Antifibrinolytics in liver transplantation: They are effective, but what about the risk-benefit ratio?

Blood loss during orthotopic liver transplantation has been identified as an important cause of morbidity and, especially in the early days, also mortality.1, 2 With increasing experience and improvements in surgical technique, anesthesiological care, and better graft preservation methods, as well as changes in patient selection, most liver transplant centers have observed a steady decrease in blood transfusion requirements. Some centers are now even reporting liver transplantation without any need for blood transfusion in up to 30% of their patients.2 Despite these improvements, most patients undergoing liver transplantation still require blood transfusions, and controlling blood loss remains an everyday clinical challenge for surgeons and anesthesiologists caring for these patients. It is well know that blood transfusions are associated with an increased risk of postoperative complications, such as infections, pulmonary complications, protracted recovery, and a higher rate of reoperations.1, 2 Even today, when some centers report median red blood cell transfusion requirements of 2 to 3 units in their liver transplant patients, a significant correlation between intraoperative blood transfusion requirement and postoperative morbidity can still be found.2 Many studies have been performed during the past decades to elucidate the mechanisms of increased blood loss in liver transplantation.3 In the late 1980s, primary hyperfibrinolysis, due the accumulation of tissue-type plasminogen activator in the circulation during the anhepatic and early postreperfusion period, was identified as an important mechanism of bleeding during liver transplantation.4 This has provided the scientific basis for the use of antifibrinolytic drugs in an attempt to reduce intraoperative blood loss. The pioneering work by Kang and his colleagues at the University of Pittsburgh in the 1980s has stimulated a growing interest for the use of antifibrinolytics in liver transplantation.5 Whereas most studies in Northern America have subsequently focused on lysine analogues, such as epsilon aminocaproic acid (EACA) and tranexamic acid, aprotinin has been used in most European studies. Although both types of drugs have potent antifibrinolytic activity, there are differences in the working mechanism and the potential additional beneficial effects.6 The synthetic lysine analogues inhibit fibrinolysis by blocking plasminogen, the precursor enzyme of plasmin. Tranexamic acid is seven times more potent and has a longer half-life than EACA, and the recommended dosage is lower that that of EACA. Aprotinin is a naturally occurring serine-protease inhibitor that inhibits fibrinolysis by forming reversible complexes with plasmin. Unlike the synthetic lysine analogues, aprotinin does not only inhibit fibrinolysis: Due to its wide acting serine-protease inhibiting activity, it also has an effect on platelet function, and it has several anti-inflammatory properties.7 Despite the significant reduction in blood transfusion requirements observed in the initial studies in the 1980s,8 the use of antifibrinolytics in liver transplantation initially remained debated due to the lack of randomized, controlled trials. During the last decade, several placebo-controlled trials have confirmed the efficacy of tranexamic acid and aprotinin in reducing blood loss during liver transplantation.9-11 For unknown reasons, EACA has been studied in only one randomized, controlled trial.12 Although there has been a large variation in the dosage of tranexamic acid or aprotinin and some variability in outcome parameters, most studies have shown a significant reduction in blood transfusion requirements in patients receiving either tranexamic acid or aprotinin, compared to placebo. A direct “head-to-head” comparison between tranexamic acid and aprotinin, however, has never been reported. It is the group of the University Hospital Bellvitge in Barcelona, Spain, that should be commended for initiating such a trial. In this issue of Liver Transplantation, Dalmau et al.13 from Barcelona report the results of a prospective, controlled study comparing the prophylactic use of tranexamic acid and aprotinin in patients undergoing orthotopic liver transplantation. Blood transfusion requirements were low in each group, and no differences were observed among patients who received tranexamic acid or aprotinin. In another randomized trial, this group had previously shown a significant reduction in blood transfusion requirements in patients treated with tranexamic acid, compared with placebo.12 The observed low transfusion requirements in the current study are not only in line with previous placebo-controlled studies, but they also suggest that tranexamic acid and aprotinin are equally effective in this respect. If confirmed by other studies, this is important information, because tranexamic acid is much less expensive than aprotinin. When the two drugs appear to be equally effective with respect to blood loss reduction, the next two questions should be answered: Are both drugs comparable when it comes to other outcome parameters, and is there a difference in safety profile? In their paper, Dalmau et al. rightly discuss the other beneficial effects of aprotinin, such as its anti-inflammatory activity, which may explain the reported beneficial effects on hemodynamics and graft injury.14, 15 Such collateral beneficial effects have not been shown, nor can they be expected from tranexamic acid, because it lacks the wide-acting anti-inflammatory activity of aprotinin. Whether this is clinically relevant still needs to be proven. The second question is more difficult to answer, because none of the randomized, controlled studies performed so far have been large enough to have enough power to detect difference in, for example, thromboembolic complications. Since the clinical introduction of antifibrinolytics to prevent blood loss during liver transplantation, it has been debated whether these drugs may be prothrombotic when given to patients. This debate is primarily fed by case reports of patients who suffered thromboembolic complications while receiving an antifibrinolytic drug. In this issue of Liver Transplantation, Ramsay et al.16 report fatal thromboembolic complications in a liver transplant patient who received prophylactic aprotinin. The observed venous and arterial intravascular thrombosis and thromboemboli in this patient were extensive and are impressively illustrated by pictures taken during autopsy. Although the administration of aprotinin in this patient cannot be confirmed as the causative agent, the authors speculate that it might have precipitated this dramatic event. Thromboembolic complications, such as intracardiac thrombus and pulmonary embolism, have been described before in patients receiving antifibrinolytic medication.17 However, these complications have also been described in patients who did not receive such medication.5, 17, 18 For example, in the original report by Kang et al.5 on the use of EACA during liver transplantation, two cases of pulmonary embolism and one hepatic artery thrombosis were observed in 77 patients who did not receive EACA. No such complications were seen in 20 patients who were treated with EACA. Case reports may have an important signaling function, but the scientific proof of a causative relationship should come from controlled clinical trials. As mentioned above, unfortunately, the studies performed so far did not have enough power to either confirm or refute such a relationship. Upon review of the literature, it appears that the cumulative experience with the use of antifibrinolytics extends to more than 1000 liver transplant recipients. A multitude of these patients will have received an antifibrinolytic drug as part of a local protocol and the results have remained unpublished. Especially in Europe, many centers use aprotinin as routine prophylaxis in their patients. Thirteen randomized, controlled trials comparing the efficacy of one drug with placebo or with another agent have been reported so far. Within these trials, 493 patients were given aprotinin (in different dose regimens), 157 patients received tranexamic acid, 42 were given EACA, and 274 patients received placebo (Table 1). Although this is not a formal meta-analysis, there seems to be no obvious difference in the cumulative incidence of thromboembolic complications among these four groups. Similar data can be obtained by studying the published results of uncontrolled or retrospective series of patients. These latter data, however, should be interpreted with caution, because thromboembolic complications were not always clearly specified in these papers. In addition to this, the group in Berlin has extensive experience with the intraoperative administration of aprotinin in more than 1500 liver transplant recipients and they have not experienced any case of clinically evident intrapulmonary thromboembolism during the operation or during the first postoperative days in these patients (U. Neumann and P. Neuhaus, personal communication, 2003). If the current literature of randomized, controlled trials does not provide evidence of or even hints for an increased incidence of thromboembolic complications in patients receiving antifibrinolytics, than how should we view case reports such as described by Ramsay et al.?16 Do they provide for us an important message that has not yet been detected in clinical trials, or do they reflect coincidence? There is no doubt that thromboembolic complications are rare in liver transplant recipients, and the presentation of the case in Ramsay et al.'s paper is quite impressive. Although the exact incidence of thromboembolic complications in liver transplant patients is not known, it is believed to be underestimated.17, 18 Subclinical pulmonary thromboembolism may be common upon graft reperfusion, and an incidence rate of 1% has been reported for postoperative pulmonary embolisms.17, 19 Several risk factors have been suggested to play a role in the development of pulmonary embolism during liver transplantation (Table 2). For most of these, a true causative relationship has not been proven, and multiple risk factors may be present in one patient. The question is, do antifibrinolytic drugs, such as aprotinin, alone induce thromboembolic complications? Aprotinin itself does not have intrinsic procoagulant activity, nor does it stimulate the coagulation system. In fact, it has been shown to have a dose-dependent anticoagulant activity due to inhibition of several proteins involved in the coagulation cascade.20 Moreover, it inhibits thrombin-induced platelet activation by preventing proteolysis of the protease-activated receptor 1.7 In this context, aprotinin can be seen as a prohemostatic but not a procoagulant drug. These anti-coagulant properties, however, have not been described for the lysine analogues. The Cochrane group has performed a meta-analysis of all randomized controlled trials of antifibrinolytic drugs in adults undergoing nonurgent surgery (a total of 7027 participants), and these investigators did not find an association between aprotinin and thromboembolic events (relative risk 0.64; 95% CI: 0.31 to 1.31).21 If thromboembolic events occur in a patient who is receiving aprotinin, theoretically at least one other factor has to activate the coagulation cascade. In particular, the widespread intravascular thrombosis as observed by Ramsay and colleagues16 is suggestive of a systemic factor causing synchronous intravascular activation of the coagulation system at various sites in the circulation. Even if the antifibrinolytic activity of aprotinin has contributed to this effect by reducing the removal of intravascular formed fibrin, it is not likely to be the factor that has caused the systemic activation of the coagulation system. It is remarkable that there is a similarity in most cases of intracardiac thrombosis or pulmonary embolism reported in the literature; the events almost invariably occurred short after reperfusion of the new liver. Although this could suggest a role for the engrafted liver as a source of the procoagulant factor, we can only speculate about the exact underlying mechanisms. Ischemia/reperfusion-injury-related prothrombotic material may be released from the reperfused liver graft into the systemic circulation, and this subject warrants further research studies. In conclusion, the paper by Dalmau et al.13 adds to the existing evidence on the blood-sparing effect of antifibrinolytics in liver transplantation and suggests that tranexamic acid and aprotinin may be equally effective in this respect. If these findings can be confirmed by others, two questions will remain: 1) Do the additional anti-inflammatory properties of aprotinin make this drug preferable above the lysine analogues, despite the higher costs? 2) Is there an increased risk of thromboembolic complications when patients are given antifibrinolytics, as suggested by cases described by Ramsay et al.?16 Unfortunately, these two questions cannot be answered by the current literature, nor by these two papers, and more clinical studies will be needed to settle these issues. EACA, epsilon aminocaproic acid; HAT, hepatic artery thrombosis.