| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2005;28:563-568
© 2005 Elsevier Science NL
Original articles |
a Department of Cardiac Surgery, 3rd Medical School of Charles University Prague, Kralovske Vinohrady University Hospital, Srobarova 50, CZ-100 34 Prague 10, Czech Republic
b Department of Biostatistics and Informatics, National Institute of Public Health, Prague, Czech Republic
Received 9 May 2005; received in revised form 11 June 2005; accepted 28 June 2005.
* Corresponding author. Tel.: +420 267 163 422; fax: +420 267 163 260. (Email: vanek{at}fnkv.cz).
Abstract
Objective: To evaluate and compare hemostatic effects of tranexamic acid vs. aprotinin vs. placebo in off-pump coronary artery bypass (OPCAB) surgery and, in addition, to assess the safety of fibrinolytic inhibitors therapies. Methods: In a prospective, randomized, double-blind study finally 91 patients undergoing OPCAB were investigated (group A, n=32, tranexamic acid 1g before skin incision and continuously 200mg/h; group B, n=29, aprotinin 1,000,000IU before skin incision and 250,000IU/h; group C, n=30, placebo). Results: Highly significant inter-group differences were found in cumulative blood loss within 4h (geometric means [95% confidence intervals]—group A: 89.3 [72.7, 109.8] mL, group B: 72.3 [49.2, 106.3] mL and group C: 192.3 [151.8, 243.5] mL) (P<0.001), within 8h (group A: 152.1 [120.7, 191.6] mL, group B: 130.3 [88.1, 192.8] mL and group C: 283.8 [226.0, 356.3] mL) (P=0.001), and within 24h postoperatively (group A: 410.3 [337.6, 498.6] mL, group B: 345.8 [256.0, 398.2] mL and group C: 619.8 [524.3, 732.8] mL) (P<0.001). At all time points, placebo group C was significantly distinct from the groups treated with fibrinolytic inhibitors (groups A and B). However, no differences between groups A and B were found. Both mean hemoglobin and hematocrit values 24h postoperatively were different between the groups (P=0.018 and P=0.077, respectively), acheiving the lowest value in group C. Number of re-transfuzed patients was highest in group C, but without statistical significance (either packed red blood cells, P=0.119 or fresh-frozen plasma, P=0.118). We observed one postoperative myocardial infarction in aprotinin treated group B and one temporary postoperative myocardial ischemia in placebo group C, no cerebrovascular or pulmonary embolism was noticed. Treated groups A and B did not demonstrate postoperative increase in mean levels of myocardial enzymes, compared with group C. Significantly higher mean values of D-dimer were found in group C 24h postoperatively (P<0.001). Conclusions: Both tranexamic acid and aprotinin seem to be similarly effective in the reduction of postoperative blood loss in OPCAB. Tranexamic acid appears to be cost-effective and safe alternative to aprotinin.
Key Words: Tranexamic acid • Aprotinin • Off-pump coronary artery bypass • Hemostasis
1. Introduction
The favourable effect of fibrinolytic inhibitors (aprotinin, tranexamic acid, aminocaproic acid) on the decrease in perioperative bleeding in on-pump coronary surgery has been confirmed in a large number of controlled trials [1–4]. In many centers these pharmacological strategies are used on a routine basis [5,6]. However, only few studies have been concerned with the use of antifibrinolytic drugs in off-pump coronary artery bypass (OPCAB) surgery [7–10], although perioperative hemorrhagic complications and the consequent need for allogenic transfusions are still one of the major problems in this type of surgery [11–13].
The aim of this prospective, randomized, double-blind, placebo-controlled study was to evaluate and compare hemostatic effects of tranexamic acid and aprotinin in OPCAB surgery. In addition, the risk of perioperative myocardial ischemia was assessed.
2. Material and methods
After obtaining the Medical Faculty Ethics Committe approval (EK/243/2003, October 1, 2003) and the informed consent from all participants, from October 15, 2003 to July 31, 2004, 100 patients scheduled for OPCAB were enrolled in the study. The criteria for non-enrollment to the study were as follows: previous cardiac surgery, myocardial infarction <7 days prior to surgery, history of hematological or liver disordes, renal insufficiency (serum creatinine>150µmol/L) and preoperative anemia (hemoglobin<11g/L, hematocrit<32). Preoperative treatment with antiaggregative/anticoagulant drugs (aspirin withdrawal <5 days before surgery, low-molecular heparin withdrawal <24h before surgery, continuous heparin infusion) was not a contraindication to the inclusion into the study, but the number of medicated patients was carefully monitored. Nobody from the study subjects was preoperatively under the influence of potent antiplatelet agents, such as ADP inhibitors and GP IIb/IIIa antagonists.
According to our predetermined exclusion criteria nine enrollees were withdrawn from the study: six of them for a conversion on cardiopulmonary bypass in the course of surgery (presence of small intramuscular arteries and heavy calcification, hemodynamic instability), three of them for the need of postoperative re-exploration for hemorrage with the finding of an evident surgical source of bleeding (once the perforation of venous bypass by the edge of chest tube, twice bleeding from the branch of internal thoracic artery, which was used as a conduit). Finally, 91 patients were assessed in the study.
2.1 Pharmacological protocol
After the enrollment into the study, the patients were randomized by an independent pharmacologist of the study into three groups (A–C), the envelope method with random numbers was used. The independent pharmacologist prepared coded infusions with the study drug/placebo and was not directly involved in the clinical treatment of randomized patients. Both the operation theater staff and that of the intensive care unit were blinded regarding the study drug. The patients from group A (n=32) were given tranexamic acid (Exacyl, Sanofi Winthrop, France) 1g before skin incision and a continuous infusion of 200mg/h during the whole surgical procedure. The patients from group B (n=29) were given aprotinin (Gordox, Gedeon Richter, Hungary) 1 million IU before skin incision and onward 250,000IU/h. The patients from group C (n=30) were infused normal saline as a placebo. The basic characteristics of the patient groups are shown in Table 1
.
|
During the procedure and postoperatively the patients were administered crystalloid solutions and 5% albumin solution if necessary, synthetic colloidal solutions were strictly not used. A red blood cell transfusion was administered when hemoglobin decreased to less than 8.5g/dL and/or hematocrit less than 26. A transfusion of fresh frozen plasma was instituted (to correct a suspected deficiency of coagulation factors) when chest drain bleeding increased to >150mL/h or to >100mL/h for two consecutive hours.
The patients were operated on from full midline sternotomy, the left internal mammary artery was harvested in all cases with possible harvest of other grafts (great saphenous vein and /or radial artery). The verticalization of the beating heart was achieved using an Axius Xpose Device (Guidant, Cupertino, CA) while an Ultima Vacuum Assist (Guidant, Cupertino, CA) was used for the stabilization of the anastomosis site. The initial dose of intravenous heparin 100IU/kg was administered after harvesting the left internal mammary artery with target activated clotting time (ACT)—Celite-activated system—(Hemochron 401, International Technidyne Corporation, Edison, NJ) over 250s. On the completion of anastomoses, heparinization was partially reversed with half-dose of protamine chloride, regardless of ACT value.
A surgical postoperative re-exploration was based on our standard criterion: chest drainage 300mL/h for two consecutive hours, or 200mL/h for 3h, or signs of cardiac tamponade verified by echocardiography.
2.3 Laboratory analyses
Blood samples for evaluation of hematological parameters (hemoglobin, hematocrit, platelet count, prothrombin time, activated partial thromboplastin time, and fibrinogen) were taken and processed by a routine way. Myocardial enzymes were assessed before the operation, and 8 and 24h postoperatively. Creatine phosphokinase (CK) and isoenzyme MB (CK-MB) levels were determined by a dry chemistry method with the Vitros 950 analyzer (Ortho-Clinical Diagnostic and Johnson & Johnson, Raritan, NJ). Troponin I levels were analyzed by a chemiluminiscence method with Immulite Turbo analyzer (DPC, Los Angeles, CA), using specific antibodies Turbo Troponin I (DPC, Los Angeles, CA). D-dimer levels were assessed by a micro-latex imunoassay procedure for a quantitative measurement of D-dimer on the Stago Compact analyzer (Diagnostica Stago, Parsippany, NJ) using Liatest Stachrom D-D antibodies (Diagnostica Stago, Parsippany, NJ).
2.4 Statistical analysis
Statistical analysis was done by statistical software Stata, release 7.0 (Stata Corporation, College Station, TX) and SPSS, version 12.0.1 (SPSS, Inc., Chicago, IL). The location of the continuous variables was characterized by arithmetic or geometric means (for normally or log-normally distributed data, respectively) and their variability was shown by 95% confidence intervals. Categorical data were described using absolute and relative frequencies (expressed as percentages).
The statistical evaluation was based on various models of the analysis of variance. If a statistically significant result was obtained, Sidak's and Dunnett's post hoc tests were used to locate the differences between the groups. For categorical data, 
2 test and Fisher's exact test were applied. The comparison of postoperative blood loss between the groups was adjusted with respect to the preoperative treatment with antiaggregative/anticoagulant drugs. Troponin I values were analyzed using a specific form of the analysis of variance suitable for left censored data. All statistical tests were evaluated at significance level of 0.05.
3. Results
From 100 patients enrolled to the study, one patient (with emergent re-exploration for massive bleeding with the peroperative findig of perforation of venous bypass) died on postoperative day 10 from multiorgan failure and sepsis. Except for three patients, who underwent a postoperative surgical revision with the finding of an apparent surgical source of bleeding, no other patient required re-exploration in the ensuing course. Altogether nine enrollees (six patients converted to on-pump surgery+3 patients with surgical cause of hemorrhage) were withdrawn from the study, so the following data concern 91 assessed patients. Median intensive care unit length of stay was similar for all regimens (P=0.691): 23 (min–max 6–84) h (group A), 22 (15–72) h (group B) and 24 (3.5–69) h (group C), respectively. The median length of in-department stay was 6 days for all groups (group A: min–max 3–11, group B: 3–10, group C: 3–41). Two patients with the longest hospitalization (12 and 41 days) originated from group C, but the in-patient difference between groups was not significant (P=0.824). In group A, the tranexamic acid total cost was estimated at EUR 2.0 per patient, vs. the aprotinin total cost in group B estimated at EUR 82.5 per patient, on average, calculated from the Czech Republic market prices.
Groups A–C showed comparable demographic, preoperative hematological and basic intraoperative characteristics, with the exception of male/female ratio. In group A, an equal numbers of male and female patients were enrolled, in contrast to groups B and C with the majority of male study subjects. However, this inequality did not reach a statistical significance (P=0.144).
The number of patients under the influence of aspirine was almost the same in all groups (P=0.836). However, a slightly (but non-significantly) higher percentage of patients in group C had a low-molecular heparin withdrawal <24h before surgery (P=0.157) and a preoperative continuous unfractioned heparin infusion (P=0.367), in comparison with groups A and B. Neverthless, additional covariates were entered into the statistical model to control a potential confounding effect of antiaggregative/anticoagulant drugs on the inter-group comparison of postoperative bleedning.
3.1 Intraoperative and postoperative blood loss
No statistically significant differences were found between the groups in the intraoperative blood loss (geometric means [95% confidence intervals]—group A: 267.2 [215.8, 330.8] mL, group B: 241.9 [198.4, 294.8] mL and group C: 319.3 [256.0, 398.2] mL) (P=0.134).
Blood loss during the first 4h postoperatively shows highly significant differences between the groups (Fig. 1 ) (geometric means [95% confidence intervals]—group A: 89.3 [72.7, 109.8] mL, group B: 72.3 [49.2, 106.3] mL and group C: 192.3 [151.8, 243.5] mL) (P<0.001). Significant inter-group differences were also found in a cumulative blood loss within the first 8h (group A: 152.1 [120.7, 191.6] mL, group B: 130.3 [88.1, 192.8] mL and group C: 283.8 [226.0, 356.3] mL) (P=0.001) and within the first 24h postoperatively (group A: 410.3 [337.6, 498.6] mL, group B: 345.8 [256.0, 398.2] mL and group C: 619.8 [524.3, 732.8] mL) (P<0.001), respectively. There were no differences between the treated groups A and B at any time, but placebo group C marks off both groups A and B. The power of ANOVA test was higher than 0.95 at all postoperative time points.
|
|
3.3 Postoperative myocardial ischemia, myocardial enzymes, and levels of D-dimer
Electrocardiographical signs of myocardial ischemia in the very early postoperative period was detected in two patients (patient J.Z. from aprotinin group B and patient M.K. from placebo group C). Patient J.Z. underwent a single coronary artery bypass grafting (a free graft of left internal thoracic artery was used as a conduit to the circumference of extremely sclerotic left anterior descending coronary artery). Patient M.K. underwent double bypass technically without any complications. In both hemodynamic stable patients, urgent echocardiographical examinations showed no new myocardial wall motion hypokinesis and the situation was solved in a conservative manner by coronary dilatating therapy. In both cases, after a few hours of treatment the electrocardiographical signs of myocardial ischemia disappeared and the ensuing course of the patients was without complications, although patient J.Z. showed postoperative peak troponin I value 27.8µg/L. No signs of low cardiac output syndrome were observed in the postoperative period of all the assesed patients, and so no Swan-Ganz catheter utilization was required.
The time course of creatine phosphokinase (CK) levels, isoenzyme MB (CK-MB) levels and the ratio (relative index) of CK-MB/CK are shown in Table 3 . No statistically significant inter-group differences were found at any time (before the operation, 8h, and 24h postoperatively). Fig. 2 presents the development of troponin I levels. Troponin I levels differed between the three groups 8h postoperatively (P=0.015). The geometric mean of troponin I level was the highest in placebo group C; post hoc tests showed first of all a statistically significant difference between groups B and C (P<0.001), and then a less noticeable difference between groups A and B (P=0.047).
|
|
4. Discussion
Drugs that preserve hemostasis through plasmin inhibition include synthetic lysine analogues, such as tranexamic acid and the most potent, naturally occuring antifibrinolytic agent aprotinin, which is called ‘a broad-spectrum antifibrionolysin’ due to its antiinflammatory and endothelial modulating properties.The efficacy of these pharmacological strategies on the decrease in the frequency of surgical re-exploration and the need of allogenic blood transfusion in on-pump cardiac surgery has been proved many times; according to meta-analysis of Levi et al. treatment with aprotinin in this type of surgery decreased mortality almost 2-folds [17]. Only a limited number of studies have been concerned with the use of antifibrinolytics in OPCAB surgery, and have demonstrated its effectiveness, although the surgical aggression in OPCAB may be as important as (or even more important than) the use of cardiopulmonary bypass in terms of coagulation–fibrinolytic pathway activation [18]. An increase in fibrinolytic activity in OPCAB is supposedly due to the release of a tissue plasminogen activator, which starts during the skin incision and sternotomy and continues through the surgical tissue manipulation. To the best of our knowledge, this is currently the first prospective, randomized, double-blind study comparing tranexamic acid vs. aprotinin vs. placebo in OPCAB surgery. In our study we have proved that both tranexamic acid and aprotinin are likewise effective in reducing postoperative bleeding in OPCAB patients in comparison with placebo. This finding is in agreement with separate considerations for tranexamic acid [7,9,10] and aprotinin [8] in OPCAB surgery, as well. Our data does not show a significant difference in blood loss between tranexamic acid group A and aprotinin group B. According to published data in on-pump coronary surgery, blood loss at 24h in the tranexamic acid treated patients was altogether higher in comparison with the aprotinin treated patients (P=0.03), but with the similar perioperative transfusion requirements [19]. In our study we have demonstrated that the mean values of hemoglobin at 24h postoperatively were significantly lowest in placebo group C and that the borderline differences have been found in mean hematocrit values. We have not found any statistically significant differences in transfusion requirements (percentage of patients who received packed red blood cells or fresh-frozen plasma), but the total number of re-transfused patients was the highest in placebo group C. The lack of statistical significance is probably due to the small number of re-transfused patients on the whole.
The question of safety associated with the use of antifibrinolytic drugs has been discussed in on-pump cardiac surgery, as well [17,20]. There is a theoretical risk of an increased thrombotic tendency during the treatment with fibrinolytic inhibitors, and several earlier sporadic reports on coronary graft occlusion in patiens receiving these therapies have been published [21,22]. These observations have not been supported by any results of large clinical studies [17,23,24]. In our study we observed only one postoperative myocardial infarction in aprotinin treated group B and one temporary postoperative myocardial ischemia in placebo group C, and no cerebrovascular or pulmonary embolism was noticed. There were no inter-group differences in the mean levels of myocardial enzymes observed, with the exception of the highest troponin I level 8h postoperatively in placebo group C. Our experience thus indicates that the use of fibrinolytic inhibitors seems to be safe in OPCAB surgery, as well, but a larger number of safety trials is required for a better assessment of possible thrombotic complications [25].
Looking into the coagulation–fibrinolytic pathway activation, we found significantly higher mean values of D-dimer in placebo group C at 24h postoperatively, compared with groups A and B, treated with antifibrinolytic drugs. The post OPCAB finding of elevated values of D-dimer and the inhibition of D-dimer levels by fibrinolitic inhibitors is in agreement with Casati et al. [7] and Englberger et al. [8].
Based on the results of our prospective, randomized, double-blind TAP study we conclude that both tranexamic acid and aprotinin significantly reduce the postoperative blood loss in OPCAB patients, and the efficacy of tranexamic acid and aprotinin, respectively, seems to be quite similar. We did not observe any statistically significant difference in the need for allogenic transfusion, although the total number of re-transfused patients was the highest in placebo group. Tranexamic acid appears to be a potent, cost-effective and safe alternative to aprotinin in OPCAB surgery.
Acknowledgments
Financial support by the Cardiovascular research project of the Charles University Prague, Czech Republic, no. MSM0021620817.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
