Effects of a high-dose 24-h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial
HALT-IT Trial Collaborators. Lancet 2020;395(10241):1927-1936.
The HALT-IT trial is an international, randomized, double-blind, placebo-controlled trial from 164 hospitals in 15 countries, enrolling adults with significant GI bleeding in whom the treating provider was unsure whether tranexamic acid (TXA) would provide a benefit.
Significant bleeding was defined clinically and indicated a risk of bleeding to death; patients with hypotension, tachycardia, and shock were included.
Patients were randomized to 1 g of TXA infused over 10 minutes followed by an infusion of 125 mg/hour for 24 hours or a normal saline placebo “as soon as possible.”
Interestingly, the primary outcome was changed from all-cause mortality to disease-specific mortality (which is considered by some to be a weaker outcome) 5 years after the trial started; the target enrollment substantially increased from approximately 8,000 to approximately 12,000.
A total of 12,009 patients were enrolled; 64% were male, and the mean age was 58 years; 72% had comorbidities, and 9% were taking anticoagulants; 89% were suspected to have upper-GI bleeding; and 43% had signs of shock.
In the primary outcome of death due to bleeding within 5 days, there was no significant difference between the groups: 222 (3.7%) for TXA vs 226 (3.8%) for placebo; RR 0.99 (95% CI 0.82-1.18).
For the original primary outcome of all-cause mortality at 28 days, there was also no significant difference: 9.5% for TXA vs 9.2% for placebo; RR 1.03 (0.92-1.16).
Similar results were obtained after adjustment for baseline covariates and in a per-protocol analysis.
The authors examined many secondary outcomes and found no differences between groups in rebleeding within 24 hours, 5 days, or 28 days of randomization; the need for surgery or radiological intervention; and blood-product transfusion.
The authors did find a difference in 1 secondary outcome: venous thromboembolic events occurred in 0.8% of the TXA group vs 0.4% of the placebo group, with an RR of 1.85 and number needed to harm of 250.
This result may be due to the time when TXA was given. The authors reanalyzed CRASH-trial data and found that the real benefit was in patients who received TXA in <3 hours, but the mean time until intervention was 21.4 hours in the TXA group and 22.5 hours in the placebo group. Only 16% of patients were randomized within 3 hours, and 58% were randomized more than 8 hours after symptom onset. This aspect is not a weakness of the trial design but instead reflects the real-world conditions in which GI bleeds are seen.
EDITOR’S COMMENTARY: In this large multinational randomized trial, the authors found no difference in disease-specific or all-cause mortality by giving TXA to sick patients with GI bleeding. Most secondary outcomes were also unchanged, except for the rate of thromboembolic events, which was two times higher in the TXA arm than the placebo arm (0.8% vs 0.4%). It is possible that the lack of benefit might have been due to the brisk nature of most GI bleeds or to very few patients being enrolled and randomized within 3 hours of bleeding onset. There may be a subset of patients that see benefit, but according to this trial, there is no value for all comers.
Copyright 2020 by Emergency Medical Abstracts – All Rights Reserved 10/20 - #01