Tranexamic Acid for Routine Use in Off-Pump Coronary Artery Bypass Surgery

Abstract

For both the patient and the surgical team, one of the most feared adverse events related to cardiac surgery is massive blood loss leading to substantial use of blood products. Because excessive fibrinolysis is one of the causes of blood loss in cardiac surgery, administration of antifibrinolytic agents has become a common strategy for blood conservation in patients at high risk for blood loss. Tranexamic acid is a synthetic lysine analog that imparts its antifibrinolytic action by competitively blocking lysine binding sites on plasminogen, thereby preventing plasminogen activation. Although inhibition of fibrinolysis may reduce bleeding, patients might be at greater risk of thrombotic events such as coronary graft occlusion, myocardial infarction, stroke, or death. Multiple randomized trials, and meta-analyses of randomized trials, have demonstrated that tranexamic acid significantly reduces blood loss and need for allogeneic red blood cell (RBC) transfusion in cardiac surgery when compared with placebo or inactive control.1 However, the majority of studies of tranexamic acid have been conducted in the setting of conventional on-pump coronary artery bypass (CCAB), which is known to be associated with inherently higher risk of allogeneic transfusion than is off-pump coronary artery bypass surgery (OPCAB). There is concern that tranexamic acid may provide less benefit to risk in patients undergoing OPCAB procedures. Although patients undergoing OPCAB experience comparatively less blood loss than CCAB, they are still at significant risk of allogeneic blood product exposure. A meta-analysis revealed that in the randomized controlled trial setting, 28% of patients undergoing OPCAB still receive allogeneic RBC transfusions, in comparison with 56% of patients undergoing CCAB.2 As a result, there remains incentive to test strategies such as antifibrinolytic agents to further reduce the need for blood transfusions during OPCAB. A number of small randomized trials of tranexamic acid in OPCAB have been published over the past decade.3–10 Each of these randomized trials compared tranexamic acid with placebo or inactive control, and each measured blood loss or transfusions as a primary endpoint. Only 1 of the 8 trials of tranexamic in OPCAB found a significant reduction in the risk of allogeneic blood product exposure.3 Synthesis of these 8 small randomized trials through meta-analysis showed a significant overall reduction in allogeneic transfusion (risk ratio 0.47; 95% confidence interval [CI], 0.33 to 0.66; P < 0.0001).11 However, none of the trials adequately reported the risk of clinically important events such as stroke, myocardial infarction, venous thromboembolism, and need for surgical re-exploration for bleeding. In addition, seizures12,13 were not reported in these studies, and it is unknown whether they occurred or were simply not reported. The small sample size of each of these trials (40 to 100 patients) and the short duration of postoperative follow-up precludes any conclusions about the safety and efficacy of tranexamic acid over the short- or long-term, even when combined through meta-analysis.11 Furthermore, a wide range of doses for tranexamic acid was used in these trials (loading dose ranged from 0.75 g to 2 g; maintenance dose ranged from 200 mg/h to 400 mg/h), which further prevents the ability to adequately assess the balance of risk versus benefit for tranexamic acid in the setting of OPCAB surgery. In this month's Journal, Wang et al. have conducted the largest known randomized trial to date of tranexamic acid in OPCAB, randomizing 231 consecutive patients undergoing elective OPCAB to tranexamic acid (1 g before incision, then 400 mg/h during surgery) or placebo.14 Patients in the tranexamic acid group experienced less blood loss from chest tubes (−237 mL at 24 hours; P < 0.001), and fewer patients received allogeneic RBC transfusions versus placebo (32% versus 47%, respectively; P = 0.019). These results are encouraging, given the 15% absolute risk reduction allogeneic blood exposure, which translates to a number needed to treat of 7 patients. This indicates that, on average, for every 7 OPCAB patients receiving tranexamic acid prophylaxis, there will be 1 less patient exposed to RBC transfusion than if no tranexamic acid had been used. Conversely, the corollary is that 6 patients received tranexamic acid without benefit. The problem is that we do not know which patient would be the one to benefit, and so we need to administer tranexamic acid to all of them if we are to benefit one of them. This raises the issue of statistical significance (yes, the P value was <0.05) versus the clinical relevance (is routine use warranted if only 1 of 7 patients benefited from transfusion avoidance?) of tranexamic acid in the setting of OPCAB. In the current study by Wang et al.,14 it is notable that 32% of patients still required allogeneic RBC transfusion despite receiving tranexamic acid, and the 47% rate of transfusion in the placebo group would be considered high in most centers that routinely perform OPCAB. This transfusion rate for OPCAB patients occurred despite the reported use of other blood conservation techniques, including discontinuation of platelet inhibitors at least 5 days prior, and intraoperative cell saver with autotransfusion of salvaged washed RBCs at completion of surgery. The threshold for transfusion was higher at hemoglobin of 9 g/dL or a hematocrit <27%. Therefore, the applicability of this study to other centers will be in question because a number of centers have reported much lower rates of transfusion in patients undergoing OPCAB.2,15 Lower baseline rates of transfusion would leave less room for tranexamic acid to impart significant reductions in need for transfusion in other settings. In addition to the concern regarding the potential for reducing allogeneic blood administration, there is also the concern that the risks of antifibrinolytics may be different for OPCAB than for CCAB.4,16–25 A number of studies have suggested that the degree and timeline of coagulopathy differs for OPCAB in comparison with CCAB.24 In particular, OPCAB has been shown in selected studies to result in greater activation of fibrinogen and acute-phase proteins that might lead to a “procoagulant” state, which has raised concerns regarding the potential for increased risk of thrombotic events, including venous thromboembolism and arterial embolism. The latter would at least theoretically have implications for stroke risk and graft patency over the short- and long-term.4 Although this procoagulant status has been suggested in a number of studies of coagulation profiles after OPCAB,4,16 some studies of OPCAB versus CCAB did not find a significant difference in procoagulant and fibrinolytic activity.21,22 Some have argued that the difference in coagulation status may be related more to the differences in heparin anticoagulation and reversal regimens used during OPCAB rather than any inherent difference in the physiologic response to the use of the bypass pump,18,23,24 and that we should refer to it as preserved hemostasis with OPCAB rather than a hypercoagulability response.21 If OPCAB truly induces a clinically relevant procoagulant state, then one would expect measurable increases in the risk of venous and arterial embolism and jeopardized graft patency in the clinical trial setting. Indeed, some studies and meta-analyses have suggested measurable reductions in graft patency with OPCAB versus CCAB,26–30 but controversy remains because other studies have not replicated this finding,31–37 and controversy surrounds the potential role for surgeon inexperience to be the reason for the reduction in graft patency.38–40 With regard to a potential procoagulant state in OPCAB leading to arterial thromboembolism, evidence suggests at least no increased risk of stroke, and perhaps even a reduction in overall risk of stroke, with OPCAB versus CCAB.2,41–43 Therefore, the jury is still out on whether the procoagulant state in OPCAB in relation to CCAB truly exists, and furthermore, what the clinical implications are. If there is a relevant procoagulant status (or simply a lesser anticoagulated status) for patients undergoing OPCAB, the fear is that the addition of antifibrinolytics such as tranexamic acid could tip the balance and amplify the risk of thrombotic events such as stroke and jeopardized graft patency. This becomes of particular concern when we consider that the magnitude of benefit to be gained from antifibrinolytics during OPCAB is already smaller to begin with because of the inherently lower risk of bleeding. Hence, the benefit– risk ratio for antifibrinolytics will be more easily tipped toward undue risk, perhaps without sufficient benefit to warrant the risks. Although the concerns of greater thrombotic risk for OPCAB remain largely theoretical, they cannot be dismissed as unimportant, because the reason they remain theoretical is that they have not yet been adequately tested. The existing clinical trials of tranexamic acid in OPCAB have, so far, failed to measure the more important outcomes of clinical relevance in adequately powered studies of sufficient duration.44 The use of tranexamic acid in OPCAB is preliminarily encouraging, but far from a “fait accompli.” More research is needed, and should be considered ethical. None of the trials of tranexamic acid in OPCAB has addressed the most important outcomes. We are interested in reducing exposure to blood products, but that should not be considered to be the end unto itself. The ultimate reason for avoiding blood products is to favorably impact on the risk of death, stroke, myocardial infarction, renal failure, reoperation for bleeding, and need for revascularization over the spectrum of months and years. Reducing blood exposure should be considered only a preliminary surrogate for these outcomes, which if achieved in early studies, should be validated in large studies adequately powered to test the real balance of clinical benefits versus risks. Seemingly paradoxically, the largest randomized trial of antifibrinolytics to date (the BART trial)45 found that, despite aprotinin's greater potency for reducing blood exposure, it did not reduce adverse events in comparison with the lysine analogs tranexamic acid and aminocaproic acid. As a result, this raises questions of the simplistic assumption that antifibrinolytics will lead to a reduction in adverse events through attenuation of blood loss.44 The study by Wang and colleagues affirmed that tranexamic acid can reduce the risk of blood exposure in OPCAB, and this confirms the results by Tagghadomi et al.3 and the meta-analysis of previous randomized trials.11 However, this preliminary knowledge of modest impact on blood loss should not be taken as sufficient to warrant uptake into routine practice. Perhaps tranexamic acid should be reserved for highest-risk OPCAB patients (i.e., reoperation, low hemoglobin, recent antiplatelet therapy, or difficulty cross-matching), but even then we lack the clinically relevant evidence that the benefits outweigh the risks. The question is no longer whether tranexamic acid can reduce blood loss for patients undergoing OPCAB. We already know that it can, on balance, reduce blood loss by about 250 mL and exposure (number needed to treat = 7).11,14 The real question is whether the reduction in blood loss or blood exposure for OPCAB patients is worth the potential added risks (seizures, and other unknown adverse events not yet discovered) and costs, and what the optimal dosage is. Currently, there is no study or meta-analysis adequately designed or powered to provide us with the answers.2,44 The best way past this uncertainty is to take stock of the current evidence, and agree as a global research community to do randomized trials of tranexamic acid in OPCAB that will move us beyond our current knowledge. Given the existing evidence,11,14 we do not need more trials with blood loss as the primary outcome. Rather, to reduce our uncertainty, we now need large trials that are powered to measure the outcomes that matter most to our patients. DISCLOSURES Name: Janet Martin, PharmD, MSc (HTA&M). Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript. Attestation: Janet Martin approved the final manuscript. Name: Davy Cheng, MD, MSc, FRCPC, FCAHS. Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript. Attestation: Davy Cheng approved the final manuscript. This manuscript was handled by: Jerrold H. Levy, MD, FAHA.