Can tranexamic acid stop a gastrointestinal bleed?

Tranexamic Acid Does NOT Stop Upper GI Bleeding

Tranexamic acid should NOT be used for upper gastrointestinal bleeding. The highest quality evidence—the 2025 French emergency medicine guidelines and 2022 EASL cirrhosis guidelines—explicitly state that tranexamic acid is ineffective for upper GI bleeding based on the HALT-IT study 1, 2.

The Evidence Against TXA in GI Bleeding

Upper GI Bleeding: Clear Inefficacy

The 2025 French guidelines provide a GRADE 1+ recommendation that tranexamic acid should be administered according to the same indications as in patients not on anticoagulants—but critically note that "in cases of upper gastrointestinal haemorrhage, the HALT-IT study concluded that tranexamic acid was ineffective whether or not an anticoagulant was present" 1.

The 2022 EASL guidelines are even more explicit: "In patients with cirrhosis and active variceal bleeding, tranexamic acid should not be used (LoE 2, strong recommendation)" 2. The HALT-IT trial of 12,009 patients with acute upper GI bleeding showed:

• No beneficial effect on death due to bleeding within 5 days
• No benefit in the subgroup with suspected variceal bleeding and liver disease
• Nearly 2-fold increase in venous thromboembolic events in the tranexamic acid group
• The risk of thrombosis was concentrated in patients with liver disease/suspected variceal bleeding 2

Why TXA Fails in GI Bleeding

The mechanism of failure is clear: variceal bleeding is primarily a mechanical/hemodynamic problem, not a coagulation problem. The limited role of hemostasis in variceal bleeding and the frequent hypofibrinolytic state in critically ill cirrhotic patients explain why an antifibrinolytic agent cannot work 2.

The Dosing Table Tells the Story

The 2025 Association of Anaesthetists guidelines provide a comprehensive dosing table for tranexamic acid across clinical settings—and gastrointestinal bleeding is explicitly listed as "Not recommended" 3.

Lower GI Bleeding: Also Ineffective

For lower GI bleeding, the evidence is similarly negative. A 2024 randomized controlled trial found that intravenous TXA had no significant effect on blood transfusion requirements in patients with acute lower GI bleeding (43/81 patients required transfusion: 22 placebo vs 21 TXA, p=0.89) 4.

The Contradictory Meta-Analyses: Why They're Wrong

Several meta-analyses 5, 6, 7, 8, 9 suggest benefit from TXA in upper GI bleeding, but these are superseded by the definitive HALT-IT trial and should be disregarded. The meta-analyses included older, smaller trials with methodological limitations. The 2022 systematic review 7 correctly identified that extended-use high-dose IV tranexamic acid does not improve mortality or bleeding outcomes and increases adverse events (DVT RR 2.01, PE RR 1.78, seizure RR 1.73).

One Exception: Post-EVL Bleeding in Cirrhosis

A single 2024 trial 10 in 600 patients with advanced cirrhosis (Child-Pugh B or C) showed that TXA reduced 5-day treatment failure (6.3% vs 13.3%, p=0.006), primarily by preventing bleeding from endoscopic variceal ligation (EVL) sites (4.9% vs 12.0%, p=0.005). However, this did not translate to mortality benefit, and this represents a very specific subpopulation and mechanism (post-procedural bleeding from EVL ulcers, not acute variceal hemorrhage itself).

Where TXA DOES Work

The FDA label 11 and guidelines 12, 3, 13 confirm TXA's validated indications:

• Trauma hemorrhage (1g IV over 10 min, then 1g over 8h, within 3 hours of injury) 12, 3
• Postpartum hemorrhage (1g IV over 10 min within 3h of bleeding onset) 3, 13
• Cardiac surgery (50-100 mg/kg IV) 3
• Major non-cardiac surgery (1g IV before incision) 3

Critical Safety Warnings

The FDA label 11 warns of:

• Thromboembolic events (arterial, venous, embolic)
• Seizures (especially with inadvertent neuraxial administration or high doses)
• Visual disturbances
• FOR INTRAVENOUS USE ONLY 11

Bottom Line Algorithm

For acute GI bleeding:

1. Do NOT give tranexamic acid for upper GI bleeding (variceal or non-variceal)
2. Do NOT give tranexamic acid for lower GI bleeding
3. Focus on standard therapy: vasoactive drugs (terlipressin/octreotide), antibiotics, endoscopic intervention, restrictive transfusion strategy (Hb <70 g/L) 2
4. The only potential exception is post-EVL ulcer bleeding in cirrhotic patients, but this requires further validation

The harm-benefit ratio is unfavorable: TXA increases thrombotic complications without improving mortality or bleeding control in GI hemorrhage.