The Warren shunt in treating bleeding esophageal varices.

In patients who have impaired hepatic reserve, the Warren shunt has been proposed as an effective operation because it decompresses the esophageal varices without disturbing portal perfusion of the liver. However, early reports of high operative mortality and technical difficulties have impeded acceptance of the procedure. The operation was done in a series of 17 patients. All patients in whom elective variceal decompression with a patent splenic vein was required and without clinical ascites were candidates for this operation. Follow-up ranged from 2 to 48 months. Six patients had alcoholic cirrhosis, two had primary biliary cirrhosis and seven had postnecrotic cirrhosis; in two the cause of the liver disease was unknown. Five patients were categorized as Child's class A, nine as class B and three as class C. No intraoperative or early postoperative deaths owing to hemorrhage occurred. However, there was one death two weeks postoperatively from pulmonary sepsis and one death five weeks postoperatively due to antigen-positive hepatitis. Two patients died from hepatic failure six weeks and five months after operation, respectively; in the first of these, chronic active hepatitis was diagnosed at the time of operation. In one patient hemorrhage recurred and transfusion was required. Although ascites, which eventually resolved, developed in eight patients after operation, the results in 76 percent of patients have been good without new episodes of hemorrhage or encephalopathy. We conclude that the Warren shunt is a safe and effective elective operation for the treatment of patients in whom hemorrhage from esophageal varices has occurred.