Liver Transplantation during Fifty-Three Years’ Experience with Randomized Controlled Trials of Emergency Portacaval Shunt for Bleeding Esophageal Varices in Cirrhosis

Abstract

Background: Currently, emergency treatment of Bleeding Esophageal Varices (BEV) consists mainly of endoscopic and pharmacologic measures, with use of TIPS when bleeding is not controlled. Surgical shunt has been relegated to salvage of survivors of failed endoscopic, pharmacologic, and TIPS treatment. We have conducted 10 prospective studies of Emergency Porta Caval Shunt (EPCS) during 46 years. BEV in cirrhosis has been considered an indication for Liver Transplantation (LT) by some respected LT programs. This important issue was examined in our two recent Randomized Controlled Trials (RCTs) in 365 unselected, consecutive patients with acute BEV conducted from 1988 to 2011. Study design: In RCT No. 1, 211 unselected, consecutive cirrhotic patients with acute BEV (“all comers”) were randomized to Emergency Endoscopic Sclerotherapy (EEST) (n=106) or EPCS (n=105). In RCT No. 2, 154 unselected consecutive cirrhotic patients with acute BEV (“all comers”) were randomized to TIPS (n=78) or EPCS (n=76). In each RCT, the two treatment groups were compared with regard to effect on survival, control of bleeding, encephalopathy, and direct cost of care. Diagnostic workup was completed within 6 hr and primary treatment was initiated within 8 to 12 hours. Regular follow-up was accomplished in 100% of patients. In RCT No. 1, 96% of patients underwent more than 10 years of follow-up, or until death. In RCT No. 2, follow-up lasted for 5 to 10 years in 85%, and 3 to 4.5 years in the remainder. Patients were evaluated for LT on admission and at regular intervals thereafter. In addition to the two RCTs, the analysis of LT was supplemented in 1300 unrandomized cirrhotic patients who previously underwent Porta Caval Shunt (PCS) with 100% follow-up. Results: Clinical characteristics of each group were similar at the time of enrollment. The distribution of patients in Child risk classes was almost identical in the two groups in each RCT. Almost one-third of patients in each group were in Child class C. Permanent control of bleeding was achieved by sclerotherapy in only 20% and by TIPS in only 22%. In contrast, EPCS permanently controlled bleeding in 97% to 100% (p<0.001). Sclerotherapy and TIPS patients required twice as many units of blood transfusion as EPCS patients. Survival rate at all time intervals and in all Child classes was significantly greater following EPCS than after sclerotherapy and TIPS (p<0.001). Median survival was over 10 years following EPCS, compared to 1.99 years following TIPS. EPCS remained permanently patent in 97% of patients. In contrast, stenosis or occlusion of TIPS was demonstrated in 84% of patients, 63% of whom underwent TIPS revision, which failed in 80%. Recurrent episodes of encephalopathy developed in 35% of sclerotherapy patients and 61% of TIPS patients. In contrast, encephalopathy occurred in 15% of EPCS patients in RCT No. 1 and 21% of EPCS patients in RCT No. 2. Direct costs of care, measured as charges per year, were 5 to 7 times greater in the sclerotherapy group ($168,100) and TIPS groups ($264,800) than in the EPCS group ($39,000) (p<0.001). Of the 365 randomized patients in the two RCTs, only 23 (6.3%) were ultimately referred for LT, mainly because of progressive liver failure. Of these, only 11 (3%) were approved for LT, and only 8 (2.2%) underwent LT. One- and 5-year LT survival rates were 0.68% and 0, compared to 81% and 73% after EPCS. In the 1300 unrandomized PCS patients, 50 (3.8%) were referred and 19 (1.5%) underwent LT. Five-year survival rate was 53% compared to 72% for all 1300 patients. Conclusions: EPCS was uniformly effective in treatment of BEV, while sclerotherapy and TIPS were disappointing. EPCS permanently stopped variceal bleeding, almost never became occluded, accomplished long-term survival that was more than 5 times survival rate following sclerotherapy or TIPS, and was much less costly than sclerotherapy or TIPS. If bleeding is permanently controlled, as occurred invariably following EPCS, cirrhotic patients with BEV seldom require LT. PCS is effective first-line and long-term treatment. Should LT be required in patients with PCS, although technically more demanding, numerous studies have shown that PCS does not increase mortality or complications. Neither EST nor TIPS are effective emergency or long-term therapeutic measures.