Portal hypertension. A new beginning for an old problem.

There are two types of shunts for the treatment of esophageal varices: total shunts or selective shunts. The various total shunts are designed to "decompress" both the portal and the variceal venous systems. In contrast, a selective shunt decompresses the varices only, and the portal pressure and presumably portal flow are not affected. The very purpose of the total shunt is the total or partial diversion of portal flow. This is the case no matter which branches of the portal and systemic venous systems are connected and whether the anastomosis is a direct one or is through a graft. In an ongoing prospective, randomized clinical trial, which now includes 46 patients, a total shunt (mesorenal H graft) and the selective distal splenorenal (Warren) shunt are compared. The operative mortality rate and the frequency of postoperative bleeding are the same. However, encephalopathy is more frequent after the total shunt (P less than 0.05), and a trend is developing which suggests a better long-term survival rate after the Warren shunt. If these trends continue, then the superiority of the Warren shunt is established in a selected group of patients who have bled at least once from their varices. If these patients continue to remain free of encephalopathy and hemorrhage will not recur, then it is probable that the Warren shunt could improve the chances of survival of cirrhotic patients with esophageal varices who did not bleed yet. The "prophylactic" value of the Warren shunt will require a multicenter, cooperative, controlled clinical trial.