Endoscopic sclerotherapy with portal-systemic shunt rescue

Abstract

This paper reports the preliminary results of a prospective randomized trial comparing endoscopic variceal sclerosis and distal splenorenal shunt (DSRS) in the management of patients with cirrhosis and variceal bleeding. Seventy-one patients have been entered; 36 have received sclerosis and 35 DSRS. Randomization of the study population was stratified on Child's A/B (56%) and Child's C (44%). Sixty-one per cent had alcoholic and 39% nonalcoholic cirrhosis. No patients have been lost to follow-up, which currently stands at a median of 26 months. Rebleeding occurred significantly (p < 0.05) more frequently in patients in the sclerosis group (19 of 36: 53%) compared to DSRS (1 of 35: 3%), but only 11 of 36 (31%) were not controlled by further sclerosis and failed that therapy. Patients in whom sclerosis failed underwent surgery. Survival was significantly (p < 0.01) improved in the sclerosis group (+ surgery in 31%), with an 84% 2-year survival compared to a 59% 2-year survival in the DSRS group. Portal perfusion was significantly (p < 0.05) better maintained in the sclerosis (95%) compared to the DSRS (53%) group. Galactose elimination capacity improved significantly (p < 0.05) in 21 patients successfully managed by sclerosis at 1 year and was significantly (p < 0.01) better maintained in the sclerosis compared to DSRS group. The authors conclude that endoscopic sclerosis: (1) has a higher rebleeding rate than DSRS, with one third of patients failing therapy from rebleeding; (2) allows significant improvement in liver function when successful; and (3) gives significantly improved survival in the management of variceal bleeding when backed up by surgical therapy for patients with uncontrolled rebleeding. The authors report the results of a randomized comparison of emergency portacaval anastomosis (PCA) (32 patients) with endoscopic sclerotherapy (EST) (32 patients). Short-term survival was similar in the two groups, respectively (44% vs. 50%). The amount of blood transfusions required and the length of the hospital admissions were significantly less in the EST group (p < 0.001). Recurrences of variceal hemorrhage, the number of transfusions required to treat them and the duration of hospital readmissions were all greater in the EST group (p < 0.001). Almost half of the EST patients who were discharged from the hospital alive subsequently rebled and required surgical treatment. The authors conclude that EST is as effective as PCA in the acute treatment of bleeding, but that when EST fails, PCA may be effective management.