Content uploaded by Ruzhen Liu
Author content
All content in this area was uploaded by Ruzhen Liu on Sep 04, 2022
Content may be subject to copyright.
Antifibrinolytics attenuate inflammatory gene expression after
cardiac surgery
Alexander F. L. Later, MD,
a
Laura S. Sitniakowsky, MSc,
b
Joost A. van Hilten, PhD,
b
Leo van de Watering, PhD,
b
Anneke Brand, MD, PhD,
b,c
Nico P. M. Smit, PhD,
d
and
Robert J. M. Klautz, MD, PhD
a
Objectives: Anti-inflammatory effects of tranexamic acid and aprotinin, used to abate perioperative blood loss,
are reported and might be of substantial clinical relevance. The study of messenger ribonucleic acid synthesis
provides a valuable asset in evaluating the inflammatory pathways involved.
Methods: Whole-blood messenger ribonucleic acid expression of 114 inflammatory genes was compared pre-
and postoperatively in 35 patients randomized to receive either placebo, tranexamic acid, or aprotinin. These
results were further confirmed by reverse transcription–polymerase chain reaction.
Results: Of the 23 genes exhibiting independently altered postoperative gene expression levels, 8 were restricted
to the aprotinin group only (growth differentiation factor 3, interleukin 19, interleukin 1 family member 7, trans-
forming growth factor a, tumor necrosis factor superfamily 10, tumor necrosis factor superfamily 12, tumor ne-
crosis factor superfamily 13B, vascular endothelial growth factor a), whereas both aprotinin and tranexamic acid
altered gene expression of 3 genes as compared with placebo (FMS-related tyrosine kinase 3 ligand, growth dif-
ferentiation factor 5, interferon-a8). In general, less upregulation of pro-inflammatory, and more upregulation of
anti-inflammatory, genes was observed for patients treated with antifibrinolytics. Gene expression affected by
aprotinin coded mostly for proteins that function through serine proteases.
Conclusions: This study demonstrates that the use of tranexamic acid and aprotinin results in altered inflamma-
tory pathways on the genomic expression level. We further demonstrate that the use of aprotinin leads to signif-
icant attenuation of the immune response, with several inhibitory effects restricted to the use of aprotinin only.
The results aid in a better understanding of the targets of these drugs, and add to the discussion on which
antifibrinolytic can best be used in the cardiac surgical patient. (J Thorac Cardiovasc Surg 2013;145:1611-6)
Supplemental material is available online.
The pharmacologic agents aprotinin and tranexamic acid
have proved to reduce substantially perioperative bleeding
and transfusion requirements in cardiac surgery.
1
Aprotinin
not only abates excess fibrinolysis, but also inhibits serine
proteases involved in triggering and signaling pathways.
2
Together with attenuation of chemotactic mediators
involved in the coagulation cascade,
3
this explains some
of the reported anti-inflammatory effects of aprotinin.
4-6
The routine use of aprotinin was questioned after results
from several observational studies indicated an increased
risk of serious adverse events.
7,8
When aprotinin-treated,
high-risk cardiac surgery patients showed a 50%increase
in 30-day mortality compared with placebo or other antifi-
brinolytics,
9
aprotinin quickly lost its dominant market
position in favor of tranexamic acid. This antifibrinolytic
is reported to be equally effective in reducing blood
loss,
10
and is thought to have similar anti-inflammatory
properties as well.
11
Although the impact of aprotinin and
tranexamic acid on the coagulation system have been stud-
ied extensively, their effects on inflammation remain
unclear. A better understanding of the signaling and regula-
tory pathways involved helps us to determine which posi-
tive and negative side effects are associated with their use.
This is even more relevant now that the European Medicine
Agency has decided that, because the results of the BART
study on which this suspension was based were deemed
unreliable, the current suspension of the marketing of apro-
tinin should be lifted.
9,12
We studied pre- and postoperative
messenger ribonucleic acid (mRNA) expression levels of
From the Department of Cardiothoracic Surgery,
a
Department of Haematology and
Blood Transfusion,
c
and Department of Clinical Chemistry,
d
Leiden University
Medical Center, Leiden, The Netherlands; and Sanquin Blood Bank Southwest,
b
Leiden, The Netherlands.
This study was funded by intramural funding only; no commercial or nonprofit fund-
ing was used for this work.
Disclosures: Authors have nothing to disclose with regard to commercial support.
Clinical Trial Registration: The original trial from which the 35 subjects described
herein were selected is registered at the Dutch Trial Register under no.
ISRCTN00157697.
Received for publication March 6, 2012; revisions received Oct 18, 2012; accepted
for publication Nov 6, 2012; available ahead of print Jan 18, 2013.
Address for reprints: Alexander F. L. Later, MD, Department of Cardiothoracic
Surgery, Leiden University Medical Center, Albinusdreef 2, 2300 RC Leiden,
The Netherlands (E-mail: a.f.l.later@lumc.nl).
0022-5223/$36.00
Copyright Ó2013 by The American Association for Thoracic Surgery
http://dx.doi.org/10.1016/j.jtcvs.2012.11.042
The Journal of Thoracic and Cardiovascular Surgery cVolume 145, Number 6 1611
Later et al Perioperative Management
PM
114 inflammatory genes in 35 patients randomized to
receive either placebo, tranexamic acid, or aprotinin. Our
primary end point was a direct comparison of aprotinin
and tranexamic acid on genomic expression of genes
involved in the inflammatory response, with effects on
relevant clinical outcomes serving as secondary outcomes.
METHODS
Patient Enrollment
We performed a single-institution case–control study in which pre- and
postoperative whole-blood samples were collected from 199 patients who
participated in a double-blind, randomized, placebo-controlled trial evalu-
ating the effectiveness of tranexamic acid and aprotinin between June 2004
and October 2006.
10
Patients were scheduled for primary coronary bypass
grafting, valvular surgery, or a combination of the 2 using cardiopulmonary
bypass (CPB). Exclusion criteria encompassed active inflammatory disease
and preoperative use of corticosteroids. A total of 35 patients (12 placebo,
12 tranexamic acid, 11 aprotinin), matched for age, gender, CPB time, the
number of red cell blood products transfused, intraoperative useof steroids,
and the incidence of postoperative multiorgan failure (MOF) were ana-
lyzed for prospective analysis of differences in pre- and postoperative
inflammatory gene expression (GE) profiles. This case–control substudy
was approved by our internal review board, and informed consent was
obtained from each participant.
Interventions
A conventional operative approach was used in all patients, including
midline sternotomy and systemic heparinization. Patients received anesthe-
sia according to a ‘‘fast track’’ protocol in which total intravenous anesthesia
was used through target-controlled infusion. At the time of the study, 2 of 6
cardiothoracic anesthesiologists employed at our hospital administered cor-
ticosteroids routinely to their patients, either 8 mg dexamethasone or 80 mg
hydrocortisone. Because anesthesiologistswere assigned randomly to the op-
erations, from a patient’s perspective, the administration of steroids was ran-
dom. Either pulsatile or nonpulsatile blood flow was established with
a centrifugalpump and hollow-fibermembrane oxygenatorprimed with crys-
talloid solution. A tip-to-tip phosphorylcholine coating of all tubing was
used. Patients were randomized to receive intraoperatively either high-dose
aprotinin (2 310
6
kallikrein inhibiting units aprotinin loading dose,
2310
6
kallikrein inhibiting units added to the CPB priming solution, and
a continuous infusion of 5 310
5
kallikrein inhibiting units/hour during
CPB),
13
full-dose tranexamic acid (1 g loading dose, 500 mg added to the
CPB system prime and a continuous infusion of 400 mg/hour during
CPB),
14
or placebo (0.9%saline solution according to an identical regime).
Blood temperature was keptbetween 35Cand37
C, and myocardialprotec-
tion was achievedby intermittent antegrade warm blood cardioplegia accord-
ing to the Calafiore protocol. Before heparinization and after reversal with
protamine, blood from the operative field wasretrieved in a cell saver device,
but was only washed and returned to the patient if blood loss exceeded 500
mL. Postoperative mediastinal chest tube blood loss was not reinfused.
Blood Sampling
Whole blood samples were withdrawn by vein puncture the day before
surgery and 24 hours after the start of CPB. All samples were stored imme-
diately at 4C and mRNA was stabilized within 24 hours using RNALater
(Applied Biosystems/Ambion, Austin, Tex). Ribonucleic acid extraction
was performed with the RiboPure-blood kit (Applied Biosystems/Ambion)
per the manufacturer’s recommendation, including the DNase digestion
step to remove residual deoxyribonucleic acid (DNA). Ribonucleic acid
quantity and quality were determined using spectrophotometry, for which
a 260/280 ratio>1.8 was deemed indicative of ribonucleic acid (RNA) of
sufficient quality. In addition, we verified RNA integrity for a random
selection of 5 samples with the BioAnalyzer (Agilent Technologies, Santa
Clara, Calif).
Arrays
Inflammatory GE analysis was performed with a commercially avail-
able oligo GE array (OHS-021; SA Biosciences, Frederick, Md) containing
128 hybridization spots; 114 spots for cytokines and 14 spots for house-
keeping genes, positive controls, and negative controls. Amplification
and labeling of RNAwas performed according to the TrueAMP 2.0 kit pro-
tocol (SA Biosciences). Complementary RNAwas hybridized for 18 hours
on the array. Chemiluminescent signal detection took place using a Chem-
iDoc XRS system (Bio-Rad Laboratories, Hercules, Calif), and quantifica-
tion of spot intensities was conducted using the GEArray Analysis Suite
(SA Biosciences). Spot intensities were corrected for the background sig-
nal by subtracting from each individual spot intensity the average value of
all spot intensities per array. Gene expression values reported are fold
changes (DGE), where values <1.0 represent a downregulation and val-
ues>1.0 represent upregulation of the gene.
Reverse Transcription–Polymerase Chain Reaction
(RT-PCR)
Based on between-treatment group differences in DGE, 8 genes were
selected for confirmation with 2-step reverse transcriptase real-time poly-
merase chain reaction (PCR). Assays were developed in our laboratory
(Table E1). Complementary DNA was constructed using RNA applied to
the arrays. Reverse transcription was performed with the SuperScript II re-
verse transcriptase, using random primers (Invitrogen, Carlsbad, Calif).The
7500 Fast RT-PCR (Applied Biosystems, Foster City, Calif) was used on
standard modus and standard program with the GE master mix(Applied Bi-
osystems). Each sample was run 4 times. In each run, a reference sample,
consisting of pooled complementary DNA of 6 volunteer blood donors,
was taken along in 4 fold. All RT-PCR output threshold cycle values were
corrected for this reference sample in the same run.
Clinical Evaluation and Measurements
Preoperative baseline characteristics such as age, weight, gender, logis-
tic EuroSCORE, use of medication, and relevant comorbidity were
recorded. The preoperative laboratory investigation included a full coagu-
lation and white blood cell profile. Intraoperatively and 24 hours postoper-
atively, the type of surgery, operation time, CPB time, crossclamp time,
Abbreviations and Acronyms
CPB ¼cardiopulmonary bypass
DNA ¼deoxyribonucleic acid
GDF ¼growth differentiation factor
GE ¼gene expression
IFN ¼interferon
IL ¼interleukin
MOF ¼multiorgan failure
mRNA ¼messenger ribonucleic acid
PCR ¼polymerase chain reaction
RNA ¼ribonucleic acid
RT-PCR ¼reverse transcription–polymerase
chain reaction
TGF ¼transforming growth factor
TNF ¼tumor necrosis factor
TNFsf ¼tumor necrosis factor superfamily
Perioperative Management Later et al
1612 The Journal of Thoracic and Cardiovascular Surgery cJune 2013
PM
blood product use, and maximum white blood cell count were recorded. To
evaluate the clinical impact of a changed postoperative inflammatory pro-
file, the incidence of the systemic inflammatory response syndrome and
MOF was evaluated on a daily basis during the complete period of recovery
in the intensive care unit, using the systemic inflammatory response syn-
drome criteria proposed by the American College of Chest Physicians
15
and the MOF criteria by Knaus and colleauges.
16
All intra- and postoper-
ative measurements were done by caretakers blinded to treatment group
assignment.
Statistical Analysis
Clinical data were expressed as median interquartile range. Demo-
graphic and perioperative clinical characteristics were compared among
the 3 treatment groups using 1-way analysis of variance for continuous var-
iables, and the c
2
test for categoric variables. To test for differences
between pre- and postoperative GE values, Wilcoxon signed rank tests
were performed. Mann-Whitney Utests were performed for between-
group comparisons. In addition, for the RT-PCR data, Mann-Whitney
Utests were performed to test the difference in RQr among the treatment
groups. Pvalues were calculated using 2-sided tests and were considered
significant when <.05. The Benjamini-Hochberg correction was used to
correct for the problem of multiple testing during the multiple comparison
gene analysis. Statistical analysis was performed using PASW Statistics
software, version 17.0.0 (SPSS, Inc, Chicago, Ill).
RESULTS
Patients
Patient demographics are described in Table 1. Preoper-
ative white blood cell count, platelet count, and coagulation
profiles were within the normal range for all patients (data
not shown). Patients were matched with respect to age, gen-
der, CPB time, the number of blood products transfused,
intraoperative use of steroids, and the incidence of postop-
erative MOF. No substantial differences in terms of the type
of surgery performed, the operation time, the time on CPB,
and the myocardial arrest time were detected. In the postop-
erative patient variables, an independent lower maximum
leukocyte count was seen in patients treated with either
tranexamic acid (P¼.040) or aprotinin (P¼.018), with
no important difference between them. No ischemic stroke
or seizures were seen.
Arrays
Quantity and quality of RNA was good for all samples
(data not shown). After hybridization and subtraction of the
background signal, 39 of 114 genes on the array differed
with respect to pre- and postoperative GE across 1 of the 3
treatment groups, with 23 genes showing significant changes
between pre- and postoperative GE levels in at least 1 of the
groups. Eleven of the 23 genes showed important between-
group differences when compared with placebo: 8 genes
showed differences in the aprotinin group only (growth dif-
ferentiation factor [GDF] 3, interleukin [IL]-19, IL-1 family
member 7, vascular endothelial growth factor adownregu-
lated, transforming growth factor [TGF]-a, tumor necrosis
factor superfamily [TNFsf] 10, TNFsf12, TNFsf13 upregu-
lated), whereas 3 genes (FMS-related tyrosine kinase 3 ligand
downregulated, GDF-5, and interferon [IFN]-a8 upregu-
lated) showed differences in both aprotinin- and tranexamic
acid-treatedpatients. Two genes (platelet-derived growth fac-
tor aand TNFsf13 upregulated) showed a trend (P<.10)
toward independence when aprotinin-treated patients were
compared with placebo (data are described in Table E2).
Gene Selection for RT-PCR Confirmation
Based on functional relevance and the amount of up- or
downregulation (Table E2), 6 genes were selected for confir-
mation with RT-PCR: FLT3-ligand, GDF-5, IFN-a8, TGF-
a1, TNFsf10, and TNFsf13. We further selected 4 additional
genes: CD40 ligand, TNF, IL-8, and TNFsf13b. CD40
ligand was chosen for its implication in platelet activation
and reports on diminished platelet activation in aprotinin-
treated patients.
17
Interleukin 8 had our special interest be-
cause array data indicated postoperative downregulation in
all patient groups, whereas increased plasma levels have
been related to patient outcome after cardiac surgery.
18,19
Tumor necrosis factor and TNFsf13B were further chosen
for their close functional connection to TNFsf13.
Confirmation RT-PCR
Reverse transcription-PCR assays were developed suc-
cessfully for 8 selected genes. Unfortunately, GDF-5 and
IFN-a8 GE could not be confirmed with RT-PCR because
of technical difficulties during assay development. Results
of the RT-PCR confirmation are presented in Table 2.
Reverse transcription-PCR results were remarkably similar
to GEArray results, the only difference being more
pronounced upregulation in RT-PCR-confirmed upregu-
lated genes. No independent DGE for IL-8, TNFsf10,
and TNFsf13B could be found for patients treated with
aprotinin. Also, TNF GE did not change substantially in
tranexamic acid-treated patients. When we tested for
between-group differences (placebo vs tranexamic acid or
aprotinin), TGF upregulation was less for patients treated
with aprotinin (DGE 1.99 vs 5.37, respectively;
P¼.019). The already noted unchanged GE of TNFsf10
in aprotinin-treated patients was confirmed and proved rel-
evant when compared with placebo (median DGE, 0.99 vs
1.98; P¼.042). For TNFsf13, a trend toward independently
lower upregulation in GE was seen in patients treated with
aprotinin (DGE, 1.68 vs 2.47; P¼.074). None of the genes
tested on RT-PCR showed important differences between
patients treated with tranexamic acid and placebo.
Differences in Array GE Between Patients
Developing MOF Versus Non-MOF Patients
We tested further for differences in GEArray patterns of
patients developing MOF (n ¼16) compared with patients
not developing MOF (n ¼19). Seventeen of the 39 genes
detectable on the array showed no substantial up- or down-
regulation after surgery, leaving 22 genes for further
Later et al Perioperative Management
The Journal of Thoracic and Cardiovascular Surgery cVolume 145, Number 6 1613
PM
analysis. Gene expression IL-12a, IL-17b, inhibin a, and in-
hibin-bsubunit A proved importantly altered after surgery,
an effect not seen in the analysis of antifibrinolytic mediated
GE. In contrast to non-MOF patients, MOF patients showed
unchanged GE inhibin a, inhibin-bsubunit A, and platelet-
derived growth factor b24 hours after surgery, whereas
IL-12a, IL-17b, and TNFsf10 GE was changed. However,
direct comparisons between MOF and non-MOF patients
yielded no independent differences (data are described in
Table E3).
The Effect of Perioperative Corticosteroids on
Array GE
Last, we tested for the effect of intraoperative administra-
tion of corticosteroids on inflammatory GE. A total of 8
patients (23%) received steroids intraoperatively, equally
distributed across the 3 treatment groups (P¼.813). Of
the 39 detectable genes, 20 genes had a substantially altered
GE in either corticosteroid- or noncorticosteroid-treated
patients (data are described in Table E5). Again, overall
differences in GE were less (DGE range, 0.46-1.73) than
the differences observed in patients treated with either anti-
fibrinolytic; only lymphotoxin-band TNFsf13 expression
differed, with stronger lymphotoxin-bdownregulation in
the corticosteroid group (DGE, 0.66 vs 0.85; P¼.003),
and less TNFsf13 upregulation (1.18 vs 1.27, P¼.023).
DISCUSSION
The effects of aprotinin and tranexamic acid on surgical
bleeding have been studied extensively, with reports indi-
cating aprotinin to be most effective,
20
whereas others con-
clude there is hardly any difference between them.
21
More
important is the issue of whether the use of antifibrinolytics
is clinically advantageous. The contrary may be true;
several studies reported higher incidences of complications
after the use of aprotinin
7-9
and tranexamic acid.
22
TABLE 1. Demographic and clinical data of 35 patients
Characteristic Placebo (n ¼12) Tranexamic acid (n ¼12) Aprotinin (n ¼11) Pvalue
Age, years 61.4 (53.3-73.4) 65.3 (54.7-75.3) 68.6 (61.0-73.6) .819
Gender, male/female, n 11/1 11/1 8/3 .331
EuroSCORE, %2.5 (1.0-3.3) 2.6 (1.6-5.4) 3.3 (1.5-4.5)
Left ventricular ejection fraction, n .585
Moderate 5 5 3
Poor 1 0 0
Diabetes mellitus, n .621
Type 1 0 1 0
Type 2 1 1 2
Preoperative medication, n
Acetylsalicylic acid 7 6 5 .820
Calcium antagonists 2 2 0 .355
Statins 6 8 7 .676
Operation
Type of surgery, n
CABG only 5 5 3 .716
Valve only 5 6 7 .570
CABG and valve 2 1 1 .777
Duration of myocardial arrest, minutes 121 (68-169) 99 (91-138) 100 (70-172) .840
Duration of CPB, minutes 167 (123-222) 139 (114-205) 152 (119-206) .712
Corticosteroids, n 3 2 3 .813
Operation and 24 h postoperative
PRBC, units/patients transfused 21/8 8/6 7/5 .586
FFP, units/patients transfused 13/5 5/2 8/4 .460
Platelets, units/patients transfused 3/3 3/2 2/2 .904
Highest leukocyte count, g/dL 16.1 (12.8-19.6) 12.5 (11.0-13.1) 11.1 (10.3-13.1) .027
Total ICU stay
Use of CVVH 1 2 0 .344
Use of IABP 0 2 0 .104
Myocardial infarction 0 2 0 .104
Incidence of SIRS, n 12 12 11 .388
Incidence of MOF, n 6 5 5 .574
Mortality, n 1 0 0 .373
Data are reported as median (interquartile range). CABG, Coronary artery bypass grafting; CPB, cardiopulmonary bypass; PRBC, packed red blood cell; FFP, fresh frozen
plasma; ICU, intensive care unit; CVVH, continuous veno-venous hemofiltration; IABP, intra-aortal balloon pump; SIRS, systemic inflammatory response syndrome; MOF,multi-
organ failure; EuroSCORE, European System for Cardiac Operative Risk Evaluation.
Perioperative Management Later et al
1614 The Journal of Thoracic and Cardiovascular Surgery cJune 2013
PM
Anti-inflammatory properties of both aprotinin and tranexa-
mic acid have been observed in studies evaluating plasma
cytokine concentrations.
23-26
Seen as messengers orchestrating
the immune response, cytokines play a key role in the
inflammatory response after cardiac surgery. Here we used
the approach to evaluate the immune modulating properties
of antifibrinolytics by studying protein synthesis at its
earliest level—namely, that of transcription of the gene.
This study compared the effects of tranexamic acid and
aprotinin on the genomic expression level in cardiac
surgery. We used an inflammatory pathway RNA gene
expression array to identify unique profiles related to the
use of antifibrinolytics. The results of RT-PCR confirmed
the array technique, indicating that the GEArray is a suitable
tool for screening for genes of interest.
We found that both antifibrinolytics alter inflammatory
GE; less formation of macrophage colonies is reflected by
a diminished upregulation of colony stimulating factor. Tra-
nexamic acid and aprotinin induce a deviation toward a stron-
ger humoral immune response through independently lower
expression of IFN-a8, further aided by less downregulation
of FMS-related tyrosine kinase 3 ligand. Common
antifibrinolytic effects were also seen; aprotinin and tranexa-
mic acid induce less upregulation of GDF-5, involved in both
plasminogen activity and migration of endothelial cells. The
independently lowered TGF-a1 expression is possibly
a result of lower thrombin concentrations, and it reaches in-
dependence in the aprotinin group only, perhaps through
aprotinin’s additional inhibition of chemotrypsin. Last, we
noticed upregulation of vascular endothelial growth factor
ainvolved in synthesis of antiplasminogen activator, thus re-
ducing fibrinolysis and clot preservation. However, most ef-
fects on GE profiles were seen in aprotinin-treated patients
only with upregulation of IL-19, a cytokine with properties
similar to anti-inflammatory IL-10, and IL-1 family member
7, thereby diminishing induction of proinflammatory IFN-g.
Only aprotinin-treated patients lack upregulation of a whole
range of TNFsf subclass proteins involved in apoptosis in-
duction, B-cell proliferation, and proinflammatory cellular
signaling pathways. Last, aprotinin-treated patients exhibit
lower upregulation of platelet-derived growth factor a, re-
flecting less platelet activation, thrombin generation, and
chemotaxis of monocytes and neutrophils.
Altogether, the use of tranexamic acid and aprotinin led
to less upregulation of proinflammatory genes and more
upregulation of anti-inflammatory genes. Considering the
GE impacted exclusively in the aprotinin group, we found
these to include genes coding for proteins that function
through serine proteases and genes with important anti-
inflammatory functioning.
Neither this study, nor the original randomized, con-
trolled trial from which this patient group was derived,
was powered to detect independent between-group differ-
ences in clinical outcomes. However, analysis of GE array
profiles revealed differences between patients who devel-
oped MOF and patients who did not. Gene expression
altered exclusively in MOF patients includes downregula-
tion of IL-12a, a cytokine that augments the cytolytic activ-
ity of natural killer cells, and IL-17b, depressing the release
of TNF-aand IL-1b(Table E3). We further noticed down-
regulation of inhibin-bA, released concurrently with
TNF-aearly in sepsis, prior to IL-6.
27
Although these dif-
ferences were independent when compared with baseline
values, no independent differences between MOF and
non-MOF patients were seen.
Our results provide insight into the cellular mechanisms
through which tranexamic acid and aprotinin exert their im-
mune modulating effects. We identified several genetic path-
ways expressed differentially in patients receiving either
tranexamic acid or aprotinin, with statistical independence
when tested against placebo. Furthermore, several GE pro-
files were unique for the use of aprotinin only, with most ef-
fects induced by aprotinin’s serine protease inhibition. Our
study demonstrates these associations despite the small sam-
ple size, reflecting the strength of the relationship between
antifibrinolytic status and particular pathway regulation.
TABLE 2. DGE RT-PCR
Gene Group n
DGE, median
(IQR)
Pvalue
pre- vs post
Pvalue vs
placebo
CD40lg PL 12 0.39 (0.26-0.51) .002
TX 12 0.39 (0.15-0.62) .008
AP 11 0.24 (0.11-0.37) .006
FLT3lg PL 12 0.26 (0.15-0.37) .002
TX 12 0.34 (0.13-0.54) .003 .133
AP 11 0.34 (0.20-0.47) .003
IL-8 PL 12 0.28 (0.02-0.58) .019
TX 12 0.31 (0.42-0.66) .012
AP 11 0.72 (0.14-1.30) .182
TGF-a1 PL 12 5.37 (3.09-7.66) .002
TX 12 4.05 (1.33-6.78) .002
AP 11 1.99 (0.66-3.33) .008 .019
TNF PL 12 0.71 (0.51-0.92) .005
TX 12 0.78 (0.36-1.19)
AP 11 0.62 (0.49-0.75) .010
TNFsf10 PL 12 1.98 (1.16-2.80) .012
TX 12 2.10 (1.35-2.85) .008
AP 11 1.00 (0.38-1.61) .042
TNFsf13 PL 12 2.47 (1.13-3.81) .003
TX 12 1.99 (1.06-2.92) .002
AP 11 1.68 (1.01-2.35) .075 .074
TNFsf13B PL 12 4.98 (2.51-7.44) .004
TX 12 5.20 (0.62-9.79) .002
AP 11 3.65 (1.02-6.28) .004
The differences in DGE after surgery between placebo and either tranexamic acid or
aprotinin were tested with the Mann-Whitney Utest. Only Pvalues<.20 are reported
here. GE, Gene expression; RT-PCR, reverse transcription–polymerase chain reac-
tion; IQR, interquartile range; CD40lg, CD40 ligand; FLT3lg, FMS-like tyrosine ki-
nase 3 ligand; PL, placebo; TX, tranexamic acid; AP, aprotinin; IL, interleukin; TGF,
transforming growth factor; TNF, tumor necrosis factor; TNFsf, TNF super family
member.
Later et al Perioperative Management
The Journal of Thoracic and Cardiovascular Surgery cVolume 145, Number 6 1615
PM
However, our study has limitations, too. Although patient
groups were well matched according to use of corticoste-
roids, type of surgery, operation duration, and other relevant
patient characteristics, this study lacked true randomization
to the intraoperative use of corticosteroids. An additional
analysis of the effects of corticosteroids on GE, however,
showed only limited effects. Second, our small sample
size makes it difficult to determine whether the attenuation
of the inflammatory gene expression by antifibrinolytics re-
sults in clinical better recovery. Third, one can argue whether
transfusions are responsible for the favorable inflammatory
response seen in antifibrinolytic-treated patients. An addi-
tional analysis concentrating on GE profiles between trans-
fused and nontransfused patient indicated small
nonindependent differences (Table E4). Last, the relatively
long time frame during which we studied alterations in GE
might be responsible for the relatively limited number of
genes with altered postoperative GE, because many plasma
proteins can be measured only shortly after cardiac surgery.
The fact that GE is affected 24 hours after randomization to
placebo, tranexamic acid, or aprotinin, however, indicates
the powerful and long-lasting effects of the use of antifibri-
nolytics in cardiac surgery.
CONCLUSIONS
This study demonstrates that the use of tranexamic acid
and aprotinin alters inflammatory GE profiles. We further
demonstrate that the use of aprotinin leads to independent
attenuation of the immune response, with several inhibitory
effects restricted to the use of aprotinin only, and not
observed in tranexamic acid-treated patients. It is unclear
from these results whether this modulated immune response
plays a causal role in presumed beneficial or detrimental
clinical effects after surgery. However, better understanding
of the targets of these drugs enable possible adjustment of
pharmacologic properties, allowing for more precise alter-
ation of hemostasis after cardiac surgery. Our results can
add further to the discussion on which antifibrinolytic, if
any, can best be used in the cardiac surgical patient.
We thank M. Bogaerts for his contribution and assistance in set-
ting up the RT-PCR confirmation of our array results.
References
1. Henry DA, Moxey AJ, Carless PA, O’Connell D, McClelland KM,
Henderson KM, et al. Anti-fibrinolytic use for minimising perioperative alloge-
neic blood transfusion. Cochrane Database Syst Rev. 1999;4:CD001886.
2. McEvoy MD, Reeves ST, Reves JG, Spinale FG. Aprotinin in cardiac surgery:
a review of conventional and novel mechanisms of action. Anesth Analg.
2007;105:949-62.
3. Levi M, Van Der Poll T. Inflammation and coagulation. Crit Care Med. 2010;38:
S26-34.
4. Day JR, Taylor KM, Lidington EA, Mason JC, Haskard DO, Randi AM, et al.
Aprotinin inhibits proinflammatory activation of endothelial cells by thrombin
through the protease-activated receptor 1. J Thorac Cardiovasc Surg. 2006;
131:21-7.
5. HsiaTY, McQuinn TC, Mukherjee R, Deardorff RL, Squires JE, Stroud RE, et al.
Effects of aprotinin or tranexamic acid on proteolytic/cytokine profiles in infants
after cardiac surgery. Ann Thorac Surg. 2010;89:1843-52.
6. Graham EM, Atz AM, Gillis J, Desantis SM, Haney AL, Deardorff RL, et al.
Differential effects of aprotinin and tranexamic acid on outcomes and cytokine
profiles in neonates undergoing cardiac surgery. J Thorac Cardiovasc Surg.
2012;143:1069-76.
7. Karkouti K, Beattie WS, Dattilo KM, McCluskey SA, Ghannam M, Hamdy A,
et al. A propensity score case–control comparison of aprotinin and tranexamic
acid in high-transfusion-risk cardiac surgery. Transfusion. 2006;46:327-38.
8. Mangano DT, Miao Y, Vuylsteke A, Tudor IC, Juneja R, Filipescu D, et al. Mor-
tality associated with aprotinin during 5 years following coronary artery bypass
graft surgery. JAMA. 2007;297:471-9.
9. Fergusson DA, Hebert PC, Mazer CD, Fremes S, MacAdams C, Murkin JM,
et al. A comparison of aprotinin and lysine analogues in high-risk cardiac
surgery. N Engl J Med. 2008;358:2319-31.
10. Later AF, Maas JJ, Engbers FH, Versteegh MI, Bruggemans EF, Dion RA, et al.
Tranexamic acid and aprotinin in low- and intermediate-risk cardiac surgery:
a non-sponsored, double-blind, randomised, placebo-controlled trial. Eur J Car-
diothorac Surg. 2009;36:322-9.
11. Robertshaw HJ. An anti-inflammatory role for tranexamic acid in cardiac sur-
gery? Crit Care. 2008;12:105.
12. DeAnda A Jr, Spiess BD. Aprotinin revisited. J Thorac Cardiovasc Surg. 2012;
144:998-1002.
13. RoystonD, Bidstrup BP, Taylor KM, Sapsford RN. Effect of aprotinin on need for
blood transfusion after repeat open-heart surgery. Lancet. 1987;2:1289-91.
14. Dowd NP, Karski JM, Cheng DC, Carroll JA, Lin Y, James RL, et al. Pharmaco-
kinetics of tranexamic acid during cardiopulmonary bypass. Anesthesiology.
2002;97:390-9.
15. AmericanCollege of Chest Phy sicians/Society of Critical Care Medicine Confer-
ence. definitions for sepsis and organ failure and guidelines for the use of inno-
vative therapies in sepsis. Chest. 1992;101:1644-55.
16. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. Prognosis in acute organ-
system failure. Ann Surg. 1985;202:685-93.
17. Mengistu AM, Rohm KD, Boldt J, Mayer J, Suttner SW, Piper SN. The influence
of aprotinin and tranexamic acid on platelet function and postoperative blood loss
in cardiac surgery. Anesth Analg. 2008;107:391-7.
18. Calafiore AM, Teodori G, Di GG, Bosco G, Mezzetti A, Lapenna D, et al. Inter-
mittent antegrade cardioplegia: warm blood vs cold crystalloid: a clinical study.
J Cardiovasc Surg (Torino). 1994;35:179-84.
19. Sablotzki A, Friedrich I, Muhling J, Dehne MG, Spillner J, Silber RE, et al. The
systemic inflammatory response syndrome following cardiac surgery: different
expression of proinflammatory cytokines and procalcitonin in patients with
and without multiorgan dysfunctions. Perfusion. 2002;17:103-9.
20. Porte RJ, Leebeek FW. Pharmacological strategies to decrease transfusion
requirements in patients undergoing surgery. Drugs. 2002;62:2193-211.
21. Casati V, Guzzon D, Oppizzi M, Bellotti F, Franco A, Gerli C, et al. Tranexamic
acid compared with high-dose aprotinin in primary elective heart operations:
effects on perioperative bleeding and allogeneic transfusions. J Thorac Cardio-
vasc Surg. 2000;120:520-7.
22. Murkin JM, Falter F, Granton J, Young B, Burt C, Chu M. High-dose tranexamic
acid is associated with nonischemic clinical seizures in cardiac surgical patients.
Anesth Analg. 2010;110:350-3.
23. Pruefer D, Makowski J, Dahm M, Guth S, Oelert H, Darius H, et al. Aprotinin
inhibits leukocyte–endothelial cell interactions after hemorrhage and reperfu-
sion. Ann Thorac Surg. 2003;75:210-5.
24. Greilich PE, Brouse CF, Whitten CW, Chi L, Dimaio JM, Jessen ME. Antifibri-
nolytic therapy during cardiopulmonary bypass reduces proinflammatory
cytokine levels: a randomized, double-blind, placebo-controlled study of
epsilon-aminocaproic acid and aprotinin. JThorac Cardiovasc Surg. 2003;126:
1498-503.
25. Tassani P, Augustin N, Barankay A, Braun SL, Zaccaria F, Richter JA. High-dose
aprotinin modulates the balance between proinflammatory and anti-
inflammatory responses during coronary artery bypass graft surgery. J Cardio-
thorac Vasc Anesth. 2000;14:682-6.
26. Jimenez JJ, Iribarren JL, Lorente L, Rodriguez JM, Hernandez D, Nassar I, et al.
Tranexamic acid attenuates inflammatory response in cardiopulmonary bypass
surgery through blockade of fibrinolysis: a case control study followed by
a randomized double-blind controlled trial. Crit Care. 2007;11:R117.
27. Jones KL, de Kretser DM, Patella S, Phillips DJ. Activin A and follistatin in
systemic inflammation. Mol Cell Endocrinol. 2004;225:119-25.
Perioperative Management Later et al
1616 The Journal of Thoracic and Cardiovascular Surgery cJune 2013
PM
TABLE E1. Sequences of primers and probes for the reverse transcription–polymerase chain reaction
Selected
genes REFSEQ ID Forward primer Reverse primer Probe
CD40lg NM_000074 CCA GGT GCT TCG GTG TTT GT CCA GTG CCA TGG CTC ACT T AAT GTG ACT GAT CCA AGC
FLT3lg NM_001459 TGG AGC GGC TCA AGA CTG T TTC ACG CGC TCC AGC AA TGG GTC CAA GAT GC
IL-8 NM_000584 CAC CGG AAG GAA CCATCT CA AGA GCC ACG GCC AGC TT TGT GTG TAA ACATGA CTT C
TGF-aNM_003236 TCC CTT GGG CCA GAT ATG TG TCC GTT GAT TGG TCT CTA AGC A TTG AGG CTT GAC TGT AGC AT
TNF NM_000594 CTT TGG GAT CAT TGC CCT GT GGA GGC GTT TGG GAA GGT T AGG AGG ACG AAC ATC C
TNFsf10 NM_003810 GCT CTG GGC CGC AAA AT AGG AAT GAA TGC CCA CTC CTT ACT CCT GGG AAT CAT
TNFsf13 NM_003808 GCA GGA ACA GAG GCG TCT TC TGG GAATGA AAA GGG AAA AGT G TTT GGC TCC CCG TTC C
TNFsf13B NM_006573 GGC CCC AAC CTT CAA AGT TC GCG TGA CTG CTC CCT TTC TG AGTAGT GAT ATG GAT GAC TCC
CD40lg, CD40 ligand; FLT3lg, FMS-like tyrosine kinase 3 ligand; IL-8, interleukin 8; TGF-a, transforming growth factor alpha; TNF, tumor necrosis factor; TNFsf, TNF super
family member; REFSEQ ID, NCBI reference sequence ID.
APPENDIX E1.
The primer search was conducted with Primer
express software (version 2.0; Applied Biosystems,
Inc, Foster City, Calif). To confirm the uniqueness of
the primer sequences for the gene of interest,
a nucleotide blast was performed at http://www.
ncbi.nlm.nih.gov/BLAST/Blast.cgi?PAGE¼Nucleotides&
PROGRAM¼blastn&MEGABLAST¼on&BLAST_PROG
RAMS¼megaBlast&PAGE_TYPE¼BlastSearch&SHOW_
DEFAULTS¼on.
Genes and sequences are depicted in Tab l e E 1 .
All probes were labeled with fluorescein amidite
dye and had nonfluorescent minor groove binder
quenchers.
Later et al Perioperative Management
The Journal of Thoracic and Cardiovascular Surgery cVolume 145, Number 6 1616.e1
PM
TABLE E2. Fold changes in cytokine gene expression on the array for
patients receiving placebo (PL), tranexamic acid (TX), and aprotinin
(AP)
Gene
Medication
group n
RQ, median
(IQR)
Pvalue,
pre vs
post
Pvalue
vs group 1
CSF1 PL 12 0.89 (0.85-1.01) .005
TX 12 0.96 (0.86-1.07) NS NS
AP 11 1.06 (0.79-1.21) NS NS
CD40lg PL 12 0.68 (0.58-0.78) .004
TX 12 0.77 (0.58-0.90) .002 NS
AP 11 0.71 (0.57-0.84) .004 NS
FLT3lg PL 12 0.52 (0.43-0.59) .002
TX 12 0.72 (0.51-0.91) .003 .033
AP 11 0.69 (0.56-0.76) .021 .027
GDF3 PL 12 0.93 (0.83-1.24) NS
TX 12 1.02 (0.79-1.19) NS NS
AP 11 1.24 (1.05-1.41) .033 .049
GDF5 PL 12 2.23 (1.64-3.44) .002
TX 12 1.38 (1.07-1.76) .005 .013
AP 11 1.11 (1.06-1.48) .003 .002
IFN-a8 PL 12 1.32 (1.13-1.40) .002
TX 12 1.08 (1.02-1.19) .010 .008
AP 11 0.98 (0.94-1.07) NS .001
TXLN-aPL 12 0.96 (0.82-1.08) NS
TX 12 1.01 (0.86-1.10) NS NS
AP 11 0.88 (0.76-0.97) .033 NS
IL-19 PL 12 0.95 (0.86-1.19) NS
TX 12 1.07 (0.92-1.21) NS NS
AP 11 1.19 (1.01-1.44) .026 .049
IL-1bPL 12 1.10 (0.95-1.36) .041
TX 12 1.24 (1.01-1.35) .006 NS
AP 11 1.33 (0.90-1.53) .026 NS
IL-1F7 PL 12 0.97 (0.90-1.10) NS
TX 12 1.05 (0.93-1.16) NS NS
AP 11 1.14 (1.02-1.26) .033 .027
IL-8 PL 12 0.50 (1.80-0.91) .034
TX 12 0.43 (0.23-0.61) .015 NS
AP 11 0.58 (0.38-0.69) .041 NS
LT-bPL 12 0.86 (0.63-1.00) .012
TX 12 0.79 (0.59-1.17) .028 NS
AP 11 0.83 (0.64-1.15) .041 NS
PDGF aPL 12 1.23 (1.13-1.33) .004
TX 12 1.12 (1.06-1.24) .019 NS
AP 11 1.10 (1.07-1.27) .013 NS
TGF-aPL 12 1.60 (1.51-1.85) .003
TX 12 1.43 (1.17-1.60) .002 NS
AP 11 1.16 (0.98-1.39) .016 .014
TNF PL 12 0.87 (0.76-0.98) .015
TX 12 0.92 (0.83-1.11) NS NS
AP 11 0.75 (0.67-0.94) .016 NS
TNFsf10 PL 12 1.40 (1.07-1.73) .041
TX 12 1.36 (1.18-1.61) .010 NS
AP 11 0.93 (0.68-1.34) NS .036
TNFsf12 PL 12 1.21 (0.99-1.47) .028
TX 12 1.04 (0.85-1.14) NS NS
AP 11 0.92 (0.77-1.03) NS .003
(Continued)
TABLE E2. Continued
Gene
Medication
group n
RQ, median
(IQR)
Pvalue,
pre vs
post
Pvalue
vs group 1
TNFsf13 PL 12 1.21 (1.08-1.35) .006
TX 12 1.18 (1.06-1.26) .015 NS
AP 11 1.08 (0.93-1.21) NS NS
TNFsf13b PL 12 1.87 (1.24-4.00) .002
TX 12 1.72 (1.49-2.27) .002 NS
AP 11 1.24 (0.74-1.67) NS .027
TNFsf14 PL 12 1.27 (1.12-1.45) .005
TX 12 1.34 (1.04-1.52) .023 NS
AP 11 1.07 (1.05-1.27) .021 NS
TNFsf9 PL 12 0.71 (0.41-1.08) .034
TX 12 0.68 (0.57-0.95) .023 NS
AP 11 0.75 (0.65-0.98) .016 NS
VEGF aPL 12 0.93 (0.90-0.98) .019
TX 12 1.02 (0.90-1.08) NS NS
AP 11 1.03 (0.93-1.08) NS .036
VEGF bPL 12 0.87 (0.72-0.93) .034
TX 12 0.84 (0.72-1.07) .023 NS
AP 11 0.70 (0.67-0.88) .006 NS
Gene expression (GE) values on the array. Pre- and postoperativevalues are expressed
as fold changes compared with the average expression on each array. The difference
in GE values are fold changes after surgery compared with before surgery. CSF1, Col-
ony stimulating factor-1; RQ, relative quantitation; IQR, interquartile range; CSF, ce-
rebrospinal fluid; PL, placebo; TX, tranexamic acid; AP, aprotinin; NS, not significant;
CD40lg, CD40 ligand; FLT3lg, FMS-like tyrosine kinase 3 ligand; GDF, growth dif-
ferentiation factor; IFN, interferon; TXLN, taxilin; IL, interleukin; LT, lymphotoxin;
PDGF, platelet-derived growth factor; TGF, transforming growth factor; TNF, tumor
necrosis factor; TNFsf, TNF super family member; VEGF, vascular endothelial
growth factor.
Perioperative Management Later et al
1616.e2 The Journal of Thoracic and Cardiovascular Surgery cJune 2013
PM
TABLE E3. Fold changes in cytokine gene expression on the array for
patients developing and not developing MOF
Gene MOF n
RQ, median
(IQR)
Pvalue,
pre vs post
Between-
group
Pvalue
FLT3lg No 19 0.67 (0.51-0.84) .001
Yes 16 0.55 (0.49-0.69) .000 NS
GDF5 No 19 1.38 (0.11-1.85) .000 NS
Yes 16 1.56 (1.11-3.44) .001
IFN-a8 No 19 1.07 (1.00-1.15) .003 NS
Yes 16 1.13 (1.07-1.36) .005
IL-12aNo 19 1.00 (0.94-1.15) NS NS
Yes 16 0.94 (0.91-1.02) .023
TXLN-aNo 19 0.59 (0.80-1.04) .018 NS
Yes 16 0.99 (0.88-1.08) NS
IL-17bNo 19 0.95 (0.86-1.23) NS NS
Yes 16 0.92 (0.76-1.02) .049
IL-1bNo 19 1.17 (0.98-1.35) .003 NS
Yes 16 1.21 (0.94-1.43) .011
IL-8 No 19 0.50 (0.24-0.64) .004 NS
Yes 16 0.48 (0.23-0.67) .011
INH-aNo 19 1.08 (1.01-1.16) .013 NS
Yes 16 1.06 (0.97-1.17) NS
INH-bA No 19 1.04 (0.99-1.16) .040 NS
Yes 16 0.95 (1.12-1.07) NS
LT-bNo 19 0.83 (0.74-1.15) .027 NS
Yes 16 0.77 (0.57-0.95) .001
PDGF-aNo 19 1.22 (1.09-1.30) .001 NS
Yes 16 1.13 (1.07-1.24) .002
PDGF-bNo 19 1.10 (0.99-1.30) .013 NS
Yes 16 1.04 (0.84-1.21) NS
TGF-aNo 19 1.30 (1.11-1.57) .001 NS
Yes 16 1.49 (1.88-1.73) .001
TNF No 19 0.90 (0.80-1.00) .016 NS
Yes 16 0.82 (0.70-1.02) .010
TNFsf10 No 19 1.30 (0.92-1.46) NS NS
Yes 16 1.34 (1.06-1.69) .034
TNFsf13 No 19 1.18 (1.00-1.23) .008 NS
Yes 16 1.14 (1.07-1.26) .001
TNFsf13B No 19 1.49 (1.07-1.76) .001 NS
Yes 16 1.81 (1.29-3.10) .002 NS
TNFsf14 No 19 1.24 (1.04-1.53) .005
Yes 16 1.17 (1.06-1.41) .001 NS
CD40lg No 19 0.69 (0.55-0.80) .000
Yes 16 0.73 (0.61-0.92) .001 NS
TNFsf9 Yes 19 0.74 (0.55-1.11) .011
No 16 0.65 (0.47-0.86) .000 NS
VEGF bYes 19 0.83 (0.70-0.91) .005
No 16 0.81 (0.69-1.05) .023 NS
Gene expression (GE) values on the array. Pre- and postoperativevalues are expressed
as fold changes compared with the average expression on each array. The difference
in GE values are fold changes after surgery compared with before surgery. MOF, Mul-
tiorgan failure; RQ, relative quantitation; IQR, interquartile range; FLT3LG, FMS-
like tyrosine kinase 3 ligand; NS, not significant; GDF, growth differentiation factor;
IFN, interferon; IL, interleukin; TXLN, taxilin; INH, inhibin; LT, lymphotoxin; PDGF,
platelet-derived growth factor; TGF, transforming growth factor; TNF, tumor necrosis
factor; TNFsf, TNF super family member; CD40lg, CD40 ligand; VEGF, vascular en-
dothelial growth factor.
TABLE E4. Fold changes in cytokine gene expression on the array for
patients transfused and not transfused with either packed red blood
cells, fresh frozen plasma, or thrombocytes
Gene Transfused n
RQ, median
(IQR)
Pvalue,
pre vs post
Between-
group
Pvalue
FLT3lg No 14 0.73 (0.54-0.88) .001
Yes 21 0.56 (0.48-0.70) .000 NS
GDF3 No 14 1.03 (0.90-1.22) .030
Yes 21 1.03 (0.79-1.33) NS NS
GDF5 No 14 1.48 (1.21-1.92) .001
Yes 21 1.38 (1.08-2.36) .000 NS
IFN-a8 No 14 1.08 (1.03-1.25) .016
Yes 21 1.13 (0.98-1.32) .001 NS
IL-19 No 14 1.07 (0.96-1.20) .022
Yes 21 1.05 (0.84-1.25) NS NS
IL-1 No 14 1.25 (1.02-1.35) .030
Yes 21 1.20 (0.96-1.43) .002 NS
IL-1F7 No 14 1.04 (0.93-1.14) .006
Yes 21 1.08 (0.93-1.18) NS NS
IL-8 No 14 0.52 (0.27-0.63) .002
Yes 21 0.42 (0.21-0.85) .013 NS
INH-aNo 14 1.04 (0.99-1.14) .008
Yes 21 1.08 (0.98-1.16) NS NS
LT-bNo 14 0.88 (0.55-1.18) .026
Yes 21 0.83 (0.65-0.95) .002 NS
PDGF aNo 14 1.21 (1.03-1.25) .004
Yes 21 1.22 (0.10-1.32) .001 NS
PDGF bNo 14 0.99 (1.57-1.26) .019
Yes 21 1.07 (0.88-1.20) NS NS
TGF-aNo 14 1.43 (1.19-1.65) .005
Yes 21 1.41 (1.09-1.76) .000 NS
TNF No 14 0.88 (0.79-1.11) NS
Yes 21 0.85 (0.72-0.99) .003 NS
TNFsf10 No 14 1.32 (1.03-1.43) NS
Yes 21 1.33 (0.82-1.75) .005 NS
TNFsf13 No 14 1.14 (1.05-1.28) .009
Yes 21 1.17 (1.03-1.26) .001 NS
TNFsf13b No 14 1.69 (1.47-2.00) .004
Yes 21 1.50 (1.07-2.76) .000 NS
TNFsf14 No 14 1.23 (1.05-1.62) .019
Yes 21 1.11 (1.04-1.44) .000 NS
CD40lg No 14 0.79 (0.67-0.86) .001
Yes 21 0.66 (0.55-0.77) .000 NS
CD70 No 14 1.05 (0.97-1.11) .035
Yes 21 1.03 (0.91-1.13) NS NS
TNFsf9 No 14 0.74 (0.58-1.02) .008
Yes 21 0.70 (0.46-0.96) .009 NS
VEGF bNo 14 0.84 (0.70-1.04) .001
Yes 21 0.75 (0.69-0.89) .007 NS
Gene expression (GE) values on the array.Pre- and postoperative values are expressed
as fold changes compared with the average expression on each array. The difference
in GE values are fold changes after surgery compared with before surgery. RQ, Rel-
ative quantitation; IQR, interquartile range; FL3lg, FMS-like tyrosine kinase 3 ligand;
NS, not significant; GDF, growth differentiation factor; IFN, interferon; IL, interleu-
kin; INH, inhibin; LT, lymphotoxin; PDGF, platelet-derived growth factor; TGF,
transforming growth factor; TNF, tumor necrosis factor; TNFsf, TNF super family
member; CD40lg, CD40 ligand; VEGF, vascular endothelial growth factor.
Later et al Perioperative Management
The Journal of Thoracic and Cardiovascular Surgery cVolume 145, Number 6 1616.e3
PM
TABLE E5. Fold changes in cytokine gene expression on the array for
patients that received corticosteroids intraoperatively and those who
did not
Gene Steroids n
RQ, median
(IQR)
Pvalue,
pre vs post
Between-
group
Pvalue
CSF3 No 27 0.93 (0.87-0.99) .037
Yes 8 1.03 (1.01-1.12) NS NS
FLT3lg No 27 0.63 (0.51-0.83) <.001
Yes 8 0.58 (0.36-0.69) .012 NS
GDF5 No 27 1.48 (1.11-2.30) <.001
Yes 8 1.35 (1.06-1.76) .012 NS
IFN-a8 No 27 1.13 (1.03-1.33) .001
Yes 8 1.05 (0.98-1.27) .012 NS
TXLN-aNo 27 0.95 (0.80-1.08) .022
Yes 8 0.91 (0.75-1.03) NS NS
IL-1bNo 27 1.28 (1.03-1.36) <.003
Yes 8 1.04 (0.94-1.48) NS NS
IL-8 No 27 0.50 (0.24-0.64) <.001
Yes 8 0.46 (0.22-1.84) NS NS
INH-aNo 27 1.05 (0.99-1.16) NS
Yes 8 1.12 (1.04-1.23) .036 NS
INH-bNo 27 1.01 (0.97-1.04) NS
Yes 8 1.06 (0.96-1.16) .017 NS
LT-bNo 27 0.85 (0.71-1.15) .003
Yes 8 0.69 (0.54-0.82) .012 .003
PDGF-aNo 27 1.13 (1.07-1.30) <.001
Yes 8 1.20 (1.11-1.24) .012 NS
TGF-aNo 27 1.42 (1.13-1.64) <.001
Yes 8 1.40 (1.08-1.71) .017 NS
TNF No 27 0.91 (0.76-1.00) .004
Yes 8 0.79 (0.72-0.85) NS NS
TNFsf10 No 27 1.35 (1.00-1.62) .016
Yes 8 1.04 (0.74-1.32) NS NS
TNFsf13 No 27 1.18 (1.02-1.26) .001
Yes 8 1.27 (1.06-1.30) .025 .023
TNFsf13B No 27 1.73 (1.37-2.70) <.001
Yes 8 1.26 (0.75-1.50) NS NS
TNFsf14 No 27 1.24 (1.05-1.50) <.001
Yes 8 1.12 (0.94-1.30) NS NS
CD40lg No 27 0.94 (0.57-0.84) <.001
Yes 8 0.64 (0.57-0.74) .017 NS
TNFsf9 No 27 0.70 (0.53-0.98) .001
Yes 8 0.81 (0.65-0.81) .036 NS
VEGF-bNo 27 0.85 (0.68-1.00) .001
Yes 8 0.71 (0.70-0.82) .017 NS
RQ, Relative quantitation; IQR, interquartile range; CSF, cerebrospinal fluid; NS, not
significant; FL3lg, FMS-like tyrosine kinase 3 ligand; GDF, growth differentiation
factor; IFN, interferon; TXLN, taxillin; IL, interleukin; INH, inhibin; LT, lympho-
toxin; PDGF, platelet-derived growth factor; TGF, transforming growth factor;
TNF, tumor necrosis factor; TNFsf, TNF super family member; CD40lg, CD40
ligand; VEGF, vascular endothelial growth factor a.
Perioperative Management Later et al
1616.e4 The Journal of Thoracic and Cardiovascular Surgery cJune 2013
PM