Background: Portal hypertensive gastropathy is a potential cause of bleeding in patients with liver cirrhosis. Studies on its natural history have often included patients submitted to endoscopic or pharmacological treatment for portal hypertension.
Patients and methods: A total of 222 cirrhotic patients with mild degree of portal hypertension (i.e., with no or small varices at entry, without previous gastrointestinal bleeding and medical, endoscopic, or angiographic treatment) were followed up with upper endoscopy every 12 months for 47 +/- 28 months.
Results: Upon enrollment 48 patients presented portal hypertensive gastropathy (43 mild and 5 severe) and the presence of esophageal varices was the only independent predictor of the presence of this gastric lesion at multivariate analysis. The incidence of portal hypertensive gastropathy was 3.0% (1.1-4.9%) at 1 yr and 24% (18.1-29.9%) at 3 yr, while the progression was 3% (1-6.9%) at 1 yr and 14% (4.2-23.8%) at 3 yr. The presence of esophageal varices and the Child-Pugh class B or C at enrollment were predictive of the incidence of portal hypertensive gastropathy, while only Child-Pugh class B or C was correlated with the progression from mild to severe, at multivariate analysis. During follow-up 16 patients bled from portal hypertensive gastropathy (9 acutely and 7 chronically) and one patient died of exsanguination from this lesion.
Conclusions: The natural history of portal hypertensive gastropathy is significantly influenced by the severity of liver disease and severity of portal hypertension. Acute bleeding from portal hypertensive gastropathy is infrequent but may be severe.