To compare the effects of endoscopic variceal ligation(EVL), central splenorenal or splenocaval shunt and port-azygos devascularization in treating the recurrent hemorrhage for hepatocirrhosis portal hypertension with esophageal varice after EVL.

Clinical data were collected from 190 patients with recurrence of bleeding after EVL for hepatocirrhosis portal hypertension with esophageal varice in the Third Xiangya Hospital of Central South University from March 2005 to March 2011. Local ethical committee approval had been received and that the informed consent of all participating subjects was obstained. All patients were randomly assigned to three treatment groups: endoscopic treatment group (n= 72), splenorenal/splenocaval shunt group(n=56) and devascularization group (n=62). Patients in the endoscopic treatment group continued to received EVL treatment. In the shunt group, the patients underwent spleen resection plus splenorenal/splenocaval shunt, in which anastomosis of near-end of splenic vein to end-to-side of left renal or splenocaval shunt was performed with the diameter of stoma less than 10 mm. Spleen resection plus port-azygos devascularization were applied in the devascularization group. Short term effect and complications of 3 groups were observed after operation and postoperative follow-up examinations were performed. The t test was applied to compare the differences of portal pressure changes, the total hospitalization time and costs of 3 groups. The chi-square test was applied to compare the differences of complications and follow-up observations.

All patients in the 3 groups stopped bleeding after treatment. One patient (1%) in the endoscopic treatment group developed massive hemorrhage after 6 days after EVL and the hemorrhage was relieved after inserting sengstaken-blakemore tube and medicine treatment. The portal pressure in the splenorenal/splenocaval shunt group and devascularization group were reduced obviously (t=17.47, 25.71, P<0.05, P<0.05); The portal pressure in the splenorenal/splenocaval shunt group was lower than that in devascularization group(t=12.20, P<0.05). The incidence of perioperative portal vein thrombosis was higher in deavascularization group compared with splenorenal/splenocaval shunt group(χ²=4.62, P<0.05). The rates of rebleeding, recurrance of esophageal varice, portal hypertensive gastropathy (PHG) and hepatic encephalopathy in splenorenal/splenocaval shunt group were lower compared with the endoscopic treatment and devascularization group. The total hospitalization time and costs in splenorenal/splenocaval shunt group were also less than those in endoscopic treatment and devascularization groups(P<0.05).

Compared with endoscopic treatment and port-azygos devascularization, splenorenal or splenocaval shunt demonstrates better effect in treating hepatocirrhosis portal hypertension with esophageal varice rebleeding after endoscopic variceal ligation.