Abstract

Background: Acute variceal bleed (AVB) from cardiofundal varices (GOV-2/IGV-1) is associated with high mortality rates in patients with liver cirrhosis. No consensus exists on the best modality to prevent rebleeding after an index episode of bleeding.

Methods: Consecutive cirrhosis patients with AVB from cardiofundal varices, after primary hemostasis by endoscopic obturation with cyanoacrylate glue (CYA), were randomized into two arms. In the ‘endoscopic intervention’ (EI) arm, endoscopic obturation with CYA was repeated at regular intervals (1, 3, 6 and 12 months); while in the ‘radiological intervention’ (RI) arm, patients underwent transjugular intrahepatic portosystemic shunt (TIPS) or balloon-occluded retrograde transvenous obliteration (BRTO); preferably BRTO, if a shunt vessel was present. Hepatic venous pressure gradient (HVPG) was measured at baseline and 1 month. Primary outcome measures included rebleed rates and all-cause mortality at 1 year.

Results: We randomized 90 patients (n=45 in each arm), median age 46 (35-55) years with mean (±SD) Child and MELD scores at baseline 7.4±1.8 and 12.3±3.2, respectively. Alcohol was the predominant etiology of cirrhosis in 33 (36.7%) patients. There were no differences in baseline characteristics between the two arms. In the RI arm, 25 patients underwent BRTO and 20 underwent TIPS. Median follow-up was 17.9 and 16.4 months, for EI and RI arms, respectively. Rebleed rates at 1 year were significantly higher in the EI arm compared to RI arm: 13 (28.9%) vs 3 (6.7%); p=0.010 (Figure 1a). Mortality at 1 year was 12 (26.7%) in the EI arm versus 7 (15.6%) in the RI arm (p=0.108) (Figure 1b). Technical success for glue injection, TIPS and BRTO was 100%, 100% and 96.2% respectively. Worsening of ascites after radiological intervention was reported by 12 (26.7%) patients versus 2 (4.4%) in EI arm; p=0.007. On sub-group analysis, patients undergoing BRTO had a statistically insignificant median rise in HVPG (2 mm versus 1 mm of Hg; p=0.715) and aggravation of esophageal varices on follow-up (24% versus 11%; p=0.150) compared to the EI arm. There was no significant difference in complications, rebleeding rates and overall mortality at 1 year between those undergoing TIPS as compared to BRTO. The probability of remaining free from all-cause rebleeding at 1 and 2 years was 70.7% versus 93%, and 52.3% versus 93% for the EI and RI arms, respectively (Figure 1a).

Conclusion: Radiological intervention for secondary prophylaxis significantly reduces rebleeding in patients with liver cirrhosis with GV hemorrhage but does not provide any survival benefit. TIPS and BRTO have comparable complications, rebleeding and mortality rates on follow-up.