| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1996;111:982-987
© 1996 Mosby, Inc.
SURGERY FOR CONGENITAL HEART DISEASE |
This work was supported in part by a grant from Kabi Pharmacia, Mississauga, Ontario, Canada. Dr. Zonis is partially supported by a grant from the American Physician's Fellowship.
Received for publication Feb. 17, 1995 Accepted for publication July 26, 1995. Address for reprints: Michael Seear, FRCPC, British Columbia's Children's Hospital, Intensive Care Unit, 4480 Oak St., Vancouver, British Columbia, Canada V6H 3V4.
Abstract
Children undergoing cardiac operations in which cardiopulmonary bypass is used are at risk of significant postoperative blood loss. The acquired coagulopathy is complex but is thought to be due, in part, to excessive fibrinolysis. We examined the possibility of reducing postoperative blood loss in children by using the antifibrinolytic drug tranexamic acid. Using a prospective, randomized, double-blind study design, we administered a single dose of tranexamic acid (50 mg/kg intravenously) or saline placebo, before skin incision, in 88 children undergoing cardiac operations. Postoperative blood loss and fluid replacement were recorded for the next 24 hours. In addition, hemoglobin, platelet counts, and coagulation measures were recorded every 6 hours. When all patients were examined, there was no significant difference in postoperative blood loss between the treated and placebo groups (21.2 ± 12 ml/kg per 24 hours, tranexamic acid, vs 27.2 ± 20.3 mls/kg per 24 hours, placebo). However, when the children with cyanosis were analyzed separately, there was a highly significant difference in blood loss between the groups during the first 6 hours (11.2 ± 3.7 ml/kg per 6 hours, tranexamic acid, vs 27.2 ± 11.4 mls/kg per 6 hours, placebo; p < 0.002), as well as the overall 24 hour study period (23.7 ± 7.5 mls/kg per 24 hours, tranexamic acid, vs 48.9 ± 27.6 mls/kg per 24 hours, placebo; p < 0.02). Also significantly less blood and blood products were administered to the treated cyanosed group. Tranexamic acid produced a significant reduction in postoperative blood loss and blood product requirements in children with cyanosis undergoing heart operations. The drug had no effect in children without cyanosis or those requiring a second thoracotomy. (J THORAC CARDIOVASC SURG 1996;111:982-7)
Children undergoing cardiac bypass operations are at risk of significant postoperative bleeding.
1,2 According to one of our own studies, 3 the average child loses 30 ml/kg, roughly half the child's blood volume, in the first 24 hours after a routine cardiac operation. Many lose significantly more than this. It is commonly claimed that children with cyanosis and those undergoing a second thoracotomy are at particular risk of excessive blood loss. 4 Apart from the cost associated with large transfusions, the patient is also exposed to the dangers of blood transfusion 5 and, in some cases, reoperation. 2 The most common finding at the second thoracotomy is a generalized ooze of blood from the surgical field, suggesting an acquired coagulopathy induced by cardiac bypass.
Many abnormalities of coagulation have been described after cardiac surgery, 1,6 including thrombocytopenia, thrombocytopathy, decreased clotting factors, increased fibrinolysis, inadequate heparin neutralization, and disseminated intravascular coagulation. Despite the difficulty of determining which abnormality is most significant, the consensus is that the main cause of nonoperative blood loss after cardiopulmonary bypass is a combination of excessive fibrinolysis 7 and an acquired platelet defect. 1 Consequently, attempts have been made to reduce blood loss with the use of antifibrinolytic drugs. 8,9
Fibrinolysis follows reversible binding between lysine residues in fibrin and specific receptor sites in plasminogen. Reversible inhibition of absorption by lysine analogs, such as epsilon aminocaproic acid or tranexamic acid, displaces plasminogen and effectively inhibits fibrinolysis. 10 Earlier studies used epsilon aminocaproic acid, but a newer drug, tranexamic acid, is roughly ten times more potent, has a longer half-life, and is now more commonly used. Both drugs have few side effects. 11 The naturally occurring inhibitor of proteolysis, aprotinin, is used less often than lysine analogs because of its high cost. Plasmin also binds to platelets and activates them. Inhibiting this step with lysine analogs has been shown to preserve platelet adenosine diphosphate levels, 12 as well as to decrease blood loss. 13
Both epsilon aminocaproic acid 14,15 and tranexamic acid 13,16 have been shown to reduce blood loss in adults, although not in all studies. 17 Only one pediatric study has shown a slight benefit that was more pronounced in cyanotic children. 18 We investigated the possible benefit of antifibrinolytic therapy in children by use of a double-blind, randomized, prospective study of tranexamic acid versus saline placebo in 88 patients undergoing cardiac bypass operations.
The primary objective of this study was to investigate whether a single preoperative dose of tranexamic acid affects the amount of blood loss after cardiopulmonary bypass in children undergoing cardiac operations. The secondary study objective was to investigate that effect in two high-risk subgroups: children with cyanotic heart disease, and those with a history of previous thoracotomy.
Methods
The study was conducted in the Intensive Care Unit at the British Columbia's Children's Hospital. All operations were performed by the same two cardiac surgeons. Postoperative management was provided by three pediatric intensivists. Ethical approval was obtained from the Clinical Screening Committee of the University of British Columbia. Eighty-eight consecutive patients requiring cardiac operations were enrolled. All legal guardians were aware of the study and gave informed consent. Preenrollment exclusion criteria consisted of a history of hematuria, renal failure, previous thrombotic episodes, or past bleeding complications.
Study design
Using a system of sealed envelopes, we allocated the patients to one of two groups: the treatment group received a single 50 mg/kg intravenous dose of tranexamic acid (Cyclokapron, Kabi Pharmacia, Mississauga, Ontario, Canada) and a control group received an equivalent volume of saline placebo. Drugs were administered by the anesthetist before the first skin incision. Only the pharmacist was aware of the patient treatment groups.
Because no literature exists concerning the use of tranexamic acid in pediatric heart surgery, the dose range had to be extrapolated from adult work. Unfortunately, literature recommendations range from 7 mg/kg 12 to 100 to 120 mg/kg. 19 It is generally claimed that 10 to 15 mg/kg of tranexamic acid inhibits fibrinolysis in normal subjects. 8,10 In discussion with the surgical service in Toronto General Hospital, who have extensive experience with the drug, 19 we decided on a single preoperative dose of 50 mg/kg given by slow intravenous infusion. Because the drug was given before the first skin incision, we did not alter the dose to account for subsequent differences in bypass priming volume.
Because estimates of intraoperative blood loss are too inaccurate for study purposes, we collected and measured total chest tube drainage in 6, 12, and 24 hour blocks, starting from the time of chest closure. No correction was made for the hematocrit value of chest tube blood. In addition, the total amount of blood products transfused, hemostatic parameters (prothrombin time, partial thromboplastin time, platelets, and fibrinogen), hemoglobin, and urine output were also recorded every 6 hours.
Five percent albumin was used to replace all chest tube losses. The indication to administer blood or other blood products was at the discretion of the attending intensivists who where blinded to the group assignment of the patients.
Sample size calculation
Using blood loss over the first postoperative 24 hours as our primary outcome, we expected the mean blood loss to be 25 ml/kg per 24 hours with a standard deviation of 75%. We wanted to be able to detect a difference of at least 50% in the treatment group.
Choosing an 
of 0.05, and a ß of 0.2 (power of 80%), the sample size calculation yields 36 patients in each group. To compensate for patient exclusions, we arbitrarily increased this number by 20% to give 44 children in each group.
Statistical analysis
All values are expressed as mean ± standard deviation. Discrete differences between the two groups were compared by unpaired t test except for within-group comparisons of preoperative and postoperative coagulation data, when a paired t test was used. Frequency data were compared by means of Fisher's exact test. Serial measurements, such as blood loss with time, were compared by means of repeated measures analysis of variance, followed by Duncan's multiple range post hoc test for significant values of F. In keeping with many pilot studies, it was necessary to make multiple statistical comparisons. To compensate for the increased chance of a type I error, the commonly accepted level of significance was halved to 2.5%.
Results
Patient exclusions
No patients met the preenrollment exclusion criteria. Of the 88 patients enrolled, six were excluded during the study period. Three of these, all from the placebo group, were returned to the operating room for mediastinal exploration during the first 24 hours because of excessive bleeding. A surgical cause for the blood loss was found in each case. In two other cases, also from the placebo group, the study code was broken at the request of the treating intensivist because of excessive blood loss. Both patients were given intravenous tranexamic acid and large infusions of blood products. Neither patient required reoperation. A sixth patient, from the tranexamic acid group, had hypotension on arrival in the intensive care unit. After the patient had been given large volumes of blood products, the surgical site was reexplored. No active bleeding was found. The child had an overt disseminated coagulopathy but subsequently made a good recovery. The reexploration rate in our study was 4.6%, similar to the proportion reported in the literature. 2 No patients died and no side effects were attributed to the treatment group.
Patient characteristics
The characteristics in the remaining 82 patients are shown in Table I. Apart from minor differences in the type of procedure, the groups are comparable with respect to age, weight, gender, and duration of bypass, crossclamp, and circulatory arrest times. Frequencies of cyanotic lesions, Down syndrome, and repeated thoracotomies were also comparable.
|
Table II. The postbypass coagulopathy in this study was typical of our earlier experience. 3 Significant within-group reductions were observed in hemoglobin value and platelet count, accompanied by significant increases in prothrombin time and activated partial thromboplastin time. However, no significant differences were seen when the treatment and placebo groups were compared, including postoperative fibrinogen levels (1.43 ± 0.31 gm/L placebo, versus 1.60 ± 0.32 gm/L, tranexamic acid; p = not significant). Postoperative blood losses in the two treatment groups are shown in Table III. No significant difference in total blood loss was seen between the treated and placebo groups when all children were analyzed together.
|
|
Table IV shows that blood loss in the subgroup with cyanosis was almost halved after treatment with tranexamic acid (23.7 ± 7.5 vs 46.9 ± 27.6 ml/kg per 24 hours; p < 0.02). Table V shows that tranexamic acid appears to have no effect at all in the acyanotic subgroup, either in total blood loss or in any of the measured time blocks (20.6 ± 12.8 mls/kg per 24 hours, tranexamic acid, vs 20.4 ± 12.7 mls/kg per 24 hours, placebo). Apart from a significant difference in hemoglobin levels between children with and without cyanosis (153.3 ± 26.9 gm/L, cyanosis, vs 125.1 ± 14.6 gm/L, without cyanosis; p = 0.01), there were no other differences in preoperative blood work, including prothrombin time, partial thromboplastin time, platelet count, or fibrinogen level. In addition to reducing blood loss in the children with cyanosis, significantly fewer patients with cyanosis required packed red cell transfusions (1/8, tranexamic acid, vs 7/10, placebo; p = 0.02) and platelet transfusions (0/8, tranexamic acid, vs 6/10, placebo; p = 0.01).
|
|
Table VI. The same table shows that this difference is largely explained by the reduced blood loss in children with second thoracotomies who had cyanosis. Patient numbers were too small to achieve significance.
|
Preoperative administration of tranexamic acid to a representative group of children undergoing cardiac operations did not produce a significant reduction in blood loss except during the first 6-hour time block. This difference was entirely due to the drug's effect on children with cyanosis, in whom tranexamic acid roughly halved postoperative blood loss and significantly reduced blood product requirements. The drug had no effect in acyanotic children or those requiring a second thoracotomy. In common with DeLeon and associates, 4 we found cyanosis, but not second thoracotomy, to be a risk factor for excessive blood loss.
When blood loss is a measured end point, the exclusion of patients because of excessive blood losses runs the risk of experimental bias. We also analyzed the results with those who bled heavily included and found that this did not alter the study's final conclusions. Of the six patients excluded for heavy bleeding, five were from the placebo group and four had cyanosis. Inclusion of these patients in the analysis simply strengthened the difference in blood loss between the subgroups with cyanosis. Because there was still no significant drug effect in the patients without cyanosis, the study's conclusions were unchanged.
Several articles have reported reduced postoperative blood loss after the use of preoperative tranexamic acid in adult patients. 8,9 Direct comparison with adult work is difficult because the study populations differ widely: the mean age is much higher, the majority of patients have coronary artery disease, and very few have cyanotic lesions or Down syndrome. In addition, there is no uniformity in drug dosage or administration. Some studies use continuous infusions of antifibrinolytic drugs, 16 and others give a single drug dose, usually before bypass 20 but occasionally after bypass. 17 Even the size of the single dose varies by an order of magnitude from 7 mg/kg 12 up to more than 100 mg/kg. 19
There is no other report on the use of tranexamic acid in pediatric heart surgery, but some comparison can be made with McClure and Izsak, 18 who reported a significant reduction in blood loss in children receiving the antifibrinolytic drug epsilon aminocaproic acid. The effect was more pronounced in those children with cyanotic lesions and was confined to the intraoperative period.
It is not immediately obvious why tranexamic acid exerts its effect on blood loss. There is a large and sometimes contradictory literature concerning blood loss after cardiopulmonary bypass. Earlier work 20 tended to stress the importance of excessive fibrinolysis, but as the technology of cardiopulmonary bypass has improved, more recent work in this area emphasizes the importance of an acquired platelet defect. 2 We performed only simple clot-based estimates of coagulation (see 
Table II) and were unable to show any significant differences between the children with and without cyanosis, except for the widely described rise in hemoglobin concentration after chronic hypoxia. 21 Suarez and associates, 22 using biochemical markers of hemostasis in children with cyanotic congenital heart disease, showed highly significant increases in thromboxane, platelet factor IV, and ß-thromboglobulin, indicating marked activation of the platelet system. They were unable to demonstrate any evidence of increased fibrinolytic activity. Tranexamic acid may improve hemostasis after surgery in two ways: It blocks plasmin-induced platelet activation, consequently preserving platelet function 12; in addition, it is an effective inhibitor of fibinolysis. Whether one or either of these mechanisms is responsible for its effect in children is not possible to tell with our present study design.
More work is required to define the exact role of tranexamic acid in the control of bleeding after cardiac operations in children. In particular, a dose-response study is necessary because it is possible that repeat postoperative doses or a continuous infusion of tranexamic acid may improve the drug's effect. In addition, detailed studies of coagulation will be necessary to define the exact mode of action of this drug in the postbypass coagulopathy of children.
Footnotes
From the Departments of Intensive Care,a Anesthesia,b Surgery,c and Pharmacology,d British Columbia's Children's Hospital, Vancouver, British Columbia, Canada. 
References
This article has been cited by other articles:
|
|
 |
|
  M. P. Eaton Antifibrinolytic Therapy in Surgery for Congenital Heart Disease Anesth. Analg., April 1, 2008; 106(4): 1087 - 1100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Guay, P. de Moerloose, and D. Lasne Minimizing perioperative blood loss and transfusions in children: [Reduire les pertes sanguines et les besoins transfusionnels en chirurgie pediatrique]. Can J Anesth, June 1, 2006; 53(6_suppl): S59 - S67. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
