, Introduction
Cardiovascular disease (CVD) is the most common cause of death in patients with end-stage renal disease (ESRD) undergoing hemodialysis (HD) [1]. Prior studies have shown that both traditional and non-traditional risk factors are multifactorial in the pathogenesis of CVD in patients with ESRD. Moreover, it is difficult to predict the occurrence of cardiovascular (CV) events in these patients [2]. ESRD patients with CVD often show no or atypical symptoms, resulting in delayed diagnosis and failure to receive appropriate management of CVD. Therefore, several studies have suggested biomarkers that can predict the occurrence of CV events in these patients [3,4], and several ongoing investigations have identified biomarkers that could support the prediction of CVD occurrence.
Endostatin is a carboxyl-terminal fragment of type XVIII collagen, which is present in various endothelial and epithelial basement membranes [5]. It exerts anti-angiogenic and anti-fibrotic effects by inhibiting the proliferation and migration of endothelial cells [6]. Endostatin can be found in the circulatory system, and circulating serum endostatin levels have been suggested to be a marker of extracellular matrix breakdown in several CV pathologic conditions [7]. Previous studies have reported that circulating endostatin is a relevant biomarker that can improve risk prediction in various disease conditions, including CV events [8]. It has also been reported that circulating endostatin is associated with the progression of atherosclerosis and vascular calcification. Serum endostatin levels are increased in patients with coronary artery disease, and significant correlations between endostatin and coronary artery calcification or aortic valve calcification have been reported in prior studies [5,9].
Patients with ESRD have a high prevalence of vascular calcification, which shows distinct pathophysiological features compared to those in the general population [2,10]. Patients with ESRD undergoing HD are at risk of atherosclerotic injury because they are constantly exposed to endothelial injury, oxidative stress, and inflammation. Repeated exposure to hemodynamic stress during dialysis worsens this atherosclerotic vascular pathology [11]. Furthermore, these patients already have various traditional risk factors for CV events that trigger the formation of atherosclerotic plaques and intimal calcification [2,12]. Therefore, patients undergoing HD have a higher prevalence of intimal vascular calcification, which is an independent risk factor for CV events and mortality in these populations [13,14]. In addition, factors related to mineral bone diseases (MBDs) in chronic kidney disease (CKD), including hyperphosphatemia, hypercalcemia, and secondary hyperparathyroidism, can cause mineral deposition on the medial layer of the vascular wall, leading to medial vascular calcification [15]. Therefore, patients with ESRD tend to have both intimal and medial layer calcification, which further increases the risk of CV events.
Considering the function of endostatin and its association with vascular calcification, it can be assumed that endostatin might have a predictive role in the occurrence of vascular calcification and CV events in patients with ESRD on HD. There is a clear need to confirm the clinical relevance of endostatin in these patients, as few studies have analyzed the association among them. In this study, we measured circulating endostatin levels and investigated their association with CV events in ESRD patients undergoing HD. We also analyzed the relationship between endostatin levels and clinical characteristics as well as circulating MBD biomarkers.
Methods
Study population and design
Participants were recruited from the K-cohort, a multicenter, internet-based, prospective cohort of patients undergoing HD in South Korea. The K-cohort study was conducted to evaluate the morbidity and mortality of patients with ESRD on HD starting in 2016 (CRIS No. KCT0003281). Patients with ESRD aged >18 years on HD were enrolled in the K-cohort if they were undergoing HD three times a week for >3 months. The exclusion criteria were pregnancy, hematologic malignancy, active or invasive solid tumor, and life expectancy of <6 months. Clinical data, laboratory results, and blood samples were prospectively collected from patients with ESRD on HD at the baseline and follow-up visits. Patient outcomes, such as CV events, comorbidities, and mortality, were analyzed. Detailed cohort design and inclusion/exclusion criteria have been previously described [16,17].
A total of 381 patients on HD with whole-plasma samples collected at the time of enrollment from June 2016 to March 2020 were included. Of these patients, nine with percutaneous transluminal angioplasty within 3 months before study enrollment were excluded because percutaneous transluminal angioplasty might affect the plasma levels of endostatin via endothelial stimulation. Therefore, 372 patients on HD were enrolled in this study. The study population was divided into three groups based on plasma endostatin levels, as follows: tertile 1, <146 ng/mL; tertile 2, 146–196 ng/mL; and tertile 3, ≥196 ng/mL. All participants were followed up prospectively after baseline characteristics were assessed. Patient follow-up was censored at the time of transfer to peritoneal dialysis, renal transplantation, follow-up loss, or withdrawal of patient consent.
The study protocol was approved by the Institutional Review Board of Kyung Hee University Hospital (No. KHNMC 2016-04-039) and conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants involved in this study.
Data collection and outcome measures
General demographic information, prior medical history, cause of renal disease, comorbid conditions, laboratory data, drug use, vascular assessment, and dialysis information were collected from the medical records and interviews. Single-pool Kt/V (spKt/V) (where K, dialyzer clearance; t, time; and V, urea distribution volume) was assessed using the conventional method [13], and body mass index (BMI) was calculated from the patient’s weight (kg) divided by the square of body height (m2) in a standard manner. Information on comorbidities was assessed using the Charlson comorbidity index score [18].
The primary study outcome was a composite of incident CV events, including both cardiac and non-cardiac vascular events. Cardiac events included acute coronary artery syndrome, heart failure, ventricular arrhythmia, cardiac arrest, and sudden death. Non-cardiac vascular events were defined as cerebral infarction, cerebral hemorrhage, and peripheral vascular occlusive diseases that required revascularization or surgical intervention. All-cause mortality was recorded and carefully reviewed.
Laboratory analysis
Blood samples for laboratory analysis were collected in a fasting state before HD during a midweek session on the same day the baseline clinical characteristics were collected. After centrifugation at 1,000×g at room temperature for 15 minutes, the samples were stored at −80 °C until analysis. Routine biochemical parameters were measured using standard laboratory methods. Plasma levels of endostatin, osteoprotegerin, and receptor activator of nuclear factor kappa-Β ligand (RANKL) were measured by enzyme-linked immunosorbent assay using Magnetic Luminex Screening Assay multiplex kits (R&D Systems, Inc.). Osteoprotegerin and RANKL levels were measured in 275 patients (73.9%), depending on sample availability.
Vascular calcification assessment
To assess the severity of vascular calcification, we used a scoring system for abdominal aortic calcification based on lateral lumbar radiography. Vascular calcification scores were estimated using a previously reported method [19] and interpreted by nephrologists who were blinded to the clinical data of the participants. The abdominal aortic walls were divided into four sections corresponding to the lumbar spine from the first (L1) to the fourth (L4) vertebrae, and each aortic wall was scored from 0 to 3 points. The abdominal aortic calcification score represents the composite score of calcification grades of the anterior and posterior abdominal aortic walls, with a maximum score of 24 points. Among all studied patients, 275 (73.9%) underwent radiographic examinations for vascular calcification assessment.
Statistical analysis
Continuous variables are presented as mean ± standard deviation values or median (interquartile range [IQR]) values, and categorical variables are presented as frequencies and percentages. Differences among the three groups were identified using analysis of variance or the Kruskal-Wallis test. The Tukey post-hoc test and Mann-Whitney U test with Bonferroni correction were used to identify intergroup differences. Categorical data were analyzed using the chi-square test or Fisher exact test. Correlations between continuous variables were evaluated using Spearman correlation analysis. A Cox proportional hazards model was constructed to identify the independent variables related to CV events and all-cause mortality. The multivariate models included parameters that were significantly associated with weight in univariate testing and clinically fundamental parameters. The parameters included in the multivariate Cox analysis were age, sex, BMI, Charlson comorbidity index, dialysis duration, hemoglobin, high-sensitivity C-reactive protein, angiotensin receptor blocker or angiotensin-converting enzyme inhibitor use, and spKt/V. Statistical analyses were conducted using the IBM SPSS version 22.0 (IBM Corp.), and p < 0.05 was considered to indicate statistical significance.
Results
Baseline demographic characteristics and laboratory data
The median circulating endostatin level was 171.0 ng/mL (IQR, 117.0–209.8 ng/mL) among all study participants. The median endostatin level was 105.0 ng/mL (IQR, 87.5–117.0 ng/mL) in tertile 1 (n = 124), 171.0 ng/mL (IQR, 160.0–183.0 ng/mL) in tertile 2 (n = 124), and 234.5 ng/mL (IQR, 209.3–259.8 ng/mL) in tertile 3 (n = 124), respectively. The baseline characteristics and laboratory results of the study population across tertiles of endostatin use are summarized in Table 1. Patients in tertile 1 had a shorter dialysis duration than those in tertiles 2 and 3. Laboratory parameters and dialysis characteristics did not differ significantly between the two groups. The plasma levels of the MBD markers did not differ across the endostatin tertiles.
Determinant factors of higher endostatin level
Univariate and multivariate logistic regression analyses of endostatin tertile 3 and the baseline parameters are shown in Table 2. Low-density lipoprotein (LDL) cholesterol levels and predialysis systolic blood pressure were marginally associated with endostatin tertile 3 in the univariate analysis. In the multivariate analysis, LDL cholesterol levels (odds ratio [OR], 1.01; 95% confidence interval [CI], 1.00–1.02; p = 0.04) and predialysis systolic blood pressure (OR, 1.01; 95% CI, 1.00–1.02; p = 0.04) were significant determinants of endostatin tertile 3.
Correlation of endostatin level with circulating mineral bone disease biomarkers and baseline characteristics
Table 3 shows the correlations between endostatin levels and circulating MBD marker levels. The plasma levels of endostatin did not significantly correlate with calcium × phosphorous, intact parathyroid hormone, osteoprotegerin, or RANKL. There was also no significant correlation between vascular calcification score in the abdominal aorta and endostatin level.
Prognostic utility of endostatin level in patients undergoing hemodialysis
During a mean follow-up of 30.5 months, 54 CV events (14.5%) and 44 cardiac events (11.8%) occurred. The cumulative event rate for CV events was significantly higher as endostatin levels increased, with endostatin tertile 3 having the highest cumulative CV event rate (p = 0.006) (Fig. 1A). Endostatin tertile 3 was associated with a greater cumulative rate of cardiac events (p = 0.04) (Fig. 1B).
Table 4 shows the hazard ratios (HRs] of plasma endostatin for the CV events. Univariate Cox regression analysis revealed that endostatin tertile 3 was significantly associated with an increased risk of composite CV events (HR, 3.89; 95% CI, 1.60–9.43; p = 0.003), and this association remained significant after adjustment for multiple variables (HR, 4.37; 95% CI, 1.79–10.67; p = 0.001). Endostatin increments per 1 ng/mL were independently associated with an increased risk of composite CV events (HR, 1.004; 95% CI, 1.001–1.007; p = 0.02). To further investigate the risk of composite CV events, the HRs for cardiac and non-cardiac vascular events were analyzed. Patients in tertile 3 had a significant risk of cardiac events even after adjusting for multiple variables (HR, 3.88; 95% CI, 1.44–10.44; p = 0.007), and endostatin increment per 1 ng/mL was also associated with a higher risk of CV events (HR, 1.005; 95% CI, 1.001–1.008; p = 0.02). However, endostatin levels were not associated with a significant risk of non-cardiac vascular events.
Discussion
In this study, we investigated the association between circulating endostatin levels and CV and MBD parameters in patients with ESRD undergoing HD using a prospective observational cohort database. Endostatin levels were significantly correlated with LDL cholesterol levels and predialysis systolic blood pressure, which are conventional risk factors for atherosclerotic CV events. Higher circulating endostatin levels were significantly associated with an increased risk of incident CV composites and cardiac events, even after adjusting for multiple variables. However, we did not observe a correlation between endostatin levels and MBD markers or vascular calcification scores in this study.
Previous studies have reported elevated plasma endostatin levels in various disease conditions and have demonstrated that plasma endostatin levels are significantly correlated with disease severity and clinical outcomes [20]. The median circulating endostatin level in our study was 171.0 ng/mL, which was relatively increased considering that the level in healthy individuals ranges from 24.6 to 136.1 ng/mL [21,22]. Chen et al. [23] documented significantly increased plasma endostatin levels in patients with non–dialysis-dependent CKD compared to controls without CKD. They also observed a concentration-dependent relationship between the severity of CKD and plasma endostatin levels. In addition, it seems that the endostatin levels in this study were higher in our study group than in the normal controls and patients with non–dialysis-dependent CKD, as measured in previous studies [23,24]. These findings suggest that endostatin levels increase as renal function decreases. Further studies are required to investigate whether increased endostatin levels with declining renal function are derived from impaired renal excretion or pathological conditions related to lower renal function.
Studies have shown that higher serum endostatin levels in patients with coronary artery disease are associated with reduced angiogenesis and poorly developed collateral vasculature [25]. Circulating endostatin levels predict the development of CV diseases such as recurrent ischemic stroke [26] and incident myocardial infarction [27]. High plasma endostatin levels in patients with stable coronary heart disease reflect CV and total mortality [28]. Two independent community-based cohorts showed a significant association between CV mortality and plasma endostatin levels [29]. Higher circulating endostatin levels may be involved in inflammatory stress and active extracellular remodeling [30]. The inflammatory process in hypoxia or ischemia, which frequently occurs in patients undergoing HD, stimulates metalloproteinase, elastase, and cathepsins, and they can modulate the extracellular matrix and release endostatin [31]. In addition, human endothelial cells incubated with high endostatin levels can show more apoptotic changes [32]. Endostatin-induced endothelial apoptosis may correlate with myocardial rarefaction and CV dysfunction [33]. Circulating endostatin levels are also associated with a higher risk of CV events in patients with CKD [34]. In line with these findings, we demonstrated the significant predictive value of plasma endostatin levels for CV events in patients with ESRD on HD. Furthermore, our results showed that the circulating endostatin level was an independent predictive factor for the occurrence of cardiac events. These findings suggest that the association between higher levels of circulating endostatin and a greater risk of atherosclerotic events remains significant in patients receiving HD treatment and that predictive values of endostatin could be expanded to several categories of diseases.
Endostatin has a complex biology. Animal studies have identified a protective role of endostatin in the development of CV disease. Endostatin neutralization by the injection of anti-endostatin antibody in myocardial infarction–induced rats deteriorated left ventricular remodeling [35], and endostatin inhibition attenuated inflammation, limited oxidative stress, and reduced plaque growth and atherosclerosis [36,37]. In addition, endostatin mediates a major blocking effect on LDL cholesterol retention in atherosclerosis-prone mice, and the infusion of endostatin lowers blood pressure and vasorelaxation [38,39]. Our results also showed that LDL cholesterol and systolic blood pressure are independent determinants of higher endostatin levels and that endostatin tertile 3 was associated with an increased risk of CV and cardiac events in patients on HD. Considering these findings, we suggest that endostatin has the therapeutic potential to prevent CV complications in patients on HD and that clinical studies are warranted to investigate the role of endostatin treatment.
In this study, we expected that endostatin levels would show a positive correlation with circulating MBD markers and vascular calcification score because increased endostatin levels are reported to be associated with coronary or aortic valve calcification [9]. However, we did not observe any correlation between endostatin levels and MBD markers or vascular calcification scores. We presumed that medial vascular calcification induced no relationship between endostatin levels and vascular calcification. MBD is the main contributor to calcific deposition on the medial arterial layer, and patients with ESRD face a considerable burden brought on by the medial type of vascular calcification [2]. While these patients may have had an intimal type of vascular calcification, the additional occurrence of medial vascular calcification from MBD may have influenced the vascular calcification scores. We speculated that circulating endostatin levels could be more closely associated with atherosclerotic intimal vascular calcification than with medial vascular calcification.
This study had some limitations. We could not perform separate analyses for specific CV events due to the limited number of events. As this was a retrospective study, there may have been unintended bias. In addition, circulating osteoprotegerin levels, RANKL levels, and vascular calcification scores were not obtained for all enrolled patients. When the study population was divided by the median level of endostatin, the results of the univariate and multivariate analyses for the composite of CV events were as follows: HR, 1.31; 95% CI, 0.76–2.26; p = 0.34 and HR, 1.21; 95% CI, 0.69–2.11; p = 0.50, respectively. Considering these findings, it is difficult to suggest the optimal cutoff value of circulating endostatin for CV event prediction based on our study. Finally, the intimal and medial types of vascular calcification were not differentiated because a simple radiological examination is limited in its capacity to assess vascular calcification type.
In conclusion, circulating endostatin levels were significantly correlated with LDL cholesterol levels and predialysis systolic blood pressure but did not correlate with MBD markers or vascular calcification. Higher circulating endostatin levels were associated with a greater risk of composite CV and cardiac events in patients with ESRD on HD. Our findings indicate that endostatin is a biomarker that can help predict CV events in patients with ESRD undergoing HD.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
