BACKGROUND: Primary prophylaxis of variceal haemorrhage with non-selective beta blockers (NSBB) or variceal band ligation (VBL) is now standard of care in patients with cirrhosis with portal hypertension. NSBB, and particularly carvedilol, may be associated with improved survival. AIM: To assess mortality in a cohort of patients previously randomised to either carvedilol or VBL. METHODS: We retrospectively analysed 152 patients recruited to a multi-centre randomised controlled trial between 7 April 2000 and 24 June 2006 designed to assess the efficacy of VBL versus carvedilol in preventing first variceal bleed. We used electronic records to undertake long-term follow-up (up to 20 years) with the primary outcome of all-cause mortality and secondary end points of liver-related mortality and decompensation events (ascites, encephalopathy, variceal bleeding). RESULTS: We included 152 patients in analysis with baseline characteristics well matched between the carvedilol (n = 77) and VBL (n = 75) groups. In the intention-to-treat analysis, carvedilol offered a significant survival advantage with median survival of 7.8 years compared to 4.2 years in the VBL group (P = 0.03). This survival benefit was maintained in per-protocol analysis when patients who crossed between treatment arms were excluded (P = 0.02). Transplant-free survival, liver-related mortality and decompensation events were similar in both groups. CONCLUSION: These data suggest that carvedilol offers a significant survival benefit for patients with cirrhosis and portal hypertension. The difference in all-cause and liver-related mortality suggests that this survival benefit may not be entirely liver-related. Prospective, studies are required to confirm these important findings.
BACKGROUND: Primary prophylaxis of variceal haemorrhage with non-selective beta blockers (NSBB) or variceal band ligation (VBL) is now standard of care in patients with cirrhosis with portal hypertension. NSBB, and particularly carvedilol, may be associated with improved survival. AIM: To assess mortality in a cohort of patients previously randomised to either carvedilol or VBL. METHODS: We retrospectively analysed 152 patients recruited to a multi-centre randomised controlled trial between 7 April 2000 and 24 June 2006 designed to assess the efficacy of VBL versus carvedilol in preventing first variceal bleed. We used electronic records to undertake long-term follow-up (up to 20 years) with the primary outcome of all-cause mortality and secondary end points of liver-related mortality and decompensation events (ascites, encephalopathy, variceal bleeding). RESULTS: We included 152 patients in analysis with baseline characteristics well matched between the carvedilol (n = 77) and VBL (n = 75) groups. In the intention-to-treat analysis, carvedilol offered a significant survival advantage with median survival of 7.8 years compared to 4.2 years in the VBL group (P = 0.03). This survival benefit was maintained in per-protocol analysis when patients who crossed between treatment arms were excluded (P = 0.02). Transplant-free survival, liver-related mortality and decompensation events were similar in both groups. CONCLUSION: These data suggest that carvedilol offers a significant survival benefit for patients with cirrhosis and portal hypertension. The difference in all-cause and liver-related mortality suggests that this survival benefit may not be entirely liver-related. Prospective, studies are required to confirm these important findings.