International trial
PeriOperative ISchemic Evaluation-3 (POISE-3) investigators at 114 hospitals in 22 countries (including countries in North and South America, Europe, and Africa; Russia; India; and Australia) randomly assigned 9,535 patients, aged 45 years or older, with or at risk for cardiovascular and bleeding complications to receive a TXA 1-g intravenous bolus or placebo at the start and end of inpatient noncardiac surgery.
Patients taking at least one long-term antihypertensive medication were also randomly assigned to a perioperative hypotension- or hypertension-avoidance strategy, which differ in the use of antihypertensives on the morning of surgery and the first 2 days after surgery, and in the target mean arterial pressure during surgery. Results from these cohorts will be presented in a separate session on April 4.
The study had planned to enroll 10,000 patients but was stopped early by the steering committee because of financial constraints resulting from slow enrollment during the pandemic. The decision was made without knowledge of the trial results but with knowledge that aggregate composite bleeding and vascular outcomes were higher than originally estimated, Dr. Devereaux noted.
Among all participants, the mean age was 70 years, 56% were male, almost a third had coronary artery disease, 15% had peripheral artery disease, and 8% had a prior stroke. About 80% were undergoing major surgery. Adherence to the study medications was 96.3% in both groups.
Secondary bleeding outcomes were lower in the TXA and placebo groups, including bleeding independently associated with mortality after surgery (8.7% vs. 11.3%), life-threatening bleeding (1.6% vs. 1.7%), major bleeding (7.6% vs. 10.4%), and critical organ bleeding (0.3% vs. 0.4%).
Importantly, the TXA group had significantly lower rates of International Society on Thrombosis and Haemostasis major bleeding (6.6% vs. 8.7%; P = .0001) and the need for transfusion of 1 or more units of packed red blood cells (9.4% vs. 12.0%; P <.0001), Dr. Devereaux noted.
In terms of secondary vascular outcomes, there were no significant differences between the TXA and placebo groups in rates of MINS (12.8% vs. 12.6%), MINS not fulfilling definition of MI (both 11.5%), MI (1.4% vs. 1.1%), and the net risk-benefit outcome (a composite of vascular death and nonfatal life-threatening, major, or critical organ bleeding, MINS, stroke, peripheral arterial thrombosis, and symptomatic proximal VTE; 20.7% vs. 21.9%).
The two groups had similar rates of all-cause (1.1% vs. 1.2%) and vascular (0.5% vs. 0.6%) mortality.
There also were no significant differences in other tertiary outcomes, such as acute kidney injury (14.1% vs. 13.7%), rehospitalization for vascular reasons (1.8% vs. 1.6%), or seizures (0.2% vs. <0.1%). The latter has been a concern, with the risk reported to increase with higher doses.
Subgroup analyses
Preplanned subgroup analyses showed a benefit for TXA over placebo for the primary efficacy outcome in orthopedic and nonorthopedic surgery and in patients with hemoglobin level below 120 g/L or 120 g/L or higher, with an estimated glomerular filtration rate less than 45 mL/min/1.73 m 2 or 45 mL/min/1.73 m 2 or higher, or with an N-terminal pro– B-type natriuretic peptide level below 200 ng/L or 200 ng/L or higher.