Abstract
BACKGROUND: Multiple biomarkers have been identified by previous studies to diagnose acute kidney injury (AKI). The combination of biomarkers with conventional criteria to define AKI substages in order to identify high-risk patients and improve diagnostic accuracy was recommended. Our study aimed to explore the incidence of AKI substages defined by serum cystatin C (CysC), determine whether AKI substages diagnosed with combined CysC criteria were associated with worse outcomes. METHODS: We prospectively included 2519 pediatric patients (<16 years) undergoing cardiac surgery with cardiopulmonary bypass (CPB) in our cohort between March 2022 and February 2023 in Fuwai Hospital. Demographic and clinical variables were collected. To define AKI substages, Kidney Disease: Improving Global Outcomes AKI definition (based on serum creatinine (SCr) or CysC) was used. The association between AKI exposure and outcomes including length of intensive care unit stay (LOIS), duration of mechanical ventilation (DMV), length of hospital stay (LOHS), and 30-day mortality was assessed. In addition, we determined areas under the receiver operating characteristic (ROC) curve and cutoff value of CysC preoperatively and postoperatively to predict AKI. RESULTS: Five hundred and seven (20.8%) patients developed SCr-AKI, with 337 (13.8%) in stage 1, 77 (3.2%) in stage 2 and 93 (3.8%) in stage 3, respectively. Of the 1925 patients without SCr-AKI, 256 (14.3%) met the definition of sub-AKI. Of the 507 patients with SCr-AKI, 281 (55.4%) patients were defined as AKI substage A, while others (226, 44.6%) were defined as AKI substage B. After adjusting for body surface area, neonates, Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery mortality score ≥ 4, previous sternotomy and CPB time > 120 min, the postoperative LOIS, LOHS, and DMV were prolonged with increasing hospitalization expense (p < .05) in patients with SCr-AKI and/or CysC-AKI. Meanwhile, only the hospitalization expense was increased in patients with SCr-AKI (p < .05) after the same adjustment. The area under curves was 0.691, 0.720, and 0.817 respectively, in ROC curves of preoperative, relative variation, or postoperative serum CysC. DeLong's test showed that postoperative serum CysC might have better diagnostic performance characteristics than preoperative or relative variation of CysC (p < .001), with cutoff point at 1.29 mg/dL (specificity, 0.77; sensitivity, 0.71). CONCLUSIONS: Our analysis indicates defining AKI with both CysC and SCr might more significantly affect clinical outcome associations in pediatric patients undergoing cardiac surgery. Moreover, the serum CysC cutoff of 1.29 mg/dL postoperatively is a valuable threshold for AKI risk assessment to define AKI subtypes.