Control of variceal bleeding. Reassessment of rationale.

The selective shunt concept has been validated both physiologically and clinically for the nonemergent patient with variceal bleeding. The patency rate of the shunt has been good and long-term control of gastroesophageal bleeding excellent. Furthermore, the transsplenic decompression of the upper stomach and lower esophagus has effectively prevented the problems related to gastritis and gastric varices seen with nonshunt procedures. Continued portal perfusion of the liver has been sustained for more than 9 years in some patients, and 9 of 11 patients have portal perfusion exceeding 7 years. The metabolic problems of the portaprival syndrome have been greatly ameliorated, and long-term survival with clinical well-being has been achieved. For the actively bleeding patient, acute renal vein hypertension needs to be prevented to provide effective control of hemorrhage. Selective splenocaval shunts have been successful in this regard in a small number of patients.