhibitory effects of aspirin and clopidogrel.
The present results have potential implications for the peri-
operative management of aspirin and clopidogrel. First, if
normal hemostasis is required, stopping of aspirin for 4 days
and clopidogrel for 10 days before surgery should ensure full
recovery of the platelet aggregation response. Second, it seems
reasonable to expect that the dose of platelets required to
reverse the anti-aggregatory effects of aspirin is lower than the
dose required to reverse the anti-aggregatory effects of
clopidogrel. However, the present study cannot provide
guidance on the dose of platelets that may be required because
our observations are based on in vitro-mixing studies. Third,
bleeding in patients receiving dual anti-platelet therapy with a
combination of aspirin and clopidogrel is of particular concern
in those who require urgent cardiac or non-cardiac surgery
[3,15]. Some of these patients are also at a high risk for
thromboembolic events, for example stent thrombosis [15]. In
such patients, one option to optimize the risk/benefit ratio
might be to stop clopidogrel 5 or more days before surgery, but
to continue aspirin until the time of surgery and to transfuse
platelets immediately before surgery to fully reverse the anti-
platelet effects of aspirin. This approach requires testing in
clinical studies.
The present study has potential limitations. First, the
number of subjects was small and we included only healthy
controls. However, in post hoc sample size calculations (based
on means and standard deviations observed in our study) we
have confirmed that our sample size provides at least 80%
power to demonstrate a significant difference between inhib-
ited and uninhibited platelets (as measured by platelet
aggregation testing) at a given time point, even after
adjustment for multiple testing; and there is no reason to
expect different results in patients with established cardiovas-
cular disease [11,13] compared with healthy controls. Second,
although our data provide the best available evidence to date
for the rate of recovery of platelet function after stopping
aspirin and clopidogrel and the reversal of their antiplatelet
effects using uninhibited platelets, there are only limited data
on the association between in vitro platelet function and
clinical events.
In conclusion, aspirin permanently inhibits platelet throm-
boxane production and clopidogrel permanently inhibits
PLADP for the life of the platelet. The platelet aggregation
response to AA recovers within 4 days of stopping aspirin,
whereas the platelet aggregation response to ADP recovers
within 10 days of stopping clopidogrel. These findings are
paralleled by the observations in in vitro mixing experiments
that the platelet inhibitory effects of aspirin are reversed with
30% non-aspirinated platelets in the mixture, compared with
the requirement for at least 90% uninhibited platelets in the
mixture to completely reverse the anti-platelet effects of
clopidogrel. The effect of clopidogrel on platelet function
lasts longer than that of aspirin because thromboxane from
non-aspirinated platelets is able to activate aspirinated
platelets whereas platelets that are still inhibited by the
active metabolite of clopidogrel do not aggregate normally in
response to ADP. Based on the present results, stopping
aspirin for 7–10 days is longer than required for the recovery
of platelet function in patients treated with aspirin and is
sufficient for the recovery of platelet function in patients
treated with clopidogrel.
Acknowledgements
This work was carried out in the Thrombosis service, Hamilton
General Hospital, Hamilton, Ontario, Canada. This work was
supported by a grant from the New Investigator Funding of
 2012 International Society on Thrombosis and Haemostasis
 Reversal of aspirin and clopidogrel effects 527

Data are expressed as mean ± standard deviation. Recovery aspirin indicates recovery of the anti-platelet effects of aspirin; Recovery clopidogrel indicates recovery of the anti-platelet effects of
PLCOLL (%)
Hamilton Health Science (NIF-08200[R]), Hamilton, Canada,

12.6
11.8
12.8
8.7
4.9
6.7
8.3
6.6
9.5
5.2
and a grant from the National Natural Science Funding of

67.1 ±
43.2 ±
53.2 ±
60.9 ±
59.1 ±
64.8 ±
65.5 ±
64.8 ±
67.9 ±
72.5 ±
China (81170181).

–
Recovery clopidogrel (n = 8)
Disclosure of Conflict of Interests

11.0
18.1
14.3
13.6
13.5

17.8
PLADP (%)

7.6

9.3

9.6
9.2
73.8 ±
24.6 ±
30.2 ±
34.3 ±
45.8 ±
59.8 ±
61.1 ±
61.4 ±
72.0 ±
74.4 ±
The authors state that they have no conflict of interest.

–
References

10.0
2.8
8.6

6.9
6.7
3.6
4.2
6.3
4.9
5.7
PLAA (%)
1 Alghamdi AA, Moussa F, Fremes SE. Does the use of preoperative

72.4 ±
59.1 ±
59.8 ±
62.6 ±
65.6 ±
68.4 ±
70.8 ±
69.9 ±
72.1 ±
71.8 ±
aspirin increase the risk of bleeding in patients undergoing coronary

–
artery bypass grafting surgery? Systematic review and meta-analysis. J
Card Surg 2007; 22: 247–56.
2 Badreldin A, Kroener A, Kamiya H, Lichtenberg A, Hekmat K. Effect

158.0

170.2
283.4
188.9
40.8
38.2
62.2
44.2
31.0
64.4

81.0
11-dH-TXB2
of clopidogrel on perioperative blood loss and transfusion in coronary

(ng mL)1)
artery bypass graft surgery. Interact Cardiovasc Thorac Surg 2010; 10:

±
±
±
±
±
±
±
±
±
±
±
Urinary
48–52.

273.0

110.7
126.6
133.0
175.2
309.5
328.8
342.1
304.3
64.2
83.9
3 Fitchett D, Eikelboom J, Fremes S, Mazer D, Singh S, Bittira B,
Brister S, Graham JJ, Gupta M, Karkouti K, Lee A, Love M,
McArthur R, Peterson M, Verma S, Yau TM. Dual antiplatelet

164.6

110.7
124.5
138.7
11.6

13.4
37.9
69.3

81.4
Serum TXB2

7.9

3.3
therapy in patients requiring urgent coronary artery bypass grafting

(ng mL)1)
surgery: a position statement of the Canadian Cardiovascular Society.

±
±
±
±
±
±
±
±
±
±
±
Can J Cardiol 2009; 25: 683–9.

3.9
197.0

100.2
168.9
215.7
240.5
192.2
19.8
15.3
45.2
70.8
4 Douketis JD, Berger PB, Dunn AS, Jaffer AK, Spyropoulos AC,
Becker RC, Ansell J. The perioperative management of antithrombotic
therapy: American College of Chest Physicians Evidence-Based Clin-

347.7

132.0
156.6
180.2
255.7
217.0
258.9
204.8
302.5
341.7
82.6
Plasma TXB2
ical Practice Guidelines (8th Edition). Chest 2008; 133: 299S–339S.
5 Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, Gardner

(ng mL)1)

±
±
±
±
±
±
±
±
±
±
±
TJ, Hart JC, Herrmann HC, Hillis LD, Hutter AM Jr, Lytle BW,

217.8
642.3
734.0
38.7
86.7
1649.6

1022.4
1227.4
1448.6
1756.6
1738.2
Marlow RA, Nugent WC, Orszulak TA. ACC/AHA 2004 guideline
Offset of the anti-platelet effects of 7-day aspirin and clopidogrel treatment (Cohort 1).

update for coronary artery bypass graft surgery: a report of the

American College of Cardiology/American Heart Association Task

11.9
18.5
20.6
19.1
Force on Practice Guidelines (Committee to Update the 1999 Guide-

8.1
9.6
6.8
3.8
4.2
6.2
5.8
lines for Coronary Artery Bypass Graft Surgery). Circulation 2004;
PLCOLL

±
±
±
±
±
±
±
±
±
±
±
110: e340–437.
80.3
40.6
54.1
56.5
77.4
77.9
83.5
86.5
84.3
86.7
80.0
(%)

6 Kapetanakis EI, Medlam DA, Boyce SW, Haile E, Hill PC, Dullum
MK, Bafi AS, Petro KR, Corso PJ. Clopidogrel administration prior
to coronary artery bypass grafting surgery: the cardiologistÕs panacea
11.6

11.3

13.2

14.4
PLADP (%)

5.6
7.6
7.3

6.4

7.1

7.1
4.6
or the surgeonÕs headache? Eur Heart J 2005; 26: 576–83.
±
±
±
±
±
±
±
±
±
±
±
7 Maltais S, Perrault LP, Do QB. Effect of clopidogrel on bleeding and
76.1
61.9
60.9
61.2
67.1
77.1
78.3
73.7
76.2
81.0
72.4
transfusions after off-pump coronary artery bypass graft surgery: im-
pact of discontinuation prior to surgery. Eur J Cardiothorac Surg 2008;
34: 127–31.
35.9
4.6
1.6
1.3

4.0
5.6
3.8
4.4
5.2
2.4
9.2
PLAA (%)

8 Hankey GJ, Eikelboom JW. Aspirin resistance. Lancet 2006; 367: 606–
17.
±
±
±
±
±
±
±
±
±
±
±

9 Krotz F, Sohn HY, Klauss V. Antiplatelet drugs in cardiological

83.6

14.7
84.6
82.3
83.2
86.2
84.8
84.5
87.4
2.4
2.1

practice: established strategies and new developments. Vasc Health

Risk Manag 2008; 4: 637–45.
81.3
73.2
49.4
25.2
30.2
Recovery aspirin (n = 7)

10 Leeksma CH, Cohen JA. Determination of the life of human blood

8.7

platelets using labelled diisopropylfluorophosphanate. Nature 1955;

657.6 ±
503.7 ±
524.9 ±
589.7 ±
635.6 ±
640.3 ±
(ARU)
RPFA

175: 552–3.
–
–
–
–
–

11 Zisman E, Erport A, Kohanovsky E, Ballagulah M, Cassel A, Quitt

M, Pizov R. Platelet function recovery after cessation of aspirin: pre-
liminary study of volunteers and surgical patients. Eur J Anaesthesiol
clopidogrel. BT, bleeding time.
2.1
2.4

1.4

1.7

1.2

2010; 27: 617–23.

BT (s)

5.9 ±
7.9 ±

7.7 ±

5.3 ±

4.5 ±

12 Feldman M, Shewmake K, Cryer B. Time course inhibition of gastric

–

–
–
–

and platelet COX activity by acetylsalicylic acid in humans. Am J

Physiol Gastrointest Liver Physiol 2000; 279: G1113–20.
Before treatment

13 Gurbel PA, Bliden KP, Butler K, Tantry US, Gesheff T, Wei C, Teng
Appendix 1

R, Antonino MJ, Patil SB, Karunakaran A, Kereiakes DJ, Parris C,

Purdy D, Wilson V, Ledley GS, Storey RF. Randomized double-blind
assessment of the ONSET and OFFSET of the antiplatelet effects
Day 10
Day 12
Day 14
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Day 8

of ticagrelor versus clopidogrel in patients with stable coronary

 2012 International Society on Thrombosis and Haemostasis

 AA-induced platelet aggregation (%) Serum TXB2 (ng/ml)
2577–85.

0 20 40 60 80 100 0 100 200 300 400 500

Appendix 3
Appendix 2

0
528 C. Li et al

Baseline 1
2
3

stopping aspirin treatment.

4 5 6
Serum TXB2 (ng/ml)

Predict PLAA value (%)

8
Thromb Hemost 1999; 25(Suppl 2): 9–14.

Plasma TXB2 (ng/ml)

10

PLAA (%)
Days after stopping aspirin (d)
12
Urinary 11-DH-TXB2 (ng/ml)

14
0 100 200 300 400 500
immediately prior to ingestion of the first dose of aspirin.

Urinary 11-DH-TXB2 (ng/ml)

200 400 600 800 1000120014001600180020002200

standard errors above and below the mean. Baseline indicates
Recovery of serum TXB2 and urinary 11-dH-TXB2 after
quiring major non cardiac surgery. Indian J Anaesth 2009; 53: 582–91.
artery disease: the ONSET/OFFSET study. Circulation 2009; 120:

Relation between PLAA and plasma TXB2 concentration after

stopping aspirin treatment (n = 7). Error bars represent 2
dose pharmacodynamics of clopidogrel in healthy subjects. Semin
14 Thebault JJ, Kieffer G, Lowe GD, Nimmo WS, Cariou R. Repeated-

15 Kiran U, Makhija N. Patient with recent coronary artery stent re-

Appendix 4
Reversal of the anti-platelet effects of 7-day aspirin and/or clopidogrel treatment by control platelets (Cohort 2).

Reversal aspirin (n = 5 + 5) Reversal clopidogrel (n = 7 + 7) Reversal aspirin and clopidogrel (n = 6 + 6)

Con-PRP/Stu-PRP PLAA PLADP PLCOLL Plasma TXB2 PLAA PLADP PLCOLL PLAA PLADP PLCOLL Plasma TXB2
(v/v) (%) (%) (%) (ng mL)1) (%) (%) (%) (%) (%) (%) (ng mL)1)

0/1.0 4.0 ± 0.7 62.0 ± 8.3 22.4 ± 7.3 11.3 ± 19.9 58.1 ± 30.0 22.6 ± 9.4 70.3 ± 16.7 4.2 ± 1.2 44.7 ± 7.5 18.7 ± 10.8 15.8 ± 19.1
0.1/0.9 34.0 ± 44.8 68.6 ± 12.6 44.0 ± 17.3 143.2 ± 100.4 74.0 ± 12.8 29.4 ± 9.3 75.4 ± 12.6 32.8 ± 34.6 48.2 ± 6.5 33.2 ± 18.2 147.6 ± 42.6
0.2/0.8 63.0 ± 38.1 71.2 ± 15.3 64.2 ± 17.7 305.1 ± 97.2 80.9 ± 8.0 37.0 ± 12.0 77.0 ± 6.9 73.7 ± 13.9 52.0 ± 5.9 57.8 ± 10.7 372.6 ± 69.1
0.3/0.7 84.4 ± 9.1 75.8 ± 13.0 64.8 ± 19.7 524.6 ± 127.7 81.6 ± 3.6 45.4 ± 10.0 76.3 ± 9.4 81.3 ± 6.0 55.7 ± 10.1 66.0 ± 9.1 532.9 ± 74.3
0.4/0.6 86.0 ± 9.2 79.6 ± 13.9 71.8 ± 11.8 – 80.9 ± 5.6 54.7 ± 14.5 83.0 ± 7.9 83.7 ± 2.3 61.8 ± 8.4 77.2 ± 7.2 –
0.5/0.5 86.8 ± 8.7 79.0 ± 17.2 75.2 ± 18.9 862.0 ± 195.0 83.3 ± 6.0 57.1 ± 12.1 85.0 ± 8.1 84.3 ± 4.1 66.5 ± 8.0 77.2 ± 3.4 841.6 ± 142.7
0.6/0.4 86.8 ± 10.1 77.4 ± 15.2 75.8 ± 14.2 – 82.6 ± 5.4 63.3 ± 10.0 82.1 ± 4.3 83.8 ± 5.8 67.8 ± 9.2 75.2 ± 1.8 –
0.7/0.3 – – – – 85.6 ± 4.1 71.6 ± 10.3 85.1 ± 4.7 86.7 ± 4.8 74.3 ± 7.6 82.0 ± 4.2 –
0.8/0.2 85.0 ± 11.2 73.8 ± 8.4 80.0 ± 11.9 1350.5 ± 573.6 83.4 ± 5.4 68.4 ± 11.1 82.1 ± 4.3 87.5 ± 5.0 74.3 ± 8.7 82.2 ± 6.5 1240.1 ± 179.2
0.9/0.1 – – – – 85.0 ± 5.7 75.7 ± 9.6 83.4 ± 6.6 86.7 ± 5.4 77.7 ± 6.7 83.7 ± 2.6 –
1.0/0 78.4 ± 13.7 70.6 ± 7.5 81.0 ± 7.1 1826.5 ± 237.5 85.6 ± 7.5 76.0 ± 15.5 82.4 ± 10.6 84.3 ± 4.8 79.2 ± 6.6 88.3 ± 7.1 1545.5 ± 196.2

Data are expressed as mean ± standard deviation. Reversal aspirin indicates reversal of the anti-platelet effects of aspirin; Reversal clopidogrel indicates reversal of the anti-
platelet effects of clopidogrel; Reversal aspirin and clopidogrel indicates reversal of the anti-platelet effects of aspirin and clopidogrel. Con-PRP, control platelet-rich plasma;
Stu-PRP, study platelet-rich plasma.

 2012 International Society on Thrombosis and Haemostasis