Improved results with selective distal splenorenal shunt in a highly selected patient population. A prospective study.

In a 5-year period 299 patients were admitted to the Heinz-Kalk Hospital with bleeding esophageal varices. Patients with acute bleeding were treated with endoscopic sclerotherapy. Sessions were performed as many times as needed for each individual case. One hundred seventy-eight patients in Child-Pugh class C were excluded from surgical treatment; the remaining 121 patients (Child AB) were selected using the following criteria: liver volume (ultrasound) between 1000 to 2500 ml, portal perfusion (sequential scintigraphy) more than 30%, no activity or progression of liver disease proved by biopsy, no stenosis of the hepatic arteries, and suitable anatomy to perform the Warren shunt. Only 32 patients fulfilled these criteria. In seven of these cases the shunt was technically impossible to perform. Operative mortality rate was 8% and the late mortality rate was 12%. No history of rebleeding, encephalopathy, and/or shunt thrombosis was recorded. Five-year survival rate, according to the method of Kaplan-Meier was 75%. We conclude that the Warren shunt is the treatment of choice for elective management of bleeding esophageal varices. The postoperative results can be improved with strict selection using the above criteria. The preoperative use of sclerotherapy has a positive influence. Prophylactic management to prevent encephalopathy is also recommended.