The Lancet NOVEMBER 09, 2019

Effects of tranexamic acid on death, disability, vascular occlusive events and other morbidities in patients with acute traumatic brain injury (CRASH-3): a randomised, placebo-controlled trial

CRASH-3 Trial Collaborators

Bottom Line

The CRASH-3 trial found that while tranexamic acid was safe, it did not significantly reduce the primary outcome of head injury-related death at 28 days among all patients, although pre-specified subgroup analysis suggested a potential benefit in patients with mild-to-moderate traumatic brain injury treated within three hours.

Key Findings

1. In the primary intention-to-treat analysis for patients treated within 3 hours, the risk of head injury-related death was 18.5% in the tranexamic acid group compared to 19.8% in the placebo group (Relative Risk 0.94; 95% CI 0.86–1.02).

2. Among patients with mild-to-moderate traumatic brain injury (GCS 9–15), tranexamic acid was associated with a reduction in head injury-related death (RR 0.78; 95% CI 0.64–0.95).

3. No significant reduction in mortality was observed in patients with severe traumatic brain injury (GCS 3–8) (RR 0.99; 95% CI 0.91–1.07).

4. The risk of vascular occlusive events (including myocardial infarction, stroke, pulmonary embolism, and DVT) was similar between the tranexamic acid group and the placebo group, demonstrating the safety profile of the intervention.

5. There was no evidence that the effectiveness of tranexamic acid was reduced by later treatment within the 3-hour window for mild-to-moderate cases, but the study emphasized that treatment should be administered as soon as possible.

Study Design

Design

RCT

Double-Blind

Sample

12,737

Patients

Duration

28 days

Median

Setting

Multicenter, 29 countries

Population Adults with traumatic brain injury within 3 hours of injury, Glasgow Coma Scale score of 12 or lower or intracranial bleeding on CT scan, with no major extracranial bleeding.

Intervention Intravenous loading dose of 1g tranexamic acid over 10 minutes followed by a maintenance infusion of 1g over 8 hours.

Comparator Matching placebo intravenous infusion.

Outcome Head injury-related death in hospital within 28 days of injury in patients randomized within 3 hours of injury.

Study Limitations

• The primary outcome failed to reach statistical significance in the overall study population, rendering the broader clinical conclusion inconclusive by strict statistical standards.

• The positive result in the mild-to-moderate traumatic brain injury subgroup is based on a subgroup analysis, which carries a risk of type I error and requires cautious interpretation.

• The exclusion of patients with major extracranial bleeding limits the generalizability of these findings to multiply-injured trauma patients.

• The trial protocol underwent a significant amendment to narrow the inclusion window to 3 hours, which potentially complicates the interpretation of the overall study design.

Clinical Significance

While the overall results were neutral, the trial suggests that tranexamic acid may be a safe and low-cost intervention for patients with mild-to-moderate traumatic brain injury when administered within three hours of injury. It does not support routine use in severe head injury.

Historical Context

The CRASH-2 trial previously established that tranexamic acid significantly reduced mortality in trauma patients with extracranial hemorrhage. Building on that, CRASH-3 sought to determine if those benefits extended to patients with traumatic brain injury by limiting intracranial bleeding, addressing a massive global burden of TBI-related mortality.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What is the pharmacological mechanism of tranexamic acid (TXA) and how does this relate to the pathophysiology of early traumatic brain injury?

Key Response

TXA is a synthetic analog of the amino acid lysine that inhibits fibrinolysis by competitively blocking the lysine-binding sites on plasminogen molecules. In TBI, early intracranial hemorrhage expansion and the 'trauma-induced coagulopathy' contribute to secondary brain injury; by stabilizing fibrin clots, TXA aims to limit hematoma growth and preserve neurological tissue.

Resident

The CRASH-3 trial identified a specific 'treatment window' and a target Glasgow Coma Scale (GCS) range. How should these findings modify your immediate management of a trauma patient in the ED?

Key Response

The trial found that TXA is most effective when administered within 3 hours of injury and provides the greatest benefit to patients with mild-to-moderate TBI (GCS 9-15). Management should prioritize early administration in these groups, as the benefit decreases by 10% for every 20-minute delay, and patients with GCS 3 or bilateral fixed pupils are unlikely to benefit.

Fellow

How do you reconcile the neutral primary outcome of the CRASH-3 trial with the significant results found in the pre-specified subgroup analysis of head injury-related deaths?

Key Response

The primary intention-to-treat analysis included patients with severe TBI (GCS 3-8) who may have already suffered irreversible primary brain injury, effectively 'diluting' the observable treatment effect. The pre-specified subgroup analysis revealed that when excluding patients with GCS 3 or bilateral fixed pupils (who have a very poor prognosis regardless of treatment), TXA demonstrated a clear survival benefit, suggesting the drug is effective but its impact is sensitivity-dependent based on injury severity.

Attending

Given the safety data in CRASH-3 regarding vascular occlusive events, should TXA be implemented as a universal standing order in pre-hospital TBI protocols despite the non-significant primary endpoint?

Key Response

CRASH-3 showed no increase in risks like pulmonary embolism or deep vein thrombosis compared to placebo. Because the drug is inexpensive, safe, and its efficacy is highly time-dependent ('time is brain'), the risk-benefit ratio supports pre-hospital administration even before a definitive GCS or CT scan is obtained, as the cost of delay outweighs the risk of treating a patient who might not strictly fall into the mild-to-moderate benefit category.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

CRASH-3 utilized a pragmatic, large-scale design. How did the inclusion of patients with a 'nil' prognosis (GCS 3 and bilateral non-reactive pupils) affect the statistical power of the study, and what does this imply for 'prognostic enrichment' in future neuro-trauma trials?

Key Response

Including patients with a near-100% mortality rate regardless of intervention introduces 'noise' that shifts the power calculation and can obscure a true biological effect in salvagable patients. Future trials should consider 'prognostic enrichment' to exclude those either too well or too sick to show a difference, thereby increasing the efficiency and likelihood of detecting a treatment effect with a smaller sample size.

Journal Editor

While CRASH-3 reports no significant increase in thrombotic complications, the pragmatic nature of the trial relied on clinical diagnosis rather than systematic screening. How might this limitation affect the internal validity of the safety profile?

Key Response

In a pragmatic international trial, the lack of standardized ultrasound or CT-angiography screening for DVT/PE means that asymptomatic or subclinical vascular events were likely missed. A reviewer would flag that while the 'clinically significant' complication rate was low, the trial cannot definitively rule out a higher incidence of silent thrombotic events in TXA-treated patients compared to placebo.

Guideline Committee

Based on the CRASH-3 evidence, should current TBI guidelines (e.g., Brain Trauma Foundation) elevate TXA to a Level I recommendation, and how does this compare to its use in extracranial trauma?

Key Response

The evidence supports a strong (Level I or IIa) recommendation for TXA in mild-to-moderate TBI treated within 3 hours. This aligns with the CRASH-2 trial for extracranial hemorrhage, though with a more nuanced application based on GCS. Committees must decide if the pre-specified subgroup benefit in a 12,000+ patient trial is sufficient to update guidelines that previously lacked pharmacological interventions for limiting hematoma expansion.

Clinical Landscape

Noteworthy Related Trials

2010

CRASH-2 Trial

n = 20,211 · Lancet

Tested

Tranexamic acid

Population

Adult trauma patients with or at risk of significant bleeding

Comparator

Placebo

Endpoint

All-cause mortality at 4 weeks

Key result: Tranexamic acid significantly reduced all-cause mortality without increasing the risk of vascular occlusive events.

2017

WOMAN Trial

n = 20,060 · Lancet

Tested

Tranexamic acid

Population

Women with postpartum haemorrhage

Comparator

Placebo

Endpoint

Death from all causes or hysterectomy within 42 days

Key result: Tranexamic acid reduced death due to bleeding when administered within 3 hours of birth.

2020

HALT-IT Trial

n = 12,009 · Lancet

Tested

Tranexamic acid

Population

Patients with acute gastrointestinal bleeding

Comparator

Placebo

Endpoint

Death due to bleeding within 5 days

Key result: Tranexamic acid did not reduce death from gastrointestinal bleeding and was associated with an increased risk of seizures.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis