International Journal of Diabetes in Developing Countries

, Volume 32, Issue 1, pp 19–24

Neutrophil Gelatinase-Associated Lipocalin (NGAL): an early marker for diabetic nephropathy

Authors

  • Mohamad Fathimah
    • Department of Pathology, Faculty of MedicineUniversity Malaya
  • Mohd Kasim Alicezah
    • Department of Pathology, Faculty of MedicineUniversity Malaya
    • Department of Pathology, Faculty of MedicineUniversity Malaya
Original Article

DOI: 10.1007/s13410-011-0061-z

Cite this article as:
Fathimah, M., Alicezah, M.K. & Thevarajah, M. Int J Diabetes Dev Ctries (2012) 32: 19. doi:10.1007/s13410-011-0061-z

Abstract

Neutrophil gelatinase-associated lipoprotein (NGAL) represents a novel biomarker for early identification of acute kidney injury. This study evaluates the usefulness of urine NGAL as a marker for the early detection of diabetic nephropathy. This is a cross-sectional study which involved ninety patients with diabetes mellitus and thirty healthy controls. The diabetic patients were categorized into three groups based on their urine albumin/creatinine ratio (UACR); normoalbuminuria (<3.5 mg/mmol), microalbuminuria (3.5–35 mg/mmol), macroalbuminuria (>35 mg/mmol). In addition to urine NGAL, HbA1C, serum creatinine, urine albumin/creatinine ratio, serum cystatin C, and urine protein were assessed to determine their correlation with urine NGAL. Data analysis was done by using SPSS and MiniTab. Urine NGAL was elevated in all groups of diabetic patients with respect to controls. It was increased proportionately to the severity of kidney function. It was also elevated in some normoalbuminuria diabetic patients. Analysis of correlation revealed that urine NGAL was not correlated with glycemic indices (HbA1C and fasting blood glucose). However, urine NGAL correlated significantly with cystatin C, serum creatinine, urine albumin/creatinine ratio, and inversely with eGFR. Besides, it is also shown to have a significant correlation with eGFR in advanced kidney disease (eGFR < 30 ml/min per 1.73 min2). Urine NGAL can be used as a non-invasive tool for the early detection and assessment of the severity of diabetic nephropathy.

Keywords

NGALDiabetic nephropathyChronic kidney disease

Introduction

Diabetes has been recognized as a worldwide health problem, affecting people at variable ages with increasing incidence and prevalence, leading to various complications [1]. Longer duration of diabetes, earlier age at diagnosis with poor glycemic control are part of the risk factors that lead to the development of diabetic kidney disease. Other risk factors include positive family history, smoking, and hypertension. Diabetic nephropathy is a major concern for the development of microvascular and macrovascular complications of diabetes and for diabetes-related and overall mortality. Besides that, with an inadequate treatment and late diagnosis, diabetic nephropathy has become the most common diagnosis at initiation of renal replacement therapy [2].

The key pathophysiologic event in diabetic nephropathy is likely to involve an interaction of metabolic and haemodynamic pathways which lead to the development of basement membrane damage [3]. In addition, active inflammation caused by the passage of macromolecules through the basement membrane will result in secondary damage to the membrane. In relation to the glomerular lesion, there will be persistent diabetic proteinuria. The sustained passage of this molecule within the tubular lumen may activate intratubular complement cascade and eventually contribute to tubular injury [4]. This last condition can lead first to tubular inflammation and then to tubulointerstitial fibrosis, which ultimately signals the appearance of an irreversible renal impairment [5].

Historically, diabetic kidney disease has been diagnosed by the presence of proteinuria in the diabetes patients with the reduction of glomerular filtration rate and increased serum creatinine.

In view of the role of the tubular injury in the progression of kidney damage in diabetes, several tubular factors have been recognized as a marker for chronic kidney disease. These include Monocyte Chemoattractant Protein-1 (MCP-1), N-acetyl-B-D glucosaminidase, cathepsin and few others [6].

Neutrophil gelatinase-associated lipocalin (NGAL) is a small 25 kDa protein which was recently discovered by nephrologists. It is a small 25-kDa protein, consists of a polypeptide chain of 178 amino acids and belongs to the ‘lipocalin’ superfamily [6]. It is expressed by neutrophils and various epithelial cells. This protein is released in blood and urine following ischemic and nephrotoxic injury from injured tubular cells after various conditions [3]. Variable degrees of NGAL gene expression is demonstrated in human tissues like uterus, prostate, salivary glands, lung, trachea, stomach, colon, and kidney [7].

NGAL is currently considered as one of the most promising biomarkers in clinical nephrology and has been extensively studied in acute kidney disease. Acute kidney disease is closely related to high morbidity and mortality. Therefore researchers have aggressively searched for a new biomarker and NGAL has been found to improve the accuracy of the current biomarkers used to detect kidney damage.

A study done by Mishra et al. has revealed that NGAL was easily detected in the urine within 2 h following ischemia [8]. This has preceded the appearance of other urinary markers such as B2 microglubulin and N-acetyl-B-D-glucosaminidase. In this study, they have confirmed that NGAL is originated from the tubule cells. Other findings from this study include the increased NGAL following the administration of nephrotoxic agent, where they have administered cisplastin to the rats. These result in the presence of early, non-invasive urinary biomarkers for ischemic and nephrotoxic renal injury [8].

Materials and methods

This is a cross-sectional study, approved by the Institutional Review Board of University Malaya Medical Centre. Sample size calculation was done by using Epi Info Version 6.0 with a power of 80% and 95% confidence interval giving a sample size of 32. With a recruitment of ninety subjects, the power of study was inferred to be more than 80%. Subjects were patients with type 2 diabetes with various degrees of renal impairment from the medical clinic in University Malaya Medical Centre. They were divided into three categories based on urine albumin/creatinine ratio (UACR); normoalbuminuria (<3.5 mg/mmol), microalbuminuria (3.5–35 mg/mmol) and macroalbuminuria (>35 mg/mmol). The macroalbuminuria group was further divided into four stages of chronic kidney disease (CKD) based on National Kidney Foundation Criteria [9]. Those with infections, neoplasias, and inflammation were excluded in order to avoid any interference in the NGAL measurement.

To minimize the confounding factors, few exclusion criteria had been considered. These include subjects with infections, neoplasia, and any inflammation that could interfere with the urine NGAL measurement [10].

Inclusion criteria included those who agreed to participate and consented to this study. They were patients in University Malaya Medical Centre with Type 2 diabetes mellitus, stable renal function and diabetic nephropathy.

The control group consisted of thirty normal healthy subjects with no history of hypertension, diabetes mellitus, inflammation, neoplastic disorder, and cardiovascular disease. None of the subjects were on medical treatment.

The following data were collected: demographic characteristics, disease duration, and renal function. Besides that, other blood markers for renal function (serum creatinine, cystatin C, eGFR, urine albumin/creatinine ratio (UACR), urine protein, and urine NGAL) were also taken to determine their correlation with NGAL.

Estimated glomerular filtration rate (eGFR) of each subject was calculated from cystatin C by using the Hook et.al formula. This eGFR was then compared with the eGFR obtained from the MDRD formula to observe the correlation between these two formulas. Patients with eGFR of less than 30 ml/min per 1.73 min2 were identified to study their correlation with urine NGAL and cystatin C.

Urine NGAL was measured by ARCHITECT (Abbott, Longford, Ireland). It was carried out in a double batch. This assay is an automated method using a Chemiluminescent Microparticle Immunoassay (CMIA) for the quantitative determination of NGAL in the human urine. It has a coefficient of variation of less than 5%.

Specimens used for urine NGAL were centrifuged at ≥ 400 RCF (Relative Centrifugal Force) for a minimum of 5 min within collection. After that, the clarified specimens were transferred to a sample cup or secondary tube for storage and were stored at −80°C if testing was delayed for more than 7 days.

Cystatin C was measured by particle-enhanced immunonephelometry using SIEMENS kit (Marburg, Germany). The Intra-Assay CV determined at 0.6 mg/L was <4.15%. All tests were carried out in a single batch. Serum creatinine was measured by using Dimension Vista (SIEMENS,Newark, USA). Urine albumin/creatinine ratio (UACR) was assayed by the ARCHITECT (Abbott, Longford, Ireland).

Statistical analysis was performed by using SPSS and MiniTab. For normally distributed values, data were presented as mean ± SD, whereas for non-normally distributed values, data were presented as median ± SD. Normality test was done by using Anderson Darling normality test. Correlation between serum NGAL and other tests were performed by using Pearson coefficient. All results were considered significant when p < 0.01. Log transformed was done to normalize the non-normally distributed data and Pearson coefficient study was then used for the correlation analysis. Analysis of the non-normally distributed data was performed with Spearman coefficient study. All results were considered significant when p < 0.01. In order to compare the three groups; (normoalbumiuria, microalbuminuria, and macroalbuminuria), log transformed of urine NGAL was analyzed by using ANOVA. Results were considered significant when p < 0.05. The ANOVA was then followed by Bonferroni test as the post hoc test. Bonferroni showed statistically significant difference between the normoalbuminuric group and macroalbuminuric group with p value of 0.01.

Results

The demographic and disease-related characteristics of the patients are presented in Table 1. This study involved ninety patients with Type 2 diabetes mellitus and thirty healthy controls. The study population consisted of 66 female patients (55%) and 54 male patients (45%) with ages ranging from 22 to 89 years old.
Table 1

Demographics and disease-related characteristics

Variables

Control

Normoalbuminuria

Microalbuminuria

Macroalbuminuria

Age

23 ± 5.2

63.3 ± 12.04

68.17 ± 9.39

62.33 ± 12.83

Gender

 Male

12(40%)

15 (50%)

17 (57%)

10 (33%)

 Female

18 (60%)

15 (50%)

13 (43%)

20 (67%)

Race

 Malay

22 (73%)

16 (53%)

11(37%)

13 (43%)

 Chinese

4 (13%)

3 (10%)

10 (33%)

11 (37%)

 Indian

4 (13%)

11 (37%)

9 (30%)

6 (20%)

eGFR (mil/min per 1.73 m2)

122.94 ± 12.7

94.92 ± 19.92

61.67 ± 51.92

49.87 ± 28.47

Serum creatinine (mmol/L)

71.9 ± 16.7

76 ± 22.33

129.5(31–442)

139.5(37–485.0)

Serum Cystatin C (mg/l)

0.631 ± 0.07

0.78 ± 0.28

1.44 ± 0.87

1.65 ± 1.12

Urine NGAL (ng/mL)

4.75 (0.1–27.5)

19.05 (1.1,–60)

26.9(3.7–603.5)

28.55 (0.7–1500)

HbA1C

5.27 ± 0.315

6.23 ± 1.56

7.62 ± 2.73

6.4 ± 3.29

Urine protein

0.12 ± 0.18

0.13 ± 0.09

0.31 ± 0.36

1.08 ± 2.33

Urine albumin creatinine ratio

0.65 ± 0.39

0.95(0.1–2.8)

13.67 ± 7.56

97 (36.4–568)

The ethnic distribution (Table 1) was 52% Malay, 23% Chinese, and 25% Indian respectively. They were categorized into normoalbuminuria (N=30), microalbuminuria (N=30), and macroalbuminuria (N=30), depending on their urine albumin/creatinine ratio.

Control group showed a perfectly conserved renal function with a mean eGFR of (122.94±12.7 ml/min per 1.73 m2 and mean creatinine of 71.9±16.7 mmol). The group with macroalbuminuria showed a reduced eGFR (49.87 ± 28.47, mean) thus signalling the presence of diabetic nephropathy. Mild reduction in renal function was seen in patients with micro and normoalbuminuria.

The levels of urine NGAL were found to be significantly elevated in patient with macroalbuminuria as compared to the control, normoalbuminuria, and microalbuminuria (mean 28.55; min 0.7, max 1500) (Table 1). However, urine NGAL was also found to be elevated in some patients with normoalbuminuria. This is an interesting finding and supports the hypothesis that urine NGAL can be used as a marker for the early detection of diabetic nephropathy [11]. Besides, the mean of urine NGAL was also observed to be increased concomittently with the degree of renal impairment. This finding indicates that urine NGAL can also be used in determining the severity of renal disease.

Urine NGAL and other renal indices

Urine NGAL did not correlate with any demographic parameters (age, gender, and race). All p values were more than 0.01(Table 1)

Figures 1, 2, 3, 4 and 5 display the statistical correlations (r) between urine NGAL and other biomarkers for renal impairment. Urine NGAL correlated significantly with cystatin C, serum creatinine, UACR, eGFR and urine protein with the r values ranging from 0.39 to 0.7 and p < 0.01.
https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig1_HTML.gif
Fig. 1

Urine Ngal and cystatin C

https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig2_HTML.gif
Fig. 2

Urine NGAL and serum creatinine

https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig3_HTML.gif
Fig. 3

Urine NGAL and urine albumin creatinine ratio

https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig4_HTML.gif
Fig. 4

Urine NGAL and eGFR

https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig5_HTML.gif
Fig. 5

Urine NGAL and urine protein

Analysis of patients with advanced stages of kidney disease (eGFR<30 ml/min per 1.73 min2), showed significant correlation of both the cystatin C and urine NGAL with eGFR. This suggests that urine NGAL may be a superior marker for advanced cases of renal impairment as compared to creatinine. However this study shows that cystatin C has higher correlation as compared to urine NGAL in patients with eGFR >30 ml/min per 1.73 min2. eGFR calculated from the cystatin C using the Hook formula shows a significant correlation with the eGFR obtained from the MDRD formula (r: 0.88) (Fig. 6).
https://static-content.springer.com/image/art%3A10.1007%2Fs13410-011-0061-z/MediaObjects/13410_2011_61_Fig6_HTML.gif
Fig. 6

MDRD and eGFR

The three groups (normoalbuminuria, microalbuminuria, and macroalbuminuria) were compared by using ANOVA. The sample means for the three groups of NGAL were: 19.05 ng/ml for normoalbuminuria, 26.9 ng/ml for microalbuminuria and 28.55 for macroalbuminuria. The difference in urine NGAL between the groups was statistically significant (p=0.012). Bonferroni test revealed statistically significant differences between the normoalbuminuric and macroalbuminuric group (p=0.01).

Discussion

Neutrophil gelatinase-associated lipocalin (NGAL) is a small 25 kDa protein recently discovered by nephrologists [12]. It is a small 25-kDa protein, consisting of a polypeptide chain of 178 amino acids and belongs to the ‘lipocalin’ superfamily. It is expressed by neutrophils and various epithelial cells [7]. This protein is released in blood and urine following ischemic and nephrotoxic injury from the tubular cells. Variable degrees of NGAL gene expression is demonstrated in human tissues like uterus, prostate, salivary glands, lung, trachea, stomach, colon, and kidney [12].

Present studies have reported that NGAL is a marker that has outperformed other recent renal markers. They demonstrated that NGAL may represent a novel early urinary biomarker particularly for the detection of acute renal injury [1316]. This is because NGAL is able to manifest in the urine after 3 h of tubular injury. In comparison, serum creatinine will only be elevated after 24 h of reperfusion [8]. In this study, we are trying to demonstrate that urine NGAL can be used as an early marker for the detection of nephropathy in these patients, particularly in those with incipient nephropathy.

However, a few studies have demonstrated that NGAL might also be elevated in some other conditions, not necessarily pertaining only to renal injury [1416]. Therefore, we have excluded those likely to interfere with our results. These factors include patients with infection, neoplasia and inflammation.

This study which involved 90 patients with diabetes and 30 controls. It has demonstrated that urine NGAL does not correlate with other diabetic indices; such as HbA1c and dermographic factors such as age, race, and gender. These findings are in accordance with reports from Bolignano et al. [11]. However, urine NGAL was noted to be correlating significantly with urine protein, urine albumin/creatinine ratio, cystatin C, and eGFR. Few studies have reported the same findings in patients with chronic kidney disease [1720], hence supporting idea that urine NGAL may be used as a marker for nephropathy.

Interestingly, in this study urine NGAL and cystatin C were elevated in some normoalbuminuria diabetic patients. The results obtained showed a few patients with normal UACR demonstrating an elevated urine NGAL and cystatin C. From this point of view, we can conclude that urine NGAL may be used as an early marker for diabetic nephropathy, particularly in incipient nephropathy. This is very important because early identification of CKD and timely detection are global priorities as it can prevent further complications and reduce the cost of treatment.

Generally, in the advanced stages of CKD, creatinine does not correlate well with eGFR [21]. However, in this study, we have managed to demonstrate that both urine NGAL and cystatin C were elevated in patients with advanced kidney disease thus further supporting the utility of urine NGAL in advance kidney disease.

However, in contrast to previous studies [22], this study showed that cystatin C has a superior correlation with eGFR as compared to urine NGA in the early detection of diabetic nephropathy. These contradictory findings may result from the different sample size, different methods and assays that has been adopted, and different populations. However, a few studies have evaluated NGAL in diabetic nephropathy as compared to acute kidney injury. This is because cystatin C may be more useful in the early detection of chronic nephropathy for some unknown reasons. Therefore, more research on urine NGAL should be done in the setting of chronic kidney disease.

In conclusion, we have demonstrated that urine NGAL has outperformed other renal markers that are currently used to diagnose diabetic nephropathy. It is able to predict renal impairment at an early stage in diabetic nephropathy. However, there are a few limitations of the present study. First of all, this is a single-centre study. The small sample size and limited number of tests may minimize the validity of the results. Secondly, as mentioned above, the eGFR was established by using cystatin C, which may have produced a biased correlation.

Further research on urine NGAL in the setting of chronic kidney disease seems warranted.

Copyright information

© Research Society for Study of Diabetes in India 2012