Abnormal portal venous flow at sonography predicts reduced survival after transjugular intrahepatic portosystemic shunt creation.

PURPOSE

To determine if ultrasound (US) findings of abnormal portal venous flow (APVF) before transjugular intrahepatic portosystemic shunt (TIPS) creation are predictive of increased mortality risk after TIPS creation.

MATERIALS AND METHODS

Retrospective review of 141 patients with US before TIPS creation was performed. APVF was defined by (i) bidirectional flow, (ii) thrombus, and/or (iii) reversed flow. Model for End-stage Liver Disease (MELD) scores were calculated. Kaplan-Meier survival curves and log-rank tests were used to detect survival differences based on the presence of APVF. Multivariate analysis included APVF, MELD, Child-Pugh class, International Normalized Ratio, creatinine level, total bilirubin level, ascites, hepatocellular carcinoma, low serum sodium level, congestive heart failure, and myocardial infarction.

RESULTS

Twenty-six percent of patients (36 of 141) exhibited APVF on US before TIPS creation. Patients with APVF had lower survival rates at 3 and 6 months after TIPS procedures in comparison with patients with normal portal flow (P = .02 at 3 months and P = .04 at 6 months). In patients with MELD scores lower than 18, there was decreased survival based on APVF at 1, 3, and 6 months (P = .04, P = .02, and P = .04, respectively). In patients with MELD scores of 18 or greater, there was a trend for lower survival rates with APVF, but it did not reach statistical significance. Multivariate analysis of patients with MELD scores lower than 18 demonstrated only APVF and low serum sodium levels as independent predictors of outcome, with APVF resulting in a greater than six-fold increased likelihood of mortality.

CONCLUSION

US findings of APVF before TIPS creation are associated with increased mortality risk and may be useful in identifying patients otherwise considered safe candidates based on MELD score alone.