CRASH 3 – A Fairy Tale of Victory of Tranexamic acid & Dr Utako Okamoto

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

Intro – 

As we know, TRANEXAMIC ACID is a synthetic analogue of the amino acid Lysine. It reduces bleeding by preventing the breakdown of fibrin clots i.e fibrinolysis.

The CRASH-2 trial showed that in patients with trauma with major extracranial bleeding, early administration (within 3 h of injury) of tranexamic acid reduces bleeding deaths by a third. Subsequent analyses showed that even a short delay in treatment reduces the benefit of tranexamic acid administration. However, patients with intra-cranial hemorrhage were excluded.

Each year, worldwide, there are more than 60 million new cases of traumatic brain injury (TBI). Despite compelling physiologic logic and preclinical data, attempts to treat patients with TBI with such diverse approaches as corticosteroids, progesterone, and brain cooling have been ineffective or worse. 

The hard truths of TBI – control bleeding, avoid hypotension, and hypoxia i.e prevent secondary brain injury. Coagulopathy is a widely acknowledged contributor to TBI mortality and is associated with TBI severity. 

Hypothesis – 

Increased fibrinolysis, as indicated by increased concentrations of fibrinogen degradation products, is often seen in patients with TBI and predicts intracranial haemorrhage expansion. Therefore, early administration of tranexamic acid in patients with TBI might prevent or reduce intracranial haemorrhage expansion and thus avert brain herniation and death.

What’s Known – 

Two small randomised trials of tranexamic acid in traumatic brain injury with a total of 510 patients. A Meta-analysis of the two trials showed a statistically significant reduction in death with tranexamic acid. But these trials provided no evidence about the effect of tranexamic acid on disability or adverse events.

Another recent RCT with 967 patients showed a mortality benefit.

Methods: 

Design: International multi-center, double-blinded, randomized controlled trial 

Setting: 175 hospitals in 29 countries, between July 20, 2012, and Jan 31, 2019.

Study population:

Inclusion criteria

Adults with TBI who were within 3 h of injury, had a Glasgow Coma Scale (GCS) score of 12 or lower or any intracranial bleeding on CT scan, and no major extracranial bleeding was eligible. The window period was initially 8 h, changed to 3 h in 2016, after about 4 years of the study (The CRASH II trial was published which suggested a lack of benefit after 3 h). 

12 737 patients to receive tranexamic acid (6406 [50·3%]) or matching placebo (6331 [49·7%]), of whom 12 561 (98·6%) received the first dose.

9202 (72·2%) patients enrolled within 3 h of injury

Outcome data available on 12,639 patients (withdrawal of consent, outcome data not available, loss to follow-up)

Patients well matched at baseline including age, gender, duration of time after injury, systolic blood pressure, pupillary reaction

Intervention: Loading dose of 1 g of tranexamic acid infused over 10 min, started immediately after randomization, followed by an intravenous infusion of 1 g over 8 h.

Control: Matching placebo (normal saline)

Statistical methodology: 

10 000 patients to detect a 15% relative reduction (20–17%) in mortality with 90% power (two¬sided α=1%). Later, the primary outcome was changed to in-hospital death within 28 days due to head injury; limited to recruitment and assignment within 3 h of injury. All analysis on an intention to treat basis. 

Results:

2560 deaths due to head injury. The median time of death was 59h after injury (IQR 20–151)

Primary outcome: death due to head injury among patients randomly assigned within 3 h of injury  

All patients: 855/4613 (18·5%) vs. 892/4514 (19·8%); RR: 0·94 CI: (0·86–1·02)

Excluding patients with bilateral unreactive pupils or GCS less than 3 (pre-specified sensitivity analysis):  485/3880 (12·5%) vs. 525/3757 (14·0%); RR: 0·89 CI: (0·80–1·00)

Head injury-related death, stratified by GCS and pupillary reaction

GCS: Mild to moderate head injury (GCS: 9–15) 5.8 vs. 7.5%; RR: 0·78 (0·64–0·95) significant

In GCS 3–8: no difference 

Both pupils reactive:  440/3820 (11·5%) Vs. 493/3728 (13·2%); RR: 0·87 (0·77–0·98)

One or both pupils unreactive: no difference

Thus, tranexamic acid seemed to improve survival in less severely injured patients. 

Time of administration after head injury made no difference. Within 1 h of injury, vs. 1 h and 3 h vs. more than 3 h after injury.

In a multivariable model adjusting for GCS, systolic blood pressure, and age.

Early treatment was more effective than later treatment in patients with mild and moderate head injury; no effect in severe head injury. 

Analysis of the impact of high income vs. low-income countries (not prespecified). The reduction in the risk of head injury­ related death with tranexamic acid was larger in high­ income countries compared to low­ income and middle ­income countries this was not significant. 

RR for death due to head injury within 24 h with tranexamic acid (this is based on the assumption that early deaths are due to worsening intracranial hemorrhage, and tranexamic acid may reduce deaths due to this): 

Overall: 0·81 (95% CI 0·69–0·95)

After excluding GCS less than 3 and bilateral unreactive pupils: RR was 0·72 (0·56–0·92)

Effect of tranexamic acid on disability among survivors: 

Using the Disability Rating Scale score

The mean disability scores among survivors were similar between groups for patients treated within 3 h of injury or after 3 h 

Adverse events: including vascular occlusive events, fatal or non-fatal stroke, seizures, similar between groups.

Tranexamic acid is more beneficial in patients with less severe injury

Safe to use; no evidence of occlusive events

Strength –

1. CRASH-3 is the first trial of a pharmacological intervention applied in the acute setting to show improved outcomes in patients with TBI.
2. Randomised control trial with Placebo control – robust trial design. Extremely well-conducted ethical study. 
3. very minimum protocol violations and no conflict of interest from the funding source. 
4. Truly international trial -29 countries-  great external validity as patients were recruited from both developed and developing countries.
5. prognostic factors were well balanced and because almost all randomly assigned patients were followed up there was little potential for bias 
6. Intention to treat analysis
7. There is some contention in testing primary outcome as All-Cause mortality vs head injury-specific mortality. This trial has wide CI but when you put it in the context of other trials on tranexamic acid, there is clear mortality benefit.
8. One of the inclusion criteria were GCS 3 with unilateral dilated pupil but not bilaterally dilated pupils- and that’s fair, as we don’t expect tranexamic acid to be very effective when herniation process has progressed that far.
9. The effect of tranexamic acid is time-dependent. Early initiation in mild-moderate TBI seemed to confer greatest benefit. This proves the hypothesis as we know that the expansion of intracranial bleed occurs in the first few hours. 
10. Also, it has proved another significant point as the effect is maximum in the first 24 hours and there is no point in administering tranexamic acid for a longer period of time. As the deaths in first 24 hours occur due to bleeding, while deaths in TBI patients post 24 hours occur due to non-bleeding related pathophysiological mechanism.
11. The reassuring fact that there was no increase in adverse events. The difference in Crash- 2 and Crash- 3 is though the difference in adverse events post 3 hour administration of Tranexamic acid in TBI patients. Thus, it’s safe to say that tranexamic acid is safe in patients with TBI and treatment within 3 hours reduces head injury-related deaths.
12. Practice changing trial.

Limitations – 

1. TXA is clot stabilising intervention which may cause an increase in the incidence of VTE, which was not captured in the trial by protocolised monitoring using USG.
2. The dosing regimen used was similar to that of CRASH 2 and WOMEN trial. There may be potential for the higher dose which may be more effective.
3. Also, whether its effect is potentiated with co-intervention such as blockade of bradykinin receptors is unknown.
4. Although it has been shown to improve mortality, it has not shown to improve the morbidity. The long term follow-up in the coming months will tell.

Summary –

Overall very well-conducted RCT, which will change our practices.