Long-term injection sclerotherapy treatment for esophageal varices. A 10-year prospective evaluation.

Long-term injection sclerotherapy after proved variceal bleeding was assessed in 245 patients. The majority had alcoholic cirrhosis and the patients were equally distributed between modified Pugh-Child's risk grades A, B, and C. Esophageal varices were eradicated in 88% of the 140 patients who survived long enough for analysis, and remained eradicated for a mean of 19.4 months. The incidence of recurrent variceal bleeding after the first hospital admission was 0.02 bleeding episodes per patient month of follow-up study and was markedly reduced after eradication of varices. The overall cumulative survival rates at 1, 5, and 10 years were 54%, 39%, and 29%, respectively. The prognosis was influenced by the risk grade and the number of variceal bleeds before entering the study and to a lesser extent by the etiology of the cirrhosis. Fifty-two per cent of the patients died during the 10-year period. Liver failure was the major cause of death. Complications were mostly of a minor nature but they became cumulative with time. Minor complications included mucosal slough and injection-site leak, although the latter had an associated mortality risk. Significant esophageal stenosis and esophageal rupture were rare. As a result of this study a more radical surgical policy is proposed for sclerotherapy failures. These are defined as patients in whom varices are difficult to eradicate or who continue to have major variceal bleeds. Such patients should be subjected to either a portosystemic shunt or a devascularization and transection procedure.