Abstract Title: Stool Sphingomyelin Analysis to Predict Parenteral Nutrition-Associated Cholestasis in Very Low Birth Weight Neonates

Authors: Esrig, A., Shenoy, A., Marrs, A., Kolling, G., Moore, S., Papin, J., Sullivan, B.

Background: Parenteral Nutrition-Associated Cholestasis (PNAC), diagnosed by elevated serum conjugated bilirubin, occurs in preterm infants with very low birthweight (VLBW, <1500g) requiring prolonged parenteral nutrition. Previous untargeted fecal metabolomics from our group identified 12 sphingomyelins as potential PNAC biomarkers. Successful identification of a stool biomarker for PNAC risk could reduce reliance on blood tests and promote preventative interventions to mitigate liver damage.

Methods We enrolled VLBW infants admitted to a level IV NICU with written consent from a parent. Stool samples were collected biweekly, and clinical data were recorded in a REDCap database. We analyzed stool samples using mass spectrometry for targeted fecal sphingomyelin quantification. Conjugated bilirubin (C. Bili) levels were obtained at the discretion of the clinical team. We identified study participants with PNAC (C. Bili > 1.0 mg/dl) and matched each PNAC case with two controls (enrolled VLBW infants without PNAC) of similar gestational age, birthweight, and sex. We used summary statistics and non-parametric statistical tests to compare clinical variables and sphingomyelin concentrations between cases and controls. The main objective is to validate and quantify fecal sphingomyelins as PNAC biomarkers through targeted metabolomics. Results

We enrolled 55 VLBW infants, of whom 3 developed PNAC. We matched two controls with these 3 cases and analyzed their biweekly stool samples for sphingomyelin concentrations. Demographics were similar for cases and controls, and cases had more days on TPN Peak sphingomyelin concentrations were higher for cases than controls, but the timing of the peaks relative to PNAC timing varied. Sphingomyelin concentrations were weakly positively correlated with TPN days.

Conclusion We identified fecal sphingomyelins using targeted mass spectrometry analysis of stool samples from VLBW preterm infants at risk for PNAC. Our preliminary results show variable timing of sphingomyelin elevations relative to PNAC diagnosis and age in controls without PNAC. PNAC is a rare disease requiring large cohorts to determine the utility of fecal sphingomyelin to predict PNAC risk. Successful identification of fecal biomarkers that predict cholestasis before the onset of PNAC could reduce blood testing & liver damage