Endoscopic variceal ligation-induced ulcer bleeding: a retrospective analysis of endoscopic therapy and patient outcomes

Aims Endoscopic variceal ligation (EVL)-induced ulcer bleeding is a rare complication, occurring in 2% – 8% of cases, but is potentially life-threatening with mortality rates ranging from 20% to 63%. Limited data exist regarding the efficacy and outcomes of the available endoscopic treatment options for this complication. Therefore, we evaluated their primary hemostasis success rate and 5-day bleeding recurrence rate as well as the in-hospital mortality rate.

Methods This retrospective study analysed all EVL procedures performed at the Charité University Clinics, Campus Virchow and Campus Mitte, in Berlin from 01.01.2016 until 30.06.2023. A total of 1864 EVLs were conducted; with 73,4% performed electively, 9,5% semi-electively during inpatient treatment for decompensated liver cirrhosis and 17,1% performed emergently due to esophageal variceal bleeding. Ulcer bleeding as a complication of EVL was observed in 61 cases (3,3%), one case was excluded due to missing data.

Results EVL-induced ulcer bleeding was a rare event after elective EVL (0,44%). However, the relative risk of this complication increased 20-fold and 37-fold after emergent or semi-elective EVL. The most commonly employed endoscopic therapy was a repeat EVL (n=24, 40%), either alone or in combination with other endoscopic modalities. This approach achieved a primary hemostasis rate of 83,3% but had a 30% 5-day bleeding recurrence rate. Fibrin glue injection was the second most often used therapy (n=16, 26,7%), either alone or in combination with other modalities. The primary hemostasis rate was 87,5% and the 5-day bleeding recurrence rate was 31,5%. Other less frequently used treatments were: no intervention, Sengstaken-Blakemore tube, epinephrine injection, fully covered self-expandable metallic stent, synthetic peptides, cyanoacrylate and clips. Primary hemostasis was achieved in 45% of these patients with one endoscopic method, whereas for 26,7% of the patients a combination of 2-4 methods was required. The in-hospital mortality for this cohort was 46,7%. Univariate analysis showed that the MELD (Model for End-stage Liver Disease) score and the CLIF-C acute decompensation score, as well as the presence of organ failure or acute on chronic liver failure at the time of the EVL along with bleeding recurrence or hospital-acquired infection were associated with increased in-hospital mortality. Based on multivariate analysis, bleeding recurrence was the strongest predictor of in-hospital mortality (OR=12,78; 95%-CI 1,43 – 114,34; p=0,02). Although the small sample size limited definitive conclusions, no endoscopic treatment demonstrated superiority regarding 5-day bleeding recurrence or in-hospital mortality.

Conclusions EVL-induced ulcer bleeding occurs more frequently following semi-elective or emergent EVLs, with multiple interventions demonstrating similar hemostasis rates. However, the risk of re-bleeding remains substantial and serves as a strong, independent predictor of mortality. Large multicenter studies are essential to develop tailored treatment algorithms.

Publication History

Article published online:
27 March 2025