Emergency. 2018; 6 (1): e2 OR I G I N A L RE S E A RC H Urinary Neutrophil Gelatinase-Associated Lipocalin (NGAL) and Cystatin C in Early Detection of Pediatric Acute Kidney Injury; a Diagnostic Accuracy Study Neamatollah Ataei1,2, Sonbol Ameli1,2, Mahmoud Yousefifard3, Alireza Oraei4, Fatemeh Ataei5, Behnaz Bazargani1,2, Arash Abbasi1,2, Mostafa Hosseini6∗ 1. Pediatric Chronic Kidney Disease Research Center, Tehran University of Medical Sciences, Tehran, Iran. 2. Department of Pediatric Nephrology, Children’s Hospital Medical Center, Tehran University of Medical Sciences, Tehran, Iran. 3. Physiology Research Center and Department of Physiology, Faculty of Medicine, Iran University of Medical Sciences, Tehran, Iran. 4. Department of Medicine, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran. 5. Department of Nuclear Medicine, Department of Nuclear Medicine, Valiasr Hospital, Zanjan University of Medical Sciences, Zanjan, Iran. 6. Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran. Received: July 2017; Accepted: November 2017; Published online: 12 January 2018 Abstract: Introduction: There is a controversy regarding accuracy of neutrophil gelatinase-associated lipocalin (NGAL) and Cystatin C in early detection of acute kidney injury (AKI). The present study aimed to compare the diag- nostic value of two biomarkers in this regard. Methods: In the present diagnostic accuracy study, all children between the ages of 1 month to 14 years were entered. Pediatric Risk, Injury Failure, Loss, End-stage renal dis- ease (pRIFLE) criteria was used for identification of children with AKI as the reference test. Blood samples were taken from all patients at baseline and 48 hours after admission to assess serum creatinine and Cystatin C level. In addition, a urine sample was obtained within 6 hours of admission in order to measure NGAL level. In the end, area under the receiving operating characteristics (ROC) curve, sensitivity, and specificity of urine NGAL (uNGAL) and Cystatin C in early detection of AKI were compared. Results: Data from 96 children with the mean age of 27.31±36.24 months were entered (56.25% girls). Area under the ROC curve of uNGAL level in diagnosis of AKI in children was 0.91 (95% CI: 0.80 to 1.00) and area under the ROC of Cystatin C was 0.90 (95% CI: 0.77 to 1.00). Both tests had the same value in diagnosis of AKI (p=0.89). The best cut-off point of uNGAL for diagnosing AKI was 125 mg/L. uNGAL had a sensitivity and specificity of 0.92 (0.62 to 0.99) and 0.69 (0.57 to 0.78), respec- tively. The best cut-off point of serum Cystatin C level was 0.4 mg/L. Cystatin C had a sensitivity and specificity of 0.92 (0.62 to 0.99) and 0.64 (0.52 to 0.74), respectively. Conclusion: The present study showed that uNGAL level has the same value as serum Cystatin C level in early diagnosis of AKI. Keywords: Cystatin C; Lipocalin-2; Acute Kidney Injury; Early Diagnosis; Dimensional Measurement Accuracy © Copyright (2018) Shahid Beheshti University of Medical Sciences Cite this article as: Ataei N, Ameli S, Yousefifard M, Oraei A, Ataei F, Bazargani B, Abbasi A, Hosseini M. Urinary Neutrophil Gelatinase- Associated Lipocalin (NGAL) and Cystatin C in Early Detection of Pediatric Acute Kidney Injury; a Diagnostic Accuracy Study. Emergency. 2018; 6(1): e2. ∗Corresponding Author: Mostafa Hosseini; Department of Epidemiology and Biostatistics School of Public Health, Tehran University of Medical Sci- ences, Poursina Ave, Tehran, Iran; Email: mhossein110@yahoo.com; Tel: +982188989125; Fax: +982188989127. 1. Introduction A cute kidney injury (AKI) is considered an important issue that if not promptly managed can lead to se- rious complications such as chronic kidney disease, end-stage renal diseases and even death (1-3). Recent stud- ies show that preventive strategies can significantly reduce the burden of the disease (4) and early diagnosis and treat- ment of AKI can significantly decrease medical costs. Unfor- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com N. Ataei et al. 2 tunately kidney diseases are asymptomatic most of the time and disease manifestations do not occur until a large amount of renal tissue is destroyed (5). Hence, the best suggestion is that nephrologists and physicians of other specialties begin to comprehend the importance of early diagnosis of kidney disease. Several diagnostic methods are introduced for iden- tification of children with AKI, but none of them could accu- rately predict the outcome of disease in the initial stages (6). Over the past decade, many studies have shown that multiple serum and urine biomarkers are useful for early detection of kidney injury and have a better prognostic role than previous diagnostic tools (7-9). After three meta-analyses we reached a conclusion indicating that biomarkers such as neutrophil gelatinase-associated lipocalin (NGAL) and Cystatin C are ac- curate predictors of AKI (10-12). However, it is yet to be stud- ied which one of these biomarkers and their sources (serum or urine) have greater value in prediction of AKI. A meta- analysis reported that serum Cystatin C has a greater diag- nostic value in detection of AKI in children and adolescents than its urine level (10). Existing evidence regarding NGAL is however inconclusive and further research needs to be car- ried out (13-15). The present study aimed to assess the value of urinary NGAL (uNGAL) and serum Cystatin C in early di- agnosis of hospitalized children with AKI. 2. Methods 2.1. Study design and setting In the present diagnostic accuracy study, the value of urine NGAL in early detection of pediatric AKI was assessed. All children brought to Children’s Medical Center of Tehran, Iran, between 2015 and 2016 were entered in a prospective manner. The protocol of the present study was approved by the Ethics Committee of Tehran University of Medical Sci- ences and all researchers adhered to the principles of the Helsinki declaration during the study period. Additionally, a consent form was obtained from patients or their parents. 2.2. Participants The study population of the present study consisted of all children between the ages of 1 month and 14 years, who were suspected to AKI according to their in-charge physi- cian’s opinion. Exclusion criteria were children with chronic kidney disease, need for renal replacement therapy and those not giving consent. Convenience sampling was used in the present study. 2.3. Test methods Blood samples were taken from all patients on admission and 48 hours after admission and were sent to the laboratory of Children’s Medical Center in order to assess serum creatinine and Cystatin C level. Creatinine measurement was done us- ing Jaffe method, a spectrophotometry kit and a Hitachi 717 device. Additionally, serum Cystatin C was measured using Cystatin C ELISA Kit from BioPorto Diagnostic of Denmark using Sandwich ELISA method. Glomerular filtration rate (GFR) was calculated based on Cystatin C formula (16). A urine sample was obtained within 6 hours of admission and was stored in -80◦c after centrifuge. Measurement of uNGAL was done using an NGAL ELISA kit from BioPorto Diagnos- tics, Denmark using Sandwich ELISA method. In addition, the laboratory and technicians who entered data were not aware of the AKI status of patients. Index test was assessed in a blind manner as uNGAL measurement was performed before GFR calculation. 2.4. Outcome AKI was considered as the endpoint of the study. Pediatric Risk, Injury Failure, Loss, End-stage renal disease (pRIFLE) criteria was used for identification of children with AKI as the reference test (17). Therefore, a decrease in estimated creati- nine clearance by at least 25% from baseline within 48 hours after admission was considered the definition of AKI. 2.5. Statistical Analysis The required sample size was estimated based on the sug- gested method of Hajian-Tilaki (18). A sensitivity of 93% for uNGAL in diagnosis of AKI and a minimum marginal error of 0.1 (19) were considered in calculating the sample size hence, 91 patients were required. In the end, data from 96 patients were entered and analyzed. Data were analyzed using STATA 11.0. Patients were divided to two groups of AKI and no-AKI based on their GFR level within 48 hours of admission. At the end, area under the receiving operating characteristics (ROC) curve, sensitivity, specificity, positive predictive value (PPV ), negative predictive value (NPV ), positive and negative likeli- hood ratios of uNGAL and Cystatin C were calculated in order to assess their discriminatory power with a 95% confidence interval (CI). Areas under the curve of the two biomarkers were compared using the method suggested by Cleves and Rock (20). General calibration was assessed by drawing cal- ibration plot, which showed that the perfect calibration was the reference line with an intercept of zero and slope of 1. As the slope and intercept of calibration plot of models be- come closer to 1 and zero, respectively, they will be consid- ered more perfect in prediction of AKI. Additionally, overall performance was calculated using Brier score in order to as- sess the predictive accuracy and predictive reliability. A p value of less than 0.05 was considered significant in all anal- yses. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 3 Emergency. 2018; 6 (1): e2 Table 1: Summary of demographic, baseline, and laboratory values based on outcome Variable No-AKI (n = 83) AKI (n = 13) Total (n = 96) P Sex (n, %) Boy 47 (56.63) 7 (53.85) 54 (56.25) 0.85 Girl 36 (43.37) 6 (46.15) 42 (43.75) Age (month) 28.66±37.58 18.69±25.58 27.31±36.24 0.64* Etiology of AKI Pneumonia 7 (8.43) 0 (0.0) 7 (7.29) 0.24# Cardiac diseases 14 (16.87) 3 (23.08) 17 (17.71) Trauma 3 (3.61) 0 (0.0) 3 (3.13) Sepsis 4 (4.82) 2 (15.38) 6 (6.25) Urologic disorders 21 (25.30) 2 (15.38) 23 (23.96) Pulmonary 14 (16.87) 2 (15.38) 16 (16.67) Liver diseases 6 (7.23) 2 (15.38) 8 (8.33) Metabolic disorders 3 (3.61) 2 (15.38) 5 (5.21) Other 11 (13.25) 0 (0.0) 11 (11.46) Height (cm) 76.48±25.30 72.08±21.55 75.88±24.75 0.63* Cystatin C on admission (mg/L) 0.35±0.45 1.97±1.10 0.57±0.80 < 0.0001* uNGAL (mg/L) 168.59±284.52 1929.92±3957.54 407.10±1554.12 < 0.0001* Creatinine (mg/L) Baseline 0.48±0.16 0.49±0.16 0.49±0.16 0.61* 48 hours 0.44±0.14 1.08±0.36 0.53±0.28 < 0.0001* eCCI (ml/min/ 1.73 m2) Baseline 91.36±37.24 34.46±10.84 83.66±39.94 0.61* 48 hours 91.36±37.24 34.46±10.84 83.66±39.94 < 0.0001* AKI: acute kidney injury; eCCI: estimated creatinine clearance by Cystatin C equation; uNGAL: urinary neutrophil gelatinase-associated lipocalin. *, based on Mann-Whitney U test; #, based on Fisher’s exact test. Table 2: Screening performance characteristics of urinary neutrophil gelatinase-associated lipocalin (uNGAL) and Cystatin C in detection of acute kidney injury (AKI) Performance Cystatin C uNGAL Cut off 0.4 mg/L 125 mg/L True positive 12 12 True negative 53 57 False positive 30 26 False negative 1 1 Sensitivity 0.92 (0.62 to 0.99) 0.92 (0.62 to 0.99) Specificity 0.64 (0.52 to 0.74) 0.69 (0.57 to 0.78) Positive predictive value 0.29 (0.16 to 0.45) 0.32 (0.18 to 0.49) Negative predictive value 0.98 (0.89 to 1.0) 0.98 (0.89 to 1.0) Positive likelihood ratio 2.55 (1.84 to 3.54) 2.95 (2.07 to 4.20) Negative likelihood ratio 0.12 (0.02 to 0.80) 0.11 (0.02 to 0.74) *, Data are presented as estimated value and 95% confidence interval. 3. Results 3.1. Baseline characteristics Data from 96 children with the average age of 27.31±36.24 months (ranging from 1 month to 156 months) were en- tered (56.25% boys). Urologic problems (26.96%) were the most common etiology for AKI. Due to pRIFLE criteria, 13 cases (13.54%) of AKI were identified. The uNGAL level in AKI group (1929.92±3957.54) was significantly higher than no-AKI group (168.59±284.52) with a p value of less than 0.0001. Similar results were observed regarding Cystatin C levels (1.97±1.10 vs. 0.35±0.45; p < 0.0001) (table 1). Area under the ROC of urine NGAL level in diagnosis of AKI in children was 0.91 (95% CI: 0.80 to 1.00) and area under the ROC of Cystatin C was 0.90 (95% CI: 0.77 to 1.00). Both tests had the same value in diagnosis of AKI (p=0.89) (figure 1). Both biomarkers had appropriate calibration levels. The slope of calibration line for uNGAL and Cystatin C were 1.02 This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com N. Ataei et al. 4 Figure 1: Comparison of the areas under the receiver operating characteristics (ROC) curve of urinary neutrophil gelatinase associ- ated lipocalin (uNGAL) and serum Cystatin C level in diagnosis of acute kidney injury in children within 48 hours of admission. and 1.001, respectively. The calculated intercept of the men- tioned biomarkers were -0.006 and -0.0006, respectively. Ad- ditionally, Brier score of uNGAL and Cystatin C were 0.07 and 0.08, respectively, which showed good overall performance of both biomarkers in diagnosis of AKI in children (figure 2). The best cut-off point of uNGAL for diagnosis AKI was 125 mg/L. uNGAL had a sensitivity and specificity of 0.92 and 0.69, respectively at this level. uNGAL level had a PPV and NPV of 0.32 and 0.98, respectively. The best cut-off point of serum Cystatin C level was 0.4 mg/L. Cystatin C had a sensi- tivity and specificity of 0.92 and 0.64, respectively at the men- tioned cut-off point. Additionally, Cystatin C had a PPV and NPV of 0.28 and 0.98, respectively (table 2). 4. Discussion The present study showed that uNGAL and serum Cystatin C level are reliable biomarkers for early detection of pedi- atric AKI. Level of these biomarkers measured early after in- jury can identify AKI much sooner than conventional meth- ods. NGAL is a 20 kDa protein expressed in renal tubules which is released into the urine in case of acute or chronic kidney injury. Therefore, uNGAL level might reflect kidney injury. Faster increase in uNGAL level compared to serum creatinine level in response to AKI is one of its benefits over creatinine. However, some biomarkers such as serum crea- tinine level indicate renal function and are indirectly indica- tive of AKI (21). In other words, GFR should decrease to a certain level in order to disturb creatinine clearance and in- crease its serum level, which is required for diagnosis of AKI. However, NGAL level increases in urine in response to kid- ney injury; hence, it can be a better marker of the severity of AKI (22). In a meta-analysis of 37 studies by researchers of the present study, it was reported that assessing uNGAL level within the first 6 hours of admission can accurately predict the incidence of acute kidney injury in children in the first 48 hours (12). Similar results were found in another meta- analysis on serum NGAL level. However, the best window period for measuring serum NGAL level was reported to be within the first 12 hours of admission (11). Due to the men- tioned findings, uNGAL level seems to have a better perfor- mance in prediction of AKI. Cystatin C is another biomarker, the predictive value of which was proven in the present study. Similarly, a meta-analysis showed that serum Cystatin C level has a reasonably good value in predicting AKI in children. Additionally, serum Cys- tatin C level has a greater value compared to its urine level (10). Hence, serum Cystatin C level was used in the present study. For the first time, the present study showed that uN- GAL and serum Cystatin C level have similar values in predic- tion of AKI in children. 5. Limitation Small sample size is one of the limitations of the present study. In the present small sized study, patients brought to an individual center were assessed and selection bias might be present as convenience sampling was used. Short follow-up period (48 hours) was another limitation of this study. Addi- tionally, misclassifications might be present as a small num- ber of patients show signs and symptoms of AKI (based on pRIFLE criteria) after the first 48 hours (during the first week of admission or even later). 6. Conclusion The present study showed that both uNGAL and serum Cys- tatin C level are highly valuable predictors of AKI within the first hours of admission. In other words, level of these biomarkers measured early after injury can identify AKI much sooner than conventional methods. Additionally, uN- GAL seems to be a better biomarker in children and adoles- cents as its assessment is less invasive than assessing serum Cystatin C level and can be measured using a single urine sample. 7. Appendix 7.1. Acknowledgements None. 7.2. Author contribution Mostafa Hosseini, Mahmoud Yousefifard and Neamatollah Ataei designed the study. Sonbol Ameli, Behnaz Bazargani and Arash Abbasi participated in acquisition of data. Mostafa Hosseini analyzed the data. Fatemeh ataei and Alireza Oraei This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com 5 Emergency. 2018; 6 (1): e2 Figure 2: Calibration plot of urinary neutrophil gelatinase-associated lipocalin (uNGAL) and Cystatin C in detection of acute kidney injury (AKI). participate in management of data. Mahmoud Yousefifard and Alireza Oraei wrote the first draft and other revising manuscript critically. All authors approved final version of the manuscript to be published and are accountable for all aspects of the work. 7.3. Funding/Support This research has been supported by Tehran University of Medical Sciences & health Services grant number: 94-03- 184-30338. 7.4. Conflict of interest The authors declare no conflicts of interest. References 1. Anavekar NS, McMurray JJ, Velazquez EJ, Solomon SD, Kober L, Rouleau J-L, et al. Relation between re- nal dysfunction and cardiovascular outcomes after my- ocardial infarction. New England Journal of Medicine. 2004;351(13):1285-95. 2. Wang C, Lv LS, Huang H, Guan J, Ye Z, Li S, et al. Ini- tiation time of renal replacement therapy on patients with acute kidney injury: A systematic review and meta- analysis of 8179 participants. Nephrology (Carlton, Vic). 2017;22(1):7-18. 3. Jha V, Garcia-Garcia G, Iseki K, Li Z, Naicker S, Plattner B, et al. Chronic kidney disease: global dimension and perspectives. The Lancet. 2013;382(9888):260-72. 4. Du Y, Zappitelli M, Mian A, Bennett M, Ma Q, Devarajan P, et al. Urinary biomarkers to detect acute kidney injury in the pediatric emergency center. Pediatric Nephrology. 2011;26(2):267-74. 5. Parikh CR, Devarajan P, Zappitelli M, Sint K, Thiessen- Philbrook H, Li S, et al. Postoperative biomarkers pre- dict acute kidney injury and poor outcomes after pedi- atric cardiac surgery. Journal of the American Society of Nephrology. 2011;22(9):1737-47. 6. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo Jr JL, et al. The seventh report of the joint na- tional committee on prevention, detection, evaluation, and treatment of high blood pressure: the JNC 7 report. Jama. 2003;289(19):2560-71. 7. Mitsnefes MM, Kathman TS, Mishra J, Kartal J, Khoury PR, Nickolas TL, et al. Serum neutrophil gelatinase- associated lipocalin as a marker of renal function in chil- dren with chronic kidney disease. Pediatric Nephrology. 2007;22(1):101-8. 8. Mishra J, Dent C, Tarabishi R, Mitsnefes MM, Ma Q, Kelly C, et al. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. The Lancet. 2005;365(9466):1231-8. 9. Nickolas TL, Barasch J, Devarajan P. Biomarkers in acute and chronic kidney disease. Current opinion in nephrol- ogy and hypertension. 2008;17(2):127-32. 10. Nakhjavan-Shahraki B, Yousefifard M, Ataei N, Baikpour M, Ataei F, Bazargani B, et al. Accuracy of cystatin C in prediction of acute kidney injury in children; serum or urine levels: which one works better? A systematic review and meta-analysis. BMC nephrology. 2017;18(1):120. 11. Izadi A, Yousefifard M, Nakhjavan-Shahraki B, Baikpour M, Mirzay Razaz J, Ataei N, et al. Value of Plasma/Serum Neutrophil Gelatinase-Associated Lipocalin in Detection of Pediatric Acute Kidney Injury; a Systematic Review and Meta-Analysis. International Journal of Pediatrics. 2016;4(11):3815-36. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com N. Ataei et al. 6 12. Izadi A, Yousefifard M, Nakhjavan-Shahraki B, Baikpour M, Mirzay Razaz J, Hosseini M. Diagnostic Value of Uri- nary Neutrophil Gelatinase-Associated Lipocalin (NGAL) in Detection of Pediatric Acute Kidney Injury; a System- atic Review and Meta-Analysis. International Journal of Pediatrics. 2016;4(11):3875-95. 13. Devarajan P, M Barasch J. Neutrophil Gelatinase- Associated Lipocalin (NGAL) as a Biomarker for Acute Kidney Injury. Recent Patents on Biomarkers. 2012;2(3):179-89. 14. Wagener G, Jan M, Kim M, Mori K, Barasch JM, Sladen RN, et al. Association between increases in urinary neutrophil gelatinase-associated lipocalin and acute renal dysfunction after adult cardiac surgery. ANESTHESIOLOGY-PHILADELPHIA THEN HAGERSTOWN-. 2006;105(3):485. 15. Wagener G, Gubitosa G, Wang S, Borregaard N, Kim M, Lee HT. Urinary neutrophil gelatinase-associated lipocalin and acute kidney injury after cardiac surgery. American Journal of Kidney Diseases. 2008;52(3):425-33. 16. Laterza OF, Price CP, Scott MG. Cystatin C: an improved estimator of glomerular filtration rate? Clinical chem- istry. 2002;48(5):699-707. 17. Soler YA, Nieves-Plaza M, Prieto M, Garcia-De Jesus R, Suarez-Rivera M. pRIFLE (Pediatric Risk, Injury, Failure, Loss, End Stage Renal Disease) score identifies Acute Kid- ney Injury and predicts mortality in critically ill children : a prospective study. Pediatric critical care medicine : a journal of the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. 2013;14(4):e189-e95. 18. Hajian-Tilaki K. Sample size estimation in diagnostic test studies of biomedical informatics. Journal of biomedical informatics. 2014;48:193-204. 19. Nickolas TL, O’Rourke MJ, Yang J, Sise ME, Canetta PA, Barasch N, et al. Sensitivity and specificity of a single emergency department measurement of urinary neutrophil gelatinase-associated lipocalin for diagnos- ing acute kidney injury. Annals of internal medicine. 2008;148(11):810-9. 20. Cleves MA, Rock L. From the help desk: Comparing ar- eas under receiver operating characteristic curves from two or more probit or logit models. The Stata Journal. 2002;2(3):301-13. 21. Askenazi DJ, Ambalavanan N, Goldstein SL. Acute kid- ney injury in critically ill newborns: What do we know? What do we need to learn? Pediatric Nephrology. 2009;24(2):265-74. 22. Haase M, Bellomo R, Devarajan P, Haase-Fielitz A. Neu- trophil Gelatinase-Associated Lipocalin (NGAL) in acute kidney injury: A systematic review and meta-analysis. In- tensive Care Medicine. 2009;35:S215. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: www.jemerg.com Introduction Methods Results Discussion Limitation Conclusion Appendix References