In cardiovascular surgery the activation of coagulation factors, inflammatory responses and fibrinolytic process often have deleterious effects on pediatric patient outcomes. Non-endothelial contact of blood activates humeral and cellular pathways making it one of the major reasons for blood activation in cardio-surgical patient. The blood is in regular contact with most of the components of a bypass machine causes multiple activation and resulting in preventable complications. Due to the unique nature of pediatric patients, hematological derangements occur at a high rate, as infant's blood volume is smaller as compared to total prime of a cardiopulmonary bypass circuit. Due to this phenomena, reduction in use of blood and blood products have received an important attention over the years. Variety of methods and pharmacological agent to encounter these deleterious effects are being constantly introduced in clinical practice. On the other hand, there are qualitative and quantitative difference in coagulation systems compared to adult patients. The smaller infant blood volume as compared to prime volume of cardiopulmonary bypass circuits produces the dilution in hemostatic components. In addition, the neonatal immune system is immature and is comprised of 30% to 70% less levels of anticoagulant and procoagulant proteins as compared to adults. There are also structural differences in compounds like fibrinogen and plasminogen. Due to the complexity of the operation, with long cardiopulmonary bypass and deep hypothermia coagulation is further deranged. The outcome of bleeding can be avoided as it is associated with extended surgical times, reoperations, hemodynamic instability and dilution of coagulation factors. Various methods have been utilized in practice to overcome this treatable outcome.
Get your grade
or your money back
using our Essay Writing Service!
Another major concern is to reduce the morbidity and mortality associated with blood product transfusion. Pharmacological intervention by using anti-fibrinolytics has received immense attention during last decade in reducing the bleeding associated with cardiopulmonary bypass. Commonly used anti-fibrinolytics are aprotinin, Tranexamic acid and aminocaproic acid.
Due to these differences, the coagulation cascade produces impaired hemostasis and interacts differently to pharmacological agents like Tranexamic acid, aprotinin etc.
The mechanism involved in preoperative bleeding is complex and involve disturbances in various mechanical and physiological systems.
Hemostasis and Fibrinolysis
Hemostasis is an integral part of the balanced mammalian vascular system, where two opposing forces act together to ensure viability of the organism. The inner lining of vascular system is made of endothelial cells that play a central role in maintaining this vascular integrity under normal circumstance. In addition, these cells secrete multiple biological components that maintain the blood fluidity and aid in normal hemostasis. Blood also contains platelets and coagulation factors that after being activated help in orchestrate the response of hemostasis and fibrinolysis. Based on involvement from various components of vascular system the hemostasis is classified in three stages.
Once the vascular injury occurs, the platelets respond by adhering to the sub-endothelial matrix and form a monolayer of cells; this process is mediated by a von Willebrand factor released by endothelial cells. After the platelets adhere, they release the contents of their granules and synthesize thromboxane A2 and other chemical attractants that attract other platelets to the site of vascular injury. Once large numbers of platelets are attracted to site of injury hemostatic plug is formed between platelets and fibrinogen mediated by the binding of bivalent fibrinogen to the platelet surface integrin heterodimer glycoprotein IIb / IIIa. In addition, this process is kept under check by production of major molecules that have platelet inhibitory effect. Vascular endothelium also produces nitric oxide and prostaglandin Ia which inhibit platelet activation, aggregation and secretion under normal circumstances.
Coagulation or secondary hemostasis is the conversion of soluble fibrinogen into soluble fibrin. This process takes place by way of various amplifying enzymatic reactions, in which the product of each reaction converts an inactive plasma protein zymogenic precursor into an active protease product. In this process, each zymogen is converted to its active form by hydrolysis of one or two peptide bonds. These linked reactions provide dramatic amplifications of small initiating stimuli that culminate in rapid and exuberant fibrin formation at the site of vascular injury. One of the important central serine proteases of the coagulation is thrombin. Conversion of fibrin monomer to fibrinogen and activation of platelets occurs once thrombin is diffused from membrane site of its generation in to blood. Further activation of platelets provide membrane surface to activate further the coagulation cascade, generate more thrombin, and thereby amplify and localize the formation of hemostatic plug. In addition, thrombin sustains the coagulation cascade by feedback activation of other coagulation factors.
Always on Time
Marked to Standard
Once fibrin monomer is formed, it assembles non-covalently in an end-to-end and side-to-side fashion to form fibrin polymers that result in the formation of a fibrin clot. In order to stabilize the fibrin clot thrombin through its activation of factor XIII catalyzes the covalent cross-linking of fibrin polymers through transamidination.
In the absence of vascular injury coagulation cascade is suppressed by several physiological antithrombotic systems. Serine protease inhibitors classes of molecules, inhibit various factors and thrombin by forming one-on-one complexes and regulating vascular fluidity.
Deposition of fibrin activated fibrinolytic system and assists maintaining an open lumen in a damaged blood vessel. Under normal circumstances, for repairing an injured vessel wall it is important to maintain a balance between the formation and lysis of fibrin. This fibrinolytic process is mediated through activation of plasminogen, the plasma precursor of the proteolytic enzyme plasmin. Plasminogen binds to lysine residues on the surface of fibrin and is converted to plasmin by a tissue plasminogen activator (t-PA) that simultaneously binds to fibrin. Plasmin then degrades fibrin into large fragments known as X and Y; which are subsequently broken down into soluble fibrin degradation products. Excessive fibrinolysis is prevented by the greater affinity of plasminogen for fibrin than for fibrinogen and increased ability of t-PA to activate plasminogen when it is bound to fibrin. In addition, plasma contains a protease inhibitor called α2-antiplasmin that rapidly inactivates any plasmin that escapes from fibrin clot.
In order to overcome the defective formation or excessively rapid dissolution of fibrin results which results in excessive or recurrent bleeding; various pharmaceutical gents are employed. Anti-fibrinolytic drugs that stabilize the fibrin structure can prevent unwanted dissolution of hemostatic fibrin. Two synthetic derivatives of the amino acid lysine, tranexamic acid [4-(aminomethyl)cyclohexanecarboxylic acid] and Æ-aminocaproid acid (EACA; 6-aminoheanoic acid) have anti-fibrinolytic activity in humans.
In cardiac surgery, the administration of high-dose heparin prevents complete thrombosis of blood entering the cardiopulmonary bypass circuit, however low level intravascular and intra-circuit coagulation continues throughout bypass. After the initiation of cardiopulmonary bypass the activation of contact phase of coagulation produces a cleavage of factor XII to XIIa and prekallikrein and Kallikren resulting in activation of inflammation and fibrinolysis.
However, pediatric patients are prone to hemostatic disturbances due to immmaturinty of
A synthetic derivative of the amino acid lysine belongs to the group of anti-fibrinolytics; its anti-fibrinolytic effects are through the reversible blockade of lysine binding sites on plasminogen molecules. This reversible blockage inhibits plasminogen molecules, further; it inhibits the interaction of plasminogen and heavy chain of plasma with lysine residues on the surface of fibrin. Chemically named, as trane-4-aminomethylcyclohexane-carboxylic acid with the empirical formulae of tranexamic acid is C8H15NO2 with a molecular weight of 157.2 (am). It is a white crystalline powder that is odorless, freely soluble in water and practically insoluble in methanol, ethanol benzene with pike of 4.3 and 10.6.
It has been found that tranexamic acid is 6 to 20 times more potent as compared to other synthetically derived anti-fibrinolytic; It is a competitive inhibitor of plasminogen activity and at much higher concentrations a non-competitive inhibitor of plasmin; it interferes in coagulation process in the same way as aminocaproic acid does.
It is also been reported that tranexamic acid induced suppression of fibrinolysis is manifested in surgical patients by reductions in blood levels of D-dimer, but the drug does not influence the activity of tranexamic acid. It shows its effect by binding considerably to both weak and strong sites of plasminogen molecule considerably strongly then aminocaproic acid.
Normally plasminogen is activated by a well-placed proteolytic cleavage that exposes the serine-histadine catalytic site. Conversion of glu-plasminogen to lys-plasminogen by plasmin promotes binding of plasminogen to fibrin or fibrinogen substrate =. Binding is affected by lock and key fit between one or more lysine binding sites on plasminogen and specific lysine residues of the substrate. Without proper binding, proteolysis cannot proceed. In vitro EACA and AMCA accelerate plasminogen activation by binding to plasminogen and altering its conformation so that it is more susceptible to activation. However lysis of a fibrin substrate is inhibited by EACA because the same binding phenomenon that induced a conformational change that accelerated activation also blocks functional activity by occupying the lysine binding sites. Thus any plasmin molecule that formed no mater how rapidly, cannot bind effectively to the fibrin substrate, thereby precluding proteolytic action by the serine enzyme site.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
The similarity in the three-dimensional structure of EACA and AMCA with lysine underscores their mode of action, namely, by steric inhibition of binding sites on plasmin (ogen). The dissociation of lys-plasminogen from the fibrin structure by the lysine analogues EACA and AMCA as the reason for their anti-fibrinolytic effect. The subtle differences in the synthetic analogues can markedly affect their inhibitory potential s, as exemplified by the approximately six to ten folds higher molar potency of AMCA in comparison to EACA.
Research suggests that after oral administration of tranexamic acid it takes maximum of 3 hours for plasma concentration to reach its maximum value. However, after a single bolus of intravenous administration that is most likely the case with pediatric patients 95% of the dose, is excreted unchanged in urine, excreted through glomerular filtration. In addition, the pharmacokinetics report that tranexamic acid remains in different tissues for about 17 hours and in serum for up to seven hours.
After reaching the therapeutic plasma concentration, the tranexamic acid is weakly bound to plasma; which is accounted for by binding plasminogen. Tranexamic acid also cross blood brain barrier and it rapidly diffuses into jointly fluid and synovial membranes.
The tranexamic acid has a low molecular weight, hydrophilic, renal elimination, little or no biotransformation and low cost.
Tranexamic acid use in cardiac surgery has increased in recent years due to the removal of aprotinin from world market. Preoperatively, tranexamic acid is transfused intravenously administered in loading doses of 10mg/kg and followed by an infusion of 1mg/kg/hour. Significant reduction in bleeding has been reported post operatively with tranexamic acid use.
PubMed, CINHAL, EMBAS, Google Scholar and Trip database were searched using following terms; bleeding in children undergoing cardiac surgery; cardiac surgery in pediatric patients and risk for bleeding; coagulation deficiency in pediatric cardiac surgical patients; pharmacological interventions in pediatric cardiac surgery for bleeding; anti-fibrinolytic agents use in surgery; anti-fibrinolytic use in cardiac surgery; risk of bleeding in pediatric cardiac patient; tranexamic acid and cardiac surgery; tranexamic acid and pediatric cardiac surgery; tranexamic acid use in neonates, infants, and children undergoing cardiac operations; tranexamic acid use in children undergoing hypothermic arrest; Cardiothoracic surgery and use of tranexamic acid. The search terms produced variety of results; appropriately, results were screened because of the relevance to the question of Tranexamic acid use in pediatric cardiac surgery. Some of the studies that were not in English and studies published before year 1993 were excluded from the discussion.
The selected studies were typically comparing tranexamic acid with other anti-fibrinolytics like aprotinin and Æ-aminocaproic acid; some were only applicable to adult cardiac patients. In addition, some studies compared tranexamic acid with placebo in randomized controlled trials. In most of the studies, the outcome was to study the effects of anti-fibrinolytics on coagulation and bleeding post cardiac surgery. In addition, some studies investigated does-response relationship of tranexamic acid and differences in plasma transfusion with anti-fibrinolytic treatment.
Furthermore, some studies were retrospective observational studies were tranexamic acid was compared with use of aprotinin or Æ-aminocaproic acid used before change of practice.
A prospective study by was done over 9 months, the study participants were exclusively children with cyanotic congenital heart disease, undergoing corrective surgery. The study reported randomization and standardization of external factors that may have had an effect on the study results. The tranexamic acid and placebo arm were comparable in terms of age, weight, and body surface area. Although cardiopulmonary bypass times, urine output, temperature and hematocrit were similar between the groups, however sternal closure times were significantly longer in one of the groups. Study reported no complication in the form of renal or cerebral dysfunction in any of the groups.
In this study by reported significant difference in blood loss between the tranexamic acid group and placebo, also a significant difference in usage of blood and blood products was apparent between both groups. In addition, coagulation tests done at 6 hours postoperatively showed considerable degradation of fibrinogen in placebo group.
In another study reported comparison of efficacy and safety between tranexamic acid and Æ-aminocaproic acid. The study analyzed the preoperative data of a 5-month period of all children weighing less than 20kg undergoing cardiac surgery who received Æ-aminocaproic acid with a previous cohort of patients treated with tranexamic acid. There were 124 children in tranexamic acid group as compared to 126 Æ-aminocaproic acid group. All the other preoperative factors and protocols were similar between the groups. There was no statistically significant trend in complexity of the operations and in primary outcome criterion of blood loss between the groups. All the other secondary outcome parameters like incidence of revision for bleeding and transfusion rates of Red blood cells, fresh frozen plasma, and platelets were statistically similar. In the safety field, there was no verifiable difference in all recorded outcome parameters, namely renal injury, renal failure, vascular thrombosis, seizure, other neurological events, cardiac events, duration of ventilation and intensive care unit stay and in-hospital mortality.
In a retrospective observational study by , data was collected for two different periods. The study included aprotinin (serine protease inhibitor) previously used at the center with tranexamic acid (lysine protease inhibitor) currently used. Other parameters that may influence results were not modified so showed no statistical significance. The results reported no difference between the groups on the basis of age and sex to receive either tranexamic acid or aprotinin. Aprotinin group showed significantly higher values for fibrinogen, anti-thrombin III and platelets; however, in tranexamic acid group values remain within normal range. There was no significant difference between the groups in intraoperative requirements for blood products or amount of erythrocyte concentrate or fresh frozen plasma transfusion. Study reported significant difference in amount of platelet concentrate transfused. There was no significant statistical difference in postoperatively bleeding, renal function and serum creatinine between the groups. Study reported two postoperative deaths in aprotinin group and one in tranexamic acid group and reoperations due to thrombosis complication in aprotinin group were not statistically significant.
In a study by to study the effects of aprotinin or tranexamic acid on proteolytic/cytokine profiles in infants after cardiac study included 37 patients. The infants ranged from 1-9 months of age undergoing isolated ventricle septal defect and tetralogy of Fallot. There were 22 children in aprotinin group and 15 children in tranexamic acid group, clinical data showed no significant statistical difference in patient characteristics. Results showed that tumor necrosis factor-alpha increased initially after CPB in both the groups, but there was considerable increased in tranexamic group at 24 and 48 hours post CPB. Interleukin-10 levels threefold higher in tranexamic acid group in comparison with aprotinin group. Plasma levels of matrix metalloproteinase (MMP) associated with inflammation was two fold higher in tranexamic acid group.
In another double blinded, randomized controlled study by to study the effects of tranexamic acid on blood loss after cardiac surgery in children. Study included (n=88) children, (n=40) in tranexamic acid group and (n=42) in placebo group. There were minor differences in the type of procedure; however, the other characteristics were similar between the groups. Preoperative and immediate postoperative coagulation results showed no significant difference between treatment and placebo group. Sub group analysis was conducted on high-risk groups; children with cyanotic heart diseases and those with the history of previous thoracotomy. The result indicated that blood loss in sub group with cyanosis almost halved after treatment with tranexamic acid; however, with acyanotic group had no significant difference was apparent. In addition, to blood loss, there was a reduction in packed red cell transfusion among cyanotic subgroup.
Meta-analysis was conducted by on the use of aprotinin, Æ-aminocaproic acid and tranexamic acid and its effects on blood loss and use of blood products in major pediatric surgery. Extended search was conducted on multiple databases like PubMed, EMBASE, and Cochrane Library, data was gathered and analyzed within the protocols set by the researchers. They reported selecting only randomized control trials involving children under going cardiac or scoliosis surgery between months of August 2006 and October 2006. Total 173 citations were screened and were selected based on the relevance to above question. Cardiac studies included 23 individual studies with 1893 children. Individually comparison of tranexamic acid with placebo was reported only in five studies included in meta-analysis. It was reported that methodologic quality of the cardiac studies included was poor as only three studies provided adequate description of allocation concealment.
In this meta-analysis, the participant size of cardiac study varies from 10 patients to 180 patients, in addition different transfusion protocols and dosing regimens were used; this resulted in considerable variation in cumulative doses between studies. Out of 28 studies, only sixteen reported various frequencies of complications or adverse events. Some of the complications were mainly cardiac in nature including arrhythmias, tachycardia, heart block, and cardiac arrest. In seven of the tranexamic, acid studies there were no complications reported.
After the meta-analysis, it was seen in tranexamic acid studies there was significant reduction in blood loss by an average of 11mL/kg (95% confidence interval 13-8mL/kg); packed cell transfusion was reduced by 7 mL/kg (95% CI 10-5 mL/kg). In addition, tranexamic acid reduced plasma transfusion by 7 mL/kg (95% CI 4-9 mL/kg) compared with placebo.
The coagulation system in infants and children is immature and continues to develop until late childhood. Studies have found that neonates have low levels of factors II, VII, IX-XII, pre-kallikrein, and high molecular weight kininogen. Due to this deficiency the protime (PT) and partial thromboplastin time (aPTT) is mildly prolonged in newborns, reflecting a slower rate of thrombin generation . In premature infants, these differences are reported to be more pronounced. It is also been reported that there is a significant difference in many coagulation factor levels between adults and children, although not clinically significant, they can have significant impact in a pediatric cardiac surgery patient when exacerbated by illness or dilution from cardiopulmonary bypass.
In a study by , they reported the levels of naturally occurring coagulation inhibitors are also altered in neonates. They found that protein C, protein S, antithrombin III, and heparin cofactor II are decreased, and levels of protein C and heparin cofactor II remain significantly lower than adult values until late childhood. On the other hand, α2-macroglobulin levels are elevated until adolescence; these altered levels of coagulation inhibitors make adequate anticoagulation necessary for cardiopulmonary bypass.
Children with congenital heart disease have been described to have variety of hemostatic abnormalities. Presence of thrombocytopenia is often inversely correlated with hemoglobin and patient age. In addition, thrombocytopenic patients are prone to have shortened platelet survival. reported dysfunction in platelets with prolonged bleeding times and impaired platelet aggregation in vitro studies among children with congenital heart diseases. Platelet function have also been affected by the perioperative medications like prostaglandins and amiodarone use. An acquired reduction of von Willebrand factor multimers has been also demonstrated in some children with congenital heart disease. In addition, level of coagulation factors like V, VIII and fibrinogen are often reduced in congenital heart disease, this in turn reflects the presence of chronic disseminated intravascular coagulation. In order to provide adequate intraoperative hemostasis during cardiopulmonary bypass, appropriate transfusion support becomes inevitable.
showed in their study that perioperative administration of tranexamic to a representative group of children undergoing cardiac surgery did not produce a significant reduction in blood loss except during first 6-hour time block.
Although anti-fibrinolytics differ in the mode of action however, their aim is to reduce the post operative bleeding due to deranged haemostasis cause by cardiopulmonary bypass. The evidence provided by the trials make the clear suggestions that they appear effective in reducing peri-operative and post-operative transfusion requirements among paediatric cardiac patients. As evident form one of the trials comparing aprotinin and tranexamic acid, it was evident that with the use of anti-fibrinolytics the rate of re-operation due to continued or recurrent bleeding rates almost half in some cases and overall were reduced. In addition, the mortality and morbidity due to the use of anti-fibrinolytics in cardiac surgery did not show any statistical significance, or did not increase with the use of tranexamic acid.
The studies also support that volume of blood transfusion decreased modestly among tranexamic acid group in comparison to placebo groups. Another study showed that significant 29% relative risk reduction in the rate of exposure to allogeneic blood transfusion in patients treated with tranexamic acid.
There have been reported adverse outcome with the use of tranexamic acid, in a multi-analysis of eight trials data on non-fatal myocardial infarction suggested no statistical significance. Out of 707 participants randomized between treatment group and control group, 391 in tranexamic acid group and 316 in control group, the relative risk of sustaining a non-fatal myocardial infarction in those treated with tranexamic acid was not elevated. There was a relative risk reduction of (RR=0.69 with 95%CI 0.21 to 2.29).
Stroke was another adverse event recorded with the participants of the trail. There have been reports of increased susceptibility to intra-operative stroke with tranexamic acid; however the data from six randomized trials showed that the number of events was small and pooled analysis of the data showed the relative risk of sustaining a stroke in those participants treated with tranexamic acid was not significantly increased.
Since coagulation profile is changes by anti-fibrinolytics there have been reports of deep vein thrombosis among patients treated with tranexamic acid. The review reported data on five trials concluding that relative risk of developing deep vein thrombosis among participants treated with tranexamic acid was not greater than the control group. The analyzed data showed relative risk of RR=0.84 (95% CI; 0.30 to 2.30).
Pulmonary embolism has also been associated with use of tranexamic acid; however, the review from five randomized trails suggests that there is no statistical significance. After the data analysis the relative risk of developing pulmonary embolus was not increased by the use of tranexamic acid (RR=0.32; 95%CI; 0.07-1.56). In addition there was no significant increase in developing thrombosis in those participants's treated with tranexamic acid.
Renal failure is one of the predicted outcomes of the use of cardiopulmonary bypass, in the review of data only two trails of tranexamic acid group reported data on renal failure or renal dysfunction. In total there were 240 participants randomized in two groups of 121 in tranexamic acid group and 119 in control group respectively. The relative risk of developing renal failure of renal dysfunction among tranexamic acid was not increased (RR=0.87; 95%CI: 0.08
Preoperative administration of tranexamic acid in cyanotic children under going cardiac surgery produces a significant decrease in blood product requirements,
Efficacy of tranexamic acid
Tranexamic acid was introduced as an anti-fibrinolytic and its primary efficacy measures are post operative bleeding, and its sequelae; transfusion, prolonged chest closure time and re exploration. These studies have shown the efficacy of anti-fibrinolytic treatment in decreasing bleeding and transfusion. The studies reported that 24hr blood loss decreased from 11%-4% and treated patients received 20%-50% less blood then controls. In addition, sternal closure times were reduced form 6-25min and re-exploration rates were improved by 50%-100% with anti-fibrinolytic treatment.
One of the studies highlighted the effect of dosage on the efficacy of anti-fibrinolytic treatment with tranexamic acid. For adults there is a dosage regimen, however for paediatric there has not been a no effective dosage regimen set for tranexamic acid. In this study by Chauhan et al. (2003) tranexamic acid versus control four different dosing regimens were studied. All does studied were effective in decreasing bleeding and transfusion requirements. However, one dosing scheme of giving single 50mg/kg bolus dose after anaesthesia and not followed by any kind of infusion. This dosing scheme did not show any benefit on the outcomes of blood loss and transfusion requirements. This is evident for the trail that dosing is an important issue for paediatric patients. Even though it was evident that range of dosage regimen was effective for the efficacy of tranexamic acid however; further research may ne need for appropriate dosing in paediatric patients.
On the other hand safety of lysine analogues in congenital cardiac study does not have sufficient power to determine safety. Theoretically, thrombosis is the primary serious risk of lysine analogue treatment. However, none of the prospective studies on efficacy have documented any increase in the complications of severe thrombosis treated with anti-fibrinolytics therapy. In literature there is case reports of severe thrombotic complications in patients treated with anti-fibrinolytics including a thrombosis of a neonate of Extra Corporeal Membrane Oxygenation (ECMO). Another study with 71 patients having modified Fontan procedure found no increase in the baffle fenestration closure with anti-fibrinolytic treatment, but it was under powered.
Larger studies are needed to asses the likely hood of various infrequent or rare adverse events related to anti-fibrinolytic use, such as stroke, renal failure, thrombosis, myocardial infarct, anaphylaxis and death. However one have to keep in mind the nature of these lysine analogues being generic products, and the manufacturers have little or no incentive in sponsoring such investigations.
Use of pharmaceutical agents to correct haemostasis, coagulation and fibrinolysis in paediatric cardiac surgery has received considerable light over the years. Various pharmaceutical agents have been introduced to overcome the deleterious effects of coagulation disorder caused by hemodilution and activation by cardiopulmonary bypass machine. On one side these agents aid in haemostasis but on the other hand they can impart adverse effects that can lead to permanent injury or death. One of the pharmaceutical agents used in paediatric cardiac surgery to correct is tranexamic acid. Over all the studies showed that tranexamic acid is at least as effective as other anti-fibrinolytic in reducing the blood loss and blood transfusion among paediatric patients in cardiac surgery. It was also evident with respect to tolerability and potential costs tranexamic acid has advantages over other anti-fibrinolytics like aprotinin. For cyanotic children the effect was profound as compared to a-cyanotic children; however the dosing regime had to be stream lined in order to increase the benefit form its use in paediatric cardiac surgery.