Risk factors for band-induced ulcer bleeding after prophylactic and therapeutic endoscopic variceal band ligation. Issue 8 (August 2015)

• Record Type: Journal Article
• Authors: Sinclair, Marie; Vaughan, Rhys; Angus, Peter W.; Gow, Paul J.; Parker, Frank; Hey, Penelope; Efthymiou, Marios
• Is Part Of: European journal of gastroenterology & hepatology. Volume 27:Issue 8(2015:Aug.)
• Publication Date: 2015-08
• Subjects: gastrointestinal haemorrhage -- gastro-oesophageal reflux -- liver cirrhosis -- oesophageal varices; Digestive organs -- Diseases -- Periodicals; Liver -- Diseases -- Periodicals; Digestive organs -- Diseases; Liver -- Diseases; Periodicals; 616.33
• DOI: 10.1097/MEG.0000000000000387
• Languages: English
• ISSNs: 0954-691X

Abstract

Abstract : Background and aims: Endoscopic variceal band ligation (EVBL) aims to eradicate high-risk oesophageal varices. There is a small risk of precipitating bleeding from EVBL-induced oesophageal ulceration, which is associated with significant mortality. We explore the risk factors and outcome of EVBL-induced ulcer bleeding. Methods: Retrospective review of our endoscopy database between 2007 and 2012 identified upper endoscopies during which EVBL was performed. Patient demographics, biochemistry and endoscopic findings were recorded as were the complications of EVBL-induced ulcer bleeding and death. Results: A total of 749 episodes of EVBL were performed in 347 patients with a mean Model for End-stage Liver Disease (MELD) score of 15.8. In all, 609 procedures were performed for prophylaxis and 140 for acute haemorrhage. There were 21 episodes (2.8% of procedures) of EVBL-induced ulcer bleeding in 18 patients, five of whom subsequently died (28%). On multivariable analysis, acute variceal haemorrhage was the only significant predictor of EVBL-induced ulcer bleeding [odds ratio (OR) 6.25 (2.57–15.14), P <0.0001]. In 609 procedures performed for prophylaxis, the EVBL-induced ulcer bleeding rate was 1.5%, with 22% mortality. In this group, higher MELD score and reflux oesophagitis were associated significantly with EVBL-induced ulcer bleeding [OR 25.53 (2.14–303.26), P =0.010 and OR 1.07 (1.01–1.13), P =0.019, respectively]. Conclusion: Our EVBL-induced ulcer bleeding rate was low, but associated with significant mortality. Highest rates were observed following EVBL for acute variceal haemorrhage, for which EVBL is unavoidable. The incidence was lower following prophylactic EVBL, with the MELD score being the predominant risk factor. Reflux oesophagitis requires further investigation as a potentially modifiable risk factor for EVBL-induced ulcer bleeding.