Potential Role of Neutrophil Gelatinase-Associated Lipocalin in Identifying Critically Ill Patients With Acute Kidney Injury Stage 2–3 Who Subsequently Require Renal Replacement Therapy Khajohn Tiranathanagul, 1,2 Sukgasem Amornsuntorn, 3 Yingyos Avihingsanon, 1 Nattachai Srisawat, 1,2 Paweena Susantitaphong, 1 Kearkiat Praditpornsilpa, 1 Kriang Tungsanga, 1 and Somchai Eiam-Ong 1 1 Division of Nephrology, Department of Medicine, 2 Excellence Center for Critical Care Nephrology, King Chulalongkorn Memorial Hopital, Thai Red Cross Society and Faculty of Medicine, Chulalongkorn University, Bangkok, and 3 Department of Medicine, Prachuap Khiri Khan Hospital, Prachuap Khiri Khan, Thailand Abstract: Delayed initiation of renal replacement therapy (RRT) in critically ill acute kidney injury (AKI) patients results in high mortality while too early RRT causes unnec- essary risks of the treatment. Current traditional indications cannot clearly identify the appropriate time for initiating RRT.This prospective cohort study was conducted to deter- mine the accuracy of using plasma neutrophil gelatinase- associated lipocalin (pNGAL) and urine NGAL (uNGAL) in early identifying of the AKI patients who subsequently required RRT. Forty-seven critically ill patients with AKI stage 2–3 who did not reach the traditional indications for RRT were enrolled in this study.The pNGAL, uNGAL, and other parameters were determined in each patient. The primary end point was RRT initiation according to the traditional indications within 3 days. The mean age of the patients was 63.0  18.1 years. pNGAL could predict subsequent RRT requirements with area under ROC 0.813 (P < 0.001, 95%CI 0.66–0.90). The cut-off point of 960 ng/mL provided sensitivity and specificity of 72.2 and 89.6%, respectively, and positive and negative predictive values of 81.25% and 83.8%, respectively. The uNGAL provided slightly lower significance of statistical parameters. The combination of pNGAL level of 960 ng/mL and APACHE II score of 20 improved statistical values. In con- clusion, pNGAL is an excellent early biomarker for RRT initiation in critically ill patients with AKI stage 2–3. The pNGAL value of 960 ng/mL, alone or in combination with APACHE II score might be used as the early new indicator for early initiation of RRT in AKI stage 2–3 and this might improve patient survival. Key Words: Acute kidney injury, Biomarker, Neutrophil gelatinase-associated lipocalin, Renal replacement therapy. Critically ill patients who adversely develop severe acute kidney injury (AKI) and require renal replace- ment therapy (RRT) are associated with substan- tially high morbidity and mortality.The mortality rate is more than 50% and increases to 80% in sepsis patients (1,2). This undesired result still persists despite the advancement in RRT modalities. One possible explanation is the proper timing of RRT initiation. Initiation of RRT according to traditional indications might be too late and results in worse outcomes. Emerging data suggest that earlier RRT initiation may attenuate kidney as well as distant non-kidney organ injury from uremia, acidemia, fluid overload, and systemic inflammation (3). On the con- trary, too early initiation of RRT causes unnecessary risks of treatment. Moreover, some early RRT- initiated patients might recover with conservative treatment alone (4). Unfortunately, traditional serum biomarkers or urine output volume cannot clearly define early/late distinction (5). Received June 2012; revised October 2012. Address correspondence and reprint requests to Dr Khajohn Tiranathanagul, Division of Nephrology, Department of Medicine, Faculty of Medicine, King Chulalongkorn Memorial Hospital, Chulalongkorn University, Bangkok 10330, Thailand. Email: khajohn.t@chula.ac.th Presented in part at the American Society of Nephrology Renal Week 2011, held November 8–13, 2011 in Philadelphia, PA, USA. Therapeutic Apheresis and Dialysis 2013; 17(3):332–338 doi: 10.1111/1744-9987.12004 © 2013 The Authors Therapeutic Apheresis and Dialysis © 2013 International Society for Apheresis 332