Introduction: Serum creatinine is a suboptimal marker for acute kidney injury (AKI), which occurs in 25% of children admitted to the pediatric intensive care unit (PICU). PICU admission urinary neutrophil gelatinase-associated lipocalin (NGAL) concentrations >125 ng/mL predict day 3 severe AKI an (area under the receiver operating characteristic curve (AUROC) = 0.83). We evaluated whether other tubular injury biomarkers (specifically admission urinary cystatin C and urinary C–C motif chemokine ligand 14 (CCL-14)) improve the predictive performance for persistent AKI in critically ill children. Methods: We conducted a prospective, opportunistic pilot study of PICU patients with prognostic enrichment. Patients admitted between October and November 2024 who were at increased risk for AKI (with respiratory or cardiovascular failure) were consecutively enrolled if they had indwelling catheters. Urine samples were collected within 24 hours of admission. Admission AKI was defined as urinary NGAL >125 ng/mL (day 1) and persistent AKI according to the KDIGO serum creatinine criteria (day 3). Urinary concentrations of NGAL, cystatin C, and CCL14 were measured using ELISA. Results: 45 patients were enrolled, with a median age of 117 months (IQR 57–172). Twenty-three (51%) were female, and the majority were White 34 (79%). The median weight was 29 kg (IQR 18–56). The most common indication for PICU admission was respiratory failure 12 (27%). Fourteen patients (31%) developed AKI by day 3. Urinary NGAL, CCL-14, and cystatin C concentrations differed significantly between patients with and without AKI. When combined, these three urinary biomarkers achieved an AUROC of 0.94 (95% CI, 0.87–1.00) for predicting day 3 AKI. Conclusions: he addition of urinary cystatin C and CCL14 to NGAL improved predictive performance for AKI compared to historical controls in high-risk critically ill children. This enhanced biomarker panel performance could enable timely interventions to potentially mitigate long-term adverse outcomes associated with AKI. Multicenter studies are needed to validate these findings.
Thadani et al. (Sun,) studied this question.