Natural history of cirrhotic patients with small esophageal varices: a prospective study.

OBJECTIVE

Contrasting data are available on the natural history and bleeding risk of small esophageal varices. The aim of this prospective study was to evaluate a large series of consecutive cirrhotics with a first endoscopic diagnosis of small varices.

METHODS

Between 1987 and 1992, 258 patients with small varices and no previous bleeding were enrolled. Patients were clinically examined every 6 months and were followed until a first episode of bleeding and/or death, or until June 1998. None received any treatment to prevent bleeding. Endoscopies were planned at 18-month intervals.

RESULTS

The cumulative risk of bleeding was low (3% at 2 yr and 8% at 4 yr) and remained low in patients in whom varices remained small at 2nd endoscopy, whereas it increased significantly when varices enlarged. The increase of varices appeared to be rather linear in time: at the 2nd endoscopy varices remained small in 79% of patients and increased in 21%; at the 3rd endoscopy varices remained small in 55%, whereas at the 4th 33% of patients still had small varices. Clinical and biochemical data at the 1st and 2nd endoscopy were included in a multiple logistic regression analysis. Only the increase in Child-Pugh score appeared to be a significant predictor of enlarged varices; the risk of aggravation increased by 37.5% for every unit of impairment of the score.

CONCLUSIONS

The present study shows that patients with small varices have a low bleeding risk. An increase in Child-Pugh score during follow-up suggests enlargement of varices, thus an increase in bleeding risk. In these patients closer endoscopic surveillance is recommended.