The Rebleeding Course and Long-Term Outcome of Esophageal Variceal Hemorrhage after Ligation: Comparison with Sclerotherapy

Abstract

Background: Endoscopic variceal ligation is widely accepted as the optimum endoscopic treatment for esophageal variceal hemorrhage. However, the rebleeding course and long-term outcome of patients with esophageal variceal hemorrhage after ligation have been poorly defined. Therefore, we conducted a long-term follow-up study to delineate the outcome of ligation and compare it with that after sclerotherapy. Methods: One hundred and eighty-five liver cirrhotic patients with endoscopically proven esophageal variceal hemorrhage were randomized to undergo endoscopic variceal sclerotherapy or ligation. These patients received regular follow-up and detailed clinical assessment. Results: Two patients developed hepatoma within 6 months of entry in each group and were excluded. Another six patients in the sclerotherapy group and seven patients in the ligation group were excluded because of poor compliance or lost to follow-up. Therefore, 84 patients in each group were analyzed. In this long-term follow-up (55.3 - 12.5 months) the rebleeding rate for ligation was lower than that for sclerotherapy, regardless of whether the rebleeding was analyzed by patient number or Kaplan-Meier analysis. With regard to the rebleeding risk of various periods, the sclerotherapy risk was higher than that of ligation within 4 weeks of the initial endoscopic treatment or before variceal eradication. Multifactorial analysis showed hematemesis, poor hepatic function, and sclerotherapy were the risk factors determining rebleeding. The annual hepatocellular carcinoma incidence was 4.9%. There was no difference in survival between sclerotherapy and ligation. Multifactorial analysis showed that poor hepatic function was the only factor determining survival. Conclusions: The rebleeding risk was higher in sclerotherapy than in ligation before variceal eradication, especially within 4 weeks of the initial endoscopic treatment. Long-term survival was dependent on hepatic reserve regardless of the treatment method.