Abstract

Background. In adult cardiac surgery, the predictive value for AKI of neutrophil gelatinase-associated lipocalin (NGAL) appears to have wide variability. The choice of definition of acute kidney injury (AKI) might, at least in part, account for such variability.

Methods. In a prospective study of 100 adult cardiac surgery patients, we assessed the value of postoperative plasma NGAL in predicting AKI according to the degree of severity used for its definition.

Results. The predictive value of plasma NGAL varied according to the AKI definition used and was higher for more severe AKI (increase in creatinine >50%: mean AUC–ROC 0.79 ± 0.01) compared to less severe AKI (>25%: mean AUC–ROC 0.65 ± 0.02); P = 0.001. The discriminatory ability of NGAL for AKI also increased with increasing RIFLE classes (AUC–ROC R: 0.72, I: 0.79, F: 0.80) or AKIN stages (AUC–ROC 1: 0.75, 2: 0.78, 3: 0.81); P = 0.015. It was highest for the prediction of renal replacement therapy (AUC–ROC: 0.83).

Conclusions. In adult cardiac surgery patients, the predictive value of NGAL increases with grade of AKI. This observation needs to be taken into account when interpreting any future studies of this biomarker.

Introduction

In experimental and clinical studies, neutrophil gelatinase-associated lipocalin (NGAL) appears to be one of the most frequently investigated and most promising biomarkers for the early diagnosis of acute kidney injury (AKI) [ 1–5 ].

NGAL was found to be an excellent biomarker for the early diagnosis of AKI in children undergoing cardiac surgery [ 1 , 4 , 5 ], in adults admitted to hospital via the emergency department [ 3 ] and in children and adults undergoing renal transplantation [ 2 ], with areas under the curve for the receiver operating characteristic (AUC–ROC) >0.9.

However, with accumulating evidence, a wide range of NGAL's predictive values for AKI after cardiac surgery have been reported in adult patients with an AUC–ROC ranging from 0.53 to 0.80 [ 6–9 ]. These observations raise concern about the robustness of NGAL as a biomarker.

Unfortunately, different definitions of AKI after cardiac surgery have been used to assess the predictive value of NGAL in different adult studies creating a definition-related confounder in our understanding of NGAL's usefulness as a biomarker. Most AKI definitions have been based on serum creatinine changes, including >25% or >50% creatinine increases from the baseline during varying periods of time and creatinine-based criteria of RIFLE (R-renal risk, I-injury, F-failure, L-loss of renal function, E-end stage renal disease) [ 10 ] or AKI Network (AKIN) [ 11 ] classification [ 1 , 4–9 ]. Such diversity of AKI definitions might have influenced the predictive value of NGAL with less stringent definitions of AKI potentially decreasing the performance of NGAL as a predictive biomarker [ 12 ]. However, the impact of the grade and/or choice of the definition of AKI on the performance of NGAL have not yet been studied.

In this investigation, we assessed the impact of the definition of AKI on the performance of plasma NGAL as renal biomarker in adult patients after cardiac surgery.

Subjects and methods

Patients

In the present study, we took advantage of a previous prospective cohort study [ 9 ], where we enrolled 100 adult patients who underwent cardiac surgery necessitating the use of cardiopulmonary bypass at a tertiary hospital. We excluded patients undergoing emergency operation (time between hospital admission to operation <24 h) or off-pump surgery, patients presenting with advanced chronic kidney disease (serum creatinine >300 μmol/L), kidney transplant patients and patients <18 years.

Plasma NGAL and serum creatinine

For the measurement of plasma NGAL, we obtained blood samples at 6 h after commencement of cardiopulmonary bypass closely corresponding to 2 h after the end of cardiopulmonary bypass measured by TriageMeter (Biosite, San Diego, CA, USA). We centrifuged the samples at 2000 g for 5 min and stored the supernatant in equal volumes at −80°C until measurement. Serum creatinine was measured preoperatively and daily from first to seventh day postoperatively utilizing the modified Jaffé method standardized by isotope dilution mass spectroscopy. The inter- and intra-assay coefficients of variation of plasma NGAL and creatinine measurement were <5%.

Endpoints

For analysis of diagnostic value of plasma NGAL, we used sustained (>2 days) increases in serum creatinine of (1) >25% and (2) >50% from the baseline to the postoperative peak value within (1) 48 h, (2) 72 h, (3) 120 h and (4) 168 h of surgery. We also assessed the value of plasma NGAL in relation to the full definition of the RIFLE [ 10 ] classes R, I and F within 7 days postoperatively and the AKIN [ 11 ] stages 1, 2 and 3 within 48 h postoperatively (Table 1 ). For the grading of patients according to the RIFLE classification, we considered the serum creatinine concentration, the estimated glomerular filtration rate (eGFR) by the Modified Diet in Renal Disease formula re-expressed for use with the serum creatinine values standardized to isotope dilution mass spectroscopy [ 13 ] and the urine output criteria. For AKIN, we considered the criteria referring to creatinine increases, use of renal replacement therapy and urine output decreases. Urine output was documented hourly. At our centre, renal replacement therapy was initiated if the patient fulfilled at least one of the following clinical criteria: oliguria (urine output <100 mL/6 h) that has been unresponsive to fluid resuscitation measures, hyperkalaemia ([K+] >6.5 mmol/L), severe acidaemia (pH <7.2) or clinically significant organ oedema (e.g. lung) in the setting of renal failure.

Table 1

Classification systems for acute kidney injury

  RIFLE criteria a (within 7 days postoperatively)    AKIN criteria b (within 48 h postoperatively)  
Class GFR Urine output Stage Creatinine Urine output 
R-Risk Crea increase ×1.5 or GFR loss  >25% <0.5 mL/kg/h × >6 h Crea increase × 1.5 or Crea  increase >0.3 mg/dL <0.5 mL/kg/h × >6 h 
I-Injury Crea increase × 2 or GFR loss >50% <0.5 mL/kg/h × >12 h Crea increase × 2 or <0.5 mL/kg/h × >12 h 
F-Failure Crea increase × 3 or GFR loss  >75% or Crea increase >4 mg/dL  (acute rise >0.5 mg/dL) <0.3 mL/kg/h × >24 h  or anuria >12 h  3 c Crea increase × 3 or Crea  increase >4 mg/dL  (acute rise >0.5 mg/dL) <0.3 mL/kg/h × >24 h 
L-Loss Persistent loss of kidney function >4  weeks  –   
E-End-stage  renal disease End-stage renal disease >3 months  –   
  RIFLE criteria a (within 7 days postoperatively)    AKIN criteria b (within 48 h postoperatively)  
Class GFR Urine output Stage Creatinine Urine output 
R-Risk Crea increase ×1.5 or GFR loss  >25% <0.5 mL/kg/h × >6 h Crea increase × 1.5 or Crea  increase >0.3 mg/dL <0.5 mL/kg/h × >6 h 
I-Injury Crea increase × 2 or GFR loss >50% <0.5 mL/kg/h × >12 h Crea increase × 2 or <0.5 mL/kg/h × >12 h 
F-Failure Crea increase × 3 or GFR loss  >75% or Crea increase >4 mg/dL  (acute rise >0.5 mg/dL) <0.3 mL/kg/h × >24 h  or anuria >12 h  3 c Crea increase × 3 or Crea  increase >4 mg/dL  (acute rise >0.5 mg/dL) <0.3 mL/kg/h × >24 h 
L-Loss Persistent loss of kidney function >4  weeks  –   
E-End-stage  renal disease End-stage renal disease >3 months  –   

a Bellomo et al . [ 10 ].

b Mehta et al . [ 11 ].

c Patients on renal replacement therapy were classified as stage 3.

Statistics

We searched Pubmed, EMBASE and Medline (as of 2 February 2009) and considered all clinical studies of paediatric and adult patients where the diagnostic value of NGAL to predict AKI after cardiac surgery was reported as the area under the curve for the receiver-operating characteristic (AUC–ROC). We performed a sensitivity analysis on the value of early plasma NGAL in predicting AKI after cardiac surgery according to AKI definitions used in previous publications in this setting [ 1–9 ] by calculation of the AUC–ROC. For AUC–ROC analysis, the control group was ‘no AKI’ and grade of AKI was the event group. The AUC–ROC value of plasma NGAL for a creatinine increase >25% was compared with that for a creatinine increase >50% by the paired student t -test. In addition, we assessed the performance of plasma NGAL in this cohort using the RIFLE and AKIN classifications. The relationship between increasing RIFLE classes and AKIN stages with the AUC–ROC value of plasma NGAL was tested by analysis of variance. Comparison of an AUC–ROC was performed as previously described [ 14 ]. For each parameter, we calculated the ‘best threshold’ (defined as the renal biomarker concentration that is closest to the point on the receiver-operating characteristic curve where sensitivity = 1-specificity = 1), sensitivity and specificity. We used SPSS Version 16.0 (SPSS Inc, Chicago, IL, USA) and MedCalc Version 9.3.9.0 (MedCalc Software, Mariakerke, Belgium).

Results

Demographic data are shown in Table 2 . Significant comorbidities included arterial hypertension, diabetes mellitus and preoperative renal impairment with eGFR <60 mL/min/1.73 m 2 [ 13 ]. The most common surgical procedures were valve surgery and coronary artery bypass grafting surgery.

Table 2

Patients’ characteristics ( N = 100)

Age (years) 69.5 ± 8.7 
Sex, n (female)  39 
Arterial hypertension, n 84 
Diabetes mellitus a , n 28 
Preoperative chronic renal impairment, n 27 
Preoperative serum creatinine (μmol/L) 90.8 ± 25.7 
Preoperative eGFR (mL/min/1.73 m 2 ) b 73.9 ± 19.7 
Congestive heart failure c , n 25 
Chronic obstructive pulmonary disease, n 14 
Peripheral vascular disease, n 
EuroScore, points d 5.3 ± 2.2 
Coronary revascularization, n 52 
Valve surgery, n 25 
Simultaneous coronary revascularization 17 
 and valve surgery, n 17 
Duration of cardiopulmonary bypass (min) 138 ± 39 
Length of stay in hospital (days) 10.8 ± 6.5 
Hospital deaths, n 
Age (years) 69.5 ± 8.7 
Sex, n (female)  39 
Arterial hypertension, n 84 
Diabetes mellitus a , n 28 
Preoperative chronic renal impairment, n 27 
Preoperative serum creatinine (μmol/L) 90.8 ± 25.7 
Preoperative eGFR (mL/min/1.73 m 2 ) b 73.9 ± 19.7 
Congestive heart failure c , n 25 
Chronic obstructive pulmonary disease, n 14 
Peripheral vascular disease, n 
EuroScore, points d 5.3 ± 2.2 
Coronary revascularization, n 52 
Valve surgery, n 25 
Simultaneous coronary revascularization 17 
 and valve surgery, n 17 
Duration of cardiopulmonary bypass (min) 138 ± 39 
Length of stay in hospital (days) 10.8 ± 6.5 
Hospital deaths, n 

a Patients on oral antidiabetic medication or on insulin.

b eGFR, estimated glomerular filtration rate according to the MDRD (Modification of Diet in Renal Disease) study equation [ 13 ].

c New York Heart Association class III/IV or impaired left ventricular function (defined as left ventricular ejection fraction <35%).

d [ 15 ]. Linear values denote mean ± standard deviation.

Seven studies reported on the predictive value of NGAL for AKI after cardiac surgery in children and adults with a wide predictive range of AUC–ROC (Table 3 ). Five studies defined AKI as an increase in serum creatinine of >50% and two used selected criteria from the AKIN classification. Also, the timing of the creatinine increase considered for AKI definition varied between studies (Table 3 ). While the timing of NGAL measurement in relation to the commencement of surgery was similar, the predictive value of NGAL ranged from 0.53 to 0.80 in adults (Table 3 ).

Table 3

Definition of acute kidney injury after cardiac surgery and the predictive value of early measured NGAL

   AKI definition   
 No. of  Creatinine Timing of postoperative Timing of NGAL measurement AUC–ROC to predict 
Reference patients Setting increase creatinine increase (after end of CPB) AKI (plasma/urine) 
Mishra et al . [ 1 ]  71 Paediatric >50% Within 5 days At 2 h 0.91/0.99 
Dent et al . [ 4 ]  120 Paediatric >50% Within 5 days At 2 h 0.96/– 
Bennett et al . [ 5 ]  196 Paediatric >50% Within 5 days At 2 h –/0.95 
Wagener et al . [ 6 ]  81 Adult >50% Within 5 days At 3 h –/0.74 
Wagener et al . [ 7 ]  426 Adult >50% or Within 2 days At 3 h –/0.60 
    >0.3 mg/dL    
Koyner et al . [ 8 ]  72 Adult >25% or need Within 3 days  At ∼ 2 ho * 0.53/0.70 
    for RRT    
Haase-Fielitz et al . [ 9 ]  100 Adult >50% Within 5 days  At ∼ 2 h * 0.80/– 
   AKI definition   
 No. of  Creatinine Timing of postoperative Timing of NGAL measurement AUC–ROC to predict 
Reference patients Setting increase creatinine increase (after end of CPB) AKI (plasma/urine) 
Mishra et al . [ 1 ]  71 Paediatric >50% Within 5 days At 2 h 0.91/0.99 
Dent et al . [ 4 ]  120 Paediatric >50% Within 5 days At 2 h 0.96/– 
Bennett et al . [ 5 ]  196 Paediatric >50% Within 5 days At 2 h –/0.95 
Wagener et al . [ 6 ]  81 Adult >50% Within 5 days At 3 h –/0.74 
Wagener et al . [ 7 ]  426 Adult >50% or Within 2 days At 3 h –/0.60 
    >0.3 mg/dL    
Koyner et al . [ 8 ]  72 Adult >25% or need Within 3 days  At ∼ 2 ho * 0.53/0.70 
    for RRT    
Haase-Fielitz et al . [ 9 ]  100 Adult >50% Within 5 days  At ∼ 2 h * 0.80/– 

RRT, renal replacement therapy.

* Six hours after the start of cardiopulmonary bypass (CPB) or the intensive care unit arrival value.

The frequencies of serum creatinine increases, RIFLE classes and AKIN stages in the study population are shown in Tables 4–6 .

Table 4

Performance characteristics of plasma NGAL at 6 h after start of CPB according to AKI severity

AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
>25% within 48 h, ( N = 36)  0.66 (0.48–0.84) 63.2 72.7 >145 
>25% within 72 h, ( N = 38)  0.64 (0.46–0.83) 61.1 69.6 >145 
>25% within 120 h, ( N = 39)  0.67 (0.48–0.85) 68.4 63.6 >145 
>25% within 168 h, ( N = 40)  0.64 (0.45–0.83) 61.1 69.6 >145 
>0.3 mg/dL or >50% within 48 h, ( N = 32)  0.66 (0.48–0.84) 63.2 72.7 >145 
>25% or RRT within 72 h, ( N = 42)  0.68 (0.51–0.85) 65.0 76.2 >145 
>50% within 48 h, ( N = 20)  0.78 (0.61–0.96) 75.0 75.9 >155 
>50% within 72 h, ( N = 21)  0.79 (0.62–0.95) 75.5 76.2 >155 
>50% within 120 h, ( N = 23)  0.80 (0.63–0.96) 79.0 78.0 >150 
>50% within 168 h, ( N = 23)  0.80 (0.63–0.96) 79.0 78.0 >150 
RRT, ( N = 4)  0.83 (0.60–0.98) 75.0 100.0 >340 
AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
>25% within 48 h, ( N = 36)  0.66 (0.48–0.84) 63.2 72.7 >145 
>25% within 72 h, ( N = 38)  0.64 (0.46–0.83) 61.1 69.6 >145 
>25% within 120 h, ( N = 39)  0.67 (0.48–0.85) 68.4 63.6 >145 
>25% within 168 h, ( N = 40)  0.64 (0.45–0.83) 61.1 69.6 >145 
>0.3 mg/dL or >50% within 48 h, ( N = 32)  0.66 (0.48–0.84) 63.2 72.7 >145 
>25% or RRT within 72 h, ( N = 42)  0.68 (0.51–0.85) 65.0 76.2 >145 
>50% within 48 h, ( N = 20)  0.78 (0.61–0.96) 75.0 75.9 >155 
>50% within 72 h, ( N = 21)  0.79 (0.62–0.95) 75.5 76.2 >155 
>50% within 120 h, ( N = 23)  0.80 (0.63–0.96) 79.0 78.0 >150 
>50% within 168 h, ( N = 23)  0.80 (0.63–0.96) 79.0 78.0 >150 
RRT, ( N = 4)  0.83 (0.60–0.98) 75.0 100.0 >340 

The increase in serum creatinine defined as an increase from a preoperative to a postoperative peak value.

NGAL, neutrophil gelatinase-associated lipocalin; CPB, cardiopulmonary bypass; AKI, acute kidney injury; RRT, renal replacement therapy.

Table 5

Performance characteristics of plasma NGAL at 6 h after start of cardiopulmonary bypass to predict AKI according to RIFLE classes.

AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
RIFLE     
 R ( N = 31)  0.72 (0.54–0.91) 70.0 71.4 >150 
 I ( N = 13)  0.79 (0.56–0.99) 85.7 75.0 >150 
 F ( N = 6)  0.80 (0.53–1.00) 66.7 100.0 >240 
 I + F ( N = 19)  0.79 (0.59–0.99) 80.0 75.0 >150 
AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
RIFLE     
 R ( N = 31)  0.72 (0.54–0.91) 70.0 71.4 >150 
 I ( N = 13)  0.79 (0.56–0.99) 85.7 75.0 >150 
 F ( N = 6)  0.80 (0.53–1.00) 66.7 100.0 >240 
 I + F ( N = 19)  0.79 (0.59–0.99) 80.0 75.0 >150 

AKI, acute kidney injury. RIFLE [ 10 ].

Table 6

Performance characteristics of plasma NGAL at 6 h after the start of cardiopulmonary bypass to predict AKI according to AKIN stages

AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
AKIN     
 1 ( N = 29)  0.75 (0.59–0.92) 71.4 73.7 >150 
 2 ( N = 11)  0.78 (0.59–0.99) 80.0 80.0 >150 
 3 ( N = 6)  0.81 (0.58–1.00) 66.7 100.0 >240 
 2 + 3 ( N = 17)  0.79 (0.57–0.99) 75.0 78.9 >150 
AKI definition (no. of patients) AUC–ROC (95% CI) Sensitivity (%) Specificity (%) Cut-off (ng/mL) 
AKIN     
 1 ( N = 29)  0.75 (0.59–0.92) 71.4 73.7 >150 
 2 ( N = 11)  0.78 (0.59–0.99) 80.0 80.0 >150 
 3 ( N = 6)  0.81 (0.58–1.00) 66.7 100.0 >240 
 2 + 3 ( N = 17)  0.79 (0.57–0.99) 75.0 78.9 >150 

AKI, acute kidney injury. AKIN [ 11 ].

Table 4 displays the performance characteristics of plasma NGAL according to grade of AKI. The predictive value of plasma NGAL varied (AUC–ROC 0.64–0.83) according to definition used and was higher for more severe AKI. Considering a creatinine increase of >50% at postoperative Days 2, 3, 5 and 7, the average predictive value of plasma NGAL was higher for AKI defined as an increase in serum creatinine of >50% (AUC–ROC 0.79 ± 0.01) compared to >25% (AUC–ROC 0.65 ± 0.02); P = 0.001. When plasma NGAL was evaluated for its ability to predict the need for renal replacement therapy, its AUC–ROC further increased to 0.83. Tables 5 and 6 present the performance characteristics of plasma NGAL according to the RIFLE classification and the AKI network definition of AKI. After pooling corresponding RIFLE classes with AKIN stages (R with 1, I with 2, F with 3, I+F with 2+3), the predictive value of plasma NGAL increased with grade of AKI; P = 0.015.

Plasma NGAL performance when using RIFLE classification was similar to that of AKIN definition (Table 5 , 6 ). The cut-off value of plasma NGAL for best sensitivity and specificity in predicting a subsequent creatinine increase, RIFLE classes or AKIN stages increased from less severe to more severe AKI (Tables 4–6 ) and ranged between >145 ng/mL and >340 ng/mL.

A considerable proportion of patients (45.5%) who did not develop subsequent AKI (as identified by a creatinine increase > 50%) showed mild postoperative increases in plasma NGAL with concentrations ranging from >100 ng/ mL [ = above normal range (normal range: 40–100 ng/mL [ 16 ])] to <150 ng/mL (just below our cut-off for best sensitivity and specificity to predict AKI). The predictive value of plasma NGAL was not related to timing of the creatinine increase ( P = 0.98).

Discussion

In a cohort of 100 adult patients, we performed a sensitivity analysis to determine the influence of AKI definition on the predictive value of plasma NGAL for AKI after cardiac surgery. We found that the predictive value of plasma NGAL increased with grade of AKI from an AUC–ROC of 0.64 for a >25% increase in serum creatinine to an AUC–ROC value of 0.83 for the prediction of the need for renal replacement therapy. We found no association with the timing of the creatinine increase.

NGAL has been reported to represent an early and highly specific biomarker for AKI in several studies in paediatric cardiac surgery with an AUC–ROC >0.9 [ 1 , 4 , 5 ]. However, more recent prospective clinical studies found a surprisingly low predictive value of early NGAL for AKI after adult cardiac surgery, with an AUC–ROC ≤0.7 [ 6–8 ] Different AKI definitions used in these studies might have contributed to these discrepancies. Most studies have analysed NGAL concentrations to predict AKI defined as an increase in serum creatinine >50% from the baseline to the postoperative peak value. However, two studies [ 6 , 8 ] have reported the predictive value of NGAL using selected criteria from the AKIN classification which excluded its urine output criteria. In these studies, even patients who did not subsequently develop AKI had a substantial early rise in NGAL concentration. This finding may reflect not only the choice of AKI grade but also the relatively short time frame during which AKI was assessed for (48 h). Such a short evaluation time may miss the diagnosis of AKI in patients who develop a substantial and clinically relevant creatinine increase 60 or 72 h after cardiopulmonary bypass [ 17 ].

In the present study, according to the AKI definition used, we found large variations in the plasma NGAL's predictive value with an AUC–ROC ranging from 0.64 (for a >25% creatinine increase) to 0.83 (for renal replacement therapy). In terms of a common classification of the value of biomarkers in general, this corresponds to a shift from ‘poor’ to ‘good’ predictive performance or from ‘no useful’ to a ‘useful’ risk predictor (as defined by AUC–ROC >0.7 [ 18 ]), respectively. The results of our study provide a possible explanation for the large range of AUC–ROC for NGAL's predictive value for AKI.

Importantly, early but mild NGAL increases might also indicate equally mild AKI or renal stress that will not be detected by a subsequent creatinine increase. Such increases may also represent a response to systemic inflammation and may not be related to AKI itself. As a consequence, a patient with increased NGAL levels (100–150 ng/mL) and a creatinine increase of <50% will be classified no AKI although tubular injury or stress might well be present. On the other hand, prerenal azotaemia, which can meet the proposed creatinine- or urine output-based definitions of AKI without any actual tubular injury, degrades the diagnostic ability and tests performance of the marker. NGAL probably does not increase in prerenal azotaemia as previously shown by Nickolas et al . [ 3 ].

Our study has limitations. This study was a post hoc analysis, yet one of the largest in adults in this field. Other factors beyond AKI definition might be of importance for the predictive value of NGAL such as the measurement of NGAL in plasma or urine. In this regard, there is speculation that due to its origin from the renal tubules, urine NGAL might be of a higher predictive value compared to plasma NGAL originating mostly from neutrophils, liver and lung cells [ 19 ]. However, to date, there is no conclusive evidence for urine NGAL being superior to plasma NGAL. As we did not measure urine NGAL, we cannot provide comparison with plasma NGAL. Also, the degree of inflammation or diseases of liver and lungs may contribute to different findings in the literature on the predictive performance of NGAL.

To address such potential confounders, we assessed NGAL in the same patient cohort using different AKI definitions using each patient as his/her own control. Furthermore, we did not only assess the most commonly used definitions of AKI in this study but we also reused those definitions previously used in the setting of the predictive performance of NGAL enabling comparability. Also, other potential confounders of the value of NGAL such as setting of AKI, comorbidities, storage conditions of samples or assay used should be investigated in separate studies and are beyond the scope of this study. Finally, as most AKI definitions are based on changes in serum creatinine and urine output—markers with many known limitations—they may not at all reflect the true value of NGAL in detecting acute kidney injury or tubular injury or tubular stress but rather the limitations of current surrogates for the presence or absence of AKI. A biomarker can only be as good as the surrogate endpoint it predicts—in the case of NGAL, serum creatinine- and urine output-based AKI, insensitive or unspecific markers of loss of GFR, respectively.

In future studies, more sensitive AKI definitions using markers other than serum creatinine or urine output should be evaluated. Once confirmed in large multicentre trials as reliable, robust and independent predictors of patient outcomes emerging novel renal biomarkers such as NGAL or fatty acid-binding protein [ 20 ] may be used to define the presence or absence or degree of AKI independent of changes in urine output or serum creatinine.

In conclusion, the results of this study support the view that, in adult cardiac surgery, the predictive value of NGAL increases with grade of AKI. This observation needs to taken into account when interpreting future studies of this biomarker.

Dr Haase holds a postdoctoral Feodor-Lynen research fellowship from the Alexander von Humboldt-Foundation, Germany. This study was partly funded by a grant from the Australian and New Zealand College of Anaesthetists and by the Austin Hospital Anesthesia and Intensive Care Trust Fund. R.B. and M.H. conceived the study. M.H., A.H.F., P.D. and R.B. participated in the design of the study, performed the statistical analysis and drafted the manuscript. D.S., D.D., M.B., G.M. and U.F. participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.

Conflict of interest statement. Drs Bellomo and Devarajan have acted as paid consultant to Abbott Diagnostics and Biosite. Dr Devarajan has received limited research grant support from Biosite and Abbott Diagnostics. Dr Haase has received lecture fees and travel expenses from Abbott Diagnostics and Biosite, both of which are involved in the development of NGAL assays to be applied in clinical practice. All other authors have nothing to declare.

(See related article by S. S. Waikar et al. Creatinine as the gold standard for kidney injury biomarker studies? Nephrol Dial Transplant 2009; 24: 3263–3265.)

References

1
Mishra
J
Dent
C
Tarabishi
R
, et al.  . 
Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery
Lancet
 , 
2005
, vol. 
365
 (pg. 
1231
-
1238
)
2
Parikh
CR
Jani
A
Mishra
J
, et al.  . 
Urine NGAL and IL-18 are predictive biomarkers for delayed graft function following kidney transplantation
Am J Transplant
 , 
2006
, vol. 
6
 (pg. 
1639
-
1645
)
3
Nickolas
TL
O’Rourke
MJ
Yang
J
, et al.  . 
Sensitivity and specificity of a single emergency department measurement of urinary neutrophil gelatinase-associated lipocalin for diagnosing acute kidney injury
Ann Intern Med
 , 
2008
, vol. 
148
 (pg. 
810
-
819
)
4
Dent
CL
Ma
Q
Dastrala
S
, et al.  . 
Plasma neutrophil gelatinase-associated lipocalin predicts acute kidney injury, morbidity and mortality after pediatric cardiac surgery: a prospective uncontrolled cohort study
Crit Care
 , 
2007
, vol. 
11
 pg. 
R127
 
5
Bennett
M
Dent
CL
Ma
Q
, et al.  . 
Urine NGAL predicts severity of acute kidney injury after cardiac surgery: a prospective study
Clin J Am Soc Nephrol
 , 
2008
, vol. 
3
 (pg. 
665
-
673
)
6
Wagener
G
Jan
M
Kim
M
, et al.  . 
Association between increases in urinary neutrophil gelatinase-associated lipocalin and acute kidney injury after adult cardiac surgery
Anesthesiology
 , 
2006
, vol. 
105
 (pg. 
485
-
491
)
7
Wagener
G
Gubitosa
G
Wang
S
, et al.  . 
Urinary neutrophil gelatinase-associated lipocalin and acute kidney injury after cardiac surgery
Am J Kidney Dis
 , 
2008
, vol. 
52
 (pg. 
425
-
433
)
8
Koyner
JL
Bennett
MR
Worcester
EM
, et al.  . 
Urinary cystatin C as an early biomarker of acute kidney injury following adult cardiothoracic surgery
Kidney Int
 , 
2008
, vol. 
74
 (pg. 
1059
-
1069
)
9
Haase-Fielitz
A
Bellomo
R
Devarajan
P
, et al.  . 
Novel and conventional serum biomarkers predicting acute kidney injury in adult cardiac surgery—a prospective cohort study
Crit Care Med
 , 
2009
, vol. 
37
 (pg. 
553
-
560
)
10
Bellomo
R
Ronco
C
Kellum
JA
, et al.  . 
Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group
Crit Care
 , 
2004
, vol. 
8
 (pg. 
R204
-
212
)
11
Mehta
RL
Kellum
JA
Shah
SV
, et al.  . 
Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury
Crit Care
 , 
2007
, vol. 
11
 pg. 
R31
 
12
Devarajan
P
NGAL in acute kidney injury: from serendipity to utility
Am J Kidney Dis
 , 
2008
, vol. 
52
 (pg. 
395
-
399
)
13
Levey
AS
Coresh
J
Greene
T
, et al.  . 
Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values
Clin Chem
 , 
2007
, vol. 
53
 (pg. 
766
-
772
)
14
Hanley
JA
McNeil
BJ
A method of comparing the areas under receiver operating characteristic curves derived from the same cases
Radiology
 , 
1983
, vol. 
148
 (pg. 
839
-
843
)
15
Nashef
SA
Roques
F
Michel
P
, et al.  . 
European system for cardiac operative risk evaluation (EuroSCORE)
Eur J Cardiothorac Surg
 , 
1999
, vol. 
16
 (pg. 
9
-
13
)
16
Axelsson
L
Bergenfeldt
M
Ohlsson
K
Studies of the release and turnover of a human neutrophil lipocalin
Scand J Clin Lab Invest
 , 
1995
, vol. 
55
 (pg. 
577
-
588
)
17
Haase
M
Haase-Fielitz
A
Bagshaw
SM
, et al.  . 
A phase II randomized controlled trial of high-dose N -acetylcysteine in high-risk cardiac surgery patients
Crit Care Med
 , 
2007
, vol. 
35
 (pg. 
1324
-
1331
)
18
Swets
JA
Measuring the accuracy of diagnostic systems
Science
 , 
1988
, vol. 
240
 (pg. 
1285
-
1293
)
19
Schmidt-Ott
KM
Mori
K
Li
JY
, et al.  . 
Dual action of neutrophil gelatinase-associated lipocalin
J Am Soc Nephrol
 , 
2007
, vol. 
18
 (pg. 
407
-
413
)
20
Portilla
D
Dent
C
Sugaya
T
, et al.  . 
Liver fatty acid-binding protein as a biomarker of acute kidney injury after cardiac surgery
Kidney Int
 , 
2008
, vol. 
73
 (pg. 
465
-
472
)

Comments

0 Comments