Platelet function tests predict bleeding and thrombotic events
after off-pump coronary bypass grafting*
Robert Postona,*, Junyan Gua, Jeffrey Manchioa, Andrew Leea, James Browna,
James Gammiea, Charles Whiteb, Bartley P. Griffitha
aDivision of Cardiac Surgery, Department of Surgery, School of Medicine, University of Maryland, N4W94 22 S. Greene St., Baltimore, MD 21201, USA
bDepartment of Radiology, School of Medicine, University of Maryland, Baltimore, MD 21201, USA
Received 9 September 2004; received in revised form 24 November 2004; accepted 13 December 2004
Objective: A balanced coagulation system after cardiac surgery minimizes bleeding and thrombotic events. However, the best method to
monitor this balance has not been established. We used a series of tests of coagulation and platelet function to define the risk of bleeding and
thrombotic events after OPCAB. Methods: In 76 patients, routine coagulation tests (i.e. prothrombin time, fibrinogen level, d-dimer, and
platelet count), thrombelastography, and whole blood aggregometry were obtained perioperatively and on days 1 and 3 after OPCAB. Intra- and
postoperative blood loss was determined. Early patency of venous bypass grafts was determined using CT angiography (Philips Medical, Corp.).
Results: Chest tube output and red cell volume loss at 24 h were 952G475 and 190G115 ml, respectively. Early graft failure developed in eight
patients. Perioperative changes in routine coagulation tests showed no correlation with either bleeding or thrombosis. However, perioperative
decline in platelet function as assessed by the area under the impedance curve for whole blood aggregometry correlated with intraoperative
blood loss (RZ0.42, P!0.05). A perioperative decline in the maximum amplitude of the thrombelastography trace showed a significant
correlation with 24 h hemoglobin loss (RZ0.45, P!0.05). Compared to those with all patent grafts, patients with early graft failure
demonstrated a reduction in platelet sensitivity to aspirin by both thrombelastography and aggregometry on day 3. Conclusions: In contrast to
standard coagulation testing, platelet function predicted both bleeding and thrombosis after OPCAB. Titration of perioperative platelet
function according to these tests may minimize thrombosis without increasing bleeding.
q 2005 Elsevier B.V. All rights reserved.
Keywords: Platelet; Saphenous vein; Coronary artery bypass grafting; Thrombosis
A postoperative hypercoagulable state that increases the
risk for thrombotic events is well established after general
surgical procedures . An ‘anticoagulating effect’ of
cardiopulmonary bypass (CPB) is thought to create a risk of
bleeding that exceeds the tendency for postoperative
hypercoagulability. Whether circumventing CPB during off-
pump coronary artery bypass (OPCAB) alters the balance of
coagulation and increases the risk for a postoperative
hypercoagulable state after OPCAB is unclear [5,6]. Although
transfusion requirements are reduced by OPCAB, post-
operative bleeding has been found to be consistently the
same between these two approaches [2–4]. Early graft
thrombosis was increased after off- compared to on-pump
CAB in one recent randomized trial  but not others .
The lack of an in vitro assay that is capable of confirming a
hypercoagulable state also adds to the controversy. There is
currently great variability in antiplatelet strategies after
OPCAB . The accurate prediction of adverse bleeding and
thrombotic events would provide a more rational basis for
therapy and reduce this variability. In this study, we tested
the hypothesis that in vitro platelet function tests provide an
accurate method for defining the risk of bleeding and graft
thrombosis after OPCAB.
2.1. Patient enrollment and study design
This study was a prospective cohort trial of patients
undergoing OPCAB at a single center. Local IRB approval was
obtained (protocol #0303108) and all subjects provided
informed consent prior to enrollment. From November 2002
until May 2004, 122 patients were screened. Forty patients
were excluded due to creatinine O2.0 mg/dl (nZ16),
conversion to on-pump CAB (nZ12), refusal of consent
European Journal of Cardio-thoracic Surgery 27 (2005) 584–591
1010-7940/$ - see front matter q 2005 Elsevier B.V. All rights reserved.
*Presented at the joint 18th Annual Meeting of the European Association
for Cardio-thoracic Surgery and the 12th Annual Meeting of the European
Society of Thoracic Surgeons, Leipzig, Germany, September 12–15, 2004.
*Corresponding author. Tel.: C1 410 328 5089; fax: C1 410 328 2750.
E-mail address: email@example.com (R. Poston).
(nZ10) and the emergent nature of the surgery (nZ2). The
primary endpoint of this study was the correlation of in vitro
platelet activity with both early graft thrombosis and
2.2. Surgical technique and patient management
Four surgeons, experienced in OPCAB, enrolled patients
during the study. After a full median sternotomy, the left
internal thoracic artery (ITA) was used as a bypass conduit in
all patients; the great saphenous vein was harvested using an
endoscopic (nZ105 venous conduits) (VasoView5 Endoscopic
Vessel Harvesting Systemw, Guidant Systems, Inc., Minnea-
polis, MN) or open (nZ23 venous conduits) approach, based
on anatomical considerations. Radial arteries were har-
vested (nZ8) using a ‘no touch’ technique. Other types of
arterial conduits were not used during the course of this
study. Conduits were stored in dilute heparinized blood until
use. Prior to distal ligation of ITA pedicle, intravenous
heparin was given at a dose calculated by the Hepcon
instrument (Medtronic, Inc., Minneapolis, MN) as sufficient
to obtain a kaolin based ACT of greater than 300 s. Further
heparin doses were given as required every 30 min to
maintain the heparin level at O2 IU/ml and ACT O300 s.
Suction-based exposure and stablizing devices (Octopus 4.3w
Tissue Stabilizer & UrchinwHeart Positioner, Medtronic,
Inc., Minneapolis, MN) were used for creating all distal
anastomosis. To minimize coagulation variability in study
subjects, patients requiring conversion to a standard, on-
pump CAB technique were excluded from analysis. At the
end of surgery, the heparin effect was reversed by half the
dose of protamine calculated by the Hepcon device.
Preoperative aspirin was continued through the date of
surgery. Postoperative aspirin (325 mg p.o. qd)—given within
6 h of arrival to the ICU—was the sole platelet inhibitor used
in all study patients. Postoperative use of blood products
followed a TEG-based algorithm as previously described .
2.3. Perioperative bleeding
Shed mediastinal blood was collected intraoperatively
using a cellsaver device (Cobe BRAT 2, Cobe Cardiovascular,
Inc., Arvada, CO), processed and retransfused. The volume
retrieved intraoperatively was measured along with the
amount of postoperative shed blood after 24 h. The red
blood cell volume lost after 24 h was calculated by the
hematocrit of the contents in the water seal chest drainage
system (Atrium Medical Corp., Hudson, NH) multiplied by the
total volume. The chest drain contents were mixed
thoroughly to obtain an average hematocrit over the 24 h
2.4. Intraoperative graft analysis
Blood flow and flow waveform were measured in each
graft using transit time ultrasound (Transonic, Inc., Ithica,
NY), a routinely available technology shown to predict
bypass graft failure on follow-up angiography . Wave-
forms were stored on a laptop computer using digital data
acquisition software (WinDaqe, DATAQ Instruments, Inc.,
Dayton, OH). Discarded vein segments were procured
intraoperatively from each bypass graft for analysis of
endothelial integrity by immunohistochemistry. A portion
was snap frozen in OCT compound and stored at K80 8C.
Sections (w5 mm thick) were stained with CD31 monoclonal
antibody (R&D System, Inc.) for calculating percentage
endothelial integrity using image analysis software (Bioquant
Nova Prime ver.6w, Nashville, TN).
2.5. Routine coagulation assays
Routine tests of coagulation (INR, activated partial
thromboplastin time, fibrinogen, platelet count, and quan-
titative d-dimer levels) were performed at baseline (prior to
skin incision), postoperatively (immediately after skin
closure), postoperative day 1 and postoperative day 3 by
2.6. Assays for platelet activity
At the same perioperative time points described above,
citrated (3.8%) whole blood was collected for the following
in vitro assays of platelet function.
TEGw. Using 360 ml whole blood per assay, the following
parameters were obtained from the TEG trace: R time—
period of time from initiation of the test to the start of the
trace and represents initial fibrin formation; a—an angle
between the line in the middle of the TEG tracing and the
line tangential to the developing TEG tracing and represents
the kinetics of fibrin cross-linking; MAkaolin—maximum
amplitude reflects strength of a clot which is dependent on
number and function of platelets and its interaction with
fibrin but is insensitive to the aspirin effect ; LY30—
measures the rate of amplitude reduction 30 min after MA
and represents the stability of the clot. Given that all
subjects were maintained on perioperative aspirin,
additional TEG assays were performed to improve the
detection of an aspirin effect. A heparinized (O2 U/ml)
sample was treated using a direct fibrinogen activator
(factor XIIIa 5 U/ml and reptilase 0.2 U/ml)Garachidonic
acid (0.14 mg/ml) and the resulting MA generated from
arachidonic acid (MAAA) or fibrin formation alone (MAF) were
recorded. The aspirin effect (MAASA) was calculated as
follows: ½ðMAAAKMAFÞ=ðMAkaolinKMAFÞ?!100% .
Whole blood aggregometry (model 592A, Chronolog,
Havertown, Pennsylvania). Whole blood diluted 1:1 with
saline was stimulated with either low (1 mg/ml) or high
(5 mg/ml) dose collagen. Impedance changes were then
measured across two electrodes immersed in the sample.
The area under the developing impedance curve (AUC) and
the maximum impedance changes (U) were assessed after
6 min. The aspirin effect was determined by comparing the
impedance change resulting from low vs. high dose collagen
as described .
2.7. Postoperative graft follow-up
The patency of bypass grafts was determined by blinded
review by a single expert reviewer (CW) of 16 detector row,
spiral computed tomography examinations (420 ms rotation,
100–150 ml contrast agent IV at 5 cm3/s). Retrospective ECG
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591 585
gating was performed for image reconstruction to minimize
cardiac motion artifact. Patency was defined as any flow
through the entire graft regardless of the presence of
stenosis. The graft was said to be nonpatent if a stump was
seen or if there was no flow by CT angiography (CTA) as
previously described .
Assumptions for the power analysis were based on prior
studies by our group showing a 6.5% rate of early graft failure
after OPCAB  and by others demonstrating the strong
correlations between platelet function testing and hemor-
rhage [8,14,18]. In addition, aspirin has been shown to
reduce graft thrombosis by 50% relative to placebo ,
suggesting an equally strong relationship between platelet
function and graft thrombosis. On the basis of these data,
enrollment of 75 patients was required to demonstrate a
correlation of platelet function with bleeding and thrombo-
sis at PZ0.05 and powerZ80% (http://calculators.stat.
Platelet hyperactivity was defined for each assay as
platelet responsive at any point of analysis O1 SD above
the mean ‘normal’ value determined from 20 healthy
volunteers . Correlations were performed using Fischer
exact test for categorical variables and linear regression
for continuous variables. The predictive value of each
platelet function test at predicting bleeding was deter-
mined by determining the area under the curve (AUC) for
Receiver Operating Characteristics curves, using O800 ml
bleeding per 24 h for the definition of a ‘bleeding’ vs.
‘nonbleeding’ patient. Calculations were performed using
the InState statistical package (GraphPad Software, San
3.1. Early thrombosis vs. patent vein grafts
Of the 82 OPCAB patients with 156 venous grafts that
completed the protocol, 19 venous grafts from six patients
were not included in the early graft patency assessment due
to prospectively determined exclusion criteria: (1) inability
to obtain postoperative angiography due to creatinine
O2.0 mg/dl (nZ3); (2) patient refusal (nZ2); and (3) graft
blood flow !10 cm3/min despite revision (nZ2). Patency of
arterial grafts was not included in this analysis, but was
100%. CTA was obtained in 97% of study subjects before
hospital discharge. Of the 137 vein grafts analyzed, eight
grafts (5.8%) in eight patients were found to be thrombosed
by CTA (Fig. 1). Four of these patients underwent further
follow-up by conventional, catheter based angiography
which confirmed the diagnosis of graft thrombosis in
No significant differences
patients with thrombosed vs. all patent grafts in demogra-
phics/preoperative risk factors or medication use (Table 1).
Intraoperatively collected data such as ejection fraction,
conduit diameter, and target size and quality and inotropic
requirements were also similar (Table 2). Intraoperative
blood loss was similar but chest tube output at 24 h was
significantly reduced in those patients that developed early
graft thrombosis (Fig. 2).
3.2. Conduit quality
Grafts that developed early failure by POD5 had less
blood flow measured intraoperatively than grafts that
remained patent (36.5G29.3 vs. 46.9G32.3 cm3/min), but
the differences were minimal and did not reach statistical
Fig. 1. Noninvasive assessment of bypass graft patency in nearly all enrolled
subjects was provided by predischarge multiple detector-row computed
tomography (Philips Medical). In these representative examples, CT angio-
graphy demonstrated complete patency of the vein and internal mammary
grafts in one patient (A) and the proximal stump of a thrombosed graft in a
second, asymptomatic patient (B).
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591586
significance. Pulsatility index (2.6G1.9 vs. 2.1G0.8) and
percentage diastolic flow (51G15 vs. 58G11%) were also
similar between groups that developed thrombosis vs. no
graft thrombosis, respectively. The percentage of endo-
thelial integrity was 16.7G21.5% in eight vein grafts that
failed early compared to 55.2G35.7% in grafts remaining
patent (PZ0.047) (Fig. 3).
3.3. Platelet hyperactivity
Prior to surgery, platelet hyperreactivity (i.e. O1 SD
above the normal value for either test) was noted in only two
patients (3%). Nearly all patients developed a significant
decrease in platelet function according to TEG and WBA
during the immediate postoperative assessment (Fig. 4). In
most patients, platelet function returned to baseline.
However, the function exceeded baseline by O1 SD on
POD#3 in 10 patients (Fig. 4 and Table 2). Only one in 10
of this hyperreactive subgroup developed early graft
thrombosis (10 vs. 5.8% of the remaining patients who
maintained normal postoperative platelet function, PZNS).
No consistent differences in platelet function were noted for
the eight patients with graft thrombosis (nZ8 grafts) vs. the
67 patients with all patent grafts (nZ129 grafts) on either
test and at any time point. However, analyzing platelet
function according to aspirin responsiveness improved the
value of these tests. A significant increase in platelet
resistance to the aspirin effect was noted in those patients
that developed early graft thrombosis (Table 2).
Preoperative demographics and postoperative course
Thrombosis No thrombosisP
Active smoker (%)
Prior stroke (%)
ACE inhibitor (%)
Length of intubation (h)
Postop low CI (%)
Postop LOS (days)
EF, ejection fraction determined on intraoperative transesophageal echo-
cardiography; low CI, cardiac index !2.0 l/min/m2; LOS, length of stay.
Aspirin resistance predicts graft thrombosis
Laboratory assayThrombosis (nZ8) No thrombosisP
Peak PT change (s)
Peak platelet count
U 6 min—high dose
% U 6 min—low vs.
high dose collagen
Fig. 2. Bloodlossduringsurgery(i.e.intraoperative)andoverthefirst24 h(i.e.
postoperative) was determined by the volume collected by the cellsaver and
chest drainage devices, respectively. Although intraoperative bleeding was
similar, patients that developed early graft failure showed preserved
Fig. 3. A discarded segment of each vein graft was procured intraoperatively
for immunohistochemical analysis of CD31, an endothelial marker that is
constitutively expressed. Endothelial integrity was determined by positive
CD31 staining (red arrows) and disruption depicted by the lack of staining on
the luminal surface (blue arrows). Using image analysis software, graft
attrition was found to be significantly associated with lesser endothelial
integrity compared to those grafts that remained patent (P!0.04 for
thrombosed vs. patent grafts) (For interpretation of the reference to colour
in this legend, the reader is referred to the web version of this article.).
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591 587
Although intraoperative blood loss did not differ, the
thrombosed group was noted to have a significant
reduction in the volume retrieved in chest drainage device
(632G366 vs. 1106G741 ml, P!0.01) and in the amount of
red blood cell volume lost (80.8G92 vs. 192G120 ml blood
for thrombosed vs. patent group, respectively, P!0.01)
over the initial 24 h after OPCAB (Fig. 3). Despite a
reduction in bleeding, no differences were noted in the use
of blood products or % receiving products between groups.
Intraoperative bleeding was found to significantly correlate
with the decrease in the pre- vs. postoperative platelet
count (RZ0.54, P!0.001) and function according to
the area under the impedance curve for WBA (RZ0.42,
P!0.05) but not for the MA of the TEG trace (RZ0.34,
PZNS). Perioperative coagulation assessed by routinely
available testing (INR, fibrinogen, platelet count) showed
no correlation with blood loss at 24 h (RZ0.10, 0.25 and
0.26, respectively, PZNS). Of all the coagulation tests
obtained, a significant correlation with 24 h hemoglobin
loss was seen only with a perioperative decline in the
maximum amplitude of the TEG trace (RZ0.45, P!0.05)
and fibrinogen levels (RZ0.43, P!0.05). ROC analysis
confirmed that only the MA demonstrated an AUC that was
O0.75, the generally accepted cutoff for an appropriately
predictive test (Fig. 5).
In this study, ‘point of care’ platelet function tests
showed a significant correlation with perioperative hemor-
rhage and graft thrombosis. The current standard of care for
perioperative coagulation monitoring consists of a platelet
count, prothrombin (PT) and activated partial thromboplas-
tin (aPTT) times. Neither these routine tests nor other
isolated indices of coagulation such as fibrinogen and
d-dimer levels have been shown to correlate with bleeding
or arterial thrombotic events after CAB [16–18]. We confirm
the findings of others regarding the value of platelet function
testing using WBA  and TEG [8,18] as a means of
predicting intra- and postoperative bleeding after cardiac
surgery. When modified to detect aspirin responsiveness,
these platelet function tests also predicted early graft
failure after OPCAB. Given the dual benefit for predicting
bleeding and thrombosis, our findings challenge the standard
of care for perioperative coagulation monitoring which
typically omits platelet function monitoring.
Transfusion protocols incorporating platelet count, fibri-
nogen level and INR have been shown to reduce the
requirement for transfusions after cardiac surgery compared
to an empiric strategy . However, these routine tests are
insensitive predictors of bleeding [16–18] and changes in
perioperative platelet counts show poor correlation with
changes in platelet function . Therefore, the mechanism
of benefit for protocols based on these assays is unclear.
Monitoring coagulation during surgery with the TEG-MA
reduces the rate of subsequent reoperation for bleeding in
part by helping the surgeon to differentiate ‘surgical’ from
‘nonsurgical’ bleeding . Postoperative TEG monitoring
also provides a rational guide for transfusions that has been
shown to minimize the rate of unnecessary transfusions .
In our study, we were not able to duplicate the work of
others  demonstrating a role for WBA as a predictor of
A hypercoagulable state after OPCAB provokes concern
about compromised patency of vein grafts. Evidence of
such a state is based mainly on plasma markers of
coagulation . However, heightened activity of this
enzymatic cascade or its inhibition with coumadin seems
to have little influence on either postoperative bleeding
Fig. 4. Assays that utilized different platelet agonists were incorporated into
the study to provide a balanced analysis of platelet function. Neither whole
blood aggregometry, (A, collagen 5 mg/ml agonist) nor thrombelastography
(B, thrombin agonist) showed a difference in the patients that developed graft
thrombosis (nZ8) compared to those with all patent grafts (nZ67).
Fig. 5. Receiver Operating Characteristic curves were calculated to depict the
value of each test used during the study for discriminating ‘bleeding’
(O180 ml red blood cell lost over 24 h). The area under the ROC curve was
greatest for the postoperative MA on TEG analysis (0.77, 95% CI: 0.57–0.91,
P!0.01) compared to other RCT.
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591588
 or vein graft thrombosis [22,24]. On the other hand,
several lines of evidence suggest that platelets play a more
central role on these outcomes after cardiac surgery.
Platelet receptor polymorphism  and antiplatelet
agents strongly correlate with thrombosis and postopera-
tive bleeding. Recent reports of increased graft attrition
after OPCAB  highlight the need for tools to monitor
perioperative platelet function.
The relationship between platelet reactivity and graft
thrombosis remains undefined due to absence of validated
assays of platelet function. Platelet aggregometry and flow
technical complexity. Commercially available tests of plate-
let function predict bleeding but have not been shown to
obtaining follow-up of graft patency. Recent advances in
multiple detector-row cardiac computed tomography pro-
vide a noninvasive means of objectively determining graft
patency that is widely accepted by patients. Using this
technique, our graft follow-up rate far exceeded that
reported in recent series that used conventional angiography
[2,3]. Because of the large diameter and predictable course
the validity of CTA to determine early patency .
Although overall platelet reactivity was not significantly
different in those with early graft failure, a hypercoagulable
state is nevertheless suggested in these patients. First,
perioperative hemostasis (i.e. hemorrhage over the first
24 h) was better preserved compared to those with all
patent grafts. Second, a significant correlation of these
assays with early graft failure was uncovered after sub-
stituting agonists shown to be more effective at determining
the aspirin effect [18,19]. These data point to a novel role
for platelet function testing as a means of defining the
therapeutic window for antiplatelet medications.
Clopidogrel is used after OPCAB  based on concerns
of a hypercoagulable state. However, this more aggressive
antiplatelet strategy increases the risk of bleeding. The
low incidence of generalized platelet hyperactivity that we
detected after OPCAB suggests that routine clopidogrel use
is likely to be associated with an unfavorable the risk-
benefit ratio in patients with preserved aspirin sensitivity.
Perioperative determination of aspirin sensitivity provides
a rational way to titrate antiplatelet medications by
establishing an individual patient’s postoperative risks.
Antiplatelet strategies other than aspirin monotherapy can
be safely avoided in those at high bleeding/low thrombosis
risk such as those with reduced platelet function or
preserved aspirin sensitivity.
The current study has several limitations. First,
although TEG and WBA predict bleeding, their sensitivity
to detect changes in platelet reactivity is not unchal-
lenged. We have recently incorporated flow cytometry, a
more established in vitro platelet assay, into our study to
address this concern. Second, the delay in the diagnosis of
aspirin resistance until POD#3 in most patients limits the
clinical utility of our efforts. We archived patient
leukocyte samples during this trial to allow for genetic
profiling of subjects with and without graft failure.
Expanding our mechanistic understanding from phenotype
to genotype may provide the best avenue to prospectively
prevent thrombosis before hypercoagulability occurs.
Third, several different surgeons enrolled patients in this
study which potentially confounded our graft patency and
postoperative bleeding results. However, bypass grafts
with inadequate blood flow (!10 cm3/min) and waveform
(PIO5) were excluded from analysis and those that
developed early failure had no difference in blood flow
vs. those that remained patent. While this does not rule
out an influence of technique on patency, it suggests that
the quality of distal anastomoses of grafts that developed
early thrombosis was similar to those that remained patent
. Adherence to a strict protocol for coagulation
monitoring and blood product use likely minimized the
surgeon’s influence on postoperative bleeding. Finally,
wide variations in the integrity of the conduit endothelium
were found to correlate with graft thrombosis and may
have hidden a more subtle effect of platelet function.
Elucidating interactions between conduit blood flow,
hypercoagulability, and endothelial disruption awaits
future analysis of larger numbers of subjects enrolled in
our ongoing trial.
In conclusion, serial assessment of platelet function
using point of care assays was found to predict both
bleeding and thrombotic events after OPCAB. In contrast,
routine coagulation tests were unable to discriminate the
risk for either of these outcomes. An analysis on platelet
function that focused on the aspirin effect was able to
identify a cohort of hyperresponsive patients that devel-
oped early graft failure despite apparently normal con-
duits. Titration of perioperative platelet function using
clopidogrel and/or other agents may minimize thrombosis
in these patients without increasing bleeding risk in the
population as a whole. Future efforts to analyze genetic
polymorphisms in this cohort may provide further avenues
to prospectively intervene at the most optimal time point
in these patients.
This work was supported by a Scientist Development
Grant from the American Heart Association (0435318N), the
Bayer Fellowship for Blood Conservation from Bayer Phar-
maceuticals, Corp. and a research grant from Phillips
Medical, Corp. Supplies were donated for the performance
of the TEG (Haemoscope Corporation) and whole blood
aggregometry (Chronolog, Inc.). The authors acknowledge
the effort provided by Dr Marcello Cardarelli as the
independent Data Safety Monitor for the study. R. Poston
is supported by grants from the American Heart Association,
Bayer Fellowship for Blood Conservation, and Women’s
Health Research Award.
Ivankovich AD. Effects of epidural anesthesia and analgesia on
coagulation and outcome after major vascular surgery. Anesth Analg
KJ,McCarthy RJ, MarchRJ,DeLaria GA,Patel RV,
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591589
 Khan NE, De Souza AD, Mister R, Flather M, Clague J, Davies S, Collins P,
Wang D, Sigwart U, Pepper J. A randomized comparison of off-pump and
on-pump multivessel coronary-artery bypass surgery. N Engl J Med 2004;
 Puskas JD, Williams WH, Duke PG, Staples JR, Glas KE, Marshall JJ,
Leimbach M, Huber P, Garas S, Sammons BH, McCall SA, Petersen RJ,
Bailey DE, Chu H, Mahoney EM, Weintraub WS, Guyton RA. Off-pump
coronary artery bypass grafting provides complete revascularization
with reduced myocardial injury, transfusion requirements, and length
of stay: a prospective randomized comparison of two hundred
unselected patients undergoing off-pump versus conventional coron-
ary artery bypass grafting. J Thorac Cardiovasc Surg 2003;125:
 Van Dijk D, Nierich AP, Jansen EWL. Early outcome after off-pump versus
on-pump coronary bypass surgery: results from a randomized study.
 Mariani M, Gu Y, Boonstra P. Procoagulant activity after off-pump
coronary operation: is the current anticoagulation adequate? Ann Thorac
 Cartier R, Robitaille D. Thrombotic complications in beating heart
operations. J Thorac Cardiovasc Surg 2001;121(5):920–2.
 D’Ancona G, Donias HW, Karamanoukian RL. OPCAB therapy survey: off-
pump clopidogrel, aspirin or both therapy survey. Heart Surg Forum
 Shore-Lesserson L, Manspeizer HE, DePerio M, Francis S, Vela-
Cantos F, Ergin MA. Thromboelastography-guided transfusion algorithm
reduces transfusions in complex cardiac surgery. Anesth Analg 1999;
 Morota T, Duhaylongsod FG, Burfeind WR, Huang C-T. Intraoperative
evaluation of coronary anastomosis by transit time ultrasound flow
measurement. Ann Thorac Surg 2002;73(5):1446–50.
 CraftRM, ChavezJJ, Bresee
correlates with optical platelet aggregation. J Lab Clin Med 2004;
 Friend M, Vucenik I, Miller M. Research pointers: platelet responsiveness
to aspirin in patients with hyperlipidaemia. Br Med J 2003;326(7380):
 Yoo KJ, Choi D, Choi BW. The comparison of the graft patency after
coronary artery bypass grafting using coronary angiography and multi-
slice computed tomography. Eur J Cardiothorac Surg 2003;24(1):
 Poston RS, Prastein D, Gu J, Lee S, Pierson R, Griffith B. Virchow’s triad,
but not use of an aortic connector device, predicts vein graft thrombosis
after off-pump bypass. Heart Surg Forum; 2004;7(5):1–6.
 Spiess B. Coagulation monitoring in the perioperative period. Int
Anesthesiol Clin 2004;42(2):55–71.
 Stein PD, Dalen JE, Goldman S. Antithrombotic therapy in patients with
saphenous vein and internal mammary artery bypass grafts. Chest 1995;
 Despotis GJ, Filos KS, Zoys TN. Factors associated with excessive
postoperative blood loss and hemostatic transfusion requirements: a
multivariate analysis in cardiac surgical patients. Anesth Analg 1996;
 Gravlee GP, Arora S, Lavender SW. Predictive value of blood clotting
tests in cardiac surgical patients. Ann Thorac Surg 1994;58:216–21.
 Essell JH, Martin TJ, Salinas J. Comparison of thromboelastography
to bleedingtimeand standard
after cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1993;7:
 Ray MJ, Hawson GA, Just SJ. Relationship of platelet aggregation to
bleeding after cardiopulmonary bypass. Ann Thorac Surg 1994;57(4):
 Avidan MS, Alcock EL, Da Fonseca J. Comparison of structured use of
routine laboratory tests or near-patient assessment with clinical
judgement in the management of bleeding after cardiac surgery. Br
J Anaesth 2004;92(2):178–86.
 Salama ME, Raman S, Drew MJ. Platelet function testing to assess
effectiveness of platelet transfusion therapy. Transfus Apheresis Sci
 Moor E, Blomback M, Silveira A. Haemostatic function in patients
undergoing coronary artery bypass grafting: peroperative perturbations
and relations to saphenous vein graft closure. Thromb Res 2000;98(1):
A novel modification
 Dietrich W, Dilthey G, Spannagl M. Warfarin pretreatment does not lead
to increased bleeding tendency during cardiac surgery. J Cardiothorac
Vasc Anesth 1995;9:250–4.
 The Post Coronary Artery Bypass Graft Trial Investigators. The effect of
aggressive lowering of low-density lipoprotein cholesterol levels and
low-dose anticoagulation on obstructive changes in saphenous-vein
coronary-artery bypass grafts. N Engl J Med 1997;336:153–63.
 Potapov EV, Ignatenko S, Nasseri BA. Clinical significance of PlA
polymorphism of platelet GP IIb/IIIa receptors during long-term VAD
support. Ann Thorac Surg 2004;77(3):869–74.
Appendix A. Conference discussion
Dr P. Sergeant (Leuven,Belgium): Atwhat time do you suggest that wedo
these measurements and how have the measurements influenced your
therapies in the immediate postoperative time frame concerning the
Dr Poston: One of problems with trying to develop an algorithm based on
this type of postoperative testing is that aspirin resistance doesn’t typically
manifest until around postop day 3. It unfortunately wasn’t present in
patients before surgery or during the immediate postop assessments on days 1
and 2. This timeframe of identifying aspirin resistant patients is likely to
extend beyond the optimal window for adding additional antiplatelet agents
to improve graft patency, making it difficult to use for guiding an algorithm.
Perhaps attempts at looking at genotype to prospectively identify patients
who most apt to develop this type of aspirin resistance would be a better way
to go about developing an optimal algorithm, and that is our current lab’s
However, I can tell you that identifying aspirin resistance in a third of our
patients after OPCAB has certainly encouraged our group to increase the use
of Plavix. These data suggest it is appropriate to do that.
Dr Sergeant: Can you give us some idea about your heparin therapy during
Dr Poston: It is guided by the kaolin based ACT of greater than 300 s.
Dr Sergeant: Have you tested this under different ACT regimens, because
300 is a rather low regimen?
Dr Poston: No. I haven’t seen any data that suggests that this is an issue
important to early graft failure, but perhaps it is.
Dr A. Wahba (Trondheim, Norway): As you know, the predictive power of
a bedside test such as TEG or others is discussed in the literature. Did you
check whether the predictive power of just recording the blood loss was
probably better than the TEG?
Dr Poston: Predictive for thromboembolic events?
Dr Wahba: Well, predictive for thromboembolic events, yes.
Dr Poston: Yes, there was a significant reduction in 24 h bleeding in
the group that developed early graft failure compared to those that
didn’t. The TEG-MA that was available immediately in the ICU after the
patient arrived from the OR accurately predicts bleeding risk over the next
24 h. In contrast, the clinical finding of reduced perioperative bleeding
was associated with graft thrombosis. Assuming that graft thrombosis
occurs early after OPCAB, 24 h bleeding is unlikely to provide a datapoint
that could be easily incorporated into an algorithm to improve graft
patency. Such difficulties are similar to our ability to predict thrombosis in
patients according to the aspirin-resistance assays. Unfortunately, most of
our efforts have resulted in findings that are associated with, but not
predictive, of graft failure.
Dr I. Kassai (Zalaegerszeg, Hungary):My question is, what isyour strategy
in the worst cases when, for example, the patient has a one-month old LIMA to
LAD graft, and the patient is on combined antiplatelet therapy because of the
one-month old or two-week old stent in the right coronary and another stent
in the carotid artery and you have to graft up to the marginal branch? Do you
stop the antiplatelet drugs?
Dr Poston: Your question gets into areas that are beyond what I can
specifically support with my data, but I am glad to speculate at the best
answer. First of all, these platelet function tests can be used to predict the
risk of bleeding as well as to predict the risk for thrombosis. If the
challenging patient that you mentioned has a reasonable TEG-MA value
preoperatively, he should be at low risk for postoperative bleeding and may
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591 590
be at risk for stent thrombosis should the agents be discontinued. This Download full-text
finding would suggest that it is safest to continue aggressive antiplatelet
agents (e.g. aspirin and plavix) through surgery. On the other hand, if the
patient that is treated with an aggressive antiplatelet therapy to prevent
stent thrombosis is found to have a low TEG-MA, discontinuing the plavix may
be important to avoid the risk of excessive bleeding.
Dr R. van Eps (Maastricht, Netherlands): A couple of studies showed that
in off-pump CAB there is, for example, more thrombin generation after like
24 h, and I was wondering if you did measurements of thrombin generation as
a function of coagulation and whether this correlated with the platelet
function test and thrombotic risk?
Dr Poston: Yes, that is a plan of future studies. We have saved serum
samples every time we have done platelet function testing, and plan on
looking at fibrinogen and thrombin levels. The coagulation cascade also plays
a role in platelet activation. Their assessment could potentially have some
additional value at predicting thrombosis as well.
R. Poston et al. / European Journal of Cardio-thoracic Surgery 27 (2005) 584–591591