Journal Club Synopsis: Tranexamic acid: From the Battlefield to the Trauma Center
Thanks to Branka for hosting at her glam club room, and to all the presenters: Nick, Paarul, Robbie, Beau, Linsey and Dave.
By way of background, tranexamic acid (TXA) is an inexpensive (about 100 $/dose) synthetic derivative of the amino acid lysine, and inhibits fibrinolysis by blocking lysine binding sites on plasminogen. In the setting of surgery, it’s been used to help decrease the need for blood transfusion. More recently, it’s been investigated in trauma patients, as the hemostatic responses to surgery and trauma are thought to be similar.
CRASH-2 Trial Collaborators. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010;376:23-32.
CRASH-2 was a RCT conducted in 274 hospitals in 40 countries, and included 20,211 adult trauma patients with or at risk for significant bleeding. Patients were randomized to either TXA or placebo. The primary outcome was all cause mortality at 4 weeks, and was significantly reduced in the TXA group (14.5% TXA versus 16.0% placebo, RR 0.91, 95% CI 0.85-0.97; p = 0.0035). This absolute risk reduction in mortality of 1.5% translates to a NNT of 67. There was also a significant reduction in the risk of death due to bleeding (4.9% TXA, 5.7% placebo). As the mechanism of action of TXA is inhibition of fibrinolysis, question possibility of increased clotting risk, but there was no significant difference in the rate of thrombosis in the 2 groups.
So what are the criticisms? First, many of the centers involved are in developing countries, and likely do not have the same level of trauma care as in the US. The CRASH-2 investigators attempted to address this by performing an outcomes analysis by continent, and stated that the positive effects of TXA persisted.
Stats pearl: CRASH-2 is a pragmatic trial, rather than an explanatory trial. We’re much more used to explanatory trials in the medical literature. They are performed under tightly controlled conditions, with many inclusion/exclusion criteria, usually in academic centers using smaller sample sizes and strictly defined control treatments. The downside is that they tend to overestimate benefits, and the results are never as good when applied to the real world (limited external validity). Pragmatic trials are the real world. Large sample sizes, broader entrance criteria, accept more variation, often performed across many centers, and with easy to generalize results. One way to think about it is that explanatory trials are specialized studies for information/efficacy, and pragmatic trials are real-world studies to help make decisions/determine effectiveness. Expect to see more pragmatic trials in the future, especially in emergency medicine.
In CRASH-2, only half the patients received blood, and only half went to the OR. The decision to enroll patients was the responsibility of the treating doctor. Again, this reflects the pragmatic nature of the trial. This relatively low percentage of patients receiving blood or an operation actually biases the results against TXA, as many patients received TXA who weren’t actually having significant bleeding. It is encouraging that they found a treatment benefit in this “real world” scenario.
Another question raised about the trial is the blood transfusion requirements between the 2 groups. If TXA helps stop bleeding, why did the TXA group have the same transfusion rate as the placebo group? This was examined in a secondary analysis of CRASH-2 (Lancet 2011), that demonstrated the importance of early administration of TXA, with a 2.4% decrease in mortality from bleeding if TXA was given in the first hour (early clot stabilization proposed as the mechanism of benefit) and a 1.3% increase in the mortality from bleeding if TXA was given after 3 hours. In addition, the transfusion rate equivalence is likely an example of survivor bias (second stats pearl!). If your primary outcome (survival) shows significant benefit, then the extra survivors in the treatment group survive to receive more blood products, decreasing differences seen in secondary outcomes such as the number of units of blood transfused.
Comments from the reviewers/audience:
Barounis: Highest benefit when given early-stressed the goal of administration within first hour after injury. Still uncertain about optimal patient population.
Kerwin favors use in hemorrhagic shock, while Den Ouden sees this as the “aspirin of trauma” J
Harwood: In EM, we ask 2 questions: how do you diagnose, and how do you treat. For therapies, we balance number needed to treat, number needed to harm, and cost. With this treatment, NNT = 67, NNH = 0, and costs per life save is < $ 7,000 (100$ per dose x 67). From a population health perspective, anything less than $20,000 per life saved is a good option.
Both Nick and Paarul were concerned about the lack of carefully defined enrollment criteria, and would like a better sense of the population that will benefit, although as this is a pragmatic trial, it was designed to rely on physician judgment when choosing eligible patients, and found a mortality benefit when using these “real world” criteria.
Sam Lam: “all cause mortality” as primary outcome is especially appropriate in this study, as many times cause of death will be unknown in the developing world.
Morrison JJ, Dubose FF, Rasmussen TE, Midwinter MJ. Military Application of Tranexamic Acid in Trauma Emergency Resuscitation (MATTERs) Study. Arch Surg. 2012;147(2):113-119.
This retrospective observational study from a surgical hospital in Afghanistan compared mortality, coagulopathy, and thromboembolic complications in 896 post-operative patients receiving at least 1 unit of pRBCS, 293 of whom also received TXA. Primary outcome was all cause mortality. Secondary endpoints included transfusion requirements, coagulation parameters, and thrombotic complications. The TXA group had lower mortality than the no-TXA group (17.4% versus 23.9%, p = 0.03, NNT = 15). The mortality benefit was even greater in patients receiving massive transfusion (14.4% versus 28.1%). Based on PT/aPTT measurements, there was a decrease in the percentage of hypocoagulable patients in the TXA group. Interestingly, transfusion requirements were higher for the TXA group, and the rate of PE/DVT was also higher in the TXA group. The TXA group had a higher Injury Severity Score and a higher percentage of patients with severe extremity injury. Also, as it was a retrospective, observational study, it is difficult to draw causal conclusions regarding the blood transfusion/thrombotic complication differences. Survivor bias may also impact the secondary outcomes.
Comments from the reviewers/audience:
Beau: external validity-patients were men in their 20’s. Matches our trauma population? Bottom line-would use TXA.
Robbie: liked that they checked coags, could see improvement in coag parameters in TXA group. ARR 6.5%, NNT = 15. For massive transfusion, even better ARR 13.7%, NNT = 7. Increased risk of 2.4% in PE group, NNH = 42 (again, retrospective study, limited information on thrombotic complications/significance). Authors hypothesize that anti-inflammatory effects of TXA also beneficial. Bottom line-would use TXA.
E. Kulstad: horse is out of the barn for any future RCT on TXA given these 2 studies.
Harwood: Mechanism of injury was explosions for majority of patients-severe force. Would have liked more specifics on how TXA administered; non-standardized dosing.
Sam Lam: Retrospective (always issues), Patients who die early don’t get TXAàaffects results (another form of survivor bias), and noted that protocol for giving TXA changed halfway through study-unknown if other components of trauma care also changed.
CRASH-2 Collaborators, Intracranial Bleeding Study. Effect of tranexamic acid in traumatic brain injury: a nested randomised, placebo controlled trial (CRASH-2 Intracranial Bleeding Study). BMJ 2011;343:d3795 doi: 10.1136/bmj.d3795.
Historically, anti-fibrinolytics have not been recommended in aneursymal SAH due to concerns about increased cerebral ischemia. Recent data are more positive, but patients with isolated head trauma were excluded from CRASH-2 due to this concern. There were still 270 patients from CRASH-2 who had multi-trauma including traumatic brain injury, and this study evaluated the intracranial hemorrhage growth in these patients as a primary outcome, comparing the 133 patients receiving TXA to the 137 patients randomized to placebo. While the study did not find significant differences in the primary or secondary outcomes, there were consistent trends favoring TXA: decreased ICH growth, decreased mortality, and decreased new focal cerebral ischemic lesions. The historical concern of increased ischemia when using TXA in patients with intracranial hemorrhage was not seen in this study.
Consensus bottom line for this Journal Club from the group:
TXA is safe and inexpensive, and likely to be of benefit to trauma patients anticipated to require blood transfusion, especially if given within the first 3 hours after injury. TXA is particularly useful in developing countries where blood products may not be safe and are less available.
What’s the future? Locally, Erik Kulstad and Nick Kettaneh will be representing the ED in the ACMC Pharmacy and Therapeutics meeting at the end of the month to discuss adding TXA to our formulary. Worldwide, look for studies evaluating the use of TXA to treat head injury (CRASH-3), post partum hemorrhage (WOMAN), and GI bleeding (HALT IT).