Immunoinflammatory biomarkers for sepsis-associated acute kidney injury: a multicenter prospective cohort study

BACKGROUND: Immune dysregulation is central to sepsis-related organ injury, yet standard clinical indicators show limited predictive value for sepsis-associated acute kidney injury (SA-AKI) in older adults(aged ≥ 65 years). This study evaluated whether immuno-inflammatory biomarkers improve early prediction of SA-AKI. METHODS: In this prospective multicenter cohort of 627 older adults with sepsis admitted to ICUs in five Beijing tertiary hospitals, the overall incidence of SA-AKI was 43.1% (270/627). Clinical variables and immune-inflammatory markers collected within 24 h were used to construct a clinical model (Model 1) and an integrated model (Model 2, combining clinical variables with immuno-inflammatory biomarkers). Model performance was examined using ten-fold cross-validation and an internal test set. RESULTS: Six clinical predictors were included in Model 1, while Model 2 additionally incorporated C3, C4, CD4% T cells, and NK cells. In the testing set, Model 2 showed markedly better discrimination than Model 1 (AUC 0.878 vs. 0.717; p < 0.001) with higher sensitivity and specificity. Immune-marker inclusion significantly improved risk reclassification (NRI 86.1%; p = 0.002). CONCLUSIONS: Adding immuno-inflammatory biomarkers substantially enhanced early prediction of SA-AKI in older adults, underscoring the key role of immune imbalance and supporting targeted immune monitoring for individualized risk assessment.