Correlation analysis between Tervaert glomerular classification and clinical indicators in patients with type 2 diabetic nephropathy

1 Introduction

Diabetic nephropathy (DN) is a serious complication of diabetes and has become the main cause of end-stage renal disease (ESRD). About 40% of people with diabetes eventually develop DN [1, 2]. It is currently believed that DN does not have an exact pathogenesis but that multiple factors are involved in disease progression, including abnormal renal hemodynamics, metabolic abnormalities caused by hyperglycemia, abnormal metabolism of vasoactive substances, and the activation of renin-angiotensin system (RAS) [3, 4]. Several studies have shown that macrophage-mediated inflammation also plays an important role in the occurrence and development of DN [5, 6]. DN has many kinds of lesions, among which the most obvious is the glomerular lesion; the most typical change that this type of lesion manifests as is the hyperplasia of extracellular matrix, which is also the main cause of ESRD. In the early stage, glomerulus manifests primarily as the thickening of basement membrane, mesangial cell proliferation, and extracellular matrix accumulation. In the end stage, glomerulus could present with irreversible tuberous sclerosis and loss of function. And there is no effective treatment for DN at present [7].

In 1989, Mogensen divided DN into five stages according to clinical stages, mainly for type 1 DN [8]. However, the pathogenesis of type 1 diabetes is completely different from that of type 2 diabetes, and the occurrence of renal damage between these two is not completely the same. Type 1 diabetes is mostly caused by autoimmune deficiency or genetic factors, resulting in low islet function. Type 2 diabetes is associated with obesity, hypertension, hyperlipidemia, and other metabolic diseases, which may lead to insufficient insulin secretion or insulin resistance [9, 10]. The classification of DN has always been controversial. Prior to 2010, Tervaert conducted a comprehensive analysis of type 1 and type 2 diabetes lesions and classified glomerular lesions into the following four classes:

• Class I: Glomerular Basement Membrane Thickening. No obvious lesions were observed in the glomeruli under light microscope, but the thickening of the basement membrane in the glomeruli was observed under electron microscope;

• Class II: Mesangial Expansion, Mild (IIa) or Severe (IIb). In Class IIa, mild mesangial cell proliferation in more than 25% of the glomeruli was observed under light microscope, and mild mesangial cell proliferation and basement membrane thickening were observed under electron microscope. In class IIb, severe mesangial hyperplasia was observed in more than 25% of glomeruli under light microscope, and diffuse thickening of basement membrane was observed under electron microscope;

• Class III: Nodular Sclerosis (Kimmelstiel-Wilson lesions). The formation of one or more K-W nodules and diffuse thickening of the basement membrane were observed under light microscope, and obvious mesangial matrix accumulation was observed under electron microscope;

• Class IV: Advanced Diabetic Glomerulosclerosis. Global sclerosis was observed in more than 50% of the glomeruli under light microscope, and accumulation of fibrous material in the mesangial region was observed under electron microscope [11].

The Tervaert classification not only emphasized the pathological changes of glomeruli but also added quantitative analysis of renal tubules and vascular lesions to make a more comprehensive assessment for DN.

Pathological assessment is of great value in the diagnosis and treatment of renal diseases. Until the present, the gold standard has always been renal biopsy [12, 13]. However, in the early stage of DN, many patients only provide microalbuminuria and do not pay enough attention to renal protection. Our study aims to evaluate the correlation between the clinical and pathological conditions of DN, to assess the degree of renal lesions through clinical indicators in the early stage, and to provide new perspectives for the prevention and treatment of DN.

2 Materials and Methods

3 Results

3.1 Renal pathology of different classifications in patients with DN

In the early stage of DN, the disease pathology was only mild mesangial hyperplasia. As the disease progressed, the mesangial matrix accumulated, leading to the formation of Kimmelstiel-Wilson nodules and eventually developing into the global sclerosis. We used the immunofluorescence and pathological staining to observe the renal change in different classifications of patients with DN. Data show that IgG and Alb were both deposited in diffuse spheric linear deposition on the basement membrane, which was consistent with the typical progression of DN (Figure 1A). PAS staining was used to observe the mesangial matrix accumulation and MASSON staining was used to observe the glomerulus sclerosis in each class (Figure 1B). We found that in Classes I and IIa, the glomerulus showed only mild mesangial proliferation, and there was no obvious fibrosis. As the disease progressed, significant extracellular matrix proliferation was observed in Class IIb. At the same time, we could observe the tuberous sclerosis in Class III. And in Class IV, more than half the glomeruli hardened, which indicated that the disease had entered an irreversible stage.

Figure 1

Renal pathological changes with different Tervaert classification in patients with DN. (A) IgG and Alb staining were detected to show the linear deposition of different classification in patients with DN. (B) PAS and MASSON staining were used to observe the mesangial matrix accumulation of different classification in patients with DN. Alb, albumin; IgG, immunoglobulin G; DN, diabetic nephropathy.

3.2 Tervaert glomerular classification is well-correlated with clinical indicators

The clinical treatment of renal diseases often requires the combination of pathological diagnosis. To further study the relationship between pathological classification and clinical indicators in patients with DN, we collected general clinical indicators, including age, HbA1c, BMI, SBP, DBP, MAP, proteinuria, serum Cr, blood urea nitrogen, eGFR, and serum uric acid (Table 1). Correlation analysis data show that Tervaert classification was positively correlated with proteinuria (r = 0.639, P < 0.001), serum Cr (r = 0.339, P = 0.012), and blood urea nitrogen (r = 0.388, P = 0.004). Tervaert classification was negatively correlated with eGFR (r = −0.325, P = 0.016) (Figure 2A–D). However, DN had no obvious correlation with age, BMI, HbA1c, SBP, DBP, and serum uric acid. These results indicate that the changes in proteinuria, serum Cr, blood urea nitrogen, and eGFR could be good indicators to predict renal lesions in patients with DN.

Table 1

Clinical characteristics of type 2 DN patients (Mean ± SD, M [P25, P75])

Patient characteristic	DN (n = 54)
Gender (male/female)	36/18
Age (years)	49.50 ± 8.10
HbA1c (%)	7.61 ± 1.62
BMI (kg/m2)	25.76 ± 3.12
SBP (mmHg)	143.94 ± 21.40
DBP (mmHg)	86.78 ± 12.83
MAP (mmHg)	105.83 ± 13.87
Proteinuria (g/24 h)	2.09 (0.09, 13.17)
Serum Cr (mmol/L)	115.13 ± 57.44
Blood urea nitrogen (mmol/L)	8.61 ± 3.20
eGFR (mL/min/1.73m2)	75.09 ± 31.18
Serum uric acid (mmol/L)	374.42 ± 90.39
Tervaert classification (n)
I	9
IIa	13
IIb	5
III	27

1. BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; SBP, systolic blood pressure; MAP, mean arterial pressure; DN, diabetic nephropathy; SD, standard deviation.

Figure 2

Correlations between Tervaert glomerular classification and clinical indicators. Tervaert classification was positively correlated with proteinuria (A), serum Cr (B), and blood urea nitrogen (C), and negatively correlated with eGFR (D) in patients with DN. P < 0.05 were considered statistically significant. DN, diabetic nephropathy; eGFR, estimated glomerular filtration rate.

3.3 The content of macromolecular protein in urine could well reflect the pathological changes of glomerulus in patients with DN

Proteinuria is an important indicator of renal disease, but its components are complex and different proteins often indicate different pathological changes. Macromolecular proteins cannot pass through the glomerular filtration membrane under physiological conditions. When renal disease occurred, the amount of these macromolecules in the urine was increased. Table 2 shows the urine protein composition from all patients and the correlations with Tervaert glomerular classification. Data show that the classification was positively correlated with TP (r = 0.563, P < 0.001), TRF (r = 0.453, P = 0.001), Alb (r = 0.544, P < 0.001), IgG (r = 0.627, P < 0.001), Alb/Cr (r = 0.587, P < 0.001), and TP/Cr (r = 0.622, P < 0.001) (Figure 3A–F). These results suggest that the content of these macromolecular proteins could well reflect the pathological changes of glomerulus.

Table 2

Urinary protein component of type 2 DN patients (Mean ± SD, M [P25, P75])

Patient characteristic	DN (n = 54)
TP (mg/L)	1274.93 (93.38, 9481.46)
TRF (mg/L)	47.53 (0.74, 356.64)
RBP (mg/L)	1.48 (0, 35.68)
NAG (U/L)	12.95 (1.66, 48.60)
Alb (mg/L)	1129.60 ± 1099.02
IgG (mg/L)	87.61 (0.62, 873.12)
Cyc (mg/L)	0.22 (0.02, 6.57)
Cr (mmol/L)	6.50 (2.00, 24.46)
β2-MG (mg/L)	0.77 (0.02, 40.91)
α1-MG (mg/L)	29.49 ± 28.16
Alb/Cr (mg/gcr)	1501.20 ± 1455.69
TP/Cr (mg/gcr)	1732.25 (81.47, 11986.00)

1. TP, total protein; TRF, transferrin; RBP, retinol binding protein; Alb, albumin; Cyc, cystatin C; NAG, N-acetyl-β-D aminoglucosidase; IgG, immunoglobulin G; β2-MG, β2-microglobulin; α1-MG, α1-microglobulin; Cr, creatinine; DN, diabetic nephropathy; SD, standard deviation.

Figure 3

Correlations between Tervaert glomerular classification and the urine level of macromolecular protein in patients with DN. Tervaert classification was positively correlated with the urine level of TP (A), Alb (B), IgG (C), TP/Cr (D), Alb/Cr (E), and TRF (F) in patients with DN. P < 0.05 was considered statistically significant. Alb, albumin; Cr, creatinine; DN, diabetic nephropathy; IgG, immunoglobulin G; TP, total protein; TRF, transferrin.

3.4 Tervaert glomerular classification was correlated with the urine level of RBP, NAG, β2-MG, α1-MG, and Cyc in patients with DN

We further analyzed the correlation between Tervaert classification and the levels of RBP, NAG, β2-MG, α1-MG, and Cyc. These small molecular proteins can physiologically be filtered through the glomeruli and eventually reabsorbed into the tubules. Data show that the classification was positively correlated with RBP (r = 0.451, P = 0.001), NAG (r = 0.326, P = 0.016), β2-MG (r = 0.385, P = 0.004), α1-MG (r = 0.397, P = 0.003), and Cyc (r = 0.446, P = 0.001) (Figure 4A–E). These results suggest that the level of these small molecular proteins can assist to assess the pathological changes of the kidneys of patients with DN.

Figure 4

Correlations between Tervaert glomerular classification and the urine level of micromolecule protein in patients with DN. Tervaert classification was positively correlated with the urine level of Cyc (A), α1-MG (B), β2-MG (C), RBG (D), and NAG (E) in patients with DN. P < 0.05 was considered statistically significant. Cyc, cystatin C; DN, diabetic nephropathy; α1-MG, α1-microglobulin; β2-MG, β2-microglobulin; RBP, retinol binding protein; NAG, N-acetyl-β-D aminoglucosidase.

4 Discussion

DN is the most serious microvascular complication of diabetes and has become the main cause of ESRD[1, 2]. DN is mainly manifested as glomerular lesions, and the most typical change is mesangial matrix accumulation [8]. At present, there are few pathogenic markers that can be used for diabetic kidney disease (DKD) progression to ESRD, and glomerulosclerosis is the only prognostic factor [14]. In 2010, the American Journal of Nephrology published the Tervaert classification standard for the pathology of DN. The advantage of this scheme is that it is applicable to both type 1 and type 2 DN [11]. Glomerular lesions are the most important pathological changes in DN. Many studies have shown that mesangial cells were significantly proliferated, and the mesangial matrix was accumulated in the kidneys of patients with DN, which is also the main cause of glomerulosclerosis [9, 10]. In the early stage of DN, mesangial cells have self-limiting proliferation, but a continuous hyperglycemic environment may lead to the stagnation of mesangial cells in the G0/G1 phase, leading to cell hypertrophy and the accumulation of the extracellular matrix [15,16,17]. A single cell sequencing of the kidneys of DKD mice also supports this supposition, and the results of a related study show that the number of mesangial cells was decreased in the kidney. This phenomenon is possibly due to the advanced glomerular sclerosis [18]. In our study, we also found that mesangial cells show mild proliferation in Classes I and IIa, but cell components in the mesangial area were significantly decreased in Classes IIb, III, and IV. The data from MASSON staining also support this result. Data show that the mesangial matrix could be excessively accumulated with the disease progression and leads to the fibrosis of the whole glomerulus.

Renal biopsy has always been the golden standard for the diagnosis of DN. However, it is a risky, invasive test, and it has indications and contraindications. Therefore, we conducted the correlation analysis between clinical indicators and renal pathological classification to assess renal lesions in patients with DN who cannot receive a renal biopsy. In our results, we found that the Tevaert classification was positively correlated with proteinuria, Cr, and blood urea nitrogen. Cr and urea nitrogen are commonly used in clinical evaluation of renal function, but their sensitivity is low and thus they cannot reflect early renal damage. The increase of Cr and urea nitrogen often indicates that the disease has entered the chronic stage. eGFR is the most important indicator to evaluate glomerular function, and can effectively reflect the changes of glomerular filtration function [19,20,21]. This was also confirmed in our results, which showed a negative correlation between eGFR and Tervaert classification.

Urine examination is an important clue for early detection and diagnosis of renal disease. Proteinuria is an important indicator of renal injury; particularly, the occurrence of microalbuminuria is an important manifestation of DN [2, 5]. The change of glomerular hemodynamics has been recognized as the initiating factor of DN. Hyperglycemia can lead to thickening of the intima of the glomerular outflow and entrainment arteries, leading to hyperperfusion and hyperfiltration in the glomerulus [3]. This process leads to increased protein filtration from the capillary walls and causes proteinuria. At the same time, the disorder of glucose metabolism will lead to the release of a large number of vasoactive factors by the inherent cells of the kidney, resulting in increased vascular wall permeability and aggravated proteinuria. Different protein components often indicate different lesions in the kidney, and the composition of urinary protein is of great significance for the diagnosis of DN [22,23,24]. We collected the urine protein composition data in patients with DN. Alb, IgG, and TRF have high molecular weight and cannot pass through the glomerular filtration barrier under physiological conditions. When the glomerular filtration barrier is damaged, these proteins can pass through the glomeruli into the urine. Some studies have also shown that the level of IgG could represent the severity of kidney damage, and the deposition of IgG in the kidney has often indicated poor prognosis of the disease [25,26,27]. In our study, we found that both IgG and Alb were deposited in the glomeruli, and their urine levels were in direct proportion to the extent of glomerular lesions, which suggests that these macromolecular protein levels may be a good indicator of glomerular lesions. The level of urinary Cr is often influenced by many factors, including age, sex, and diet. Therefore, we used the ratio of TP and Alb to Cr to reduce this effect. We found that the classification was strongly correlated with the ratio of TP/Cr and Alb/Cr, which indicates that these ratios may serve as good indicators for renal function assessment.

Cyc is a small molecule protein, and it can be filtered through the glomeruli and reabsorbed in the tubules. Because of its stable metabolism, it has become an ideal marker to reflect glomerular filtration rate. Some studies in the literature suggest that eGFR equations incorporating cystatin C are superior to eGFR based on Cr alone for detecting kidney injury in the early stage. And Cyc could be used to predict worsening renal function after computed tomography coronary angiography [28, 29]. NAG, RBP, α1-MG, and β2-MG have small molecular weights, and they can be freely filtered through the glomerulus under physiological conditions and finally absorbed into the proximal tubule [30,31,32]. These indicators are commonly used to assess tubular damage. Interestingly, we found that the levels of these proteins were also associated with glomerular lesions. This is consistent with previous research [11]. The evaluation of DN often needs to be combined with glomerular, tubular, and vascular lesions. The level changes of these small molecular proteins can be better used to assist the clinical evaluation of DN renal lesions.

Currently, there is no effective treatment for DN. In the early stage of diabetes, strict control of hyperglycemia and hypertension, correction of intraglomerular hypertension, and improvement of lipid level can delay or prevent the occurrence and progression of DN. The glomerular damage of DN is irreversible, and thus all patients should pay adequate attention to renal protection once diabetes is diagnosed. Urine composition analysis is a non-invasive test with good operability. Therefore, our study has important clinical significance and can provide ideas for the prevention and treatment of early DN. However, this study has some limitations. In identifying renal lesions in patients with DN, patients with impaired functioning of renal tubules and renal vessels because of reasons not connected with renal lesions were also unavoidably included, owing to the nature of the mechanism of evaluation; and this is a limitation that is expected to be addressed in future studies.

5 Conclusion

In this study, we found that Tervaert classification was positively correlated with proteinuria, serum Cr, and blood urea nitrogen and negatively correlated with eGFR in patients with DN. Moreover, Tervaert classification was positively correlated with TP, TRF, RBP, NAG, Alb, IgG, Cyc, β2-MG, α1-MG, Alb/Cr, and TP/Cr. Our results show the important correlation between Tervaert classification and clinical indicators, which provides ideas for the early prevention and treatment of DN.