Derivation and Validation of an Optimal Neutrophil Gelatinase-Associated Lipocalin Cutoff to Predict Stage 2/3 Acute Kidney Injury (AKI) in Critically Ill Children

Abstract

IntroductionAcute kidney injury (AKI) defined by changes in serum creatinine (SCr), or oliguria is associated with increased morbidity and mortality in critically ill children. We derived and validated a clinical cutoff value for urine neutrophil gelatinase-associated lipocalin (NGAL), in a prospective multicenter study of critically ill children. We report the clinical performance of urine NGAL (uNGAL) to aid in pediatric AKI risk assessment. MethodsEligible subjects were aged > 90 days to < 22 years, admitted to an intensive care unit (ICU), and had one or more of the following: mechanical ventilation, vasoactive medication administration, solid organ or bone marrow transplantation, or hypotension within 24 hours of admission. uNGAL assessed within 24 hours of admission. The primary outcome was SCr-based Stage 2/3 AKI presence at 48-72 hours. ResultsTwenty-five (12.3%) derivation study patients had Stage 2/3 AKI at 48-72 hours. uNGAL concentration of 125 ng/mL was the optimal cutoff. Forty-seven (9.1%) validation study patients had Stage 2/3 AKI at 48-72 hours. The AUC-ROC for uNGAL performance was 0.83 (95%CI 0.77-0.90). Performance characteristics were sensitivity 72.3% (95%CI 57.4-84.4%), specificity 86.3% (95%CI 82.8-89.3%), positive predictive value 34.7% (95%CI 28.5-41.5%), and negative predictive value 96.9% (95%CI 95.1-98.0%). ConclusionsThese prospective, pediatric, multicenter studies demonstrate uNGAL in the first 24 hours performs very well to predict KDIGO Stage 2/3 AKI at 48-72 hours into an ICU course. We suggest that a urine NGAL cut point of 125 ng/mL can aid in the risk assessment for Stage 2/3 AKI persistence or development.