Tranexamic acid as a stopgap for low platelets?

In this edition of Blood, Gernsheimer et al¹ report the outcomes of a randomized trial of tranexamic acid vs placebo to prevent hemorrhage in patients with hematologic malignancies during prolonged episodes of severe thrombocytopenia. Tranexamic acid is an antifibrinolytic agent that was first developed in the 1960s by Okamoto et al² specifically to reduce postpartum hemorrhage. Randomized data have since demonstrated the utility of tranexamic acid to reduce bleeding in a number of indications including postpartum hemorrhage, heavy menses, traumatic injury, and cardiac surgery.³ An indication of keen interest to the hematology community is the use of tranexamic acid to prevent bleeding in patients with severe thrombocytopenia. Patients with hematologic malignancies often have extended episodes of thrombocytopenia, leading to an increased risk of hemorrhage, in excess of 70% in some prospective studies.^4,5 Tranexamic acid is an appealing adjunct medication for prolonged thrombocytopenia based on the potential to not only prevent hemorrhage but to reduce the number of platelet and red blood cell transfusions administered (assuming less hemorrhage). In a small randomized study of 38 patients, the use of tranexamic acid reduced the number of episodes of bleeding and platelet transfusions following consolidation therapy for acute leukemia.⁶ Retrospective studies similarly suggested benefit in patients with cancer and thrombocytopenia experiencing hemorrhage.⁷ 

The trial conducted by Gernsheimer et al evaluates the superiority of tranexamic acid vs placebo during a period of anticipated prolonged thrombocytopenia (defined as platelet count ≤ 10 000/μL for at least 5 days).¹ Tranexamic acid (1.3 g orally or 1 g IV every 8 hours) or placebo was administered when the platelet count dropped below 30 000/μL. The primary outcome was the incidence of World Health Organization grade 2 or higher hemorrhage that was not significantly lower with tranexamic acid (73 of 145, 50.3%) compared with placebo (78 of 144, 54.2%). Similarly, there was no statistical difference in the number of platelet or red cell transfusions over the first 30 days. Notably there were relatively few thrombotic events in both arms (6 with tranexamic acid and 9 with placebo), although 23% of potentially eligible patients were excluded because of a history of thrombosis. The incidence of central line occlusion was higher in the tranexamic group (16.6% vs 6.7%).

These results follow a string of somewhat disappointing trial results that have sullied the unmitigated enthusiasm of tranexamic acid across a spectrum of hemorrhagic conditions. Although tranexamic acid improved outcomes in women with postpartum hemorrhage after cesarean delivery,⁸ it did not reduce the rate of postpartum hemorrhage after vaginal delivery.⁹ Tranexamic acid also failed to demonstrate improvement in bleeding-associated mortality in the setting of gastrointestinal hemorrhage, appears to have a fairly tight 3-hour window of efficacy following trauma or delivery, and did not meet the noninferiority definition for cardiovascular complications following noncardiac surgery.^3,10 Context clearly matters, and to that end, there are currently more than 500 studies listed on clinicaltrials.gov evaluating different indications, doses, and routes of tranexamic acid administration such as nebulized administration during sinus surgery or to patients with osteogenesis imperfecta undergoing orthopedic surgery.

Whether a clear pattern of efficacy will emerge across studies for different indications is uncertain. Much of the success of tranexamic acid has been in the trauma or surgical setting, but even those data are not uniform. Some of the differences in efficacy outcomes can be ascribed to variations in dosing or timing of administration. In the case of thrombocytopenia, as the authors hypothesize, there may be a tight window of “fibrinolytic activation” whereby antifibrinolytic therapy is most effective. To that end, it would be interesting to see if baseline fibrinolytic biomarkers such as D-dimer could predict therapeutic efficacy. Alternatively, it may just be that antifibrinolytics are a poor substitute for a lack of platelets.