In critical care patients  or those undergoing major surgery (e.g. cardiac surgery), excessive blood loss is associated with increases in mortality, morbidity and ICU stay [2,3]. Indeed, the most common cause of significant intraoperative bleeding is inadequate surgical haemostasis, known as the silk deficiency . Moreover, current risk stratification is only partially successful and is unable to account for the striking interpatient variability in postoperative blood loss . Perioperative blood loss may vary according to the anesthetic agent used and the type of anesthesia . However, in some cases, other factors such as disseminated intravascular coagulation, heparin overdose or hyperfibrinolysis may play an important role [7,8]. Whatever the cause, uncontrolled bleeding can lead to a combination of haemodilution, hypothermia, consumption of clotting factors and acidosis, which in turn exert their own negative influences on the clotting process, creating a vicious circle .
The effect of antifibrinolytic drugs on PAI-1 activity or secretion during cardiopulmonary bypass (CPB) has not been studied extensively. Kojima et al. showed that tranexamic acid effectively suppressed fibrinolysis by inhibiting tissue plasminogen activator and plasmin activity, but no effects on PAI-1 or alpha2-antiplasmin were demonstrated. More recently, Kang et al. reported that, in patients treated with aprotinin during CPB, PAI-1 secretion was significantly lower soon after CPB was stopped only when compared with controls consisting of patients who did not receive aprotinin.
In contrast, the impact of specific genetic variants has not been systematically explored in the context of blood loss in the field of perioperative cardiac surgery, although very interesting papers have been published . PAI-1 is the primary inhibitor of plasminogen activators in plasma. A common single-nucleotide insertion/deletion in the PAI-1 promoter has been identified 675 bp upstream from the start codon. This polymorphism induces two alleles containing either four or five sequential guanosines (GGGG or GGGGG, respectively). Individuals homozygous for the deletion allele (4G/4G) have activated gene transcription, significantly elevated plasma PAI-1 activity and significantly diminished fibrinolytic activity (hypofibrinolysis) compared with individuals homozygous for the 5G allele (5G/5G) or those who are heterozygous (4G/5G), in which a transcriptional repressor results in decreased PAI-1 activity . It has been reported that individuals with 4G/4G have significantly higher plasma PAI-1 levels (25% higher) than either individuals heterozygous (4G/5G) or homozygous for five guanines (5G/5G). For the clinician, it is important to highlight that inhibition of fibrinolysis through elevated circulating concentrations of PAI-1 may lead to a procoagulant state [14,15].
Pharmacological approaches to reduce bleeding and transfusion aim at either preventing or reversing the defects associated with the coagulopathy. Inhibition of fibrinolysis either with lysine analogues (epsilon aminocaproic acid or tranexamic acid) or with aprotinin (a broad-spectrum serine protease inhibitor) is a therapeutic approach that has been thoroughly reviewed in the literature and evaluated in comprehensive meta-analyses .
In contrast, pharmacogenetics holds the promise that drugs might one day be tailor-made for individuals and adapted to each person's own genetic makeup. Pharmacogenomics is a rapidly emerging field that aims to elucidate the genetic basis for interindividual differences in drug response, using genome-wide approaches to identify genetic polymorphisms that govern an individual's response to specific drugs .
To our knowledge, the drugs employed for preventing or reducing blood loss have not been analysed from the pharmacogenetic perspective. Therefore, the aim of this pilot study was to investigate whether the 4G/4G genotype of the PAI-1 gene was associated with less bleeding after cardiac surgery (when compared with other genotypes not related to a procoagulant state, 4G/5G and 5G/5G) and whether this genotype may influence the prophylactic use of antifibrinolytic drugs, specifically aprotinin and tranexamic acid.
This case–control association study was aimed at comparing the distribution of genotypes of the 4G/5G polymorphism 675 bp from the initiation site of transcription of the PAI-1 gene (genotypes 4G/4G, 4G/5G and 5G/5G) among cardiac surgery patients scheduled to undergo elective cardiac surgery with CPB at the surgical ICU of the 12 de Octubre University Hospital (n = 260) and nonhospitalized age-matched controls (n = 111) recruited from the same geographical area as the patients. Then, in the subset of cardiac surgery patients, we evaluated the possible association of the homozygous 4G/4G genotype (considered procoagulant) in two different cohorts of patients (treated with aprotinin or tranexamic acid) according to postoperative blood loss and transfusion requirements. The data shown here are a secondary analysis of another study reported elsewhere .
The study was approved by the ethics committee of the 12 de Octubre University Hospital, and written informed consent was obtained from each patient enrolled. All patients were Spanish whites. Exclusion criteria were emergency operations, heart transplantation, ventricular assist device placement, chronic dialysis, hepatic failure, respiratory insufficiency, haemorrhage and active infection. Patients taking daily aspirin were asked to discontinue this therapy at least 4 days before the operation. All care providers were blinded to patient genotypes.
Blood loss and time points
Chest tube output, an adequate surrogate for blood loss , was measured at three time points (6 h and 24 h after ICU arrival and total blood output). Blood loss higher than 2 l was considered to be massive bleeding . In this study, intraoperative blood loss was not included, and postoperative bleeding represents blood loss only from chest tubes. Also in this study, total blood loss represents the blood output from the patient's arrival on the ICU until the tubes were removed.
A prognostic scoring system, European system of cardiac-operative risk evaluation score (EuroScore), was used  for risk assessment. Scores of at least 6 were considered high risk .
Operative techniques, anticoagulation and antifibrinolytic drug
Anaesthesia management and CPB were conducted in accordance with the institution's protocol. Briefly, patients received general endotracheal anaesthesia, consisting of induction with a combination of thiopental, midazolam, fentanyl or etomidate and maintenance with isoflurane, pancuronium and fentanyl. During CPB, the haematocrit was maintained between 20 and 25%, pump flow rates between 2.4 and 4.8 l/min/m2 and mean arterial pressures between 50 and 70 mmHg. Before cannulation, heparin (3.5 mg/kg) was given. Anticoagulation for CPB was achieved to maintain an activated clotting time (CT) above 480 s. At the end of CPB, anticoagulation was reversed by protamine sulphate (1–1.5 mg/100 IU of heparin administered in the previous hour). The systemic temperature was kept between 28 and 32°C.
Institutional guidelines, which are in accordance with the recent literature , were followed for the use of antifibrinolytic drugs. Briefly, aprotinin is indicated to reduce the number of patients requiring blood transfusion, to reduce total blood loss and to limit reexploration in high-risk patients. According to these guidelines, tranexamic acid is considered to be a slightly less potent blood-sparing drug. Hence, aprotinin was reserved for patients who were undergoing complex procedures in whom prolonged CPB support was expected or had at least two previous sternotomies.
The patient's clinical situation and disease status were important factors in determining the need and indication for transfusion. Transfusion decisions in the ICU are based on the patient's age, medical history, myocardial performance, presumed cause of bleeding and evidence of end-organ dysfunction. In this context, the final decision on transfusion was taken by the attending physician. In general, patients with acute blood loss (acute blood loss volume >20% of total blood volume, regardless of haematocrit) were transfused. In addition, patients without acute blood loss with a haematocrit less than 30% were transfused if there was a risk of ischaemia (age >65 years, history of coronary artery disease, history of stroke or transient ischaemic attack). Finally, patients without any risk for ischaemia but with signs or symptoms of acute anaemia (syncope, tachycardia, angina, dyspnoea, pulse oximetry <90% or paO2 <70 mmHg) were transfused if they were under 40 years old with a haematocrit less than 24%, if they were between 40 and 65 years old with a haematocrit less than 27% or if they were over 65 years with a haematocrit less than 30%. Packed red blood cells were transfused in sets of at least two units. Platelets were transfused in four-unit sets for microvascular haemorrhage continuing after normalization of activated CT or for platelet dysfunction or low platelet count in the setting of ongoing clinical bleeding. Platelets were not administered simply for low platelet count in the absence of bleeding. Plasma was transfused in sets of at least two units when bleeding continued after platelet transfusion.
DNA was isolated and the 4G/5G polymorphism was typed as described elsewhere . Samples were genotyped at 12 de Octubre University Hospital and Joan XXIII University Hospital.
Two comparative groups were formed: 4G homozygous (4G/4G, genotype procoagulant) vs. non-4G homozygous [heterozygous (4G/5G) and 5G homozygous (5G/5G), both considered genotypes that were not procoagulant]. Statistical analyses were performed with the SPSS statistical package version 11.0 (SPSS Inc., Chicago, Illinois, USA). Categorical variables were compared by chi-squared test or Fisher's exact test for discrete data when necessary. Univariate analysis of chest tube output and number of units of packed red blood cells transfused was conducted using the t-test after Levene's test for homogeneity of variances. Significance was defined as a P value of 0.05 or less. For multivariate analysis, we evaluated chest tube output at 6 and 24 h and total blood output as dependent variables using stepwise linear regression (probability of F to enter ≤0.05, probability of F to remove ≥0.1). The following were entered stepwise as independent variables for chest tube output: age older than 70 years, sex, duration of CPB, and preoperative and EuroScore of at least 6. All these variables were found to be significant independent variables on earlier research focusing on bleeding after cardiac surgery [25,26]. In addition, the use of antiaggregant or anticoagulant drugs was associated with higher total blood output in the univariate analysis (P ≤ 0.2) (data not shown). Moreover, on the basis of pharmacogenetic studies, other variables, such as the use of aprotinin or tranexamic acid and carriers of homozygous 4G/4G genotype, which may be associated with the end point of the study, were also included in the model .
Two hundred and sixty consecutive patients undergoing cardiac surgery with CPB were enrolled in this study. General, historical and operative characteristics of the patients have been reported elsewhere  and are shown in Table 1. The rate of coronary artery bypass grafting surgery was significantly higher in the non-4G homozygous subgroup. The distribution of preoperative drugs did not differ significantly between the groups. Aprotinin and tranexamic acid were used in 117 (45%) and 106 patients (40.8%), respectively. The remainder of the patients received other haemostatic drugs or no haemostatic drugs at all for clinical reasons. The distribution of haemostatic drugs between comparative groups did not show any significant differences. Table 2 shows the baseline characteristics of the nonhospitalized age-matched controls in whom mean age (SD) was 62.4 (11.6) years, 37.8% were over 70 years of age and 62% were men.
Genetic analysis identified the genotype 4G/4G in 47 patients (18.1%), 4G/5G in 131 (50.4%) and 5G/5G in 82 (31.5%). In the group of nonhospitalized age-matched controls, the distribution of genotypes was 4G/4G in 29 patients (26.1%), 4G/5G in 59 (53.2%) and 5G/5G in 23 (20.7%). No significant differences in genotype distribution were found between the two groups, indicating that the allele frequencies in these patients were not associated with a predisposition to cardiac surgery diseases. Allele frequencies were 0.43 for 4G and 0.57 for 5G in patients and 0.53 for 4G and 0.47 for 5G in controls. The distributions did not differ significantly from the Hardy–Weinberg equilibrium law.
The distribution of postoperative blood loss by genotypes in the overall cohort of patients did not show any significant differences. The distribution of blood loss by genotypes in the subset of patients treated with aprotinin is shown in Table 3. Interestingly, significantly less blood loss in homozygous 4G/4G patients was documented at 6 and 24 h. A trend towards lower total blood loss (total blood output) was also found in the subset of homozygous 4G/4G patients. The distribution of blood loss by genotypes for patients treated with tranexamic acid is shown in Table 4. In this subset, we found no differences in the distribution of postoperative blood loss at the three time points considered.
In our study, on the basis of linear regression models for blood loss, the homozygous 4G/4G genotype does not protect against bleeding in the general population of patients. Moreover, considering the two antifibrinolytic drugs separately, the model does not show significant changes. Nevertheless, antiaggregant drugs and a EuroScore of at least 6 were significantly associated with total blood loss [odds ratio (OR), 2.6; 95% confidence interval (CI), 46–331; P = 0.02 and OR, 2; 95% CI, 5–284; P = 0.04, respectively].
Interestingly, in the subset of homozygous 4G/4G patients, aprotinin was independently associated with less total blood loss (Table 5). In addition, in the subset of homozygous 4G/4G patients treated with tranexamic acid, the genotype was not a specific protector of blood loss and, as in the general population of patients, antiaggregant drugs and EuroScore of at least 6 were significantly associated with total blood loss (Table 6).
Only 10 patients (3.8%) had massive blood loss. There were no significant differences between groups. Surgical reexploration due to bleeding was necessary in 12 patients (4.6%). Of those, only two patients were homozygous 4G/4G.
Fifty-one of the patients treated with aprotinin (42.7%) received blood components (packed blood red cells, platelets, fresh frozen plasma or all), whereas 40 of those administered tranexamic acid (37.7%) were transfused. In addition, 35.6% of homozygous 4G/4G patients received blood components compared with 40% of patients who were not homozygous 4G/4G. In the univariate analysis, we found a trend towards a significant difference in transfusion, as 26.3% of homozygous 4G/4G patients treated with aprotinin received transfusion compared with 47.2% of patients who were not homozygous 4G/4G [relative risk (RR), 0.4; 95% CI, 0.1–1.2; P = 0.2]. However, 31.6% of homozygous 4G/4G patients treated with tranexamic acid required transfusion compared with 32.9% of patients who were not homozygous 4G/4G (RR, 0.9; 95% CI, 0.3–2.7; P = 0.6), respectively. Finally, for homozygous 4G/4G patients treated with aprotinin, the mean number of units of packed red blood cells was 0.8 (SD 1.8) vs. 1.5 (SD 3.5), P = 0.2, for patients who were not homozygous 4G/4G.
To our knowledge, this is the second study to evaluate a potential association between the 4G/5G promoter polymorphism and the risk of bleeding after cardiac surgery with CPB from the pharmacogenetic perspective. The main result is that the 4G/4G genotype of the PAI-1 gene was not associated with less bleeding after cardiac surgery except in the subgroup of patients treated with aprotinin, who showed a trend towards lower transfusion requirements. In addition, only a EuroScore of at least 6 and therapy with platelet antiaggregants were related to bleeding after cardiac surgery in the patient population as a whole.
In a previous study , we showed the possible role of the 4G/4G genotype in neurological complications after cardiac surgery with CPB owing to its procoagulant properties. This polymorphism has also been associated with poorer outcome in patients with meningococcal disease  and trauma  due to microthrombosis. For this reason, these polymorphisms may be well suited to investigate the possible protective impact on blood loss after cardiac surgery using a candidate gene approach.
Little is known of the role of genetic factors in bleeding following cardiac surgery. Donahue et al. showed that factor V Leiden (FVL), a polymorphism that consists of a glutamine substitution for arginine 506 producing a variant that is resistant to inactivation by activated protein C, protects against blood loss. Welsby et al. identified seven genetic polymorphisms associated with bleeding after cardiac surgery (GPIaIIa−52C>T and 807C>T, GPIbα 524C>T, tissue factor −603A>G, prothrombin 20210G>A, tissue factor pathway inhibitor −399C>T and angiotensin converting enzyme deletion/insertion) but, perhaps because of differences in surgical populations and in the use of antifibrinolytic prophylaxis, did not confirm the protective role of FVL against blood loss. More recently, Boehm et al. found that FVL carriers treated with aprotinin did not have reduced blood loss compared with noncarriers. In our study, in the overall population, we did not find an association between the 4G/4G genotype and reduced bleeding after cardiac surgery. Nevertheless, our study had a significantly higher rate of valvular surgery patients than the other studies. Unfortunately, the first two studies only described bleeding until 12 h and 24 h after surgery, respectively, whereas our study reports data on bleeding at 6 h and 24 h and total bleeding.
An important pharmacogenetic question is whether the effect of the 4G/4G genotype depends on the haemostatic drugs used. Interestingly, patients with the 4G/4G genotype who were treated with aprotinin showed significantly less bleeding than those treated with tranexamic acid at 6 h and 24 h, and therefore required less transfusion. Accordingly, a trend toward lower transfusion requirements was found. The mechanisms by which aprotinin exerts its haemostatic effects are not fully understood, but it is known to be a powerful inhibitor of plasmin, trypsin, chymotrypsin, kallikrein, thrombin and activated protein C through the formation of reversible enzyme–inhibitor complexes . For its part, tranexamic acid is a competitive inhibitor of plasminogen activator and, at much higher concentrations, a noncompetitive inhibitor of plasmin. One possible pharmacogenetic explanation for this higher rate of bleeding in 4G/4G patients treated with tranexamic acid might be that tranexamic acid blocks the lysine-binding site on plasminogen, thereby preventing the activation of plasminogen on the surface of fibrin. Therefore, patients with the 4G/4G genotype, who have higher plasma levels of PAI-1, also inhibit plasminogen activation. Nevertheless, as aprotinin has multiple sites of activity, other mechanisms of antifibrinolysis may be involved in addition to the effect obtained by higher levels of PAI-1 in homozygous 4G/4G patients. Clearly, the replication of findings in multiple settings and populations (in even larger samples) would lend the greatest credibility to the genetic association identified and is necessary to definitively rule out potential deleterious effects of the combination of 4G/4G carrier–tranexamic acid recipient.
Our study has several important limitations. First, although the priority of the study was to enrol two homogeneous subsets of cardiac surgery patients on the basis of antifibrinolytic treatment, the small sample size of the cohort of patients treated with aprotinin and tranexamic acid allowed limited analysis of the association of the genotype with the treatment and, finally, with the outcome. In a post-hoc analysis based on our data, we estimated that at least 60 homozygous 4G/4G patients treated with aprotinin or tranexamic acid, respectively, would be needed to demonstrate lower total blood loss in the subset of patients treated with aprotinin, with 80% power and an alpha error of 0.05. This means that approximately 350 patients per arm (aprotinin or tranexamic acid) would be required. In addition, it may lack sufficient power to prove a definitive association between the homozygous 4G/4G genotype and bleeding after cardiac surgery; with a larger sample size, we could have applied more robust linear regression models. Second, quantification of intraoperative blood loss and sites of postoperative bleeding other than chest tubes would have helped to define a better profile of blood loss. Third, more detailed information regarding transfusional requirements, particularly during surgery and hospital stay, would probably have allowed us to present a more accurate estimate of this parameter. Fourth, the evaluation of postoperative complications such as myocardial infarction might have served as indicators for severe ischaemic complications during hospital stay after cardiac surgery, particularly in homozygous 4G/4G patients treated with aprotinin. Fifth, measurement of PAI-1 plasma levels in patients would have helped to establish whether homozygous 4G/4G patients do, in fact, show higher PAI-1 levels in this setting. Finally, it is unlikely that any single polymorphism in isolation would be sufficient to cause significant bleeding; probably, several polymorphisms contribute to the final outcome. Caution is needed when interpreting genetic association studies; genetics in critical care and single nucleotide polymorphism (SNP) complex phenotype association studies are moving towards studies involving some form of haplotyping (because difficulty in distinguishing a causal genetic variation from correlation because of linkage disequilibrium is another limitation of genetic associations and a possible explanation for inconsistencies between studies) or whole-genome scanning. Single gene variation may, in fact, provide limited information on this complex phenomenon.
In summary, although this study did not investigate a causal relationship between the polymorphism and bleeding, our findings suggest that the 4G/4G genotype of the PAI-1 gene was not associated with less bleeding after cardiac surgery except in the subgroup of patients treated with aprotinin, who presented a trend towards lower transfusional requirements. Moreover, although selection of one drug might be a confounder for other conditions, this study should be considered to be a pilot study and, on the basis of the design, a randomized controlled trial study with a pharmacogenetic approach. Our data suggest that approximately 350 patients per arm would be required to demonstrate a significant reduction in blood loss in the subset of 4G/4G homozygous patients treated with aprotinin. In these patients, we recommend adding the analysis of SNPs associated with bleeding and the determination of PAI-1 levels.
The authors want to thank Mike Maudsley for advising on the English revision of the manuscript.
Sources of support: MAPFRE Cardiovascular, CIRIT 2005SGR920, ISCIII CM 04/00023, PI 040691 and CIBER Enfermedades Respiratorias (CD06/06/0036).
1 Leal-Noval SR, Rincon-Ferrari MD, Marin-Niebla A. Transfusion
of erythrocyte concentrates produces a variable increment on cerebral oxygenation in patients with severe traumatic brain injury. A preliminary study. Intensive Care Med 2006; 32:1733–1740.
2 Hall TS, Sines JC, Spotnitz AJ. Haemorrhage related reexploration following open heart surgery: the impact of preoperative and postoperative coagulation testing. Cardiovasc Surg 2002; 10:146–153.
3 Merlani PG, Chenaud C, Cottini S. Point of care management of heparin administration after surgery. A randomized, controlled trial. Intensive Care Med 2006; 32:1357–1364.
4 Bevan DH. Cardiac bypass haemostasis: putting blood through the mill. Br J Haematol 1999; 104:208.
5 Stover EP, Siegel LC, Parks R. Variability in transfusion
practice for coronary artery bypass surgery persists despite national consensus guidelines: a 24-institution study. Institutions of the Multicenter Study of Perioperative Ischemia Research group. Anesthesiology 1998; 88:327–333.
6 Albertin A, La Colla L, Gandolfi A, et al
. Greater peripheral blood flow but less bleeding
with propofol versus sevoflurane during spine surgery: a possible physiologic model? Spine 2008; 33:2017–2022.
7 Marietta M, Facchini L, Pedrazzi S. Pathophysiology of bleeding
in surgery. Transplant Proc 2006; 38:812–814.
8 Paparella D, Brister SJ, Buchanan MR. Coagulation disorders of cardiopulmonary bypass: a review. Intensive Care Med 2004; 30:1873–1881.
9 Lynn M, Jeroukhimov I, Klein Y. Updates in the management of severe coagulopathy in trauma patients. Intensive Care Med 2002; 28:S241–S247.
10 Kojima T, Gando S, Morimoto Y, et al
. Systematic elucidation of effects of tranexamic acid on fibrinolysis and bleeding
during and after cardiopulmonary bypass surgery. Thromb Res 2001; 104:301–307.
11 Kang H-M, Kalnoski MH, Frederick M, Wayne WL. The kinetics of plasmin inhibition by aprotinin in vivo. Thromb Res 2005; 115:327–340.
12 Iribarren JL, Jimenez JJ, Hernández D, et al
. Postoperative bleeding
in cardiac surgery
. The role of tranexamic acid in patients homozygous for the 5G polymorphism of the plasminogen activator inhibitor-1 gene. Anesthesiology 2008; 108:596–602.
13 Dawson S, Hamsten A, Wiman B, et al
. Genetic variation at the plasminogen activator inhibitor-1 locus is associated with altered levels of plasminogen activator inhibitor-1 gene. Arterioscler Thromb 1991; 11:183–190.
14 Eriksson P, Kallin B, Van't Hoof FM. Allele-specific increase in basal transcription of the plasminogen activator inhibitor-1 gene is associated with myocardial infarction. Proc Natl Acad Sci U S A 1995; 92:1851–1855.
15 Burzotta F, DiCastelnuovo A, Amore A. 4G/5G promoter PAI-1
gene polymorphism is associated with plasmatic PAI-1
activity in Italians: a model of gene-environment interaction. Thromb Haemost 1998; 79:354–358.
16 Sedrakyan A, Treasure T, Elefteriades JA. Effect of aprotinin on clinical outcomes in coronary artery bypass graft surgery: a systematic review and meta-analysis of randomized clinical trials. J Thorac Cardiovasc Surg 2004; 128:442–448.
17 McLeod HL, Evans WE. Pharmacogenomics: unlocking the human genome for better drug therapy. Annu Rev Pharmacol Toxicol 2001; 41:101–121.
18 Sirgo G, Pérez-Vela JL, Morales P, et al
. Relation between 4G/5G polymorphism of the plasminogen activator inhibitor 1 gene with stroke or encephalopathy after cardiac surgery
. Intensive Care Med 2006; 32:668–675.
19 Filsoufi F, Castillo JG, Rahmanian PB. Effective management of refractory postcardiotomy bleeding
with the use of recombinant activated factor VII. Ann Thorac Surg 2006; 82:1779–1783.
20 Karkouti K, Beattie S, Wijeysundera DN. Recombinant factor VII for intractable blood loss after cardiac surgery
: a propensity score-matched case control analysis. Transfusion
21 Gogbashian A, Sedrakyan A, Treasure T. EuroSCORE: a systematic review of international performance. Eur J Cardiothorac Surg 2003; 23:360–367.
22 Calafiore AM, Di Mauro M, Canosa C. Early and late outcome of myocardial revascularization with and without cardiopulmonary bypass in high risk patients (EuroSCORE ≥6). Eur J Cardiothorac Surg 2003; 24:953–960.
23 Ferraris VA, Ferraris SP, Saha SP, et al
. Perioperative blood transfusion
and blood conservation in cardiac surgery
: the Society of Thoracic Surgeons and The Society of Cardiovascular Anesthesiologists clinical practice guideline. Ann Thorac Surg 2007; 83(Suppl 5):S27–S86.
24 Benza RL, Grenett HE, Bourge RC. Gene polymorphism for plasminogen activator inhibitor-1/tissue plasminogen activator and development of allograft coronary artery disease. Circulation 1998; 98:2248–2254.
25 Parr KG, Patel MA, Dekker R. Multivarite predictors of blood product use in cardiac surgery
. J Cardiothorac Vasc Anesth 2003; 2:176–181.
26 Karthik S, Grayson AD, McCarron EE. Reexploration for bleeding
after coronary artery bypass surgery: risk factors, outcomes and the effect of time delay. Ann Thorac Surg 2004; 2:527–534.
27 Aquilante CL, Langaee TY, Lopez LM. Influence of coagulation factor, vitamin K epoxide reductase complex subunit 1, and cytochrome P450 2C9 gene polymorphisms on warfarin dose requirements. Clin Pharmacol Ther 2006; 79:291–302.
28 Hermans PWM, Hibberd ML, Booy R. 4G/5G promoter polymorphism in the plasminogen-activator-inhibitor-1 gene and outcome of meningococcal disease. Lancet 1999; 354:556–560.
29 Menges T, Hermans PW, Little SG. Plasminogen-activator-inhibitor-1 4G/5G promoter polymorphism and prognosis of severely injured patients. Lancet 2001; 357:1096–1097.
30 Donahue BS, Gailani D, Higgins MS. Factor V Leiden protects against blood loss and transfusion
after cardiac surgery
. Circulation 2003; 107:1003–1008.
31 Welsby IJ, Podgoreanu MV, Phillips-Bute B. Genetic factors contribute to bleeding
after cardiac surgery
. J Thromb Haemost 2005; 3:1206–1212.
32 Boehm J, Grammer JB, Lehnert F. Factor V Leiden does not affect bleeding
in aprotinin recipients after cardiopulmonary bypass. Anesthesiology 2007; 106:681–686.
33 Wells PS. Safety and efficacy of methods for reducing perioperative allogeneic transfusion
: a critical review of the literature. Am J Ther 2002; 9:377–388.
34 Matsubara Y, Hayakawa T, Tsuda T. Angiotensin converting enzyme insertion/deletion polymorphism is associated with plasma antigen levels of plasminogen activator inhibitor-1 in healthy Japanese population. Blood Coagul Fibrinolysis 2000; 11:115–120.