Combined detection of urinary biomarkers noninvasively predicts extent of renal injury in patients with early diabetic kidney disease with kidney qi deficiency syndrome: A retrospective investigation

1 INTRODUCTION

Diabetic kidney disease (DKD) is a common health problem in China and is one of the major microvascular complications of diabetes mellitus (DM) (Wang et al., 2019). Approximately 30–40% of patients with DM develop DKD, which is the dominant cause of end-stage renal disease (ESRD) (Umanath & Lewis, 2018). Therefore, clinical incipient diagnosis and noninvasive detection are important for preventing the progression of DKD to ESRD. It has been reported that albuminuria is one of the most typical clinical changes in the early stages of DKD, and the level of albuminuria excretion can be determined as a way to screen patients with DKD (Groop et al., 2018). However, growing evidence has recently suggested that detection of albuminuria alone is not comprehensive and sensitive for patients with DKD, especially for those with inchoate and latent injuries in the glomeruli and renal tubules (Colhoun & Marcovecchio, 2018). In contrast, although renal biopsy may be the best method for diagnosing DKD at the incipient stage (Furuichi et al., 2018), it cannot be performed in all cases because of its invasion. Given these shortcomings in the clinic, more workable biomarkers in urine other than urinary albumin (UAlb) are crucially required to be explored for early diagnosis and prediction of renal injury in patients with DKD (Persson & Rossing, 2018).

Glomerular dysfunction is a major cause of DKD development (Abdel-Rahman et al., 2012). However, impaired absorption of filtered proteins from the renal tubular epithelium might also play many important roles in incipient patients with DKD (Vallon & Thomson, 2012). Increasing evidence has demonstrated that, as biomarkers, some tubular injurious indicators in urine have clinical implications (Fiseha, 2015). These urinary tubular biomarkers (UTBs) include urinary cystatin C (UCysC) (Khosravi et al., 2018; Kim et al., 2013), urinary N-acetyl-β-d-glucosaminidase (UNAG) (Vaidya et al., 2011), urinary kidney injury molecule (UKim)-1 (Moresco et al., 2018; Vaidya et al., 2011), urinary liver-type fatty acid-binding protein (UL-FABP) (Watanabe et al., 2018), urinary retinol-binding protein (URBP), urinary neutrophil gelatinase-associated lipocalin (UNGAL), urinary β2-microglobulin (Uβ2-MG), among others (Caplin & Nitsch, 2017; Curovic et al., 2018). Thereinto, UNAG and URBP have been reported to be associated with the progression of Type 2 DM (Curovic et al., 2018). In contrast, Maclsaac et al. (2014) considered that there was neither a clinical assay nor an adequate study focused on defining the prognostic value of urinary biomarkers in DKD. Meanwhile, Hsu et al. (2017) have reported that urinary biomarkers of tubular injury did not improve in a clinical model predicting chronic kidney disease (CKD) progression. Therefore, to date, the clinical significance of urinary biomarkers in predicting renal injury in patients with DKD has remained controversial.

In traditional Chinese medicine (TCM), DKD is recognized as Xiaoke (a disease with symptomatic polydipsia)-related nephropathy. According to the fundamental principles of TCM theory, the main pathogenesis of DKD is kidney asthenia (Lu et al., 2019; Sun et al., 2016). More importantly, in the clinic, Chinese herbal medicines (CHMs) formulas focused on nourishing kidney such as Tangshen Formula (TSF) (Yang et al., 2016) and Liuwei Dihuang Pills (LDP) (Lin et al., 2016) play many important roles in the treatment of Xiaoke and its related renal complications. Yang et al. (2016) have reported that TSF combined with conventional therapy may be effective in reducing urinary protein and UL-FABP, which is may be associated with the severity of DKD as a new urinary renal tubular biomarker. Correspondingly, our previous study based on 108 patients with Stage III Type 2 DKD in China, a unicentral observation with a cross-sectional design, has shown that 79 patients with deficiency states of both the kidney and spleen showed different increased UCysC, UNAG, and URBP levels, which are specific markers of renal tubular dysfunction rather than glomerular damage. An increase in UCysC has been closely related to UAlb in 30 patients with kidney deficiency syndrome (Wan et al., 2013). These results strongly suggest that the combined detection of urinary biomarkers can be used as a practical clinical technique to noninvasively forecast the extent of renal injury in patients with Type 2 DKD with kidney asthenia syndrome. Additionally, in patients with DKD with different TCM syndrome, Wang et al. (2019) and Xiao and Song (2021) also found that UTBs can be affected by CHM treatment, respectively. Hence, we hypothesized that UTBs might be related to the specific TCM syndromes in patients with DKD.

In this study, we aimed to further evaluate the clinical significance of combined detection of urinary biomarkers in noninvasively predicting the extent of renal damage in patients with early Type 2 DKD with kidney qi deficiency syndrome in an integrated traditional and western medical center, and preliminarily explore the potential biological link between UTBs and TCM syndrome in DKD.

2 MATERIALS AND METHODS 2.1 Study population

We retrospectively investigated urine samples from 92 patients with Type 2 DM who were consecutively enrolled at the Department of Endocrinology and TCM, Nanjing Drum Tower Hospital from March 2012 to April 2013. The study was approved by the Ethics Committee of the Affiliated Hospital of Nanjing University Medical School (Nanjing Drum Tower Hospital). The clinical trial was registered, and the data were published on the registration website (trial registration site: http://isrctn.org. Registration number: ChiCTR2000029784). All patients were categorized into three groups according to their albuminuria levels based on urinary albumin-to-creatinine ratio (UACR), including 20 patients with normoalbuminuria (<30 mg/g creatinine), 50 patients with microalbuminuria (30–300 mg/g creatinine), and 22 patients with macroalbuminuria (>300 mg/g creatinine) (Figure 1).

image

Trial flow diagram

The diagnostic criteria for DKD in Western medicine were developed with reference to “The Chinese guidelines for the diagnosis and treatment of DKD (Chen et al., 2021).” The eligible subjects meeting the following criteria were included in this study: (a) conforming to the diagnostic criteria for Type 2 DKD, (b) serum creatinine (Scr) <106 μmol/L, (c) stable renal function status without two-fold elevation of Scr for at least 5 months, and (d) no history of administration of renin-angiotensin system (RAS) inhibitors, including angiotensin-converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARB). Accordingly, the exclusion criteria were as follows: (a) renal disease other than DKD, (b) active urinary tract infection, (c) thyroid dysfunction, (d) severe liver dysfunction, (e) neoplastic disorders, (f) active or chronic persistent inflammatory disorders, (g) pregnancy, and (h) acute myocardial infarction.

2.2 Collection of kidney deficiency syndrome information

The diagnostic criteria for DKD in TCM syndrome were developed with reference to “Development and application of TCM syndrome differentiation guidelines for diabetic nephropathy (Liu et al., 2018).” Therefore, signs and symptoms of kidney deficiency syndrome of TCM diagnostic methods, including kidney qi deficiency syndrome, kidney yang deficiency syndrome, kidney yin deficiency syndrome, and kidney essence insufficiency syndrome (Chen et al., 2012; Mou et al., 2013; Wang & Zhang, 2017) were collected, respectively. The clinical manifestations include aching and weakness of loins and knees, dispiritedness and lassitude, frequent micturition, dripping urination, incontinence of urine, light-colored tongue with whitish fur and thin pulse, dizziness and tinnitus, insomnia and amnesia, flushed cheeks in the afternoon, bone-steaming tidal fever, night sweating, dry mouth and throat, emaciation, yellowish and scanty urine, reddish tongue with scanty fur and thin and rapid pulse, cold limbs and body, loose stool, early morning diarrhea, clear and profuse urine, profuse nocturnal urine, bright whitish complexion, and light-colored tongue with white fur as well as sinking and deep and weak pulse. All kidney deficiency syndrome information was collected using a unified questionnaire and confirmed by TCM physicians.

2.3 Clinical and laboratory measurement

All eligible subjects provided 20 ml second voided clean-catch urine samples in the early morning after an overnight fast. These samples were immediately frozen and stored −80°C to prevent protein degradation before testing. Urine samples were centrifuged at speed of 3,000 rpm for 10 min. The supernatants were collected for the sample test in 2 hr. Twenty-four hours UAlb (24hUAlb) was tested using an immunoturbidimetric assay (Yong Ye, Shanghai, China). Subsequently, baseline albuminuria status was determined according to the UACR. URBP, UNAG, and UCysC levels were measured using the immune turbidimetric method (KangTe, Zhejiang, China), colorimetry (Mei Kang, Zhejiang, China), and particle-enhanced nephelometric immunoassay (YanJing, Shanghai, China), respectively. All procedures were performed according to the manufacturer's instructions.

All patients donated fasting venous blood (2 ml) in the morning. The serum was separated by centrifugation. The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease (MDRD) formula: MDRD = 186 × (serum creatinine [mg/dl])−1.154 × (age in years)−0.203 × 0.742 (a correction factor of 0.742 was used for women) (Andrassy, 2013).

2.4 Statistical analysis

All analyses were performed using SPSS software (version 16.0, SPSS Inc., Chicago, IL, USA, 2007) and GraphPad Prism 5. Data are expressed as the mean ± SD for normally distributed values and median (interquartile range) for nonparametric values. Qualitative data are described as frequencies and analyzed using the chi-square test. Differences among groups were analyzed by one-way ANOVA followed by Bonferroni's test for normally distributed values, and by the Kruskal–Wallis test for nonparametric values. The Student–Newman–Keuls (SNK) or least significant difference (LSD) method was used for multiple comparisons. To test correlations between different urinary markers, Pearson's correlation coefficient was employed for normally distributed values and Spearman's correlation coefficient for skewed distributed values. To determine the association between UACR, 24hUAlb, and urinary markers, with the exception of bivariate correlation analysis, we performed a linear regression analysis with urinary markers as independent variables and UACR and 24hUAlb as dependent variables to investigate urinary markers related to albuminuria. Clinical parameters related to elevated urinary markers were analyzed using multivariate logistic regression according to the “Enter” procedure. The p-value report was two-sided, and p < .05 was considered significant.

3 RESULTS 3.1 Clinical characteristics in all patients

The study population of Type 2 DM consisted of 20 patients with normoalbuminuria, 50 patients with microalbuminuria, and 22 patients with macroalbuminuria. The baseline characteristics of all patients are summarized in Table 1. The mean age and DM duration of the patients were 63.12 ± 12.07 years and 12.62 ± 1.17 years, respectively, and there were 47 men and 45 women. According to the baseline data, no significant differences were found in sex, age, DM duration, systolic blood pressure (SBP), diastolic blood pressure (DBP), fasting blood glucose (FBG), 2 hr postprandial blood glucose (2hPBG), glycated hemoglobin (HbA1c), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), blood urea nitrogen (BUN), and uric acid (UA) in the normoalbuminuria, microalbuminuria, and macroalbuminuria groups. Although the levels of body mass index (BMI) and total cholesterol (TC) in the normoalbuminuria, microalbuminuria, and macroalbuminuria groups showed a stepwise increase with albuminuria level, no statistical significances were observed among the three groups. Moreover, the Scr level in the macroalbuminuria group was significantly increased than that in the microalbuminuria group (p = .017), and the eGFR level in the macroalbuminuria group was obviously decreased than that in the microalbuminuria group (p = .037), whereas no significant differences were noted between the microalbuminuria and normoalbuminuria groups.

TABLE 1. Baseline characteristics of laboratory parameters in all patients Normoalbuminuria (n = 20) Microalbuminuria (n = 50) Macroalbuminuria (n = 22) p-value Sex (male/female) 12/8 22/28 13/9 .332 Age (years) 64.80 ± 13.90 61.38 ± 11.57 65.55 ± 11.33 .318 Duration of DM (years) 12.40 ± 1.10 12.68 ± 1.10 12.68 ± 1.39 .641 BMI (kg/m2) 23.29 ± 1.41 25.17 ± 3.87 25.70 ± 4.38 .077 SBP (mmHg) 136.10 ± 14.80 136.58 ± 18.18 143.23 ± 17.10 .278 DBP (mmHg) 78.20 ± 12.67 80.04 ± 10.72 79.09 ± 12.57 .827 FBG (mmol/L) 9.04 ± 3.00 10.18 ± 2.38 9.16 ± 2.56 .135 2hPBG (mmol/L) 16.07 ± 5.13 15.22 ± 5.98 15.44 ± 4.87 .847 HbA1c (%) 8.75 ± 2.01 9.30 ± 2.12 8.57 ± 1.21 .273 TC (mmol/L) 4.31 ± 0.95 4.66 ± 1.12 5.07 ± 1.05 .072 TG (mmol/L) 1.34 ± 0.77 1.74 ± 1.02 2.74 ± 4.89 .168 LDL-C (mmol/L) 2.50 ± 0.78 2.68 ± 0.89 2.80 ± 0.71 .505 BUN (mmol/L) 6.17 ± 1.55 5.95 ± 1.54 6.50 ± 2.09 438 Scr (μmol/L) 69.58 ± 16.06 64.85 ± 20.27 78.82 ± 17.07a .017 eGFR (ml/min/1.73 m2) 99.38 ± 24.16 111.67 ± 47.65 86.12 ± 23.76a .037 UA (μmol/L) 286.65 ± 75.28 311.82 ± 95.71 324.95 ± 81.89 .365 Note: Data are expressed as the mean ± SD for parametric variables and median (interquartile range) for nonparametric variables. Abbreviations: 2hPBG, 2 hr postprandial blood glucose; BMI, body mass index; BUN, blood urea nitrogen; DM, diabetes mellitus; eGFR, estimated glomerular filtration; FBG, fasting blood glucose; HbA1c, glycosylated hemoglobin A1c; LDL-C, low-density lipoprotein cholesterol; SBP and DBP, systolic and diastolic blood pressure; Scr, serum creatinine; TC, total cholesterol; TG, triglyceride; UA, uric acid. 3.2 Distribution of kidney deficiency syndrome in all patients

TCM syndrome can be validated by the corresponding diagnostic standards for kidney asthenia (Wang & Zhang, 2017). Kidney deficiency syndrome of all patients was divided into the following types: kidney qi deficiency syndrome, kidney yang deficiency syndrome, kidney yin deficiency syndrome, and kidney essence insufficiency syndrome. As shown in Table 2, among the 92 patients with Type 2 DM, the syndromes that occurred at least 30% included: aching and weakness of loins and knees (78.3%), dispiritedness and lassitude (78.3%), frequent micturition (47.8%), dripping urination (33.7%), dizziness and tinnitus (31.5%), light-colored tongue with whitish fur (86.9%), and thin pulse (30.4%). These TCM syndrome distributional characteristics in all patients were exactly consistent with the clinical manifestations of kidney qi deficiency syndrome.

TABLE 2. Distribution of kidney asthenia syndrome distribution in all patients Normoalbuminuria (n = 20) Microalbuminuria (n = 50) Macroalbuminuria (n = 22) Aching and weakness of loins and knees 14 (70%) 40 (80%) 18 (81.8%) Dispiritedness and lassitude 12 (60%) 42 (84%) 18 (81.8%) Frequent micturition 4 (20%) 29 (38%) 11 (50%) Dripping urination 3 (15%) 18 (36%) 10 (45.5%) Incontinence of urine 0 0 0 Light-colored tongue with whitish fur 18 (90%) 45 (90%) 17 (77.3%) Thin pulse 8 (40%) 10 (20%) 10 (45.5%) Dizziness and tinnitus 3 (15%) 18 (36%) 8 (36.4%) Insomnia and amnesia 1 (5%) 2 (4%) 1 (4.5%) Flushed cheeks in the afternoon 0 1 (2%) 0 Bone-steaming tidal fever 0 0 0 Night sweating 2 (10%) 1 (2%) 0 Dry mouth and throat 2 (10%) 4 (8%) 2 (9%) Emaciation 0 0 0 Yellowish and scanty urine 0 0 0 Reddish tongue with scanty fur 1 (5%) 1 (2%) 2 (9%) Thin rapid pulse 0 1 (2%) 1 (4.5%) Cold limbs and body 0 0 0 Loose stool 1 (5%) 2 (4%) 2 (9%) Early morning diarrhea 1 (5%) 1 (2%) 0 Clear and profuse urine 1 (5%) 1 (2%) 1 (4.5%) Profuse nocturnal urine 2 (10%) 1 (2%) 1 (4.5%) Bright whitish complexion 0 0 1 (4.5%) Light colored tongue with white fur as well as sinking 0 0 0 Deep and weak pulse 0 0 0 3.3 Differences of Scr, eGFR, 24hUAlb, and UACR in all patients

Figure 2 shows the differences in the Scr, eGFR, 24hUAlb, and UACR levels among the normoalbuminuria, microalbuminuria, and macroalbuminuria groups. We found that the Scr level in the macroalbuminuria group was significantly higher than that in the microalbuminuria group (p = .017), the eGFR level in the macroalbuminuria group was significantly lower than that in the microalbuminuria group (p = .037), although no significant differences were noted between the microalbuminuria and normoalbuminuria groups. Additionally, we concluded that the 24hUAlb and UACR levels showed a stepwise increase in the normoalbuminuria, microalbuminuria, and macroalbuminuria groups (p = .000).

image

Comparisons of Scr, eGFR, 24hUAlb, and UACR in all patients. The differences of the Scr, eGFR, 24hUAlb, and UACR levels among the three groups are shown. Each value is expressed as the mean ± SD. **p < .01 versus the normoalbuminuria group; #p < .05, ##p < .01 versus the microalbuminuria group. 24hUAlb, 24 hr urinary albumin; eGFR, estimated glomerular filtration; Scr, serum creatinine; UACR, urinary albumin-to-creatinine ratio

3.4 Changes of UCysC, UNAG, and URBP in all patients

Table 3 and Figure 3 show the changes in the UTBs levels, including UCysC, UNAG, and URBP, in the normoalbuminuria, microalbuminuria, and macroalbuminuria groups according to the albuminuria level. We found that the UNAG, URBP, and UCysC levels synchronously showed a gradual and significant increase consistent with the degree of albuminuria in the three groups, and a significant difference was observed between the normoalbuminuria and microalbuminuria groups (p = .000), as well as between the microalbuminuria and macroalbuminuria groups (p = .000).

TABLE 3. Changes of urinary tubular biomarkers in all patients Normoalbuminuria (n = 20) Microalbuminuria (n = 50) Macroalbuminuria (n = 22) p-value UNAG (U/L) 16.21 ± 2.52 23.78 ± 3.88a 38.90 ± 6.07a,b .000 URBP (mg/L) 0.72 ± 0.33 1.77 ± 0.84a 3.36 ± 1.67a,b .000 UCysC (mg/L) 0.20 ± 0.09 0.45 ± 0.07a 0.75 ± 0.13a,b .000 Note: Data are expressed as the mean ± SD for continuous variables. The p-values were obtained using ANOVA. Abbreviations: UCysC, urinary cystatin C; UNAG, urinary N-acetyl-β-d-glucosaminidase; URBP, urinary retinol-binding protein. image

Comparisons of urinary tubular biomarkers in all patients. The changes of the urinary tubular biomarkers levels, including UNAG, URBP, and UCysC, in the three groups according to albuminuria are revealed. Each value is expressed as the mean ± SD. **p < .01 versus the normoalbuminuria group; ##p < .01 versus the microalbuminuria group. UCysC, urinary cystatin C; UNAG, urinary N-acetyl-β-d-glucosaminidase; URBP, urinary retinol-binding protein

3.5 Associations of 24hUAlb and UACR with UCysC, UNAG, and URBP in all patients

Table 4 and Figure 4 illustrate the associations of 24hUAlb, UACR with UNAG, URBP, and UCysC in all 92 patients with Type 2 DM. We concluded that 24hUAlb and UACR were positively correlated with UNAG, URBP, and UCysC, respectively (p = .000, respectively).

TABLE 4. Correlation analysis of 24hUAlb, UACR, and urinary tubular markers in all patients UNAG (U/L) UCysC (mg/L) URBP (mg/L) 24hUAlb (mg) r = 0.548, p = .000 r = 0.613, p = .000 r = 0.488, p = .000 UACR (mg/g) r = 0.693, p = .000 r = 0.677, p = .000 r = 0.599, p = .000 Note: p < .05 is considered significant. Abbreviations: 24hUAlb, 24 hr urinary albumin; UACR, urinary albumin-to-creatinine ratio; UCysC, urinary cystatin C; UNAG, urinary N-acetyl-β-d-glucosaminidase; URBP, urinary retinol-binding protein. image

Associations of 24hUAlb, UACR with UCysC, UNAG, and URBP in all patients. The associations of 24hUAlb, UACR with UNAG, URBP, and UCysC in all patients are illustrated. 24hUAlb, 24 hr urinary albumin; UACR, urinary albumin-to-creatinine ratio; UCysC, urinary cystatin C; UNAG, urinary N-acetyl-β-d-glucosaminidase; URBP, urinary retinol-binding protein

3.6 Correlations of UCysC, UNAG, and URBP in 72 patients with Type 2 DKD

The correlations of UNAG, URBP, and UCysC in 72 patients with Type 2 DKD with microalbuminuria and macroalbuminuria were analyzed in Table 5. There was a positive correlation between UNAG and URBP (p = .000). UCysC was also positively correlated with UNAG and URBP, respectively (p = .000, respectively).

TABLE 5. Correlation analysis of urinary tubular markers in 72 patients with Type 2 DKD UNAG (U/L) UCysC (mg/L) URBP (mg/L) UNAG (U/L) — r = 0.744, p = .000 r = 0.611, p = .000 UCysC (mg/L) r = 0.744, p = .000 — r = 0.556, p = .000 URBP (mg/L) r = 0.611, p = .000 r = 0.556, p = .000 — Note: p < .05 is considered significant. Abbreviations: UCysC, urinary cystatin C; UNAG, urinary N-acetyl-β-d-glucosaminidase; URBP, urinary retinol-binding protein. 3.7 Associations of clinical baseline parameters with UCysC, UNAG, and URBP in all patients

In Table 6, using multivariate logistic regression analysis, we incidentally found that clinical baseline parameters, such as BMI and FBG, were positively correlated with UCysC, respectively (p = .012 or p = .046) in all 92 patients with Type 2 DM.

TABLE 6. Logistic regression analysis of clinical baseline parameters with UNAG, UCysC, and URBP in all patients UNAG (U/L) UCysC (mg/L) URBP (mg/L) Age (years) OR = 1.457, p = .227 OR = 0.364, p = .546 OR = 0.789, p = .374 BMI (kg/m2) OR = 0.253, p = .615 OR = 6.360, p = .012 OR = 3.174, p = .075 SBP (mmHg) OR = 0.774, p = .379 OR = 1.938, p = .164 OR = 0.151, p = .697 DBP (mmHg) OR = 0.016, p = .898 OR = 0.193, p = .661 OR = 0.049, p = .824 FBG (mmol/L) OR = 0.450, p = .503 OR = 3.986, p = .046 OR = 1.220, p = .269 2hPBG (mmol/L) OR = 0.280, p = .597 OR = 1.627, p = .202 OR = 0.491, p = .484 HbA1c (%) OR = 2.336, p = .1

留言 (0)

沒有登入
gif