LAUSR

BACKGROUND Transjugular intrahepatic portosystemic shunt (TIPS) placement is an effective intervention for recurrent tense ascites. Some studies show an increased risk of acute on chronic liver failure (ACLF) associated with TIPS placement. It is not clear whether ACLF in this context is a consequence of TIPS or of the pre-existing liver disease. AIM To better understand the risks of TIPS in this challenging setting and to compare them with those of conservative therapy. METHODS Two hundred and fourteen patients undergoing their first TIPS placement for recurrent tense ascites at our tertiary-care center between 2007 and 2017 were identified (TIPS group). Three hundred and ninety-eight patients of the same time interval with liver cirrhosis and recurrent tense ascites not undergoing TIPS placement (No TIPS group) were analyzed as a control group. TIPS indication, diagnosis of recurrent ascites, further diagnoses and clinical findings were obtained from a database search and patient records. The in-hospital mortality and ACLF incidence of both groups were compared using 1:1 propensity score matching and multivariate logistic regressions. RESULTS After propensity score matching, the TIPS and No TIPS groups were comparable in terms of laboratory values and ACLF incidence at hospital admission. There was no detectable difference in mortality (TIPS: 11/214, No TIPS 13/214). During the hospital stay, ACLF occurred more frequently in the TIPS group than in the No TIPS group (TIPS: 70/214, No TIPS: 57/214, P = 0.04). This effect was confined to patients with severely impaired liver function at hospital admission as indicated by a significant interaction term of Child score and TIPS placement in multivariate logistic regression. The TIPS group had a lower ACLF incidence at Child scores < 8 points and a higher ACLF incidence at ≥ 11 points. No significant difference was found between groups in patients with Child scores of 8 to 10 points. CONCLUSION TIPS placement for recurrent tense ascites is associated with an increased rate of ACLF in patients with severely impaired liver function but does not result in higher in-hospital mortality.