Background: The aim of this study was to determine the prognostic relevance of the portal pressure gradient (PPG) before and after transjugular intrahepatic portosystemic stent shunt (TIPS) insertion in patients with liver cirrhosis and recurrent oesophageal variceal bleeding.
Methods: 118 cirrhotic patients (Child A/B/C, 41/56/21; Child score, 7.7+/-2.0; baseline PPG, 21.8+/-4.7 mmHg) underwent TIPS for the prevention of variceal rebleeding. A multivariate logistic regression analysis was applied to identify the independent determinants of rebleeding and survival. The estimated rebleeding rate and the estimated survival were compared by log-rank testing.
Results: TIPS insertion reduced the PPG by 53.2+/-17.7%. During follow-up 21 patients suffered significant rebleeding (17.8%); bleeding-related mortality was 3.4% (four patients). The median survival [95% confidence intervals (CI)] was 48.2 (39.8; 60.8) months. The multivariate Cox model identified creatinine as the only independent predictor of survival, and the initial decrease of the PPG after TIPS as the only independent predictor of rebleeding. PPG before TIPS (21.8+/-4.7 mmHg) and the gradient at the time of rebleeding (22.0+/-2.9 mmHg) did not differ significantly. Patients with an initial decrease of the PPG after TIPS <30% were at the highest risk for rebleeding. Patients with an initial decrease of the PPG >60% rarely suffered from rebleeding.
Conclusions: The initial decrease in the PPG after TIPS is a predictor for the risk of rebleeding but not for survival after TIPS. For that reason, in patients undergoing TIPS placement for the prevention of recurrent bleeding from oesophageal varices, an initial reduction of the PPG of 30-50% should be attempted.