Nutritional and clinical predictors of intestinal maturation and feeding tolerance in the preterm infant

Project Summary/Abstract: In 2020, 10% of U.S. infants were born preterm and ~2%, or 60,000 infants, were born very preterm (VPI; <32 weeks PMA). VPI infants are at high risk for of substantial medical complications, including necrotizing enterocolitis (NEC). In VPI, advancing and maintaining nutritional support reduces disease risk and improves neurodevelopmental outcomes; however, up to 25% of preterm infants demonstrate feeding intolerance, which may be benign or may progress to NEC. However, precise measures and clinical tools that reliably differentiate benign, intestinal immaturity from life-threatening symptoms are lacking. Therefore, the overall objective of this application is to establish intestinal host and microbial biomarkers of intestinal function from an existing, longitudinal, prospective cohort of 400 analyzable VPI and to relate those biomarkers to the spectrum of intestinal function, from consistent enteral nutrition tolerance to intermittent intolerance to ischemic injury. For this purpose, we will utilize our novel non-invasive (exfoliated mucosal cell) methodology to simultaneously assess host- microbiome interactions in the VPI gut. Our central hypothesis is that the transgenomic cross-talk between intestinal mucosal cells and the fecal metagenome and metabolome will provide mechanistic insight into the spectrum of clinical presentations ranging from normal gut developmental biology to abnormal pathophysiology. Three specific aims will test our central hypothesis. Aim 1 will annotate the host exfoliated mucosal cell transcriptome and fecal bacterial metagenome and metabolome profiles to identify biomarkers for preterm infants who have consistent tolerance to enteral feeding or who are diagnosed with feeding intolerance. Aim 2 will annotate the host exfoliated mucosal cell transcriptome and fecal bacterial metagenome and metabolome profiles to identify biomarkers for preterm infants who are diagnosed with feeding intolerance compared to those who develop ischemia. Aim 3 will utilize machine learning algorithms to construct putative diet-health outcome driven Artificial Neural Networks (ANNs). Completion of these aims will provide the necessary data to develop predictive algorithms to enable identification of at-risk VPI who would benefit from precision medicine/nutrition guided interventions targeting specific risk factors.