Background/aims: The incidence and natural history of small esophageal varices (EV) in cirrhotics may influence the frequency of endoscopies and the decision to start a pharmacological treatment in these patients.
Methods: We prospectively evaluated 206 cirrhotics, 113 without varices and 93 with small EV, during a mean follow-up of 37+/-22 months. Patients with previous gastrointestinal bleeding or receiving any treatment for portal hypertension were excluded. Endoscopy was performed every 12 months.
Results: The rate of incidence of EV was 5% (95%CI: 0.8-8.2%) at 1 year and 28% (21.0-35.0%) at 3 years. The rate of EV progression was 12% (5.6-18.4%) at 1 year and 31% (21.2-40.8%) at 3 years. Post-alcoholic origin of cirrhosis, Child-Pugh's class (B or C) and the finding of red wale marks at first examination were predictors for the variceal progression. The two-years risk of bleeding from EV was higher in patients with small varices upon enrollment than in those without varices: 12% (95% CI: 5.2-18.8%) vs. 2% (0.1-4.1%); (P<0.01). Predictor for bleeding was the presence of red wale marks at first endoscopy.
Conclusions: In patients with no or small EV, endoscopy surveillance should be planned taking into account cause and degree of liver dysfunction.