Identification of biomarkers associated with acute lung injury after cardiopulmonary bypass by integrative transcriptomic analysis and clinical validation

Pediatric cardiopulmonary bypass (CPB) induces profound systemic inflammation and immune dysregulation that contribute to postoperative complications. However, conserved transcriptional programs underlying CPB-associated immune responses remain incompletely characterized. This study aimed to define reproducible immune transcriptional signatures associated with postoperative acute lung injury in pediatric CPB. Three complementary transcriptomic datasets, including whole-blood bulk RNA sequencing, peripheral blood mononuclear cell single-cell RNA sequencing, and neutrophil bulk RNA sequencing, were integrated to characterize immune responses. Differential expression, functional enrichment, co-expression network analysis, protein-protein interaction analysis, and machine-learning-based feature selection were applied to identify robust genes. Key candidates were validated in pediatric cohorts using quantitative real-time PCR. Cross-dataset integration revealed consistent activation of innate immune and myeloid programs after CPB. Integrated network and machine-learning analyses converged on a three-gene signature comprising CD163 , IL10 , and PPARG . Clinical validation demonstrated significant postoperative upregulation of all three genes, which correlated with the oxygenation index, indicating an association with postoperative ALI severity. This integrative transcriptomic analysis identifies a reproducible three-gene immune signature associated with postoperative acute lung injury following pediatric CPB. These findings provide insight into CPB-induced immune dysregulation and support the potential relevance of compact immune-related gene signatures for early postoperative risk stratification.