867: Serum Exosomal Proteomics for Characterizing and Predicting Sepsis-Associated Ards and Aki

Introduction: Sepsis is a life-threatening condition characterized by dysregulated host responses to infection, often accompanied by complications such as acute kidney injury (AKI) and acute respiratory distress syndrome (ARDS), which markedly increase mortality and treatment complexity. Early identification of these complications is critical for improving patient outcomes. Methods: This study included 96 patients with sepsis (excluding ARDS and AKI), 46 with sepsis-associated ARDS, 48 with sepsis-associated AKI, and 48 healthy controls. Serum exosomes were isolated and analyzed using high-throughput proteomic sequencing. Differentially expressed proteins (DEPs) were identified through comparative analyses among groups. Gene Ontology (GO) and KEGG pathway enrichment analyses were performed to elucidate the biological functions of DEPs. Common DEPs between the ARDS and AKI comparison groups were selected and used to construct a random forest–based predictive model with five-fold cross-validation. Results: Distinct sets of DEPs were identified between groups. Inflammatory pathways were predominantly enriched in the sepsis-associated ARDS group, while metabolic processes were highlighted in the sepsis-associated AKI group. A 21-protein panel, including ADAMTS2, FLT4, and GALNT7, was selected for model construction. The predictive model demonstrated good discrimination performance: the area under the ROC curve (AUC) was 0.77 for sepsis vs. complications, 0.68 for sepsis vs. healthy controls, 0.79 for sepsis vs. AKI, and 0.72 for sepsis vs. ARDS. Conclusions: This study reveals distinct exosomal proteomic signatures in sepsis-associated ARDS and AKI and establishes a robust predictive model for early diagnosis. These findings may contribute to improved risk stratification and clinical management of sepsis-related organ dysfunction.