Combination of carvedilol with variceal band ligation in prevention of first variceal bleed in Child-Turcotte-Pugh B and C cirrhosis with high-risk oesophageal varices: the 'CAVARLY TRIAL'

Objectives: Beta-blockers and endoscopic variceal band ligation (VBL) have been preferred therapies for primary prophylaxis of variceal bleeding. However, the choice of therapy in patients with advanced liver disease with high-risk varices is not clear. A comparison of these therapies alone or in combination to prevent the first variceal bleed in advanced cirrhosis patients was carried out.

Design: 330 Child-Turcotte-Pugh (CTP) B and C cirrhosis patients, with 'high-risk' varices were prospectively enrolled (n=110 per group) to receive carvedilol (group A), VBL (group B) or combination (group C). Primary endpoint was reduction in the incidence of first variceal bleed at 12 months. The secondary endpoints included overall mortality, bleed-related mortality, new-onset decompensation, change in hepatic vein pressure gradient (HVPG) and treatment-related adverse events.

Results: The patients were predominantly males (85.2%), aged 51.4±10.5 years with CTP score of 8.87±1.24, MELD score 15.17±3.35 and HVPG-16.96±3.57 mm Hg. The overall incidence of variceal bleed was 23.8% (n=78) at 1 year. Intention-to-treat analysis showed that the combination arm (group C) significantly reduced the incidence of first variceal bleed by 62.9% as compared with group B (HR 0.37, 95% CI 0.192 to 0.716, p<0.003) and by 69.3% as compared with group A (HR 0.31, 95% CI 0.163 to 0.578, <0.001). The overall mortality was 13.6% (45/330). The 1-year mortality in group C was lowest among the three groups (A, B, C=20%, 14.5%, 6.3%, p=0.012). Reduction in HVPG (20.8% vs 25.1%, p=0.54) and the rate of non-response to carvedilol (53.4% vs 41.25%, p=0.154) were not different between group A and C patients. The incidence of new-onset ascites, spontaneous bacterial peritonitis, shock, and acute kidney injury and postbleed organ failure was also comparable between the groups.

Conclusion: In CTP B and C cirrhosis patients with high-risk varices, combination of carvedilol and VBL is more effective than either therapy alone, for primary prevention of variceal bleeding.

Trial registration number: NCT03069339.