Thursday, March 28, 2013

What's the MATTER(s) with Tranexamic Acid?

The Gist:  The anti-fibrinolytic, tranexamic acid (TXA), may be lifesaving in the context of hemorrhagic trauma with fewer thrombotic complications than other agents; however, perfect data is lacking.  Use and availability of TXA is varying within the U.S. - it just hasn't caught on in many places.  Is this due to weakness in the data? Lack of pharmaceutical promotion? Lack of education?  Should we wait for more studies?  I only have guesses for answers..  Follow the evolving literature on this topic and check out the MATTERs study by Morrison et al (full text).  Although it's not an RCT, it offers additional, if imperfect, data upon which to inform our knowledge and discussions on TXA, especially regarding patients may benefit the most.

Many centers in the United States don't have TXA or don't incorporate it in their massive transfusion protocol.  Some of this may be the result of skepticism regarding the small absolute mortality benefit in CRASH-2, lack of additional prospective studies, appropriate indications,doses, time frame, and side effect profile.  Yet, many physicians don't know about the drug as the inexpensive drug lacks a marketing campaign from the pharmaceutical industry.  This is where free, open access, medical education (FOAM) may have a role in the knowledge translation gap regarding TXA.

In the FOAM world, chatter abounds about the CRASH-2 study and subsequent sub-group analyses (see PulmCCM, St. Emlyn's, Trauma Professional, Resus.Me).  In fact, some even suggest pre-hospital TXA to maximize the apparent benefit of early administration.  Also, a 2012 Cochrane Review (CRASH-2 authors) recommends TXA within 3 hours of injury in the bleeding trauma patient.  Aside from a reference on this EMCrit podcast, other studies on TXA seemed non-existent and drew from the massive cohort in CRASH-2.  In fact, I wrote a post on some of the barriers to TXA knowledge translation last year and initially neglected the MATTERs trial.

Why does this study matter?
  • Study population:  Actively bleeding patients versus those suspected of actively bleeding in CRASH-2.
    • CRASH-2 had more subjective inclusion criteria, part of which had the uncertainty principle at play.
  • Confirmed results of CRASH-2, showing a mortality benefit with TXA
    • 1.5% ARR in CRASH-2 vs 6.5% ARR in MATTERs
  • Demonstrated that the most severely injured cohort might benefit the most
  • Retrospective cohort, single center (military hospital in Afghanistan). 
  • Included: n = 896 patients who received at least 1 unit of PRBCs within 24 hours of admission after combat injury
    • Hospital protocol:  all trauma patients receiving emergency transfusion with evidence of hyperfibrinolysis (rotational thrombelastography) also received a 1g bolus of TXA and additional TXA at the discretion of the physician.
  • n=293 received a mean of 2.3 g TXA within 1 hour of injury (125/293 also had massive transfusion)
  • n=603 didn't get TXA (195/603 had massive transfusion)
Massive Transfusion
  • Absolute in-hospital mortality reduction in TXA group - 6.5%
  • Absolute reduction in the TXA + massive transfusion (10+ units of blood products in 24 hours) group was 13.7%
  • Sickest patients received TXA (and did better)
  • More VTE in TXA cohort (n=15 in TXA cohort vs n=3), but too few to assess for
    • Pulmonary embolism - 2.7% (TXA) v 0.3%
    • Deep venous thrombosis: 2.4% (TXA) v 0.2% 
  • No reduction in blood products in TXA cohort
    • Note: This may be due to a survivorship phenomenon.
  • This study was conducted at a single military hospital so generalizability may be limited.
  • No randomization
  • Retrospective cohort
  • Really quick time to administration of TXA
  • Thrombelastography is not regularly used in most EDs in the US
Things to keep your eye on in the future:
  • CRASH-3 Trial -international, multicenter, pragmatic, double-blind RCT to quantify the effects of the early administration (<8 h of injury) of TXA on death and disability in patients with a traumatic brain injury. (end of follow up in 2017)

1 comment:

  1. CRASH-2 was far better evidence than most of what we do in medicine. Compare and contrast tPA (a "core measure" with NO proof of help and can harm) to TXA (non-industry sponsored RCT with n=20k and lots of experience with the drug w/o evidence of harm). There is no reason to wait longer. Push the TXA.