Tranexamic acid is an antifibrinolytic agent that has been shown to reduce the incidence and severity of periprocedural bleeding events or blood loss in both cesarean section and cardiac surgery without increasing thrombosis risk.1, 2 A recent meta-analysis suggested that tranexamic acid does not increase the risk of arterial or venous thrombosis; however, at total daily doses greater than two g it could increase the risk of seizures.3, 4 Therefore at this juncture, there remains equipoise on whether tranexamic acid would be safe and effective for patients undergoing noncardiac surgeries.
The POISE-3 study attempts to answer whether the periprocedural use of intravenous tranexamic acid safely reduces the risk of bleeding complications in patients with risk factors for bleeding and cardiovascular complications.
This was a randomized trial in which 9,535 patients were assigned to either periprocedural tranexamic acid (1 g intravenous bolus at both the start and at the end of surgery) versus placebo. The types of surgeries included general surgery, orthopedic, vascular, urologic, and others. Major comorbid conditions included coronary artery disease (29.6%), peripheral vascular disease (15.0%), and atrial fibrillation (10.0%). The primary efficacy outcome was a composite bleeding outcome (life-threatening bleeding, major bleeding, bleeding into critical organ), and the primary safety outcome was a composite cardiovascular outcome (myocardial injury, nonhemorrhagic stroke, peripheral arterial thrombosis, symptomatic venous thromboembolism). Outcome assessments were done at 30 days for both composite efficacy outcome (for superiority compared with placebo) and safety outcome (for noninferiority compared with placebo).
The investigators found that the likelihood of the composite bleeding outcome was significantly lower with tranexamic acid (9.1%) compared with placebo (11.7%; hazard ratio [HR], 0.76; 95% CI, 0.67-0.87; p<0.001 for superiority), with the difference composed mostly of major bleeding events. Meanwhile, the composite cardiovascular event occurred in a similar proportion of patients receiving tranexamic acid (14.2%) compared with placebo (13.9%; HR, 1.02; 95% CI, 0.92-1.14; p=0.04 for noninferiority), though the noninferiority of tranexamic acid was not established by prespecified definitions. There was a numerically greater increase in seizures in the tranexamic acid group (10 vs. 3 events in the placebo group) but the absolute numbers were small, and CIs were large (HR, 3.35; 95% CI, 0.92-12.20). The study was ultimately stopped early at 95 percent of the original planned sample size owing to financial limitations during the COVID pandemic.
In Brief
This is a compelling study that demonstrates a potential benefit for the use of tranexamic acid in reducing bleeding complications during noncardiac surgery. It is notable that the difference in bleeding events was composed of major bleeding events (by International Society on Thrombosis and Haemostasis criteria),5 and not life-threatening or critical organ bleeds. Nonetheless, major bleeding events still carry clinical relevance particularly in the acute postoperative setting. These potential benefits must be counterbalanced by the fact that in this study, noninferiority for cardiovascular events could not be established. The upper bound of the 95% CI of the HR would suggest the up to 14 percent risk could not be ruled out. Additionally, although the number was small, there was a numeric trend toward more seizures in patients receiving tranexamic acid.
So, which patients should receive tranexamic acid, and how should it be administered? This will continue to be an individualized consideration per patient that accounts for the bleeding risk of the procedure, patient-specific bleeding risk factors, and the cardiovascular and seizure risks for each patient. If it is administered for noncardiac surgery, it should be administered in a lower-dose fashion (1 g intravenous at the start and end of each procedure), rather than high-doses used in cardiovascular surgery.2, 6 Until additional perioperative data are available, I will continue to use this strategy on a case-by-case basis. The net clinical benefit may be most apparent in individuals with postpartum hemorrhage and major trauma.7, 8
Competing Interests
Dr. Tseng indicated no relevant conflicts of interest.
