Introduction: Sepsis is a global burden affecting 3 million infants and accounts for >400,000 deaths annually. Infection-related inflammation contributes to long-term adverse neurodevelopmental outcomes in infants that survive sepsis. Sepsis is a frequent complication among extremely preterm infants (<28 weeks gestational age), with ~25% developing sepsis during their NICU stay. Early diagnosis of sepsis is critical to minimise inflammation and facilitate antibiotic therapy, but diagnosis is complicated by slow (24-36 hours) and variable diagnostic tests. Consequently, 2/3 of uninfected infants receive antibiotic therapy, which is in preterm infants is associated with unintended adverse events, including death. Thus, there is an urgent and unmet need for accurate and more rapid adjunct diagnostics to reduce the high prevalence of antibiotic use in this vulnerable population. Proteome differences in response to infection can be used to identify protein biomarkers to improve the current diagnostic approach.
Aim: To use targeted and untargeted approaches to explore the plasma proteome for biomarker discovery in infants with sepsis.
Methods: Untargeted label-free quantitative diaPASEF mass-spectrometry based proteomics and targeted immunoassays were used to explore the plasma proteome of preterm human infants born <29 weeks gestational age with and without sepsis. Plasma was processed from peripheral blood samples collected at first suspicion of sepsis.
Results: In very preterm infants with and without sepsis (n=15 and n=39, respectively), >500 plasma proteins were identified. Over 70 differentially expressed proteins are associated with sepsis. A panel of 7 protein biomarkers can identify sepsis with high accuracy.
Conclusion: We identified known and novel proteins that are associated with sepsis. On-going analysis suggests that a subset of proteins may have clinical utility as biomarkers for early diagnosis of sepsis in very preterm infants.