Objectives and methods: The prognostic factors have not yet been fully evaluated in patients with cirrhosis and gastric fundal varices (FV). We investigated the natural history of 145 patients with cirrhosis and FV with no history of bleeding. Various possible prognostic factors, which include clinical, biochemical, and endoscopical variables, were analyzed using Cox's proportional hazard model. Results: Among the 145 patients with cirrhosis and FV, there were 76 patients in class A, 45 in class B and 24 class C according to Child's classification. Sixty-five patients had concomitant hepatocellular carcinoma at the time of enrollment. Seventy deaths and 34 episodes of the hemorrhage from FV occurred during the mean follow-up period of 26.4 months. The cumulative survival rates at 1, 3, and 5 years were 75, 53 and 34%, respectively. The cause of death was related to gastrointestinal hemorrhage in 18 patients (15 deaths were related to FV hemorrhage), hepatic failure in 22, hepatocellular carcinoma in 22, and other causes in eight patients. In patients with small-, medium-, and large-sized FV, the deaths related to FV hemorrhage were 4, 21 and 54%, respectively. Overall, the death related to FV hemorrhage was 21%. A multiple regression analysis using Cox's model showed hemorrhage from FV, the presence of hepatocellular carcinoma and poor Child's status were all highly significant prognostic factors. Conclusion: The natural history of the patients with cirrhosis and FV was adversely modified by the hemorrhage from FV, concomitant hepatocellular carcinoma and poor hepatic functional reserve. Since the number of deaths related to FV hemorrhage was great in patients with large-sized FV, it is important to identify high-risk large FV and its prophylactic obliteration. Further studies are needed to elucidate the efficacy of prophylactic obliteration of large-sized FV.
Objectives and methods: The prognostic factors have not yet been fully evaluated in patients with cirrhosis and gastric fundal varices (FV). We investigated the natural history of 145 patients with cirrhosis and FV with no history of bleeding. Various possible prognostic factors, which include clinical, biochemical, and endoscopical variables, were analyzed using Cox's proportional hazard model. Results: Among the 145 patients with cirrhosis and FV, there were 76 patients in class A, 45 in class B and 24 class C according to Child's classification. Sixty-five patients had concomitant hepatocellular carcinoma at the time of enrollment. Seventy deaths and 34 episodes of the hemorrhage from FV occurred during the mean follow-up period of 26.4 months. The cumulative survival rates at 1, 3, and 5 years were 75, 53 and 34%, respectively. The cause of death was related to gastrointestinal hemorrhage in 18 patients (15 deaths were related to FV hemorrhage), hepatic failure in 22, hepatocellular carcinoma in 22, and other causes in eight patients. In patients with small-, medium-, and large-sized FV, the deaths related to FV hemorrhage were 4, 21 and 54%, respectively. Overall, the death related to FV hemorrhage was 21%. A multiple regression analysis using Cox's model showed hemorrhage from FV, the presence of hepatocellular carcinoma and poor Child's status were all highly significant prognostic factors. Conclusion: The natural history of the patients with cirrhosis and FV was adversely modified by the hemorrhage from FV, concomitant hepatocellular carcinoma and poor hepatic functional reserve. Since the number of deaths related to FV hemorrhage was great in patients with large-sized FV, it is important to identify high-risk large FV and its prophylactic obliteration. Further studies are needed to elucidate the efficacy of prophylactic obliteration of large-sized FV.