Studies examining the use of tranexamic acid (TXA) inescapably seem to pit our rigorous methodological demons against our practical clinical angels.
The CRASH-2 trial randomized 20,211 adult trauma patients presenting to 274 hospitals in 40 countries to receive 1 g of TXA over 10 minutes and an infusion of an additional 1 g over eight hours or matching placebo. (Lancet. 2010;376:23; http://bit.ly/2CoZJB4.) The authors found a 1.5 percent absolute difference in their primary outcome, in-hospital mortality at 28 days. But questions about the trial's external validity arose because many of the sites enrolling patients did not have modern trauma systems.
Do the benefits of TXA remain if hemorrhage control is achieved rapidly? This question remains unanswered nine years after the publication of CRASH-2. This trial was a massive undertaking, one that we are unlikely to see validated in a modern trauma system. We are tasked with deciding whether to accept its observed benefits or whether the methodological shortcomings invalidate its findings.
The WOMAN mega-trial examined using TXA in postpartum hemorrhage. (Lancet. 2017;389:2105; http://bit.ly/2skflDp.) The authors included 20,060 women diagnosed clinically with postpartum hemorrhage, randomized to TXA or placebo. The authors reported no difference in the primary outcomes, death within 42 days or hysterectomy (TXA 5.3% v. placebo 5.5%). Nor did they find a difference in all-cause mortality. They did note, however, a statistically significant difference in death from postpartum hemorrhage, 1.5 percent in the TXA group v. 1.9 percent in the placebo group (p=0.045).
TXA may be efficacious in postpartum hemorrhage, but the signal of benefit was diluted in this cohort because of the low severity of illness. On the other hand, a trial of 20,000 patients with a negative primary outcome and no change in all-cause mortality is about as negative a study as you can get.
Similar to CRASH-2, the WOMAN trial is a massive study. Despite its size and quality, we are left without a convincing answer. We would like to know the efficacy of TXA in the subset of patients with serious bleeding. Unfortunately, the authors do not present a subgroup analysis examining those patients. Many physicians will still give TXA despite these negative results when faced with a patient dying from postpartum hemorrhage.
This brings us to the publication of CRASH-3. (Lancet. 2019;394:1713; http://bit.ly/36xP2sn.) The authors enrolled adult patients with TBI within three hours of injury with a GCS score of 12 or under or any intracranial bleeding on CT scan without signs of major extracranial hemorrhage. The patients were randomized to receive 1 g of TXA over 10 minutes followed by an infusion of an additional 1 g over the next eight hours or matching placebo.
The authors enrolled 12,737 patients, 6406 (50.3%) to the TXA arm and 6331 (49.7%) to the placebo arm, from July 2012 to January 2019 in 175 hospitals in 29 countries. A total of 9202 (72.2%) patients were treated within three hours of injury.
Similar to the WOMAN trial, the authors reported no difference in their primary outcome, the rate of head injury-related death in the hospital within 28 days of injury—18.5 percent in the TXA group v. 19.8 percent in the placebo group (RR 0.94 [95% CI 0.86-1.02]). Once again, the authors present us with a caveat. They noted a statistically significant 1.5 percent absolute reduction in head injury-related death (12.5% in the TXA group v. 14.0% in the placebo group) in the TXA subset of patients, excluding those with a devastating neurological injury (patients with a GCS score of 3 or bilateral unreactive pupils at baseline). This reduction did not translate into a difference in all-cause mortality at 28 days or functional status at follow-up, both of which were similar in the TXA and placebo groups.
Despite a massive, well-done randomized controlled trial, we are left with a sense of dissatisfaction. There may be a signal of benefit for TXA in the subset of patients with survivable head injury, but at best it is small and does not translate into true patient-important outcomes of all-cause mortality and disability. Like CRASH-2 and the WOMAN trial, it is unlikely that efforts to replicate these findings will be conducted in the near future.
We are tasked with figuring out how to apply this evidence to our clinical practice. Should we administer TXA the next time a patient presents to the ED with a TBI? That, of course, is a choice each of us has to make. Given the nominal benefit observed in CRASH-3, we should not make the mistake of prioritizing its administration above other more important, more efficacious interventions.
Dr. Spiegelis an assistant professor of emergency medicine and critical care at Washington Hospital Center in Washington, D.C. Visit his blog athttp://emnerd.com, follow him on Twitter@emnerd_, and read his past articles athttp://bit.ly/EMN-MythsinEM.