medwireNews: Use of a pre-emptive transjugular intrahepatic portosystemic shunt (p-TIPS) significantly reduces the risk for rebleeding or death compared with glue obliteration and nonselective beta blockers (NSBB) in patients with cirrhosis and acute bleeding from fundal varices, concludes the GAVAPROSEC trial.
“The results of the present study strongly support the use of p-TIPS in the management of acute gastric variceal bleeding and add an additional argument in favor of TIPS, which improves prognosis by effectively treating the hemodynamic disorders associated with cirrhosis,” write Christophe Bureau (Université de Toulouse, France) and colleagues in The Lancet Gastroenterology and Hepatology.
The open-label trial included 101 patients aged 18–75 years with cirrhosis who were admitted to 17 French tertiary care centers for acute fundal variceal bleeding. Eligible patients had bleeding related to non-type 1 gastroesophageal varices according to the Baveno criteria that was controlled for at least 12 hours after an initial glue injection.
The patients were a mean of 58.2 years old, 80% were men, and 90% had alcohol-related cirrhosis. They had a mean Child–Pugh score of 8.4 points, indicating moderate liver dysfunction, and a mean Model for End-Stage Liver Disease score of 14.3 points, suggesting moderate severity.
The researchers randomly assigned the participants to receive either p-TIPS under general anesthesia (n=47) or continue with glue obliteration combined with NSBB (n=54). They were then followed up at 7 days post-initial glue injection or hospital discharge, and again at 42, 90, 180, and 365 days’ post-treatment, with a mean follow-up time of 365 days.
For patients receiving p-TIPS, the researchers inserted a covered stent within 72 hours of an initial endoscopic glue injection. The patients also received vasoactive drugs including octreotide, somatostatin, or terlipressin according to hospital protocols for 3–5 days until the p-TIPS was in place.
Patients in the glue obliteration plus NSBB group were given further injections when needed until the varices were fully closed plus band ligation for esophageal varices where necessary. They also received propranolol, nadolol, or carvedilol from day 5, at average doses of 124 mg/day, 14.8 mg/day, and 80 mg/day, respectively, to achieve a heart rate of 50–65 bpm.
Bleeds were most commonly caused by gastroesophageal varices type 2 followed by isolated gastric varices type 1, with respective rates of 56% and 42%.
A total of 18% of patients died, with 56% of the deaths related to liver disease, and 27% experienced rebleeding during the study. Rescue TIPS was required by 37% of patients in the glue plus NSBB group, mostly due to recurrent bleeding.
The p-TIPS group was significantly more likely to meet the primary endpoint of staying alive and free from rebleeding at 1 year than the glue obliteration plus NSBB group, with corresponding rates of 77% versus 37% and a hazard ratio (HR) of 0.25.
Early TIPS significantly reduced the risk for rebleeding compared with glue obliteration, with respective rates of 6.4% and 44.4% by 365 days and a sub-HR of 0.11. Two of the three rebleeds in the p-TIPS group were due to return of portal hypertension that was corrected with shunt-related interventions, and one was from a glue-related ulcer, treated endoscopically.
There were no significant differences between the groups in terms of other secondary endpoints including hepatic encephalopathy, new or worsening ascites, length of hospital stay, and overall survival.
The survival results when considering only the patients who received TIPS within 72 hours versus controls were significantly in favor of p-TIPS, with rates of 76% versus 37%. However, survival among patients given TIPS at any time point in the study was only significantly in favor of p-TIPS when those who switched from glue obliteration plus NSBB, as a rescue treatment for rebleeding, were also considered, giving an HR of 0.40.
There were no treatment failures or treatment-related deaths. The p-TIPS group experienced no major complications linked directly to the procedure, while the glue obliteration plus NSBB group saw 22 glue-related complications including 13 cases of bleeding and eight of glue migration.
There were no significant differences between the p-TIPS and glue obliteration plus NSBB groups in terms of rates of adverse events, at 38% and 46%. The most common adverse events were infections, affecting 26% and 28% of each group, respectively.
In an editorial related to the study, Michael Praktiknjo and Jonel Trebicka, both at the University of Münster in Germany, say that the researchers “are to be commended for this important study, which contributes to closing the knowledge gap and provides a strong argument for p-TIPS in gastric variceal bleeding.”
However, they caution that “important questions about patient selection, [spontaneous portosystemic shunt] embolisation, and long-term disease modification remain unanswered.”
They conclude: “An important first step has been made, but there is a long way to go until the goal of a personalised approach to gastric variceal bleeding management is reached.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2025 Springer Healthcare Ltd, part of Springer Nature
Lancet Gastroenterol Hepatol 2025; doi:10.1016/S2468-1253(25)00156-6
Lancet Gastroenterol Hepatol 2025; doi:10.1016/S2468-1253(25)00162-1