Abstract
This study aimed to evaluate the diagnostic potential of hemocyte parameters and procalcitonin (PCT) in detecting bloodstream infections (BSI) in neonates and explore the contribution of hemocyte activation-related genes to pediatric sepsis through bioinformatics analysis. A cohort of 419 neonates, categorized as BSI (positive blood culture) and control, underwent comparative analysis of PCT and hemocyte parameters. A predictive model for neonatal BSI was established, demonstrating an impressive area under the receiver ROC curve of 0.968 with remarkable sensitivity (92%) and specificity (87.3%). Hemocyte parameters, including lymphocyte and neutrophil percentages, platelet distribution width (PDW), platelet to lymphocyte ratio (PLR), and PCT, emerged as independent predictors of neonatal BSI. Furthermore, bioinformatics analysis utilizing Gene Expression Omnibus (GEO) datasets yielded significant insights. Differential gene expression (DEGs), gene ontology (GO), pathway enrichment, gene set enrichment analysis (GSEA), and protein-protein interaction (PPI) networks were explored. The differentially expressed genes and hub genes were notably enriched in the activation of neutrophils, lymphocytes, and platelets. Notably, elevated expression levels of SPI1, TYROBP, and FCER1G were observed in pediatric sepsis or septic shock, with positive correlations between SPI1, FCER1G, and TYROBP. In summary, the combination of lymphocyte, PDW, PLR, and PCT effectively diagnosed neonatal BSI. Bioinformatics analysis underscored the pivotal role of activated hemocytes in diagnosing pediatric sepsis, with SPI1, TYROBP, and FCER1G co-expression influencing the disease's pathophysiology by modulating neutrophil and platelet activity.