Tranexamic acid saves lives, reduces transfusions. So why does no one use it? (Review)
Tranexamic Acid: Underused for Uncontrolled Bleeding?
Tranexamic acid is a simple little molecule, just a synthetic derivative of the amino acid lysine. But it's also a potent pro-hemostatic drug that binds plasminogen and plasmin and stops the degradation of fibrin (the stuff in blood clots).
In the U.S., tranexamic acid is sold as Lysteda (oral) and Cyklokapron (I.V.) The drug has long been known to reduce the need for blood transfusion in surgery patients. However, almost no one uses tranexamic acid in the U.S., where it had only $17 million in sales for 2011, mostly for the reduction of bleeding during dental procedures.
It's not for a lack of evidence: The U.K.-based CRASH-2 trial showed tranexamic acid given early after trauma saves lives. In this meta-analysis in BMJ, Katharine Ker, Phil Edwards, Pablo Perel, Haleema Shakur, and Ian Roberts argue it's time to stop doing placebo-controlled randomized trials to see if tranexamic acid reduces transfusion requirements, because it's clear it does. The point now should be to demonstrate its safety and whether tranexamic acid improves clinical outcomes as a treatment for trauma and other bleeding causes.
They analyzed 129 trials enrolling 10,488 patients between 1972 and 2011. As you could guess, the older studies were methodologically poor by today's standards, and often unblinded. Nevertheless, tranexamic acid reduced the need for blood transfusion by one third (risk ratio 0.62, 95% confidence interval 0.58 to 0.65; P<0.001). After restricting the analysis to blinded trials, tranexamic's 33% transfusion-sparing effect was virtually unchanged.
The trials were inconclusive as to whether tranexamic acid also caused blood clots, or prevented them, but the safety trends were favorable. For example:
The risk ratio for myocardial infarction with use of tranexamic acid was 0.68 (0.43 to 1.09; P=0.11);
For deep vein thrombosis 0.86 (0.53 to 1.39; P=0.54);
and pulmonary embolism was 0.61 (0.25 to 1.47; P=0.27),
but for stroke it was 1.14 (0.65 to 2.00; P=0.65).
Tranexamic acid's effect on mortality was likewise unclear, but encouraging -- a 0.67 risk ratio (0.33 to 1.34, p=0.25). Cumulative meta-analysis showed that reliable evidence that tranexamic acid reduces the need for transfusion has been available for over 10 years.
Why Don't Doctors Use Tranexamic Acid?
So tranexamic acid stops bleeding, reduces transfusion requirements, and saves lives in bleeding trauma patients. And unlike NovoSeven, it doesn't cost $1,200,000 a gram (or about $10,000 per 8 mg dose, given every 2 hours). So why is almost no one in the U.S. prescribing tranexamic acid -- for G.I. bleeding patients, intracranial hemorrhages, trauma, and surgical misadventures?
Ian Roberts, principal investigator of the CRASH-2 trial demonstrating tranexamic acid's benefit in trauma, chalks it up to the recency of CRASH-2 and unawareness among physicians of the drug's benefit. While conducting other trials testing tranexamic acid in postpartum hemorrhage (which kills 100,000 women each year globally) and G.I. bleeding, they're promoting the CRASH-2 findings in film, a creepy, yet catchy jingle and a manga cartoon, all of which are clickable from the CRASH-2 website.
The U.S. military already gives tranexamic acid to soldiers with severe bleeding (their tranexamic acid dose: 1 gram in 100 mL crystalloid, administered over 10 minutes within 3 hours of injury, followed by another gram after fluid resuscitation has begun), and included tranexamic acid in its revised protocol for combat casualties. Tranexamic acid is almost absurdly cheap at $100 per gram in the U.S., and less in other countries.
Enough data is in that tranexamic acid's should become a standard treatment for trauma and possibly other causes of severe hemorrhage, Tim Coats (CRASH-2 co-author) tells Scott Weingart in this EMCrit podcast:
Katharine Ker, Phil Edwards, Pablo Perel, Haleema Shakur, and Ian Roberts. Effect of tranexamic acid on surgical bleeding: systematic review and cumulative meta-analysis. BMJ 2012; 344:e3054.