Can tranexamic acid (TXA) stop a gastrointestinal (GI) bleed?

Tranexamic Acid Does Not Stop GI Bleeding and Should Not Be Used

High-dose intravenous tranexamic acid does not reduce mortality or rebleeding in gastrointestinal bleeding and significantly increases thromboembolic complications—it should not be used outside of clinical trials. 1, 2

Evidence Against TXA in GI Bleeding

The HALT-IT trial (2020), the largest and highest-quality study on this topic with 12,009 patients, definitively demonstrated that tranexamic acid provides no benefit in GI bleeding 2:

• No mortality reduction: Death due to bleeding occurred in 4% of both TXA and placebo groups (RR 0.99,95% CI 0.82-1.18) 2
• No reduction in rebleeding: RR 0.92 (95% CI 0.82-1.04) 3
• No reduction in need for surgery: RR 0.91 (95% CI 0.76-1.09) 1

Significant Harm from TXA

TXA increases thromboembolic events in GI bleeding patients 2, 3:

• Deep venous thrombosis: 2-fold increase (RR 2.01,95% CI 1.08-3.72) 3
• Pulmonary embolism: 78% increase (RR 1.78,95% CI 1.06-3.0) 3
• Seizures: 73% increase (RR 1.73,95% CI 1.03-2.93) 3

Guideline Recommendations

The American College of Gastroenterology explicitly recommends against high-dose IV TXA for gastrointestinal bleeding due to lack of benefit and increased thrombotic risk 1.

The British Society of Gastroenterology states that TXA use in acute lower GI bleeding should be confined to clinical trials only 4.

The European Association for the Study of the Liver provides a strong recommendation against TXA in cirrhotic patients with variceal bleeding, as it showed no benefit in controlling esophageal variceal hemorrhage and increased venous thromboembolism risk 5, 1.

Why TXA Fails in GI Bleeding

The mechanism of GI bleeding differs fundamentally from trauma or surgical bleeding 6:

• GI bleeding involves vascular injury from ulceration, portal hypertension, or mucosal damage—not the systemic hyperfibrinolysis seen in trauma 6
• Patients with cirrhosis often have hypofibrinolytic states, making antifibrinolytics mechanistically inappropriate 5
• Historical trials showing benefit were conducted before modern endoscopic therapy and high-dose proton pump inhibitors, making them non-generalizable to current practice 4, 6
• When meta-analyses are limited to low-risk-of-bias trials, the apparent benefit of TXA disappears entirely 4

Special Populations

Variceal Bleeding in Cirrhosis

TXA is contraindicated 5, 1:

• The HALT-IT trial showed nearly 50% of patients had suspected variceal bleeding, with no benefit from TXA and concentrated thrombotic risk in patients with liver disease 5
• Standard therapy with vasoactive drugs (octreotide/terlipressin), antibiotics, and endoscopic band ligation should be used instead 1

Lower GI Bleeding

TXA should only be used within clinical trials 1, 4, as evidence is insufficient and the HALT-IT trial showed no benefit across all GI bleeding types 2.

Clinical Approach

For all GI bleeding, prioritize standard management 1:

• Resuscitation with restrictive transfusion strategy (target hemoglobin 7-9 g/dL for upper GI bleeding) 1
• Endoscopic therapy for source control 1
• High-dose proton pump inhibitors for non-variceal upper GI bleeding 4
• Vasoactive drugs and antibiotics for variceal bleeding 1

Do not use high-dose IV TXA due to lack of benefit and increased harm 1, 2, 3.

Common Pitfall

Do not extrapolate TXA's proven efficacy in trauma or surgical bleeding to GI bleeding—the pathophysiology is fundamentally different, and disease-specific evidence shows no benefit and significant harm 1, 4, 6.