CRASH-3 (2019): Tranexamic acid in acute traumatic brain injury

In Review

Roberts I, et al. "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". The Lancet. 2019. 394(10210):1713-1723.
PubMed • Full text • PDF • ClinicalTrials.gov

Clinical Question

In adult patients with traumatic brain injury with a Glasgow coma scale greater than 12, did tranexamic acid within 3 hours of injury decrease mortality?

Bottom Line

Early treatment of traumatic brain injury within 3 hours of injury with tranexamic acid led to a non-significant reduction in mortality in all comers and a significant reduction in those with mild-to-moderate injury.

Major Points

Hemorrhage is a leading cause of death among patients with major trauma. Trauma leads to hemorrhage both from primary injury to organs and blood vessels, but also to transient hyperfibrinolysis that is incompletely understood. Tranexamic acid, a synthetic derivative of the amino acid lysine, appears to interfere with this hyperfibrinolysis, prompting investigators to study whether the agent may reduce the morbidity and mortality of traumatic hemorrhage. It has previously been used in controlling bleeding complications in those with hemophilia, those undergoing cardiopulmonary bypass, and those with menorrhagia.

The 2010 Clinical Randomization of an Antifibrinolytic in Significant Hemorrhage 2 CRASH-2 randomized patients to receive tranexamic acid or matching placebo within 8 hours of presentation. At 4 weeks of follow-up, tranexamic acid was associated with a 1.5% absolute reduction in mortality (14.5% vs. 16%). A separately published but prespecified subgroup analysis demonstrated that early administration of tranexamic acid (within one hour of injury) was associated with greater reductions in death due to bleeding, while delayed administration (>3 hours from injury) was associated with increased bleeding deaths. All-cause mortality was reduced in the <1 hour and 1-3 hour strata, but not in the >3 hour stratum. This benefit was balanced without an increase in vascular occlusive events (1.7% vs. 2.0%). The results of CRASH-2 were concordant with those of the subsequent MATTERs trial (2012).

CRASH-3, generally a negative study, did not demonstrate that tranexamic acid reduces mortality from head injury-related death at 28 days in all comers. This was balanced with no increase in vascular occlusive or other adverse events. For patients seen within 3 hours with mild to moderate head injury (GCS 9 – 15), the risk of death related to head injury was significantly reduced (Risk Ratio .78; 95% CI .64-.95). Tranexamic acid reduced the risk of head injury-related death within 24 hours of injury, but this benefit was not shown at 28 days. The mortality effect was most pronounced when patients with a GCS 3 or bilateral unreactive pupils (Risk Ratio .89, 95% .8-1). When this data was meta-analyzed with data from previous trials, mortality related to head injury favoured tranexamic acid (Risk Ratio .89, 95% CI .8–.99). The disability measurements, arguably the most important for many patients, demonstrated no difference between groups.

Guidelines

Tactical Combat Casualty Care (US Military) 2011

• Early administration of 1 gram of tranexamic acid to casualties expected to receive blood transfusions as soon as possible (class I, level B).
• Administration should be no more than 3 hours after initial trauma.

Design

• Multicenter, double-blind, parallel-group, randomized, controlled trial
• N=12 737 randomized patients
  • Intervention (n=6406)
  • Placebo (n=6331)
• Setting: 175 hospitals in 29 countries
• Enrollment: July 20, 2012 to January 31, 2019
• Follow-up: 28 days
• Analysis: Intention-to-treat
• Primary Outcome: head injury-related death in hospital within 28 days of injury

Population

Inclusion Criteria

• Adults with TBI who were within 3 h of injury (originally 8hours, this is reduced to 3 hours in 2016
• GCS ≤ 12
• No major intracranial bleeding on CT scan
• Responsible clinician was substantially uncertain as to the appropriateness of tranexamic acid therapy

Exclusion Criteria

• Major extra-cranial bleeding

Baseline Characteristics

‘’Tranexamic Acid Group displayed’’

• Demographics: 19% female, mean age 42 years,
• Physiologic parameters: SBP 90-119 32%, SBP 120-139 31%
• Pupil reaction: none 9%, one 8%, both 80%, unable to assess 3%

Interventions

• Tranexamic acid (TXA) loading dose 1g over 10 minutes, followed by 1g infusion over 8h
• Placebo

Outcomes

Comparisons are tranexamic acid vs. placebo.

Primary Outcomes

Head injury-related death in patients randomly assigned within 3 h of injury

18.5% vs. 19.8% (Risk Ratio 0.94; 95% CI 0.86-1.02)

Secondary Outcomes

Head injury-related death in patients with GCS mild-moderate (9-15)

5.8% vs. 7.5% (Risk Ratio 0.78; 95% CI 0.64-0.95)

Head injury-related death in patients with GCS Severe (3-8)

39.6% vs. 41.1% (Risk Ratio 0.99; 95% CI 0.91-1.07)

Head injury-related death in patients with both reactive pupils

11.5% vs. 13.3% (Risk Ratio 0.87; 95% CI 0.77-0.98)

Head injury-related death in patients with any non-reactive pupils

52.3% vs. 50.8% (Risk Ratio 0.94; 95% CI 0.94-1.13)

Head injury-related death in patients randomized within 1 hour of injury

Risk Ratio 0.96 (95% CI 0.79–1.17)

Head injury-related death in patients randomized between 1-3 hours of injury

Risk Ratio 0.93 (95% CI 0.85–1.02)

Head injury-related death in patients randomized greater than 3 hours from injury

Risk Ratio 0.94 (0.81–1.09)

Subgroup Analysis

Head injury-related death in patients within 3 h of injury, excluding patients with GCS = 3 or bilateral unreactive pupils

12.5% vs. 14% (Risk Ratio 0.89; 95% CI 0.8-1)

Adverse Events

No statistically significant differences observed between groups

All vascular complications

1.6% vs. 1.6% (Risk Ratio .98, 95% CI 0.74-1.28)

Stroke

0.7% vs. 0.7% (Risk Ratio 1.08, 95% CI 0.71–1.64)

Myocardial Infarction

0.3% vs. 0.3%, (Risk Ratio 0.89, 95% CI 0.47–1.68)

Renal Failure

1.6% vs. 1.3% (Risk Ratio 1.18, 95% CI .88-1.57)

Seizure

3.2% vs. 3% (Risk Ratio 1.09, 95% CI 0.90–1.33)

Criticisms

• Change in study protocol due to new evidence but lacked an interim analysis to this change
• Statistically significant improvement seen largely in secondary outcomes
• Quality of life, while not a hard outcome like mortality and more challenging to measure, may be a more important outcomes to patients, in this trial the intervention showed no difference in disability.
• 28-day mortality may have been too early of an outcome as injury-related-mortality could have occurred later due to withdrawal of life-support
• Adverse events may have been under-reported given the criteria for them to be clinically obvious
• Dose and route optimization could be considered for future study, possibly to allow earlier intervention

Funding

• Run in phase (first 500 patients): The JP Moulton Charitable Trust
• Main phase: the National Institute for Health Research Health Technology Assessment (NIHR HTA; 14/190/01), and Joint Global Health Trials, Medical Research Council, Department for International Development, Global Challenges Research Fund, and the Wellcome Trust (MRM0092111).
• New Brunswick Trauma Program grant of $10 000 CAD to support the trial in Canada