Early hemodynamic changes following selective distal splenorenal shunt for portal hypertension: Comparison of surgical techniques

Ninety patients with cirrhosis undergoing elective distal splenorenal shunt (DSRS) for variceal bleeding between January, 1977 and September, 1988 comprised the study group. In 63 cases, the original technique of Warren was used and, in 15, the modified Britton procedure was employed. Twelve patients had a DSRS plus splenopancreatic disconnection. Thirtyfour had alcoholic cirrhosis and 56 had nonalcoholic cirrhosis. Intraoperative portal pressure remained high after the shunt (29.4 cm H₂O)even if its initial value was probably decreased by the loss of the splenic flow. Splenic pressure was reduced to 21 cm H₂O. The hepatic artery diameter enlarged even after selective shunt (from 6.5 to 7.1 mm). The persistence of a high portal pressure allowed for the preservation of hepatopedal portal flow in 87% of cases. Disconnection between the high-pressure mesenteric area and the low-pressure splenic area seemed to be ideal in only 17% of cases. Fifty-five percent of cases had the early development of minimal or moderate portomesenteric gastrosplenic (PM-GS) collateral pathways. In 33%, the PM-GS collaterals were generally abundant and often allowed visualization of the splenic and caval veins during the venous phase of the superior mesenteric arteriograms. In this group, portal flow was generally highly reduced and even abolished. The incidence of portal thrombosis was 11%. Early angiographic checks after DSRS did not show a different hemodynamic behavior between alcoholics and nonalcoholics. Splenopancreatic disconnection seems to prevent the development of collaterals and the loss of portal perfusion after shunt surgery.