Long-term assessment of bilirubin and transaminase trends in pediatric intestinal failure patients during the era of hepatoprotective parenteral nutrition

Abstract

PurposeThis study aimed to characterize the relationship between hepatoprotective parenteral nutrition (PN) dependence and long-term serum liver tests in children with intestinal failure (IF). MethodsA retrospective review was performed of children with severe IF (> 90 consecutive days of PN) who were followed from 2012 to 2019 at a multidisciplinary intestinal rehabilitation program. Patients were stratified into three groups based on level of PN dependence at most recent follow up: EN (achieved enteral autonomy), mixed (parenteral and enteral nutrition), and PN (> 75% of caloric intake from PN). PN at any point for this cohort was hepatoprotective, defined as soy-based lipids < 1.5 g/kg/day, combination (soy, medium chain fatty acid, olive and fish oil) lipid emulsion, or fish oil-based lipid emulsion. Kaplan-Meier analysis and a generalized estimating equation (GEE) model were utilized to estimate time to normalization and trends, respectively, of two serum markers of liver health: direct bilirubin (DB) and alanine aminotransferase (ALT). ResultsThe study included 123 patients (67 EN, 32 mixed, 24 PN). Median follow up time was 4 years. Based on the Kaplan Meier curve, 100% of EN and mixed group patients achieved normal DB levels by 3 years, while 32% of the PN group had elevated DB levels (Fig. 1). At 5 years, 16% of EN patients had elevated ALT levels compared to 73% of PN patients (p < 0.001, Fig. 2). The PN group's ALT levels were 1.76-fold above normal at 3 years (95%CI 1.48–2.03) and 1.65-fold above normal at 5 years (95%CI 1.33–1.97, Fig. 3). ConclusionsWhile serum bilirubin levels tend to normalize, long-term PN dependence in the era of hepatoprotective PN is associated with a persistent transaminase elevation in an overwhelming majority of patients. These data support continued vigilant monitoring of liver health in children with intestinal failure. Level of evidenceIII