Early Posthepatoportoenterostomy Predictors of Native Liver Survival in Biliary Atresia

• Published in: Journal of pediatric gastroenterology and nutrition Vol. 64; no. 2; pp. 203 - 209
• Main Authors: Nightingale, Scott, Stormon, Michael O., O'Loughlin, Edward V., Shun, Albert, Thomas, Gordon, Benchimol, Eric I., Day, Andrew S., Adams, Susan, Shi, Edward, Ooi, Chee Y., Kamath, Binita M., Fecteau, Annie, Langer, Jacob C., Roberts, Eve A., Ling, Simon C., Ng, Vicky L.
• Format: Journal Article
• Language: English
• Published: United States by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology 01.02.2017
• Subjects: Biliary Atresia - complications; Biliary Atresia - surgery; Child, Preschool; Clinical Decision-Making; Decision Support Techniques; End Stage Liver Disease - diagnosis; End Stage Liver Disease - etiology; End Stage Liver Disease - surgery; Female; Follow-Up Studies; Health Status Indicators; Humans; Infant; Infant, Newborn; Kasai procedure; liver transplantation; Liver Transplantation - statistics & numerical data; Logistic Models; Male; neonatal cholestasis; Portoenterostomy, Hepatic; Prognosis; Retrospective Studies; ROC Curve; Treatment Outcome
• ISSN: 0277-2116; 1536-4801
• DOI: 10.1097/MPG.0000000000001289

ABSTRACT Objectives: Most infants with biliary atresia (BA) require liver transplantation (LT) after hepatoportoenterostomy (HPE), including those who initially clear jaundice. The aim of the present study was to identify clinical and routine laboratory factors in infants with BA post‐HPE that predict native liver survival at 2 years. Methods: A retrospective cohort study was conducted in 217 patients with BA undergoing HPE in Sydney, Australia and Toronto, Canada between January 1986 and July 2009. Univariate and multivariate logistic regression using backwards‐stepwise elimination identified variables at 3 months after HPE most associated with 2‐year native liver survival. Results: Significant variables (P < 0.05) on univariate analysis included serum total bilirubin (TB) and albumin at 3 months post‐HPE, bridging fibrosis or cirrhosis on initial liver biopsy, ascites of <3 months post‐HPE, type 3 BA anatomy, age at HPE of >45 days, change in length z scores within 3 months of HPE, and center. On multivariate analysis, TB (P < 0.0001) and albumin (P = 0.02) at 3 months post‐HPE, and center (P = 0.0003) were independently associated with native liver survival. Receiver operating characteristic analysis revealed an optimal cut‐off value of TB <74 μmol/L (4.3 mg/dL; area under the receiver operating characteristic curve 0.8990) and serum albumin level >35 g/L (3.5 mg/dL; area under the receiver operating characteristic curve 0.7633) to predict 2‐year native liver survival. TB and albumin levels 3 months post‐HPE defined 3 groups (1: TB ⩽74 μmol/L, albumin >35 g/L; 2: TB ⩽74 μmol/L, albumin ⩽35 g/L; 3: TB >74 μmol/L) with distinct short‐ and long‐term native liver survival rates (log‐rank P < 0.001). Length z scores 3 months post‐HPE were poorer for group 2 than group 1 (−0.91 vs −0.30, P = 0.0217) with similar rates of coagulopathy. Conclusions: Serum TB and albumin levels 3 months post‐HPE independently predicted native liver survival in BA when controlling for center. Serum albumin level <35 g/L in infants with BA who were no longer jaundiced at 3 months post‐HPE was a poor prognostic indicator. Poorer linear growth and absence of significant coagulopathy suggest a role for early aggressive nutritional therapy in this group.