ABSTRACT
INTRODUCTION AND OBJECTIVES: TIPS placement is an effective, possibly life-saving, treatment for complications of portal hypertension. The pressure shift induced by the stent can lead to cardiac decompensation (CD). We investigated the incidence of CD, possible variables associated with CD and the validity of the Toulouse algorithm for risk prediction of CD post-TIPS. PATIENTS AND METHODS: A total of 106 patients receiving TIPS for variceal bleeding (VB, 41.5%) or refractory ascites (RA, 58.5%) with available echocardiography and NT-proBNP results were included and retrospectively reviewed. Development of CD between time of TIPS placement and occurrence of liver transplantation, death or loss-to-follow-up was recorded. Competing risk regression analysis was performed to assess which baseline variables predicted occurrence of CD post-TIPS. RESULTS: A total of 12 patients (11.3%) developed CD after a median of 11.5 days (IQR 4 to 56.5) post-TIPS. Multivariate regression showed age (HR 1.06, p = 0.019), albumin (HR 1.10, p = 0.009) and NT-proBNP (HR 1.00, p = 0.023) at baseline predicted CD in the RA group. No clear predictors were found in those receiving TIPS for VB. Correspondingly, the Toulouse algorithm successfully identified patients at risk for CD, however only in the RA population (zero risk 0% vs. low risk 12.5% vs. high risk 35.3% with CD; p = 0.003). CONCLUSIONS: CD is not an infrequent complication post-TIPS occurring in 1/10 patients. The Toulouse algorithm can identify patients at risk of CD, though only in patients receiving TIPS for RA. Allocation to the high-risk category warrants close monitoring but should not preclude TIPS placement.