High MELD Score Has No Effect On the Magnitude of the Mortality Increase Observed With Transjugular Intrahepatic Portosystemic Shunt (TIPS) Placement.

Abstract

Purpose: Both MELD ≥18 and a need for transjugular intrahepatic portosystemic shunt (TIPS) placement identify patients at increased mortality risk. We evaluated if mortality risk is independently increased by the placement of a TIPS, in patients with both high MELD and need for TIPS. Materials and Methods: Retrospective single center study of patients receiving an elective TIPS for refractory ascites, between 1999 and 2011. Study Group: 88 patients with refractory ascites who underwent an elective TIPS. Control group: 109 matched controls with refractory ascites who did not have a TIPS. We did a Cox multivariable analysis adjusting for etiology, age and MELD score, dichotomized to MELD <18 and MELD ≥18. We determined if there was an interaction between the performance of a TIPS and MELD score on mortality in the first 6 months after TIPS and ≥6 months after TIPS, to establish the impact of TIPS on mortality. Results: The performance of a TIPS had no impact on the magnitude of increased risk of death in patients with MELD <18 or ≥18, either within 6 months of the TIPS (p=0.765) or after 6 months (p=0.307). We found no significant interaction between TIPS and MELD ≥18 during either time period. As expected, we did show that both MELD ≥18 (HR 10.2, CI 4.7-22.2, p<0.0001) and the need for a TIPS (HR 6.6, CI 3.0-14.4, p=0.001) were associated with increased mortality risk after adjusting for age and etiology of cirrhosis. Conclusion: It has been assumed in the literature that placement of a TIPS increases mortality in high MELD patients, often excluding these patients from consideration. Our study shows that mortality is increased in both patients with high MELD (≥18) (HR 10.2) and those in need of a TIPS (HR 6.6), but that there was no significant interaction between need for TIPS and high MELD score on mortality, either in the short term (<6 months, p=0.765) or long term (≥6 months, p=0.307). This supports that the higher mortality seen in patients with high MELD scores undergoing TIPS is due to the underlying disease, not the performance of the TIPS itself. High MELD score (≥18) had no effect on the magnitude of the mortality increase observed with TIPS placement, suggesting that excluding these patients from this potentially beneficial procedure warrants re-examination with controlled studies that include higher MELD patients.