Role of Serum Chromogranin A in Early Diagnosis of Diabetic Nephropathy

Background: Kidney is the main site for the removal of Chromogranin A (CgA). Previous studies have found that patients with renal impairment displayed elevated concentrations of CgA in plasma and the CgA concentrations reect deterioration of renal function. In this study, we aimed to estimate the serum CgA levels and to evaluate the role of serum CgA in early diagnosis of diabetic nephropathy (DN). Methods: A total of 219 patients with type 2 diabetes mellitus (T2DM) were included in this cross-sectional study. They were classied into normoalbuminuria (n = 121), microalbuminuria (n = 73), or macroalbuminuria (n = 25) groups based on their urine albumin to creatinine ratios (UACR). A control group consisted of 45 healthy subjects. The serum CgA levels were measured by ELISA and other key parameters were assayed. Results: The serum CgA levels were higher in patients with T2DM compared to control subjects and a statistically signicant difference among studied subgroups regarding CgA was found (P<0.05). Levels of serum CgA were increasing gradually with the degree of DN (P<0.001). Serum CgA levels were moderate intensity positively correlated with UACR (P<0.001). A cut-off level of 3.46 ng/ml CgA showed 69.86% sensitivity and 66.12% specicity to detect DN in early stage. Conclusions: Levels of serum CgA increased gradually with the degree of DN, and can be used as a biomarker in early detection of DN. [2]. It is appreciated that up to 40% of the patients with type 1 and type 2 diabetes mellitus (DM) present DN [3, 4]. Early detection and appropriate treatment are essential to prevent disability and death. with T2DM. Considering the interaction between sCr and eGFR, and their similar clinical implication, only sCr would be used for multivariate logistic regression. Spearman rank correlation test was performed to analyze the relationship between 2 quantitative parameters. Statistical analyses were conducted using SPSS software (version 23.0; SPSS, Inc., Chicago, IL, United States). Statistical signicance was considered as 2-tailed, with a p-value of < 0.05.

Spot urine samples were collected at rst morning void. Blood samples were collected before eating in the morning, they were centrifuged (3,000 rpm, 4 ° C, 10 min) to acquire serum and stored at − 80 ° C until use. Samples for measurements of serum creatinine (sCr), serum uric acid (sUA), blood urea nitrogen (BUN), fasting blood glucose (FBG), hemoglobin A1c (HbAlc), triglyceride (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), urinary microalbumin to creatinine ratio (UACR) levels were examined at the department of clinical examination using standard procedures. The estimated glomerular ltration rate (eGFR) was calculated using the abbreviated MDRD equation [12]. eGFR (mL/min per 1.73m 2 ) = 186 Pcr − 1.154 X age − 0.203 X 0.724 (if female), where Pcr is in mg/dl and age is in years. The concentration of CgA was measured using Human-CgA ELISA kits (Cat# ab196271) from Abcam.

Statistical Analysis
Continuous variables with normal distribution were reported as mean ± standard deviation (SD), whereas skewed distributed variables were expressed as median (interquartile range). Categorical data were summarized as proportions with frequencies. Continuous variables were compared between groups using unpaired t test, one-way ANOVA test, Kruskal-Wallis test or Jonckheere-Terpstra test where appropriate, while the chi-square test or Fisher's exact test was used to analyze the differences in categorical variables. Logistic regression analyses were used to ascertain the predictive value of CgA level for the presence of early DN in patients with T2DM. Considering the interaction between sCr and eGFR, and their similar clinical implication, only sCr would be used for multivariate logistic regression. Spearman rank correlation test was performed to analyze the relationship between 2 quantitative parameters. Statistical analyses were conducted using SPSS software (version 23.0; SPSS, Inc., Chicago, IL, United States). Statistical signi cance was considered as 2-tailed, with a p-value of < 0.05.

Clinical Characteristics
All enrolled T2DM patients were categorized into three subgroups according to UACR. Demographic and laboratory features of the patients as well as the healthy subjects are reported in Table 1. Our results showed that there was signi cant difference among the subgroups of patients regarding to SBP, Known diabetes duration, UACR, sCr, BUN and eGFR. Meanwhile, there was no signi cant difference among the subgroups of patients regarding to gender, age, DBP, FBG, HbAlc, sUA, TC, TG, HDL-C and LDL-C. Compare the clinical and laboratory data among the subgroups of patients and the healthy subjects, there was no signi cant statistical difference only in gender, age, TC and LDL-C levels. a P-value: comparisons among the healthy controls group, the normoalbuminuria group, the microalbuminuria group and the macroalbuminuria group; b P-value: comparisons among the normoalbuminuria group, the microalbuminuria group and the macroalbuminuria group SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; UACR, urine albumin to creatinine ratio; sCr, serum creatinine; BUN, blood urea nitrogen; sUA, serum uric acid; eGFR, estimated glomerular ltration rate; TC, total cholesterol; TG, triglyceride; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; CgA, Chromogranin A.
Comparison of serum CgA levels among subgroups of patients with T2DM and healthy subjects A comparison of serum CgA levels among subgroups of patients divided according to UACR and healthy subjects was performed using the Kruskal-Wallis test. The serum CgA levels were higher in patients with T2DM compared to healthy subjects and a statistically signi cant difference among studied subgroups regarding CgA was found (P < 0.05). A non-parametric Jonckheere-Terpstra test showed that CgA was increasing gradually with the degree of DN (Z = 6.567, P < 0.001). The results are shown in Fig. 1 and Table 1.

Correlation between serum CgA levels and clinical and laboratory data
Spearman rank correlation test demonstrated that there was moderate intensity positive correlation between serum CgA levels and Known diabetes duration, UACR in all subjects (P < 0.001). Weak correlation was found between serum CgA levels and FBG, HbAlc, sCr, BUN, eGFR, HDL-C (all P < 0.05), and there was no correlation between serum CgA levels and age, SBP, DBP, sUA, TC, TG, LDL-C in all subjects. There was moderate intensity positive correlation between serum CgA levels and UACR in patients with T2DM (P < 0.001). Weak correlation was found between serum CgA levels and Known diabetes duration, DBP, HbAlc, sCr, BUN, eGFR (all P < 0.05), and there was no correlation between serum CgA levels and age, SBP, FBG, sUA, TC, TG, LDL-C, HDL-C in patients with T2DM. The correlation coe cients between serum CgA levels and clinical and laboratory data are presented in Table 2.

ROC analysis
ROC analysis was performed to assess the sensitivity and speci city of serum CgA as the potential biomarkers in prediction of early DN. The ROC analysis of serum CgA yielded an AUC of 0.714 (95% CI, 0.639-0.788; P < 0.001) in the differentiation of T2DM patients with early DN (Table 4 and Fig. 2). The diagnostic sensitivity and speci city of serum CgA for early DN was 69.86% and 66.12% when the cut off value was 3.46 ng/mL.

Discussion
DN is an important microvascular complication of diabetes mellitus. Up to now, there is no satisfactory method to prevent early DN. According to as a diagnostic marker for (early) DN. The change of glomerular basement membrane structure may be earlier than the increase of albuminuria [14]. About 30-45% T2DM patients who develop kidney disease associated with decrease in glomerular ltration are observed no increased albuminuria [15,16]. Meanwhile, albuminuria is not speci c for DN. Hypertension or obesity may also affect the ltration barrier of the glomeruli leading to increased albuminuria in patients with T2DM [17]. However, treatment of hypertension usually uses renin angiotensin aldosterone inhibitors, which reduce glomerular hydrostatic pressure and make proteinuria in the normal range. These factors affect the accuracy of diagnosis based on current guidelines. When the kidney of DN is damaged, the decrease of eGFR occurs quite late. The early damage is often accompanied by hyper ltration [18]. Therefore, both routine markers of DN (albuminuria and eGFR) re ecting glomerular damage has certain limitations.
CgA is the main member of the chromogranins family consisting of water soluble acidic glycoproteins. It was rst found in the secretory granules from adrenal medullary chroma n cells, and released into circulation after splanchnic nerve stimulation together with cathecholamines [19]. CgA also exist in a lot of endocrine and neuroendocrine cells [20], nerve cells [21] and immune cells [22]. Elevated serum CgA levels have been found in many cancers [23] and neurodegenerative diseases such as Alzheimer's disease [24] or Parkinson's disease [25]. Previous study has showed that CgA, like many other low MW proteins, is handled by the kidney [26]. However, serum CGA increases in renal failure more than creatinine and the other studied low MW proteins, β2-MG(β2 microglobulin) and TATI (tumor associated trypsin inhibitor) [26]. The study by Chen et al. suggests CgAstimulation of endothelial cell exocytosis of endothelin as a possible mechanism for regulation of renal function in health and disease [27].
Our study aimed to investigate the role of serum CgA in early diagnosis of DN in patients with T2DM. Because our groups were matched regarding age and sex, the effect of these factors on the results of serum CgA in our study were excluded. In our study we found there was a signi cant difference among normoalbuminuria, microalbuminuria and macroalbuminuria groups regarding known diabetes duration. This is consistent with other previous studies [28,29].
In our study, the level of serum CgA in patients with T2DM was much higher than that in the healthy people. Previous study has documented the clinical value of measurements of CgA as a potential marker for diabetes [30]. The concentration of plasma and salivary CgA was signi cantly higher in diabetic groups compared to the control [31]. Our results are consistent with their ndings.
Serum concentrations of CgA correlated strongly with eGFR [32]. Its relationship with GFR is very similar to that of other substances eliminated by the kidney like creatinine and TATI [26]. In our result, Spearman rank correlation test demonstrated that serum CgA was negatively correlated with eGFR, not only in patients with T2DM, but also in all subjects. There was moderate intensity positive correlation between serum CgA levels and Known diabetes duration, UACR in all subjects.
We found for the rst time that serum CgA was increasing gradually with the degree of DN, serum CgA level was associated with the occurrence of early DN. One study showed that serum CgA was a poor indicator of DN since patients with T2DM and reduced glomerula ltration rate failed to show any signi cant increase in serum CgA [33]. In their research, patients were divided into those without DN with normoalbuminuria (n = 27), patients with DN with microalbuminuria (n = 8), or macroalbuminuria (n = 42). It is noteworthy that in their study P value of the difference between patients with normoalbuminuria and DN was 0.07. The failure to reach the statistical difference may be due to the small sample size. Moreover, other studies have found that patients with renal impairment displayed elevated concentrations of CgA in plasma and the CgA concentrations re ect deterioration of renal function [26,34]. CgA contains multiple amino acid motifs prone to endoproteolytic cleavage resulting in multiple processing fragments. This constitutes a biochemical challenge for accurate quanti cation of CgA. Several assays for measurements of CgA have been developed, but since the used antibodies detect different epitopes, the results from the assays vary considerably [35]. The assay kits we used were different from those used in mojiminiyi OA's team [33].
Roc analysis showed diagnostic accuracy of serum CgA in differentiation between T2DM patients with early DN and without DN. Using cut-off 3.46 ng/mL for serum CgA, we found sensitivity 69.86%, speci city 66.12%, and diagnostic accuracy 71.4% in predicting early DN.
The ndings are limited by the sample size and the crosssectional design of the study therefore this cannot be used to determine the direction of causality sequence. Additional prospective studies are needed to examine associations between CgA with early DN risk.

Conclusion
The results of this study show that serum CgA levels were higher in patients with T2DM compared to healthy subjects, and increased gradually with the degree of DN. It also can be used as a biomarker in early detection of DN.
Declarations HY, HpW, JF and HW participated in the conception and design of the study, and analyzed the data of study. HY drafted the initial manuscript, JF and HW revised the manuscript. XS, AlC, XmY, YmW and BbZ participated in data collection. All authors have read and approved the nal manuscript.

Funding Sources
This work was supported by the Key Medical Discipline Project of Shanghai Municipal Health Bureau (ZK2019A12).

Availability of data and materials
The current data are available to all interested researchers upon reasonable request. Requests for access to data should be made to the principal investigators of the study.
Ethics approval and consent to participate All subjects have given their written informed consent. The study protocol has been approved by the institutional committee on human research.

Consent for publication
Not applicable.