A novel score to predict outcome after transjugular intrahepatic portosystemic shunt for refractory ascites (TAS score)

Background:

Treatment of refractory ascites with transjugular intrahepatic portosystemic shunt (TIPS) is well established and associated with increased survival and reduced hospitalization.

Severe TIPS-associated complications, e.g. hepatic encephalopathy, demand for optimal selection of candidates. We aimed to identify routine laboratory features predictive for mortality and liver transplantation after TIPS and to develop a risk-stratifying point-score.

Methods:

We retrospectively analyzed 82 TIPS-placements with the indication “refractory ascites” at the University Hospital of Graz. Laboratory parameters measured 1 – 48h prior to TIPS were considered. Score variables were determined by stepwise multivariate Cox-regression. Cut-off-values were calculated using Youden-Index. For the scoring-points of the variables, the rounded hazard ratios were utilized.

Results:

We identified three parameters as independent predictors of mortality and liver transplantation within the first year after TIPS; Bilirubin > 1,4 mg/dl (hazard ratio (HR) 5.7; p < 0.001), urea > 39 mg/dl (HR 3.1; p = 0.04) and aspartate-aminotransferase (AST) > 34 U/L (HR 3.2; p = 0.05). We suggest the TIPS-Ascites-Score (TAS) with 0 – 4 points: bilirubin > 1,4 mg/dl implies +2 points, urea > 39 mg/dl and AST > 34 U/L +1 point, respectively.

In our cohort, a lower TAS was associated with significantly better 1-year-survival (p = 0,001) as well as 1-year-transplantation-free-survival (p < 0.001). In the highest possible TAS-group of 4 (n = 10) after 118 days every patient was either transplanted (n = 5) and/or had died (n = 7). Serum urea levels of > 39 mg/dl were predictive for transplantation/mortality (p = 0.02) as well as for development of hepatic encephalopathy (p = 0.04) on its own within the first year.

Discussion:

With the TAS-score, we developed a decision-making tool that helps identifying a subgroup that may not benefit from TIPS. Moreover, we could show that serum urea is a valuable prognostic marker in TIPS candidates, possibly because it summarizes disturbances either in renal function or amino-acid/ammonia metabolism, all of which are relevant for prognosis in cirrhosis. The score needs to be validated in external cohorts.