Hemostasis and massive transfusion


  • 6C-PL7-02

Marcel Levi, Department of Medicine, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands
E-mail: m.m.levi@amc.uva.nl


Blood loss may occur either spontaneously or upon an invasive procedure. Bleeding may be minor, major or excessive, for which classification various criteria are used. Commonly, excessive blood loss is (arbitrarily) defined as bleeding at a rate of > 200 ml/h, and or bleeding leading to hemodynamic instability, and/or bleeding leading to transfusion of > 5 U red cell concentrate within 3 h. For major bleeding, a more consensual definition is used, i.e. any bleeding leading to a clinically significant intervention [such as transfusion, hospital admission, a procedure (e.g. endoscopy)] and any intracranial, intra-abdominal, retroperitoneal, or intrathoracic bleeding.

The leading causes of major blood loss (20% of total blood volume or more) are surgical operations, among which the most frequently implicated are cardiovascular operations, liver transplantation and resection, major orthopaedic procedures, including hip and knee replacement and spine surgery [1]. Non-surgical causes of excessive blood loss may also occur. For instance, bleeding is the second most important cause of death in patients with trauma, contributing to approximately 30% of trauma-related mortality. There are also situations in which bleeding poses a major clinical challenge because of its critical localization, as illustrated by intracerebral haemorrhage.

Severe bleeding often results in the need of allogeneic blood transfusion. Even if benefits of transfusion outweigh the still existing risks (allergic reactions, transmission of infections, acute lung injury and immunosuppression), the implementation of strategies meant to minimize the use of a limited community resource is mandatory (see further chapter on Blood transfusion in the ICU). The most obvious and probably most effective strategy is the improvement of surgical and anaesthesiological techniques. A successful example is liver transplantation, which required huge amount of blood products in the past but is now performed with minimal transfusion requirements.

In the bleeding patient, it is first important to rule out abnormalities of hemostasis that can usually be corrected by replacement of the defective components. However, there are cases for which no surgical or hemostatic cause can be identified and yet excessive blood loss warrants the adoption of pharmacological strategies, broadly divided in perioperative prophylaxis during at-risk operations or intervention should bleeding occur.

Pro-hemostatic treatment

Most experience with pro-hemostatic therapy has been accumulated in the prevention and treatment of bleeding in patients with congenital and acquired coagulation defects. Indeed, specific correction of a hemostatic defect is highly effective in this situation, as for example has been shown in the management of haemophilia with coagulation factor concentrates. There is, however, increasing evidence that also in patients with less specific abnormalities or even a normal coagulation status and who encounter severe bleeding or are at high risk for bleeding, promoting hemostatic function may be of benefit. Interestingly, there seems in general not to be a strong need to specifically target a factor or pathway in the coagulation or fibrinolytic system that is causally related to the hemostatic defect, because interference in one part of the system may be able to compensate for a defect in another part.

The safety of pro-hemostatic therapy also deserves some consideration. Interfering in the balance between coagulant and anticoagulant mechanisms can indeed result in undesirable adverse effects. The best illustration may be the higher risk of bleeding in patients receiving anticoagulant therapy. Conversely, pro-hemostatic agents may, at least theoretically, predispose for thrombotic complications. The occurrence of such complications, which are fortunately relatively rare, seems to be very much dependent on considerate clinical use of this therapy. Obviously, the expected benefit of the application of pro-hemostatic agents in distinct clinical situations should be balanced with the risk of thrombosis in that particular patient population. Ideally, the benefit/risk ratio should be evaluated in properly controlled clinical trials.

Desmopressin (DDAVP)

De-amino d-arginine vasopressin (DDAVP, desmopressin) is a vasopressin analogue that despite minor molecular differences has retained its antidiuretic properties but has much less vasoactive effects. DDAVP induces release of the contents of the endothelial-cell-associated Weibel Palade bodies, including von Willebrand factor. Hence, the administration of DDAVP results in a marked increase in the plasma concentration of von Willebrand factor (and associated coagulation factor VIII) and (also by yet unexplained additional mechanisms) a remarkable augmentation of primary hemostasis as a consequence. A rare but important adverse effect of DDAVP is the occurrence of acute coronary syndromes, such as myocardial infarction and unstable angina, particularly in patients with pre-existing unstable coronary artery disease, probably because of the remaining vasoactive effect of the drug. Hence, in these patients, the use of DDAVP is contraindicated. The antidiuretic effect of a single or once repeated dose of DDAVP is clinically not very significant and may be dealt with by fluid restriction for some time. An exception to this rule are children, who may experience a severe dilution hyponatremia after the administration of DDAVP, which should be monitored clinically for 24 h after its administration.

Recombinant activated factor VII

Based on the current insight that activation of coagulation in vivo predominantly proceeds by the tissue factor/factor VII(a) pathway, recombinant factor VIIa (NovoSeven®, Novo Nordisk, Bagsvaerd, Denmark) has been developed as a pro-hemostatic agent and is now available for clinical use. Indeed, recombinant factor VIIa appears to exert potent pro-hemostatic effects. Most experience with recombinant factor VIIa has been accumulated in patients with severe coagulation defects that are difficult to treat, such as patients with antibodies to coagulation factors (e.g. so-called inhibitors to factors VIII or IX) and excessive bleeding. In addition, in patients with severe thrombocytopenia or disorders of primary hemostasis that fail to respond to conventional treatment, recombinant factor VIIa has been applied. In most of these situations, administration of recombinant factor VIIa was shown to be effective in controlling bleeding, although most of the reports are uncontrolled series. Recombinant fVIIa has been licensed for the prevention and treatment of bleeding in patients with antibodies to coagulation factors and complex platelet disorders. Nevertheless, there is an abundant number of case reports and case series on the hemostatic efficacy of rFVIIa in patients undergoing surgery. However, virtually all these reports did not include adequate controls, which makes it very hard to draw any clinical conclusion from them. Notwithstanding this lack of sound evidence, there is widespread off-label use of rFVIIa in patients with excessive blood loss that is hard to treat by conventional measures. The first controlled trial of recombinant fVIIa was carried out in patients undergoing prostatectomy. This study demonstrated that the administration of recombinant factor VIIa was associated with a 50% reduction in perioperative blood loss, thereby completely eliminating the need for blood transfusion. Preliminary trials in patients with liver cirrhosis undergoing laparoscopic liver biopsies or liver transplantation indicated that the use of recombinant factor VIIa may limit blood loss and prevent transfusion and subsequently large trials in major liver surgery were performed. In parallel, based on successful case reports and encouraging preclinical studies, administration of recombinant factor VIIa to trauma patients with excessive blood loss has also been evaluated in large controlled multicenter clinical trials. Generally, all trials show that administration of relatively high doses of recombinant factor VIIa is effective in reducing blood loss and reducing (excessive) transfusion; however, the intervention is not significantly effective at clinically directly more relevant outcomes, including mortality. A recent phase II dose-escalation and placebo-controlled study in 172 patients with major blood loss after cardiac surgery showed that the administration of rFVIIa resulted in less blood loss, a significant reduction in the need for reoperation (25% in the placebo group, 14% in the group receiving rFVIIa 40 μg/kg and 12% in the group receiving rFVIIa 80 μg/kg), and a significant larger proportion of patients not needing any transfusion after the administration of rFVIIa (10% in the placebo group, 28% in the group receiving rFVIIa 40 μg/kg and 32% in the group receiving rFVIIa 80 μg/kg). However, there were more serious adverse events in the patients treated with rFVIIa (13% vs. 7% in the placebo group, not significant). Indeed, in view of its pro-hemostatic potency, the safety of rFVIIa, in particular as related to the potential occurrence of thrombosis, has been the subject of attention and surveillance. In controlled clinical trials, administration of this agent resulted in a relatively low incidence of thrombotic complications, comparable to placebo-treated patients [2,3]. However, most of these studies were carried in patients with impaired coagulation or at low risk for thrombosis. In the trial carried out in patients with a much higher risk, such as those with intracerebral haemorrhage, serious thromboembolic events, mainly myocardial or cerebral infarction, occurred in 7% of patients treated with rFVIIa, when compared with 2% of placebo-treated patients. Hence, there is some indication that rFVIIa may heighten the risk of thrombotic complications, and this needs to be offset to its potential benefit in patients with severe blood loss. More randomized controlled trials are needed to establish efficacy and safety in patients undergoing cardiac surgery.

Antifibrinolytic treatment

Agents that exert anti-fibrinolytic activity are aprotinin and the group of lysine analogues. The pro-hemostatic effect of these agents proceeds not only by the inhibition of fibrinolysis (thereby shifting the procoagulant/anticoagulant balance towards a more procoagulant state), but also because of a protective effect on platelets, as has been demonstrated at least for aprotinin [1]. Aprotinin is a 58 amino acid polypeptide, mainly derived from bovine lung, parotid gland or pancreas. Aprotinin directly inhibits the activity of various serine proteases, including plasmin, coagulation factors or inhibitors and constituents of the kallikrein-kinin and angiotensin system. This rather non-specific mode of action of aprotinin is frequently considered as a disadvantage for its use; however, the interactions of aprotinin with proteases other than plasmin have never been demonstrated to cause clinically important adverse effects. The clinically most important side-effect of aprotinin is a rarely occurring but sometimes serious allergic or anaphylactic reaction. The use of aprotinin is contraindicated in case of ongoing systemic intravascular activation of coagulation, as in disseminated intravascular coagulation (DIC), and in patients with renal failure. Although many meta-analyses of controlled clinical trials with aprotinin have confirmed the potency of this agent to reduce (perioperative) blood loss and transfusion requirements [4], the safety of aprotinin was questioned by a study in 4374 patients who underwent elective coronary-artery bypass surgery. The study was observational and non-randomized but used a propensity score method to balance the covariates. Compared with untreated controls, aprotinin (but neither aminocaproic acid nor tranexamic acid) doubled the occurrence of severe renal failure, increased the incidence of myocardial infarction or heart failure by 55% and was associated with a nearly twofold increase in stroke or other cerebrovascular events. Subsequently, similar results were found in two other studies and in an observational survey conducted by the manufacturer of aprotinin in 67 000 patients undergoing cardiac surgery. A prospective randomized trial comparing aprotinin and lysine analogues in 2331 high-risk cardiac surgery patients confirmed a higher 30 day mortality in the aprotinin group (6·0%) in comparison with 3·9% in the tranexamic acid group and 4·0 in the aminocaproic acid group. Based on all these findings, the FDA has suspended the license of aprotinin in the United States, and the manufacturer has stopped the distribution of the agent in the rest of the world.

Lysine analogues, i.e. ε-aminocaproic acid and tranexamic acid are potent inhibitors of fibrinolysis. The antifibrinolytic action of lysine analogues is based on the competitive binding of these agents to the lysine-binding sites of a fibrin clot, thereby competing with the binding of plasminogen. Impaired plasminogen binding to fibrin delays the conversion of plasminogen to plasmin and subsequent plasmin-mediated fibrinolysis, which then proceeds at an inefficient and slow rate. Subtle molecular variations between different lysine analogues may have important consequences for their fibrinolysis-inhibiting capacity. Indeed, tranexamic acid (Cyklokapron®, Pfizer, New York, USA) is at least 10 times more potent than ε-aminocaproic acid (Amicar®, Xanodyne Pharmaceuticals, Newport, KE, USA). The use of lysine analogues is contraindicated in situations with ongoing systemic activation of coagulation (such as in DIC) and furthermore in case of macroscopic hematuria, because the inhibition of urinary fibrinolysis because of the high concentrations of the antifibrinolytic agent in the urine may result in deposition of urinary tract-obstructing clots. In view of the studies showing an efficacy in reducing blood loss of tranexamic acid that is similar to that of aprotinin, tranexamic acid (most frequently used total dose 3–10 g, usually divided in a loading dose of 2–7 g and a maintenance dose of 20–250 mg/h, or given as bolus doses of 1 g, four times daily) is the most appropriate antifibrinolytic agent for use in patients with major blood loss or patients undergoing surgery that is high risk for bleeding.