Trauma
Approximately one-third of all severely injured patients are coagulopathic on arrival in hospital, and mortality in trauma patients increases >4-fold if acute traumatic coagulopathy is present. Pathogenesis depends on various factors, such as haemodilution, shock, tissue injury, inflammation, hypothermia, and acidosis. Endothelial injury and ischaemia initiate both, the coagulation system, via collagen and tissue factor, and the fibrinolytic system by releasing tPA. In the presence of major trauma and shock, an imbalance of the two systems can occur and may lead to a generalized hyperfibrinolytic state. The availability of point-of-care tests, like thromboelastography, has raised awareness for this condition in recent years, and trauma patients have been treated successfully after identifying hyperfibrinolysis with the help of viscoelastic tests.
The Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage 2 (CRASH-2) trial was a very large, multinational and multicentre trial, during which 20 211 adult patients with significant bleeding after trauma were randomly assigned to receive tranexamic acid (1 g bolus+1 g infusion >8 h) or placebo within 8 h of trauma. All-cause mortality was significantly lower in the treatment group (RR 0.91; 95% CI 0.85–0.97, P=0.0035). There was also significantly less bleeding with no increase in thromboembolic events in the tranexamic acid group. A later analysis of the mortality because of bleeding, as opposed to overall mortality, revealed that there was a significant reduction of death because of bleeding only if tranexamic acid was started within 3 h after trauma. In fact, later administration was associated with significantly increased risk of death because of bleeding compared with placebo (RR 1.44, 95% CI 1.12–1.84; P=0.004).
Inclusion criteria for this trial were generous, and based only on basic clinical assessment and uncertainty about the benefit of antifibrinolytic treatment in the individual patient. Therefore, a large proportion of patients may have been included who would not have fulfilled laboratory criteria for acute traumatic coagulopathy or even hyperfibrinolysis. However, there was no increase in thromboembolic events in the tranexamic acid group. Considering the complexity of factors affecting outcome in trauma and the great variety in type of trauma and management of the patients enrolled, it is remarkable that CRASH-2 was able to demonstrate a survival benefit with respect to a single, relatively simple intervention. The most recent Cochrane review on antifibrinolytic drugs in trauma found a risk-reduction for death after acute traumatic injury of 10% (RR=0.90, 95% CI 0.85–0.97; P=0.0035). However, the conclusions are largely based on the CRASH-2 data. Only four trials were included in total, with just one further trial studying tranexamic acid. The two other studies, investigating aprotinin, provided insufficient data.
As a by-product of the main trial, the CRASH-2 investigators conducted a randomized controlled study in a subset of trauma patients, investigating tranexamic acid in traumatic brain injury. Two hundred and seventy patients with traumatic brain injury in addition to fulfilling the inclusion criteria for the main trial were randomized into a tranexamic acid and in a placebo group. They underwent computed tomography of the head before randomization and 24–48 h later, and the increase of total intracranial haemorrhage was the primary outcome. The growth of haemorrhage was reduced by −3.8 ml (95% CI −11.5–3.9, P=0.33) in the tranexamic acid group and the adjusted OR for death was 0.47 (95% CI 0.21–1.04, P=0.06). The number of new ischaemic lesions was lower in the tranexamic acid group (adjusted OR 0.51, 95% CI 0.18–1.44, P=0.2). As none of the results reached statistical significance the authors rightly conclude that neither benefit nor harmful effects can be excluded based on their data. The study design however makes it difficult to judge on tranexamic acid in traumatic brain injury because all patients enrolled also had significant (extracranial) haemorrhage leading to hypotension and possible coagulopathy. This is likely to impact heavily on overall outcome and on intracranial pathology, both haemorrhagic and ischaemic. Currently, the CRASH-3 trial (ISRCTN15088122) is recruiting patients with traumatic brain injury and no significant extracranial haemorrhage to investigate the effect of tranexamic acid in isolated brain injury.
In conclusion, the existing evidence, based on one large multicentre randomized controlled trial, strongly supports the use of tranexamic acid in bleeding trauma patients as a cost-effective empirical treatment with a good risk–benefit ratio, even without evidence of on-going hyperfibrinolysis. A multidisciplinary European guideline for the management of bleeding after major trauma also recommends the use of antifibrinolytic agents, ideally combined with monitoring of fibrinolytic activity by thromboelastometry.