It’s Time to Break Up with TXA
BY RICHARD PESCATORE, DO
We have been fed false promises of efficacy and manufactured optimism for success with the use of tranexamic acid in all manner of acute and subacute bleeding conditions for far too long. Beginning with the publication of CRASH-2 in 2010, emergency physicians across the world quickly fell in love with TXA, entranced by the drug’s low price and benign side effects across the spectrum of emergency department presentations. As observational data sets and anecdotal reports flourished, TXA quickly became a darling of emergency medicine, championed as the panacea to all types of hemorrhage.
It’s time, however, to recognize that this wonder drug has done nothing but fall short time and again, betraying our confidence and leaving us repeatedly stranded at the evidence-based altar. In all likelihood, TXA has little utility in the ED, much less the vaunted status it has undeservedly enjoyed.
The betrayal began with the publication of the WOMAN trial. (Lancet. 2017;389:2105; https://bit.ly/3lfnsbQ.) Designed primarily to identify a composite outcome of all-cause mortality and hysterectomy within 42 days of giving birth, the trial’s primary endpoint was eventually adjusted to investigate TXA’s effect on death from postpartum hemorrhage alone, justified by the authors as a needed methodological change but nonetheless a red flag by even the loosest evidence-based standards. Even with the change, however, this trial fell short in its primary outcome, finding no statistically significant difference in outcomes except in a secondarily dissected time-based endpoint. Nonetheless, the trial was widely interpreted as positive, and TXA remains in postpartum hemorrhage guidelines and algorithms to this day.
Three years ago, I wrote that the much-anticipated HALT-IT trial would settle the debate, but we were left in the meantime with consistent and compelling evidence that TXA should be routinely incorporated into the resuscitation of the critically ill GI bleeder. Our patients deserve TXA, I said. (EMN. 2018;40:1; https://bit.ly/3tic020.)
The preliminary literature was indeed compelling. TXA showed a relative mortality improvement of 39 percent in seven randomized, controlled trials of 1385 patients with GI bleeds. (Aliment Pharmacol Ther. 2008;27:752; http://bit.ly/2xgvIAL.) This was confirmed by a 2014 Cochrane Review, where eight RCTs and more than 1700 patients with GI bleeding found a relative risk for mortality of 0.6, with no difference in venous thromboembolic events. (Cochrane Database Syst Rev. 2014;2014:CD006640; http://bit.ly/2NbRxvc.) The 2020 publication of the HALT-IT trial once again uncovered TXA’s silent duplicity. (Lancet. 2020;395:1927; https://bit.ly/3qR36qI.)
Patients in this well-structured randomized trial received a loading dose of 1 g of tranexamic acid, which was infused over 10 minutes, followed by a maintenance dose of 3 g of tranexamic acid infused over 24 hours or placebo (sodium chloride 0.9%). Patients who received TXA had no improvement in the primary outcome (death due to bleeding within five days of randomization), but did suffer a higher rate of venous thromboembolic events. TXA was not content just to betray us; it began to hurt us as well.
There have been plenty of other hints to which we have perhaps simply turned a blind eye. Consider the CRASH-3 trial, where TXA fell short of a statistically significant improvement in head injury-related death, or the TICH-2 trial, where no benefit was found in three-month functional status among patients with spontaneous intracranial hemorrhage. Nonetheless, many emergency physicians walked away from these publications with a positive outlook on using TXA in intracranial bleeding, bamboozled once again by a drug we seem to stick by no matter the accruing literature justifying its abandonment.
Following the release of a 2017 experiment showing the superior performance of intranasal TXA to nasal packing in epistaxis, emergency physicians (myself included) flocked to the drug for nosebleeds and recalcitrant HEENT bleeding. (Acad Emerg Med. 2018;25:261; https://bit.ly/2PQiZR9.) Intranasal TXA became a mainstay of epistaxis management for many, a promise finally fulfilled. The publication of the NoPAC trial, however, showed TXA’s true colors once again.
This well-done trial randomized patients to saline- or TXA-soaked dental pledgets after first-line therapy (generally, intranasal phenylephrine) had failed. Despite nearly two-thirds of patients taking anticoagulants, TXA was no more effective than placebo at controlling bleeding and reducing the need for anterior nasal packing. There is plenty of room for maneuvering—a startlingly high 43 percent of patients needed packing, for example—but the conclusion is fairly clear: TXA had betrayed us again.
We have placed our confidence in tranexamic acid again and again to bring patients back from the bleeding brink, only to be recurrently jilted by perennially paltry p-values. TXA, once thought to be a simple solution to bleeding from nearly any source, has failed to deliver one too many times.
Dr. Pescatore is the chief physician for the Delaware Division of Public Health and an emergency physician at Einstein Healthcare Network in Philadelphia. He is also the host with Ali Raja, MD, of the podcast EMN Live, which focuses on hot topics in emergency medicine: http://bit.ly/EMNLive. Follow him on Twitter @Rick_Pescatore, and read his past columns at http://bit.ly/EMN-Pescatore.