Topical tranexamic acid and sucralfate prevent delayed bleeding after endoscopic sphincterotomy: a randomized controlled trial (with video)

Endoscopic retrograde cholangiopancreatography (ERCP) is a complex procedure that combines endoscopy and fluoroscopy for the diagnosis and treatment of pancreaticobiliary ductal disorders, such as choledocholithiasis (bile duct stones), biliary strictures, and cholangitis (bile duct inflammation).¹ Although ERCP is essential for both diagnosis and treatment, it carries a higher risk of complications than do other endoscopic procedures. Common complications include post-ERCP pancreatitis (PEP), bleeding, perforation, and biliary tract infections. The overall incidence rate of ERCP-associated complications is approximately 10%, with severe complications occurring in approximately 24% of all patients who have complications.^2,3 The rate of ERCP-associated mortality ranges from 0.2% to 0.7%.^2,4 Despite technological advancements and enhanced safety protocols, the rate of ERCP-associated complications remains high.

Among all complications, delayed bleeding after endoscopic sphincterotomy (EST) represents a major challenge for clinicians and endoscopists. This complication often manifests as tarry or bloody stools and anemia, contributing to increased mortality risks and prolonged hospital stays. The incidence rate of clinically meaningful post-EST bleeding is approximately 2% in the general population and 12% in high-risk patients.^5–7 Risk factors for delayed bleeding include comorbidities such as liver cirrhosis and end-stage renal disease (ESRD), antiplatelet or anticoagulant agent use, intraprocedural bleeding, and endoscopist’s low technical experience (<200 ERCP procedures).⁵

Although interventions such as nonsteroidal anti-inflammatory drugs, pancreatic stents, and intravenous fluids can prevent PEP, few standardized methods or medications are currently available for preventing delayed post-EST bleeding.¹ Given the severity of delayed post-EST bleeding, which can lead to life-threatening conditions necessitating endoscopic hemostasis, transcatheter arterial embolization, or surgery, reliable preventive strategies should be urgently developed.

Tranexamic acid (TXA) is an antifibrinolytic agent that mitigates bleeding by inhibiting fibrin degradation.⁸ It exerts its effect by binding to tissue plasminogen, preventing blood clot lysis and thereby reducing excessive bleeding. Numerous studies have demonstrated the effectiveness of TXA in managing bleeding across various organs, except the gastrointestinal tract.⁹ Topical application of TXA reduces the risk of intraoperative bleeding and improves the quality of the surgical field in various surgeries, without inducing the major side effects associated with systemic administration.¹⁰ We previously demonstrated the efficacy of topical TXA powder in treating bleeding peptic ulcers.¹¹ For patients with chronic renal disease, the recommended maximum daily dose of TXA is 1 g.

Sucralfate is an aluminum hydroxide–sucrose octasulfate complex that binds to the wound base,¹² forming a protective barrier that shields the wound site from environmental damage, promotes angiogenesis, and facilitates mucosal healing.¹³ It has been widely used to treat various wounds and ulcers, such as skin wounds, oral ulcers, and peptic ulcers.^14,15 The standard daily dose of sucralfate is 2 g.

Given the hemostatic properties of TXA and the mucosal protective effects of sucralfate, we hypothesized that targeted endoscopic application of 1 g of TXA and 2 g of sucralfate to post-EST wound sites would enhance hemostasis and prevent delayed bleeding. To test this hypothesis, we evaluated the efficacy and safety of topical TXA and sucralfate application in preventing delayed post-EST bleeding.

From December 2023 to December 2024, we screened 178 potentially eligible patients for enrollment. Patients were excluded if they had unsuccessful CBD cannulation (n = 2), had a history of EST or EPBD (n = 8), were scheduled for biliary stent insertion without a sphincterotomy (n = 30), underwent EPBD instead of EST (n = 7), or had any known biliary or pancreatic malignancies (n = 11). The final cohort comprised 120 patients, 60 per group. No participant underwent EST plus EPBD. Figure 2 presents a flowchart depicting patient selection.

Baseline characteristics, pre-ERCP blood test results, comorbidities, and antiplatelet or anticoagulant use were similar between the intervention and control groups (Table 1). Between-group similarity was also noted in the presence of juxta-papillary diverticulum, the maximum size of CBD stones, and the distribution of papilla types. Transpancreatic precut sphincterotomy was performed in two patients, one from each group. All patients who had immediate post-EST bleeding achieved successful endoscopic hemostasis by epinephrine injection or heater probe coagulation. No patient received a metal biliary stent for hemostasis.

Five patients in the control group (4.1%) reported tarry stools after ERCP and received endoscopic hemostasis, whereas no patient in the intervention group had delayed post-EST bleeding. Active oozing from major papilla wounds occurred during endoscopy, and hemostasis was achieved by epinephrine injection and heater probe coagulation. Kaplan–Meier survival curve analysis revealed that freedom from delayed post-EST bleeding within 14 days was longer in the intervention group than in the control group (p = 0.023; Figure 3). The incidence rate of delayed post-EST bleeding was lower in the intervention group than in the control group (0% vs 8.3%; p = 0.029; Table 2). The risk difference was 0% (0/60) in the intervention group versus 8.33% (5/60) in the control group, yielding a risk difference of −8.33% favoring intervention. The number needed to treat was 1/0.0833 = 12 patients. To address 0 events in the intervention group in relative risk estimation, we applied the Newcombe–Wilson method. The relative risk for the intervention group was 0.09 (95% confidence interval (CI): 0.006–0.18; p = 0.03).

On post-ERCP day 2, hemoglobin loss was significantly lower in the intervention group than in the control group (0.63 ± 0.86 vs 1.15 ± 1.11 g/dL; p = 0.006; Table 2). In addition, the incidence rate of substantial blood loss (hemoglobin loss >2 g/dL) was significantly lower in the intervention group than in the control group (1.6% vs 16.6%; p = 0.004). Notably, all 5 patients who had delayed post-EST bleeding were among the 11 who had a hemoglobin loss of >2 g/dL. No significant between-group difference was observed in total bilirubin reduction, PEP incidence, or cholangitis incidence. No patients required transcatheter arterial embolization or surgery. Moreover, no cases of bowel perforation or mortality were reported.

Logistic regression was performed to identify key risk factors for substantial blood loss (Table 3). Univariate analysis revealed that topical TXA and sucralfate application was associated with a reduced risk of substantial blood loss (odds ratio: 0.08; 95% CI: 0.01–0.68; p = 0.021). Multivariate analysis confirmed that topical TXA and sucralfate application was independently associated with a reduced risk of substantial blood loss on post-EST day 2 (odds ratio: 0.07; 95% CI: 0.008–0.62; p = 0.017).

To the best of our knowledge, this study is the first to evaluate the efficacy of hemostatic powders in preventing delayed post-EST bleeding. Topical TXA and sucralfate application not only prolonged the duration of freedom from delayed post-EST bleeding but also reduced the risks of delayed bleeding, hemoglobin loss, and substantial blood loss. The small amount of powder applied to the post-EST wound did not increase PEP risk or hinder total bilirubin reduction, reflecting a strong safety profile with no severe side effects.

Despite history and clinical evidence supporting the efficacy of TXA, certain misconceptions persist. Large-scale trials have demonstrated the efficacy of TXA in managing bleeding in various organs, except the gastrointestinal tract.^9,24 Although systemic administration of TXA does not prevent gastrointestinal bleeding or associated mortality, pharmacodynamic analyses suggest that local administration of TXA exerts a strong hemostatic effect.^25,26 A study reported that administering TXA solution through a nasogastric tube to a patient with upper gastrointestinal hemorrhage conferred no additional benefit.²⁷ By contrast, a clinical trial revealed that intravenous plus nasogastric TXA infusion reduced the need for urgent endoscopy, highlighting potential benefits of nasogastric TXA delivery.²⁸ The inconsistency may be attributable to the free-flow delivery of liquid TXA. Precise application of TXA powder, which adheres directly to the bleeding site, yields favorable outcomes.¹¹ In the present study, topical TXA application on the post-EST wound effectively reduced the risks of delayed bleeding and hemoglobin loss after the procedure.

Sucralfate rapidly interacts with duodenal mucosal fluid, forming a protective mucosal barrier.²⁹ It accelerates wound healing by binding to epidermal growth factors and stimulating epithelial proliferation.³⁰ Research has highlighted the advantages of sucralfate in the treatment of gastrointestinal ulcers, underscoring its potential to promote healing by increasing the bioavailability of growth factors and prostaglandins and reducing the production of reactive oxygen species.³¹ The barrier formed by sucralfate shields the post-EST wound from gastric acid interference.³²

Endoscopists have consistently focused on reducing the risk of post-EST bleeding, which has led to major advancements in sphincterotomy. Factors such as incision direction and electrocautery settings strongly influence outcomes. Sphincterotomy along the 11 to 1 O’clock axis reduces the risk of post-EST bleeding by avoiding regions with an increased arterial concentration.³³ Although the pure cut mode in EST is associated with an increased risk of intraprocedural bleeding, the effects of electrocautery settings on delayed post-EST bleeding remain unclear.³⁴

Pharmacologic strategies for preventing delayed bleeding remain inadequate. Although local injection of diluted epinephrine can control intraprocedural bleeding, its short duration of action does not ensure long-term prevention of delayed bleeding.³⁵ Similarly, oral prophylactic proton pump inhibitors exhibit limited efficacy in preventing delayed post-EST bleeding. A clinical trial reported no significant efficacy of proton pump inhibitors in preventing post-EST bleeding.²² By contrast, the present study revealed that combining TXA with sucralfate prevented delayed post-EST bleeding.

This study has several limitations. First, its single-center design and small sample size may limit the generalizability of the findings. Although randomization mitigated some of the effects of study design and sample size, large-scale studies are required for further validation. Second, delayed post-EST bleeding occurred in only five patients (4.1%), which precluded extensive subgroup and multivariate Cox regression analyses. To address this limitation, an analysis was performed in 11 patients (9.1%) who had a hemoglobin loss of >2 g/dL. In addition, multivariate logistic analyses were performed to adjust for potential confounders. Third, during their spraying, some drug powders may remain in the delivery catheter, but the amount is usually small (<0.1 g) and can be ignored. Because this study sought to combine the hemostatic effect of TXA and the wound protection effect of sucralfate, the individual contributions of TXA and sucralfate could not be distinguished. Fourth, this study was not a double-blinded trial. To minimize bias, we ensured that the endoscopists were unaware of group allocation before immediate bleeding was stopped through endoscopic hemostasis. Notably, patients who reported immediate bleeding were successfully managed, and the incidence of immediate bleeding was similar between the two groups. Furthermore, the care provider at the ward was unaware of group allocation, which ensured a low risk of bias. Fifth, none of the enrolled patients had ESRD, which raises concerns regarding the applicability of our findings to patients with ESRD. Further research is required to evaluate the efficacy of TXA and sucralfate application in preventing delayed bleeding in patients with ESRD. Similarly, patients who continued their antiplatelet or anticoagulant therapy were excluded from this study. Thus, additional studies are required to determine whether topical TXA and sucralfate application exerts protective effects in antiplatelet or anticoagulant users. Finally, the duration for which the drug powders remained adhered to the post-EST wound was not determined. All patients remained nil per os for 6 h after ERCP, after which they switched to a clear liquid diet and then gradually to a regular diet. Whether oral intake and constant flow of gastric, biliary, and pancreatic secretions influence the retention of drug powders remains unclear. Future studies should determine the duration of drug powder retention.

In conclusion, topical TXA and sucralfate powder application significantly prolongs freedom from delayed post-EST bleeding and potentially reduces the risks of delayed bleeding and hemoglobin loss after EST. This prophylactic approach appears to be a safe and effective hemostatic strategy with no major side effects. Additional large-scale, multicenter studies are required to validate and expand these findings.

This manuscript was edited by Wallace Academic Editing.