The role of serum and urinary biomarkers in the diagnosis of early diabetic nephropathy in patients with type 2 diabetes

Introduction

Diabetic nephropathy (DN), also known as diabetic kidney disease, is a common and severe microvascular complication of type 2 diabetes that can result in end-stage kidney disease. With the increasing incidence of type 2 diabetes worldwide, DN has become a global health concern (Wan, Xu & Dong, 2015). The effective treatment of DN requires glycemic control and antihypertensive measures (Wan, Xu & Dong, 2015). It is important to identify DN during its early stages as prompt treatment can reduce the medical and economic burden of this disease (Tziomalos & Athyros, 2015; Gudehithlu et al., 2018).

Currently, microalbuminuria is the most widely investigated biomarker for the diagnosis of DN. However, its diagnostic value in early-stage DN is limited as renal injury commonly precedes proteinuria. In a previous study, Perkins et al. (2010) demonstrated that only 52.2% of patients with advanced-stage DN presented with proteinuria during a 12-year follow-up study. In another study, only 33% of patients with type 1 diabetes having confirmed microalbuminuria developed clinically-apparent kidney disease (Rossing, Hougaard & Parving, 2005). Other studies have reaffirmed these findings, demonstrating that a considerable portion of diabetic patients with renal dysfunction do not have proteinuria (An et al., 2009; Kramer et al., 2003; Zachwieja et al., 2010). Therefore, there is a dire need for more sensitive and specific biomarkers for the diagnosis of early-stage DN.

It has been well-established that glomerular damage, tubular injury, inflammatory responses, and oxidative stress contribute to the development of diabetic kidney disease (Gurley et al., 2018). In this study, we evaluated 13 serum and urinary biomarkers involved in glomerular damage, tubular injury, inflammation, and oxidative stress, for their potential application in the diagnosis of early-stage DN.

Materials and Methods

Results

The parameters of each group are shown in Table 1. The Tf, IgG, PCX, CysC, NAG, α1MG, 8-OHdG, IL-18, and UA levels of patients in the macroalbuminuria group were higher than those in the other three groups (P < 0.05 for UA, P < 0.001 for other parameters). The NGAL, TNF-α, and NLR levels in the macroalbuminuria group were higher than those of the healthy control and normoalbuminuria groups (P < 0.001). The NLR levels in the macroalbuminuria group were higher than those in the microalbuminuria group (P < 0.05). TBIL in the macroalbuminuria group was lower than the normoalbuminuria and microalbuminuria groups (P < 0.001). The Tf, IgG, PCX, NGAL, NAG, α1MG, 8-OHdG, TNF-α, IL-18, and NLR levels of patients in the microalbuminuria group were higher than the healthy controls and normoalbuminuria groups (P < 0.05 for NLR, P < 0.001 for other parameters). The Tf, PCX, NGAL, NAG, α1MG, TBIL, and IL-18 levels in patients with normoalbuminuria were higher than the healthy controls (P < 0.001 or P < 0.05).

The multivariate linear regression was performed for the biomarkers correlated with UACR, and adjustments were made to account for risk factors, such as gender, age, diabetes duration, HbA1c, eGFR, LDL, systolic blood pressure and diastolic blood pressure. Urinary Tf, IgG, NGAL, and TNF-α were found to be significantly related to the UACR in patients with type 2 diabetes (P < 0.001) (Table 2). The diagnostic value of urinary Tf, IgG, NGAL, and TNF-α were analyzed by ROC curves (Fig. 1) and the results showed that, in terms of AUC values, IgG > NGAL > Tf > TNF-α; in terms of sensitivity, NGAL > IgG > Tf > TNF-α; and in terms of specificity, IgG > Tf > TNF-α > NGAL (P < 0.001) (Table 3).

The diagnostic value of the combination of urinary Tf, IgG, NGAL, and TNF-α were analyzed with ROC curves (Fig. 2). The results showed that the combination of urinary Tf, IgG, NGAL, and TNF-α had the highest AUC when compared with the individual biomarkers (P < 0.001) (Table 3).

Discussion

Diabetic nephropathy is a chronic and progressive kidney disease characterized by microalbuminuria and decreased glomerular filtration rates, which can eventually lead to end-stage renal disease and death. However, its early diagnosis and intervention could allow for the prompt treatment of DN, which would delay the pathological progression of the disease. In this study, we evaluated several potential biomarkers for their diagnostic potential for detecting early-stage DN. The primary finding of this study was that urinary Tf, IgG, NGAL, and TNF-α levels, both individually and in combination, are valuable biomarkers for assessing early-stage DN in patients with type 2 diabetes.

While some highly sensitive and specific biomarkers were previously identified for the diagnosis of DN (Hara et al., 2012), this study covered multiple parameters involved in the four major mechanisms of DN and evaluated the diagnostic value of them individually and in combination. The pathogenic mechanisms of DN include glomerular damage, tubular injury, inflammation, and oxidative stress (Wada & Makino, 2013; Arora & Singh, 2013). Therefore, biomarkers involved in these processes may potentially serve as clinically-relevant diagnostic parameters of DN (Gluhovschi et al., 2016).

Glomerular damage leads to increased urinary excretion of plasma proteins, such as Tf, IgG, and PCX (Currie, McKay & Delles, 2014). Tf is a plasma protein with a similar molecular weight as albumin. However, Tf is less anionic than albumin, which makes it more easily filtered via the glomerular barrier. Previous studies found that high urinary Tf levels in type 2 diabetic patients with and without microalbuminuria (Kanauchi, Akai & Hashimoto, 2002; Narita et al., 2004). A 5-year follow-up study found that elevated urinary Tf levels could be used to predict the progression to microalbuminuria in patients with type 2 diabetes (Narita et al., 2006). While urinary IgG is also elevated in some patients prior to microalbuminuria (Narita et al., 2004). This suggests that urinary IgG levels may be a clinically-relevant biomarker for the early assessment and diagnosis of DN in patients with type 2 diabetes (Narita et al., 2006). In the current study, we found that urinary Tf and IgG may be used alone or in combination with other biomarkers for the diagnosis of early-stage DN. PCX, a negatively charged sialoglycoprotein, is expressed in podocytes and plays an essential role in maintaining the function of glomerular podocytes (Lee & Choi, 2015). Urinary PCX levels were found to be associated with UACR levels and were also increased in diabetic patients with normoalbuminuria (Shoji et al., 2016). While PCX may be an excellent marker for podocyte injury (Hara et al., 2012), it is not considered to be an excellent diagnostic parameter for DN by the current study.

Renal tubular injury is a critical characteristic of DN (Kumar, Robertson & Burns, 2004). Urinary tubular injury markers may increase in diabetic patients, even before the onset of microalbuminuria (Russo et al., 2009). NGAL, also named lipocalin-2, is a secretory protein that is released during renal injury and serves as a marker of acute kidney injury (Bolignano et al., 2009). CysC, an inhibitor of the cysteine protease, is freely filtered through the glomeruli and reabsorbed by the proximal tubule (Jeon et al., 2011). Previous studies demonstrated that urinary NGAL and CysC were higher in diabetic patients and positively associated with UACR (Assal et al., 2013; Mahfouz, Assiri & Mukhtar, 2016; Rao et al., 2014). Our study found that urinary NGAL was increased in diabetic patients with microalbuminuria, as well as diabetic patients with normoalbuminuria when compared with the healthy controls. However, CysC was only higher in patients with macroalbuminuria. Similarly, another study showed that CysC was only elevated in advanced-stage diabetic kidney disease and remained at normal levels during the early stages of the disease (Ogawa et al., 2008). NAG is a lysosomal enzyme expressed in proximal tubules and was previously found to be a sensitive marker of tubular injury (Bazzi et al., 2002). Urinary NAG was increased in type 2 diabetic patients compared to non-diabetic subjects, with higher levels in patients with microalbuminuria than those with normoalbuminuria levels (Sheira et al., 2015; Bouvet et al., 2014). The α1MG protein is freely filtered and reabsorbed through renal tubules. Urinary α1MG was also increased before the onset of microalbuminuria (Shore, Khurshid & Saleem, 2010; Hong et al., 2003). Although we failed to identify NAG and α1MG as early diagnostic markers, NAG was significantly elevated in the macroalbuminuria and microalbuminuria groups. Interestingly, α1MG levels differed among the four groups, suggesting that these two parameters might play a role in the assessment of disease severity and progression.

Oxidative stress is one of the mechanisms for DN (Forbes, Coughlan & Cooper, 2008). 8-OHdG is a stable oxidative product and a marker of DNA damage (Valavanidis, Vlachogianni & Fiotakis, 2009). Urinary 8-OHdG was increased in patients with type 2 diabetes and was found to be a marker for monitoring the progression of DN (Waris et al., 2015; Hinokio et al., 2002). However, another study found that urinary 8-OHdG levels were similar in patients with DN and diabetic patients with normoalbuminuria (Serdar et al., 2012). In our study, although elevated urinary 8-OHdG levels were found in in the macroalbuminuria and microalbuminuria groups, urinary 8-OHdG was not considered to be a valuable diagnostic biomarker for early-stage DN. Bilirubin is an endogenous antioxidant, and low serum bilirubin levels are considered to be a risk factor for diabetic kidney damage (Okada et al., 2014). UA was reported to be associated with oxidative stress and endothelial dysfunction (Jalal et al., 2011), and was previously considered as a risk factor for DN (Jalal et al., 2010). However, another study found that UA levels were not associated with the progression of DN (Ahola et al., 2017). In the current study, TBIL and UA levels in the macroalbuminuria group differed from the normoalbuminuria and microalbuminuria groups, but there were no differences between the normoalbuminuria and microalbuminuria groups. Therefore, these two biomarkers were not considered as potential candidates for the detection of early-stage DN. However, they may be useful for assessing the severity of renal damage.

Diabetic nephropathy is a low-grade inflammatory disease (Wada & Makino, 2013), yet this study revealed a strong connection with TNF-α, which is a well-known inflammatory cytokine associated with renal injury (Navarro et al., 2006). Compared to the normoalbuminuria group, a 90% increase in urinary TNF-α excretion was present in type 2 diabetic patients with microalbuminuria, which was further correlated with the UACR (Navarro et al., 2008). Urinary IL-18 was positively associated with urine protein excretion and may be a predictive factor for assessing the progression of DN (Liu et al., 2015; Nakamura et al., 2005), yet one study disagreed with this finding (Nadkarni et al., 2016). NLR is a novel inflammatory marker known to be increased in patients with type 2 diabetes when compared with the healthy controls. NLR was also higher in patients with microalbuminuria group when compared with the normoalbuminuria group (Huang et al., 2015; Khandare et al., 2017). Our study found elevated levels of TNF-α, NLR, and IL-18 in patients with microalbuminuria and macroalbuminuria that were significantly higher than those detected in the healthy controls and type 2 diabetic patients with normoalbuminuria. However, only TNF-α was identified as a candidate biomarker for the detection of early-stage DN.

Using the ROC analysis, four biomarkers (urinary Tf, IgG, NGAL, and TNF-α) showed diagnostic value in terms of high AUC, in type 2 diabetic patients with DN. Urinary IgG showed a high specificity of 92.3%, yet the sensitivity was relatively low. Similarly, urinary NGAL was highly sensitive but suffered from its low specificity. Further analyses revealed the combination of these four biomarkers had better diagnostic value, in terms of AUC values, than those of the biomarkers individually. The AUC was as high as 0.922 for the combined biomarkers, indicating a superior value in the detection of early-stage DN.

Conclusions

In conclusion, this study found that urinary Tf, IgG, NGAL, and TNF-α levels may be used as biomarkers for the diagnosis of early-stage DN in patients with type 2 diabetes. However, the combination of these four biomarkers showed higher sensitivity and specificity in comparison to the biomarkers when used individually. The four biomarkers assessed in this study are involved in glomerular damage, tubular injury, and inflammation, respectively, which represents different aspects of the pathogenesis of DN. The combined use of these novel biomarkers may improve the detection of early-stage DN. However, further studies using larger sample sizes should be performed to establish the diagnostic value of other biomarkers for detection of early-stage DN.

Supplemental Information