Abstract
BACKGROUND: Cardiac surgery-associated acute kidney injury (CSA-AKI) is commonly observed after cardiac surgery and has been shown to be associated with increased morbidity and mortality. This study was conducted using the Kidney Disease Improving Global Outcomes (KDIGO) criteria to analyze potential perioperative biomarkers of CSA-AKI. METHODS: Blood was collected from patients intraoperatively on entry into the pericardium and at 4, 12, 24, and 48 hours postoperatively. Repeated-measures, mixed-model analysis was conducted to determine which cytokines and/or chemokines were associated with postoperative CSA-AKI. LASSO regression and random forest modeling were used for variable selection and incorporation into a multivariable regression model. RESULTS: There were no demographic or preoperative differences between patients with CSA-AKI and patients without CSA-AKI except for preoperative diabetes status, hemoglobin concentration, and CKD status. Additionally, there were no significant differences in preoperative medications between the 2 groups. Ten of the 40 biomarkers were statistically significant (P < .05) for the between-group main effect after repeated measures analysis: myoglobin, growth/differentiation factor 15 (GDF-15), neutrophil gelatinase-associated lipocalin (NGAL), haptoglobin, tumor necrosis factor alpha (TNFα), monocyte chemoattractant protein 1 (MCP-1), interleukin (IL)-1RA, IL-8, IL-6, and C-reactive protein. Multivariable stepwise regression showed the earliest independent predictors of postoperative AKI were 4-hour myoglobin (adjusted odds ratio [aOR], 1.61; 95% confidence interval [CI], 1.12-3.12; P = .036), 4-hour IL-1RA (aOR, 1.001; 95% CI, 1.000-1.001; P = .48), and 4-hour haptoglobin (aOR, 1.07; 95% CI, 1.03-1.18; P = .001). CONCLUSIONS: Multiple cytokines were significantly elevated between the CSA-AKI group and the CSA-non-AKI group. Myoglobin, haptoglobin, and IL-1RA are potential blood biomarkers for AKI after cardiac surgery. Further research is needed to investigate the roles of these biomarkers and their associations with CSA-AKI.