Antifibrinolytic therapy: new data and new concepts.

Activation of the fi brinolytic system is an integral part of vascular haemostatic mechanisms to maintain vascular patency. The basis of fi brinolysis is the conversion of the inactive substrate plasminogen to plasmin, an enzyme that cleaves fi brin but also has pleiotropic eff ects. Multiple mechanisms are responsible for generating plasmin, including endothelial activation and release of tissue plasminogen activator, and contact activation and kallikrein-mediated plasmin activation. Tissuetype and urokinase-type are the two major plasminogen activators expressed in many cell types and tissues. As part of the haemostatic balance, plasmin generation and activity are also modulated by multiple inhibitors that include plasminogen activator inhibitor 1, thrombinactivatable fi brinolysis inhibitor, and α2-antiplasmin. Thus fi brinolysis involves several regulatory mechanisms under physiological conditions. However, after the extensive tissue injury that occurs with trauma or surgery, the equilibrium is shifted and fi brinolysis that occurs is considered to be an important contributor to bleeding and coagulopathy. In surgical patients, many studies reported the use of antifi brinolytic agents to decrease bleeding and need for allogeneic transfusions. The agents most commonly used are the lysine analogues, ε-aminocaproic acid and tranexamic acid, and aprotinin. Lysine analogues interfere with the binding of plasminogen to fi brin, necessary for activating plasmin, whereas aprotinin is a direct plasmin inhibitor. Thus inhibition of fi brinolysis with antifi brinolytics reduces bleeding after tissue injury, as has been extensively studied in surgical patients. In The Lancet today, the CRASH-2 investigators report the use of tranexamic acid in trauma patients with or at risk for substantial bleeding. CRASH-2 evaluated an impressive 20 211 trauma patients randomised and treated within 8 h of injury with either 2 g tranexamic acid (1 g load, then 1 g over 8 h) or placebo. In-hospital mortality within 4 weeks of injury was the primary outcome, while vascular occlusive events, transfusions, or surgical interventions were secondary outcomes. Allcause mortality was 14·5% in the tranexamic acid group (1463/10 060) compared with 16·0% with placebo (1613/10 067; relative risk 0·91, 95% CI 0·85–0·97; p=0·0035). Bleeding-related mortality was also reduced (4·9% vs 5·7%, respectively), without an increase in fatal or non-fatal vascular occlusive events. Despite the reduction in mortality, there were no statistically signifi cant differences in transfusion requirements in patients receiving tranexamic acid or placebo. A crucial aspect of the original idea for the study was to reduce bleeding, an important cause of mortality after trauma, by use of an antifi brinolytic agent. Because tissue injury in trauma and surgery are similar, the investigators hypothesised that tranexamic acid could reduce mortality. Although there were no statistical diff erences in transfusion between the groups, how inhibition of fi brinolysis might have reduced mortality is important. The study did not show an antifi brinolytic eff ect on the basis of laboratory values; however, the tranexamic acid dose of 2 g administered over 8 h is suffi cient to inhibit fi brinolytic activity. However, there