A simple model for predicting survival in cirrhotic patients undergoing portosystemic shunt embolization

초록

Background and aims: Portosystemic shunt embolization (PSSE) is an effective treatment for hepatic encephalopathy and gastric varix in patients with portosystemic shunt. However, some of the patients rapidly progress to hepatorenal syndrome and hepatic failure after PSSE. The aim of this study was to develop a prognostic model for predicting survival in patients treated with PSSE. Method: We included 188 patients with portosystemic shunt who received PSSE for gastric varix or recurrent hepatic encephalopathy to develop a prediction model for 1-year mortality using a Cox proportional-hazard model incorporating significant prognostic factors, which was validated in a separate cohort of 184 patients. Results: Using multivariable analysis, baseline serum levels of total bilirubin, albumin, and international normalized ratio (INR) were the independent prognostic factors that were significantly associated with the 1-year survival of the patients. An Albumin-Bilirubin-INR (ABI) score was developed assigning 1 point for ≥1.5 mg/dL of total bilirubin, <3.0 g/dL of albumin, and ≥1.5 for INR, respectively. Time-dependent area under the curve of the ABI score for predicting 6-month and 1-year survival were 0.851 (95% confidence interval [CI], 0.778–0.923) and 0.857 (95% CI, 0.793–0.921) in the development cohort and 0.777 (95% CI, 0.664–0.829) and 0.727 (95% CI, 0.624–0.829) in the validation cohort, which was comparable to the MELD score and the Child-Pugh score, indicating excellent discrimination performance. The ABI score showed a batter calibration performance than other scores, especially in patients at high-risk. Conclusion: The novel ABI score, which is easier to calculate than MELD or Child-Pugh scores, may help identify a proper indication for PSSE in cirrhotic patients with portosystemic shunt.