Eighty interposition mesocaval shunts, using a knitted Dacron large diameter prosthesis, have been performed during the past five and one-half years. Patients were evaluated from the standpoint of protection from recurrent esophageal hemorrhage, shunt patency, encephalopathy and cumulative survival analysis. In a selected group of patients, hemodynamic measurements were also obtained in the pre, intra, and postoperative periods. These included measurements of wedged hepatic vein pressure, superior mesentric venous blood flow, and residual superior mesenteric, hepatic sinusoidal and inferior vena cava pressures following the shunt procedure. Additionally, direct shunt flow measurements utilizing a square wave of electromagnetic flowmeter were also performed. Results indicate that the shunt patency is 95%; adequate decompression of the portal system was accomplished; recurrent variceal hemorrhage did not occur if the shunt remained patent; the incidence of encephalopathy was low (11%); and the operative mortality for the entire series was 9%. Continued perfusion of the liver was documented in 44% of patients and appears to be a function of the residual total portal resistance largely controlled by inferior vena caval pressure at the level of graft replacement. Life survivhat the interposition mesocaval shunt appears to be an effective technique for the control of variceal hemorrhage, has important hemodynamic advantages and can be applied to most patients for the control of variceal hemorrhage due to portal hypertension.

Since 1980, the authors have managed 19 patients with operative injuries to their biliary tracts. Eleven patients (58%) incurred their injuries during cholecystectomies for acute cholecystitis (average age--56 years); seven patients (37%) received their injuries during elective cholecystectomies (average age--24 years); and in one patient (5%) the injury occurred during gastrectomy. In group I were eight patients in whom injuries were recognized and repaired intraoperatively during their initial operations. Seven of these patients (88%) had primary duct reanastomoses, and one patient had a choledochoduodenostomy. All healed without further surgery, and none later had cholangitis develop. In group II were 11 patients diagnosed and reoperated later after surgery (mean time until diagnosis, 12 days). Seven of these patients (64%) were managed with Roux-en-Y hepaticojejunostomies or choledochojejunostomies. Four patients had cholangitis develop after surgery: two had demonstrable anastomic stenosis and two had no stenosis. Three of these four patients (75%) who had cholangitis develop did not have stents used in their repairs. The overall mortality rate in this series of 19 patients was 11%. The major risk factors for biliary tract injury were the presence of acute cholecystitis and of anatomically small biliary ducts. For this latter reason, younger patients who had elective cholecystectomies were particularly at risk. In delayed repair, the use of internal stents appeared to be useful in preventing the later development of stenosis and/or cholangitis; however, recurrent cholangitis developed in two patients who did not have demonstrable anastomotic stenoses.