Early-onset sepsis (EOS) due to invasive bacterial infection is a leading cause of morbidity in infants and disproportionately affects those born preterm. Accurate diagnosis of EOS remains inadequate. The low sensitivity and delayed time to culture results combined with the poor predictive values of available laboratory markers leads to our inability to rule-out EOS. Thus, infants are exposed to extended duration broad-spectrum empiric antibiotics, which have serious associated adverse consequences such as necrotizing enterocolitis, antibiotic resistant infections, and an altered microbiome. The proposed research will apply proteomics to identify gestational age-specific umbilical cord blood markers toward development of a proposed diagnostic tool with strong negative risk prediction for EOS. Cord blood is promising to identify EOS as it reflects the intrauterine environment where infection originates in preterm birth. In previous work, we found that inflammatory proteins serum amyloid A, C-reactive protein and haptoglobin are substantially elevated in cord blood of preterm infants with culture-confirmed EOS. These data inform our hypothesis that a combination of cord blood proteins provides a signature to differentiate impending EOS and uninfected states. Archived cord blood from infants in an existing longitudinal cohort will be used in a nested case-control design.
Aim 1 will measure the cord blood proteome of uninfected infants (controls) across gestational ages using an untargeted proteomics approach to characterize the developmental spectrum.
Aim 2 (a) will then determine the cord blood proteome signatures of infants with confirmed EOS (cEOS) and culture-negative presumed sepsis (PS). cEOS will be compared to matched controls to identify differentially expressed proteins and candidate markers. PS proteomes will be analyzed to help delineate true infection versus inflammatory subtypes in this heterogeneous group. Bioinformatics pathway analysis may provide novel insights into fetal immune response.
Aim 2 (b) will determine the best combination of markers to exclude EOS through machine learning decision analysis of proteomics data. Paired placental proteomics in Aim 2(c) will explore origins of fetal inflammation.
Aim 3 (a) will quantify candidate proteins by an orthogonal immunoassay method. In a validation set of new subjects, Aim 3(b) will measure candidate markers by targeted proteomics and test application of the proposed diagnostic tool. With formal didactics and a mentorship team at Northwestern University with expertise in prematurity, sepsis, biomarker discovery, biostatistics, and decision modeling led by Patrick Seed MD Ph.D., the candidate will gain advanced translational research experience and skills in perinatal research study design, advanced immunology, proteomics, and bioinformatics to further her research agenda. The proposed research and career development will be fundamental towards achieving the candidate?s career goals to: 1) lead a clinical-translational research program on the pathophysiology and host responses of perinatal infection and 2) bring new, reliable sepsis diagnostics to clinical practice, to improve neonatal outcomes through targeted treatment of EOS and antibiotic stewardship.
Invasive bacterial infections in newborns carry a high risk of complications and death, yet current diagnostic methods are inaccurate, resulting in the harmful overuse of antibiotics. The proposed work will develop a multi- factor diagnostic model for early-onset neonatal sepsis using umbilical cord blood proteins as biomarkers and enhance our understanding of fetal immune response. These are critical steps toward improving infant health outcomes through both targeted treatment of infection and reduced unnecessary early-life antibiotic exposures.