Neutrophil Gelatinase-associated Lipocalin: As a Predictor of Early Diabetic Nephropathy in Type 2 Diabetes Mellitus

Introduction

Diabetic nephropathy (DN) is a major microvascular complication of diabetes mellitus, accounting 20%–40% of populations requiring renal replacement therapy.[12] Pathologically, it is a diffuse process involving glomerular endothelial cells, tubular epithelial cells, and interstitium. It evolves through glomerular hyperfiltration, silent phase (normoalbuminuria), incipient nephropathy (microalbuminuria), overt nephropathy (macroalbuminuria), and established renal failure.[3] Whether tubular epithelial and interstitial injury in diabetics is primary, that precedes glomerular endothelial injury or secondary to albuminuria, is a subject of debate.[4] Chronological age of appearances of glomerular basement membrane and tubular basement membrane thickening on histopathology of type 1 diabetes suggests that tubular injury is independent of glomerular injury as both appears after similar duration of disease.[56] As with other renal diseases, outcome of diabetic kidney disease (DKD) is better determined by tubulointerstitial changes than glomerular changes.[789] Classically, albuminuria, which is considered as a hallmark of DN, is a marker of glomerular injury (glomerular phase). However, Pun et al. in 1990 and later by many other researchers suggested that tubular injury can occur even at stage that precedes microalbuminuria (tubular phase).[1011] Low Vitamin-D and erythropoietin in normoalbuminuric diabetic patients imply early involvement of tubulointerstitial component rather than glomerular.[12]

Chronic hyperglycemia leads to increased production of oxidative free radicals, advanced glycosylated end-products,[13] vasoactive amines (angiotensin-II), and cytokines (transforming growth factor-β), which result in functional followed by structural damage to tubulointerstitium.[89] Functional damage reflected as tubular hypertrophy due to increased absorption of glucose,[14] dysfunction of transporter proteins such as megalin and cubilin,[15] and defective activity of lysosomal system that leads to increased excretion of protein which can serve as biomarker to diagnose while glomerular function is still normal. Hence, now, investigators are in search of ideal biomarkers to identify DN in the stage of functional damage so that measures can be taken to prevent the progression or retard the disease process.

Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker of renal tubular injury, upregulated in distal tubules and collecting duct, is extensively evaluated for acute kidney injury (AKI).[16] It is a 25 kDa glycoprotein with 178 amino acid belonging to lipocalin superfamily.[17] It is a constituent of specific granules and exists in neutrophil as a part of NGAL-gelatinase complex.[18] It is involved in antimicrobial defense mechanism and upregulated in systemic bacterial infection.[17181920] It plays a protective role in epithelial injury by its anti-apoptosis effect.[21] As it is not produced by burnt out nephron, it is supposed to be a marker of active injury and represents mass of salvageable nephrons.[22] Its usefulness as a biomarker in chronic kidney disease (CKD) and later in DN has been suggested by Bolignano et al.[2324] Other than renal parenchyma disease, heart failure, endothelial dysfunction, preeclampsia, infection, inflammation, and malignancy are associated with increase in NGAL level independent of glomerular filtration rate (GFR).[172526272829]

In our current study, we aimed to compare markers of glomerular injury (urinary albumin) with that of tubular injury (NGAL), in a subset of type 2 diabetics mellitus (T2DM) those never been exposed to angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) and to establish diagnostic role of the later in nonalbuminuric DN. This is to avoid bias as ACEIs or ARBs known to reduce albuminuria and tubular biomarkers.[30] Again, there is no study based on Indian population to decide diagnostic accuracy of NGAL in diabetic populations.

Materials and Methods

Results

Comparison of baseline characteristics among different groups

Table 1 shows comparison of baseline characteristics between the study groups. Of 198 study participants, there were 36 normoalbuminuric, 59 microalbuminuric, 49 macroalbuminuric, and 54 control populations. In the baseline characteristics among the four groups, there was no significant difference in the distribution of sex, age, and BMI while significant difference in regard to duration of disease, SBP, DBP, HbA1c, eGFR, total cholesterol (TCH), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglyceride (TG), serum albumin, antihypertensive drugs (AHDs), lipid-lowering agents (LLAs), and diabetic retinopathy (DR). Duration of diabetes and HbA1c were progressively higher among normoalbuminuria, microalbuminuria, and macroalbuminuria while reverse was true for eGFR and HDL-C. There was no significant difference between the control and normoalbuminuria with respect to SBP. Difference between the normoalbuminuria and microalbuminuria groups with regard to TCH and LDL-C was also nonsignificant. Macroalbuminuria group had significantly low serum albumin and higher incidence of DR than other groups. Usage of AHDs and LLAs was more frequently observed in the macroalbuminuria group than in the normoalbuminuria and microalbuminuria groups.

Table 1 Comparison of clinical characteristics among different groups

Comparison of serum neutrophil gelatinase-associated lipocalin and urine neutrophil gelatinase-associated lipocalin values according to albuminuria

Table 1 and Figure 1 show comparisons in values of uNGAL, sNGAL, uNGAL/urine creatinine (uCr), and other markers among different groups. The sNGAL, uNGAL, and uNGAL/uCr were significantly higher in diabetic individual than in the control. Multiple comparisons indicated the significant difference in sNGAL, uNGAL, and uNGAL/uCr values between the groups (P < 0.05). Difference of above values between control value and normoalbuminuria was also significant (P < 0.05).

Close

Figure 1

Showing median value of various biomarkers and eGFR in different stage of diabetic nephropathy. (a) Serum NGAL (ng/ml), (b) Urine NGAL/urine creatinine –(ng/mg), (c) Serum creatinine (mg/dl) and (d) GFR (ml/min/1.73m²)

Export to PPT

Associations of albuminuria with serum and urinary markers in type 2 diabetic patients

Spearman's rank correlation coefficient (rho; ρ) was calculated between the various serum and urinary marker with each other. The sNGAL showed good (P < 0.05) positive correlation with UACR (ρ = 0.92), uNGAL (ρ = 0.77), uNGAL/uCr (ρ = 0.93). Similarly, uNGAL and uNGAL/uCr showed good (P < 0.05) positive correlation with UACR (ρ = 0.0.77, 0.88) and sNGAL (ρ = 0.77, 0.93).

Univariate binary logistic regression analysis showed that predictors, i.e., ACR, serum creatinine, GFR, BMI, duration of diabetes, SBP, DBP, HbA1c, AHD, and uNGAL/uCr, were found individually significant with sNGAL (P < 0.05). In the multivariate analysis, variables included were ACR (adjusted odds ratio [AOR] = 1.04, 95% confidence interval [CI] = 1.02–1.07, P = 0.002), GFR (AOR = 0.92, 95% CI = 0.84–1.00, P = 0.054), and uNGAL/uCr (AOR = 1.05, 95% CI = 1.01–1.10, P = 0.023). Similarly, univariate analysis, for uNGAL/uCr, predictors, i.e., ACR, serum creatinine, GFR, BMI, duration of diabetes, SBP, DBP, HbA1c, AHD, sNGAL, TG, TCH, HDL, and LDL, were found individually significant (P < 0.05), but in the multivariate model, variables included were ACR (AOR = 1.03, 95% CI = 1.02–1.04, P < 0.001), SBP (AOR = 1.10, 95% CI = 0.99–1.23, P = 0.074), AHD (AOR = 0.40, 95% CI = 0.08–2.14, P = 0.284), and LDL (AOR = 1.02, 95% CI = 1.01–1.04, P = 0.025) [Table 2].

Table 2 Univariate and multivariate binary logistic regression analysis

Receiver operating characteristics curve analysis of serum neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin/urine creatinine, and urinary albumin-creatinine ratio

Receiver operating characteristics (ROC) curve analyses were performed to define the diagnostic value of sNGAL, uNGAL, uNGAL/uCr, and UACR in identifying DN with or without proteinuria [Table 3 and Figure 2]. The uNGAL showed a very good diagnostic profile, describing an AUC of 0.996 (95% CI: 0.991–1.000, P < 0.001) with a best cutoff value of 21.31 ng/ml (sensitivity 95.1%; specificity 100%). This is followed by sNGAL, showing an AUC of 0.993 (95% CI: 0.998–1.000, P < 0.001) and a best cutoff value of 78.73 ng/ml (sensitivity 95.1%; specificity 100.0%) [Table 3]. Similarly, uNGAL/uCr had a good diagnostic profile, showing an AUC of 0.996 (95% CI: 0.992–1.00, P < 0.001) and a best cutoff value of 24.96 ng/g (sensitivity 93.1%; specificity 100.0%). UACR had least diagnostic profile among four with an AUC of 0.926 (95% CI: 0.89–0.96, P < 0.001) and a best cutoff value of 29.5 mg/g (sensitivity 75.5%; specificity 100.0%).

Table 3 Receiver operating characteristics curve analysis of serum neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin/urine creatinine and urinary albumin-creatinine ratio

Close

Figure 2

Receiver operating characteristics curve analysis of serum neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin, urine neutrophil gelatinase-associated lipocalin/urine creatinine, and urinary albumin-creatinine ratio

Export to PPT

Discussion

Diabetics are continuously exposed to metabolic and hemodynamic stress. Metabolic stress is related to hyperglycemia and hyperlipidemia. Hemodynamic stress includes hypertension. Hyperglycemia, hyperlipidemia, and hypertension together lead to inflammation and atherosclerosis resulting endothelial dysfunction and tubulointerstitial damage. While tubular damage leads to increase in tubular biomarkers, endothelial dysfunction progresses to albuminuria. In 1993, Yaqoob et al. concluded that endothelial dysfunction and tubular injury may precede microalbuminuria.[31] Hence, even diabetics with preserved eGFR are at increased risk for AKI in comparison to healthy individuals, indicating decreased renal reserve. Hence, several tubular markers such as NGAL, KIM-1, NAG, MCP, and urinary liver-type fatty acid-binding protein (uLFABP) may act as clinical indicators of onset and progression of DN.[32]

In our study, we observed sNGAL and uNGAL values across nondiabetic control and three groups of diabetics with normal serum creatinine – normoalbuminuria, microalbuminuria, and macroalbuminuria. There was significant difference in median value of sNGAL and uNGAL between diabetes patients compared to well-matched control, suggesting that NGAL can be a useful marker of onset of DN even when serum creatinine is normal range. This may be due to protective response of NGAL in response to metabolic and hemodynamic stress. Intergroup comparison among three groups of T2DM patients shows that difference was significant with highest median value in macroalbuminuria and lowest in normoalbuminuria, which suggest that sNGAL and uNGAL can be used to predict progression of DN. Again, sNGAL and uNGAL correlates positively with albuminuria suggesting that former correlates with severity of renal involvement. This concludes that NGAL can be used as a marker to stratify diabetes nephropathy into different stages.

Results obtained above were in agreement with previous studies which reported similar tendency for NGAL[33] and other biomarkers of tubular injury such as NAG, MCP, uLFABP, and KIM-1 whose level corresponds with degree of proteinuria[2434353637] but contradicted by Kim et al., where there were no significant differences in NGAL value between normoalbuminuria and microalbuminuria groups.[38]

Normoalbuminuric diabetes patients had significantly higher median sNGAL and uNGAL value than control group, suggesting that tubular injury may precede glomerular injury and NGAL can serve as an early laboratory marker of DN, supporting the previous study.[243237394041] This also suggested that tubular dysfunction in early DN is not secondary to albuminuria, but consequences of metabolic and hemodynamic stress are associated with chronic hyperglycemia contradicting previous hypothesis.[479] As NGAL is a marker of distal tubule injury, it is not going to be affected by albuminuria or proteinuria, which again supports our earlier hypothesis that tubular injury may precede glomerular injury.[36] Markers of tubular injury predicting onset of albuminuria and progression of disease have been observed in various other studies using other biomarkers such as uLFABP, KIM-1.[303242] In 2012, Kim et al. suggested in their trial that differences in NGAL value between control, normoalbuminuria, and microalbuminuria were nonsignificant, contradicting NGAL as an early marker of DN.[38]

In our study, sNGAL, uNGAL, and uNGAL/uCr show strong positive correlation with one another and that with albuminuria and duration of DM but negative correlation with eGFR which is supported by other study also.[33374043] However, in multivariate binary logistic regression, only UACR and uNGAL/uCr were predictors of sNGAL whereas UACR and LDL-C were predictors of uNGAL/uCr. A strong correlation between NGAL and albuminuria suggests glomerular and tubular injury may occur as parallel processes. Lacquaniti et al. and Zachwieja et al. in their studies found no correlation between sNGAL and uNGAL. They also found no correlation NGAL with renal function such as eGFR and creatinine.[3941] Bolignano et al., Zachwieja et al., and Fathimah et al. concluded that NGAL value does not correlate with glycemic control (HbA1c) which is supported by our study but contradicted to the study by Lacquaniti et al.[2433373941] Nauta et al. concluded that it is not NGAL but uLFABP that predicts decline in renal function (eGFR) in diabetic individuals, whereas Nielsen et al. suggested that none of the tubular markers can predict decrease in renal function (true GFR-51Cr-EDTA).[3644]

Advantage of our study is that we have included patients with better control of blood sugar and blood pressure; normal serum creatinine and GFR >60 ml/min/1.73 m². As NGAL might be elevated in nonrenal causes, we have excluded those with cardiac dysfunction and malignancy. On the contrary to other study, we have excluded those who are already on ACEI and/or ARB as protective role of both of them on glomerular and tubulointerstitial compartments has already been proved in both animal and human study.[304445] Some trials conducted also showed protecting role of ACEI and/or ARB in DN using NGAL as a marker.

Major limitation of our project was small sample size, single-center study with cross-sectional design. Small sample size is probably related to strict criteria of excluding those who are already on ACEI or ARBs. Other than DN, AKI, infection, inflammation, and malignancy are responsible for raise sNGAL and uNGAL value suspecting specificity of NGAL as diagnostic marker of DN. Elevated circulating NGAL could also be due to endothelial dysfunction of glomerular capillaries or extra-renal blood vessels.[26] As diabetes is a metabolic disease affecting multisystem and endothelium of blood vessel is main target organ, extra-renal source of NGAL due to endothelial dysfunction leading to elevated circulating NGAL cannot be ruled. Again, high uNGAL could be the consequences of elevated circulating NGAL. Hence, along with NGAL, we should use other marker of tubular injury for better assessment of renal function. We can interpret our limitation in better way that NGAL, particularly sNGAL, can be marker of endothelial dysfunction that can be predictor of microvascular and macrovascular complication in diabetes individuals in early stage. Other limitation of our study was that we did not collect 24 h sample which was due to technical infeasibility to collect sample.

Conclusion

Tubular injury may precede glomerular injury in diabetic patients, and NGAL can be used as biomarker to diagnosed DN even earlier to incipient nephropathy. Both sNGAL and uNGAL can predict albuminuria and be used as a noninvasive tool for diagnosis, staging, and progression of DN.