Article Text
Abstract
Introduction International guidelines recommend consideration of early-TIPSS for high-risk patients (Childs-Pugh grade B and C) presenting with oesophageal variceal bleeding (VB), following initial haemostasis with endotherapy and drugs. However, this is based on a single randomised controlled trial (RCT).
Methods An open-label, parallel-group, RCT was performed at 2 experienced UK centres. Cirrhotic patients with a Childs Pugh (CP) score ≥8 presenting to either centre with VB were considered for inclusion in the study. Following haemostasis with endoscopic band ligation (EBL) and vasoactive drugs, consenting participants were randomised to either continued EBL ± non-selective beta-blockers (standard care) or early-TIPSS (ie within 72hrs). All patients were followed up for 1 year, or until death. The primary endpoint was survival at 1 year. Secondary endpoints included early and late re-bleeding (<6 weeks and 6 weeks - 1 year). The study was completed between April 2012 – January 2018.
Results 216 patients were screened with 58 eligible participants ultimately included (29 per group). EBL and early-TIPSS were well matched for age (53.9 vs 49.4); CP score (10.1 vs 9.8); and MELD score (16.8 vs 16.4) respectively.
On intention to treat, 1-year survival rates were 75.8% in the EBL group versus 79.3% in the early-TIPSS group (p=0.79) (see figure 1). 1 patient in the EBL group received TIPSS as rescue therapy, but later died. 4 patients who were randomised to early-TIPSS did not receive this intervention due to time delays and therefore underwent standard care. However, on per-protocol analysis, no survival benefit was observed between groups (p=0.85). In patients with Childs-Pugh C cirrhosis (n=33) no difference in survival was seen between groups (p=0.81). At 6 weeks, rebleeding occurred in 2 patients in both the EBL and early-TIPSS groups. At 1-year, rebleeding occurred in 8 patients in the EBL group compared with 6 patients in the early-TIPSS group. No significant differences in rates of rebleeding were seen at the specified time points. Additionally, no significant differences were observed between the two treatment groups with respect to serious adverse events.
Conclusions Contrary to the previous study, our findings do not support the use of early-TIPSS in the reduction of mortality or rebleeding in patients with cirrhosis who are hospitalised for acute variceal bleeding. Given the challenges of resourcing the widespread provision of early-TIPSS, further multi-centred studies with larger patient numbers are required to resolve this important issue.