Can TXA Stop a GI Bleed?
No, tranexamic acid should not be used to stop acute gastrointestinal bleeding based on the definitive HALT-IT trial, which showed no mortality benefit and an increased risk of venous thromboembolism.
The Definitive Evidence
The HALT-IT trial (2020) is the largest and most recent high-quality randomized controlled trial specifically addressing this question, enrolling 12,009 patients across 164 hospitals in 15 countries 1. This trial directly contradicts earlier smaller studies and provides the strongest evidence available:
- No reduction in death from bleeding: 4% mortality in both TXA and placebo groups (RR 0.99,95% CI 0.82-1.18) 1
- Increased venous thromboembolism risk: 0.8% in TXA group vs 0.4% in placebo group (RR 1.85,95% CI 1.15-2.98) 1
- No arterial thromboembolism difference: Similar rates of MI/stroke between groups 1
Why Earlier Studies Were Misleading
Prior meta-analyses suggested potential benefit, but these were based on small, older trials with significant limitations 2, 3, 4:
- A 2008 systematic review of 7 trials (1,754 patients) suggested mortality reduction (RR 0.61), but only 1 trial included modern endoscopic treatments or proton pump inhibitors 2
- A 2021 meta-analysis showed reduced bleeding rates and mortality, but was dominated by small heterogeneous studies conducted before the HALT-IT trial 3
- These older trials excluded 21% of randomized patients and had unclear adverse event reporting 2
Clinical Bottom Line
The HALT-IT investigators explicitly concluded that tranexamic acid should not be used for gastrointestinal bleeding outside of randomized trials 1. The harm from increased venous thromboembolism, combined with zero mortality benefit, makes this a clear recommendation against use.
Important Caveats
- The European trauma guidelines recommend TXA for trauma-related bleeding (including within 3 hours of injury), but this does not apply to spontaneous GI bleeding 5
- One smaller trial (410 patients) showed TXA reduced urgent endoscopy rates from 30% to 10-14%, potentially allowing conversion to elective procedures, but this surrogate outcome does not justify use given the HALT-IT mortality and thrombosis data 6
- WHO guidelines support TXA for post-partum hemorrhage, but this is a completely different clinical context 5