Clinica Chimica Acta 452 (2016) 161–166
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
MMP-2 serum concentrations predict mortality in hemodialysis patients: a 5-year cohort study Kuang-Chih Hsiao a,b,1, Jen-Pi Tsai c,1, Shun-Fa Yang a, Wen-Chin Lee b, Jong-Yu Huang b, Shun-Chi Chang b, Chun-Shuo Hso b, Horng-Rong Chang a,d,e,⁎ a
Institute of Medicine, Chung Shan Medical University, Taichung, Taiwan Division of Nephrology, Department of Medicine, Show Chwan Memorial Hospital, Changhua, Taiwan Department of Nephrology, Buddhist Dalin Tzu Chi General Hospital, Chiayi, Taiwan d Division of Nephrology, Department of Medicine, Chung Shan Medical University Hospital, Taichung, Taiwan e School of Medicine, Chung Shan Medical University, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 11 September 2015 Received in revised form 17 November 2015 Accepted 19 November 2015 Available online 26 November 2015 Keywords: Albumin Matrix metalloproteinase Hemodialysis Survival
a b s t r a c t Background: We evaluated the ability of matrix metalloproteinase (MMP)-2, MMP-9, myeloperoxidase, osteopontin and stromal cell-derived factor 1 to predict mortality in hemodialysis (HD) patients. Methods: One hundred forty HD patients were enrolled and followed from December 2007 until December 2012. At the end of this 5-year period, data were compared between the patients who were alive and those who had died. Results: The patients who alive were younger (56 vs. 63 y), with lower frequency of diabetes mellitus (34.34% vs. 58.53%), higher concentrations of albumin (4.13 vs. 3.91 mg/dl) and lower concentrations of MMP-2 (430.76 vs. 521.59 ng/ml). Multivariate analysis showed that age (HR = 1.03, p = 0.02), diabetes mellitus (HR = 2.395, p = 0.012), albumin (HR = 0.475, p = 0.047) and MMP-2 (HR = 1.003, p = 0.005) were independent factors predicting mortality in HD patients. Receiver operating characteristic curve analysis showed that albumin (AUC = 0.628, p = 0.027) and MMP-2 (AUC = 0.643, p = 0.004) had a similar ability (p = 0.76) to predict survival of HD patients. Conclusions: Compared with albumin, serum MMP-2 is a non-inferior prognostic marker for predicting the survival of HD patients. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Serum albumin concentration is a powerful predictive marker of mortality in hemodialysis (HD) patients [1]. It is influenced by nutrition and other factors including inflammation and depression, and comorbid conditions including cardiac failure, ischemic heart disease, and cardiac mortality in HD patients [2]. The Dialysis Outcomes and Practice Patterns Study (DOPPS) reported that patients with a serum albumin concentration of b3.5 g/dl had a higher risk of mortality [3], and another 10-year cohort study reported an increased risk of mortality in patients with a serum albumin concentration below 3.8 g/dl [4]. A meta-analysis of 38 studies including a total of 265,330 patients showed an inverse relationship between albumin and all-cause and cardiovascular mortality [5]. Hypoalbuminemia predisposes dialysis patients to vascular events
⁎ Corresponding author at: Division of Nephrology, Department of Medicine, Chung Shan Medical University Hospital, No. 110, Sec. 1, Jianguo N Road, South District, Taichung 40201, Taiwan. E-mail address: [email protected] (H.-R. Chang). 1 Contributed equally to this article.
http://dx.doi.org/10.1016/j.cca.2015.11.019 0009-8981/© 2015 Elsevier B.V. All rights reserved.
(acute myocardial infection, hospitalization of heart failure), and this accounts for the increase in mortality [6]. Hemodialysis patients are at a significantly higher risk of cardiovascular events than the general population [7], and atherosclerotic cardiovascular disease (CVD) appears to increase mortality and morbidity in dialysis patients [8]. Long-term HD treatment leads to abnormalities in metabolic, humoral, inflammatory and hemodynamic factors (e.g., chronic volume overload, oxidative stress, hyperphosphatemia, upregulation of cytokines and C-reactive protein), which can promote the activation of matrix metalloproteinases (MMPs) [9,10]. MMPs are a large family of zinc-dependent proteases that regulate tissue remodeling, cell proliferation and angiogenesis by cleaving many components of the extracellular matrix [11]. MMP genes expressions are regulated by inflammatory cytokines and their excessive or inappropriate synthesis have been associated with atherosclerotic CVD in patients with uremia [10,12,13]. We previously reported increased MMP-2 and decreased MMP-9 concentrations in patients with chronic kidney disease (CKD) [14], and more recent studies have suggested that increased MMP-2 concentrations correlate with proteinuria, oxidative stress, and CVD in dialysis patients [10,15]. MMP-9 activation promotes transforming growth factor-beta signaling and a sequence of events that leads to
162
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
elastocalcinosis and vascular stiffness [16]. Upregulated MMP-9 disrupts elastin and affects normal vascular smooth muscle cells allowing the extracellular matrix to expand and calcify, eventually leading to overt hydroxyapatite deposition and vascular dysfunction [17]. Other biomarkers have been associated with atherosclerosis in dialysis patients in addition to MMP-2 and MMP-9. Myeloperoxidase (MPO) is an abundant hemoprotein, released mainly by activated neutrophils into the blood stream when neutrophils become activated [18]. Previous studies have shown that MPO concentrations in the blood predict the risk of clinical events in patients with acute coronary syndromes, demonstrating the potential usefulness of MPO to identify patients at high risk [19,20]. Osteopontin is a phosphoprotein expressed in mineral tissues which inhibits mineralization by blocking hydroxyapatite formation and activating osteoclast function [21]. Osteopontin has been associated with vascular calcification and may be a risk factor for all-cause and cardiovascular mortality in patients undergoing dialysis [22]. Furthermore, stromal cell-derived factor 1 (SDF-1) encodes an inflammatory chemokine expressed in cells of relevance to CVD [23]. Mehta et al. reported that a higher concentration of plasma SDF-1 was associated with CVD risk factors and prevalent CVD as well as the risk of incident myocardial infarction and mortality in CKD patients [24]. Taken together, we hypothesized that MMP-2, MMP-9, MPO, osteopontin and SDF-1may contribute to adverse outcomes in HD patients. Therefore, the aim of this study was to investigate the ability of MMP2, MMP-9, MPO, osteopontin and SDF-1 to predict long-term mortality in HD patients in a 5-y cohort study. 2. Materials and methods
loading buffer containing 0.01% SDS, but with no reducing agents or boiling of the samples before loading. After electrophoresis, the gel was washed twice with 100 ml distilled water containing 2% Triton X100 for 30 min at room temperature. The gel was then incubated in 100 ml reaction buffer (40 mmol/l Tris–HCl, pH 8.0, 10 mmol/l CaCl2, 0.02% NaN3) for 16 h at 37 °C, stained with Coomassie brilliant blue R250 and destained with methanol-acetic acid-water (50/75/875, v/v/v).
2.4. Measurement of plasma concentration of osteopontin, MPO and SDF-1 with Enzyme-Linked Immunosorbent Assay method (ELISA) The osteopontin, MPO and SDF-1 concentrations in the plasma samples were analyzed using human ELISA kits (DOST00, DMYE00B and DSA00 respectively; R&D Systems). Fifty microliters of each plasma sample was directly transferred to the microtest strip wells of the ELISA plate and then assayed according to the manufacturer's instructions. Absorbance was measured at 450 nm in a microtest plate spectrophotometer. Osteopontin, MPO and SDF-1 concentrations were quantified with a calibration curve using human osteopontin, MPO and SDF-1 as a standard. The detection concentrations of biomarkers are as follows: Osteopotin (0.024 ng/ml): MPO (0.618 ng/ml) and SDF-1 (47 pg/ml).
2.5. Outcome The study outcome was 5 y all-cause mortality. The causes of death were ascertained by review of the patients' medical records and death announcement registration.
2.1. Patients and laboratory measurements From December 2007 to December 2012, 140 HD patients were recruited. Approval for this study was granted by the Institutional Review Board of Chung Shan Medical University Hospital. Baseline characteristics were obtained from the database of Chung Shan Medical University Hospital, including age, gender, body mass index, diabetes mellitus (DM), hypertension, hepatitis B, hepatitis C, smoking, hemodialysis mode (low/high flux), urea reduction rate, Kt/V and ultrafiltration rate. Laboratory data were measured from fasting blood samples using an autoanalyzer (Roche Diagnostics GmbH, Cobas Integra 400). Laboratory measurements included pre-dialysis serum concentrations of albumin, urea nitrogen, serum creatinine, sodium, potassium, calcium, phosphate, uric acid, intact parathyroid hormone (iPTH), alanine aminotransferase, aspartate aminotransferase, white blood cells, hemoglobin, hematocrit, cholesterol, triglycerides and fasting blood glucose. Intact PTH chemiluminescence immunoassay (Roche) with a reported reference range of 15–88 pg/ml was used to assess iPTH concentrations. 2.2. Dialysis procedure All patients received regular HD for 4 to 5 h three times per week at a blood flow rate of 250 to 350 ml/min via arteriovenous fistulas or graft. A bicarbonate based dialysate was used at a flow rate of 500 to 700 ml/min in each patient. All HD sessions were performed using a high-flux or low-flux polysulfone membrane dialyzer (APS; Asahi Medical). The efficiency of dialysis was assessed based on the delivered dose of dialysis (Kt/V) calculated using a single-pool urea kinetic model. 2.3. Measurement of serum MMP-2 and MMP-9 concentrations Activities of MMP-2 and -9 were determined by gelatin zymography according to the protocol developed by Kleineret et al. [25]. Eight percent SDS polyacrylamide gels (acrylamide/bis-acrylamide = 30/1.2) containing 0.1% gelatin was used for electrophoresis. Plasma samples (containing 10 μg protein) were prepared with a standard SDS-gel-
2.6. Statistical analysis Continuous variables were expressed as mean ± SD, and categorical variables as number or percentage. To determine the differences between the patients who were still alive and those who had died, the Student's t-test or the Mann–Whitney U test was used for continuous variables, and the chi-square test for categorical variables. A Cox proportional hazards model was used to assess all variables and determine the significance of variables for predicting all-cause 5-y mortality. To determine the risk of mortality, hazard ratios (HRs) and 95% confidence intervals (CIs) were obtained using the Cox proportional hazards model. A univariate Cox proportional hazards regression model was used to calculate the risk of mortality, and stepwise multivariate Cox regression analysis was used to identify the risk factors for mortality in these patients. The positive predictive values for MMP-2 and albumin concentration were also analyzed using receiver operating characteristic (ROC) curve analysis by calculating the area under the ROC curve (AUC). A p b 0.05 was considered to be statistically significant. All data were analyzed using SPSS statistical software ver. 14.0.
3. Results 3.1. Patient characteristics Of the 140 HD patients, 99 were still alive and 41 had died (Table 1). The patients who were still alive were younger (56 ± 14 vs. 63 ± 13 y, p = 0.005), had a lower frequency of DM (34.34% vs. 58.53%, p = 0.014), more high flux dialysis (50.00% vs. 13.88%, p = 0.019), higher concentrations of serum creatinine (10.37 ± 2.37 vs. 8.99 ± 1.61 mg/dl, p = 0.001), albumin (4.13 ± 0.35 vs. 3.91 ± 0.48 mg/dl, p = 0.18), and iPTH (244.22 ± 305.15 vs. 114.39 ± 117.23 pg/ml, p = 0.045), but lower concentrations of MMP-2 (430.76 ± 181.92 vs. 521.59 ± 138.72 ng/ml, p = 0.008) and MPO (62.61 ± 23.36 vs. 73.33 ± 25.51 ng/ml, p = 0.018).
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166 Table 1 Demographic and clinical characteristics of the patients who survived and those who died.
Patient number Age (years) Gender (M/F) BMI Diabetes mellitus Hypertension HBV HCV Smoking Dialysis mode (low/high) BUN (mg/dl) Creatinine (mg/dl) URR Kt/V (Daugirdes) Kt/V (Gotch) Ultrafiltration rate Cholesterol (mg/dl) Triglyceride (mg/dl) Glucose [AC] (mg/dl) Albumin (mg/dl) Uric acid (mg/dl) Na (meq/l) K (meq/l) Ca(mg/dl) P (mg/dl) iPTH (pg/ml) ALT (IU/L) AST (IU/L) WBC (×1000/ul) Hemoglobin (g/dl) Hematocrit (%) MMP-2 (ng/ml) MMP-9 (ng/ml) MPO (ng/ml) Osteopontin (ng/ml) SDF-1 (pg/ml)
Alive
Dead
p value
99 56 ± 15 45/54 22.09 ± 3.37 34 (34.34%) 20 (20.20%) 11 (11.11%) 6 (6.06%) 16 (16.16%) 66/33 (50.00%) 72.3 ± 17.95 10.37 ± 2.37 0.74 ± 0.05 1.65 ± 0.26 1.37 ± 0.21 4.55 ± 1.70 163.88 ± 37.03 154.36 ± 127.93 112.16 ± 66.76 4.13 ± 0.35 7.18 ± 1.42 137.26 ± 2.71 4.61 ± 0.67 9.38 ± 0.84 4.68 ± 1.25 244.22 ± 305.15 16.24 ± 12.72 16.82 ± 8.52 6.37 ± 1.95 10.57 ± 1.24 32.98 ± 3.61 430.76 ± 181.92 331.9 ± 163.44 62.61 ± 23.36 20.54 ± 38.7 2884.3 ± 1330.72
41 63 ± 13 21/20 22.82 ± 3.41 24 (58.53%) 5 (12.19%) 7 (17.07%) 1 (2.43%) 9 (21.95%) 36/5 (13.88%) 68.35 ± 16.73 8.99 ± 1.61 0.73 ± 0.05 1.57 ± 0.22 1.32 ± 0.19 4.36 ± 1.57 158.85 ± 40.52 128.54 ± 62.89 117.46 ± 89.43 3.91 ± 0.48 6.95 ± 1.33 136.51 ± 3.61 4.74 ± 0.78 9.21 ± 0.63 4.47 ± 1.44 114.39 ± 117.23 18.39 ± 11.95 19.42 ± 11.67 6.16 ± 2.06 10.35 ± 1.26 33.45 ± 5.48 521.59 ± 138.72 344.76 ± 213.69 73.33 ± 25.51 26.31 ± 50.9 3306.96 ± 1236.85
0.005⁎ NS NS 0.014⁎ NS NS NS NS 0.019⁎ NS 0.001⁎ NS NS NS NS NS NS NS 0.018⁎ NS NS NS NS NS 0.045⁎ NS NS NS NS NS 0.008⁎ NS 0.018⁎ NS NS
HBV, hepatitis B; HCV, hepatitis C; URR, urea reduction ratio; iPTH, intact parathyroid hormone; MMP-2, matrix metalloproteinase-2; MMP-9, matrix metalloproteinase-9; MPO, myeloperoxidase; SDF-1, stromal cell-derived factor 1. ⁎ p b 0.05 indicates significance.
3.2. Univariate and stepwise multivariate Cox proportional hazards regression analysis to predict mortality in the HD patients In univariate Cox proportional hazards regression analysis, the significant factors predicting 5-year mortality of the HD patients included: age (HR = 1.032, 95% CI = 1.008–1.057, p = 0.009), DM (HR = 2.358, 95% CI = 1.27–4.387, p = 0.007), dialysis mode (HR = 0.319, 95% CI = 0.126–0.808, p = 0.017), serum creatinine (HR = 0.773, 95% CI = 0.661–0.905, p = 0.001), albumin (HR = 0.324, 95% CI = 0.167– 0.629, p = 0.001), iPTH (HR = 0.998, 95% CI = 0.996–1, p = 0.02), MMP-2 (HR = 1.003, 95% CI = 1.001–1.005, p = 0.004) and MPO (HR = 1.015, 95% CI = 1.003–1.027, p = 0.019) (Table 2).
Table 2 Univariate Cox proportional hazards regression analysis.
Age Diabetes mellitus Dialysis mode (low/high) Creatinine (mg/dl) Albumin (mg/dl) iPTH (pg/ml) MMP-2 (ng/ml) MPO (ng/ml)
HR
95% C.I.
p value
1.032 2.358 0.319 0.773 0.324 0.998 1.003 1.015
1.008–1.057 1.27–4.378 0.126–0.808 0.661–0.905 0.167–0.629 0.996–1 1.001–1.005 1.003–1.027
0.009⁎ 0.007⁎ 0.017⁎ 0.001⁎ 0.001⁎ 0.02⁎ 0.004⁎ 0.019⁎
iPTH, intact parathyroid hormone; MMP-2, matrix metalloproteinase-2; MPO, myeloperoxidase. ⁎ p b 0.05 indicates significance.
163
After conducting stepwise multivariate Cox proportional hazards regression analysis by adjusting for markers with a p value less than 0.1 in univariate analysis (age, DM, gender, mode of dialysis, serum creatinine, urea reduction rate, KT/V, iPTH, aspartate aminotransferase, alanine aminotransferase, albumin, MPO, MMP-2, and SDF-1), the independent factors predicting mortality in the HD patients were age (HR = 1.03, 95% CI = 1.005–1.056, p = 0.02), DM (HR = 2.395, 95% CI = 1.217–4.714, p = 0.012), albumin (HR = 0.475, 95% CI = 0.229–0.987, p = 0.047) and MMP-2 (HR = 1.003, 95% CI = 1.001–1.005, p = 0.005) (Table 3). 3.3. ROC and Kaplan–Meier curve analysis to predict mortality in the HD patients ROC curve analysis showed that albumin (AUC = 0.628, p = 0.027) and MMP-2 (AUC = 0.643, p = 0.004) could predict mortality in the HD patients, and that both had a similar ability (p = 0.76) to predict mortality in these patients (Fig. 1). The cut-off value of MMP-2 was ≥ 492.9 ng/ml (sensitivity = 68.9%, 95% CI = 51.9–81.9; specificity = 56.6%, 95% CI = 46.2–66.5) and albumin was ≤3.7 mg/dl (sensitivity = 40%, 95% CI = 24.9–57.7; specificity = 93.9%, 95% CI = 87.7–97.7). Fig. 2 was Kaplan–Meier analysis curve in HD patients at risk for allcause mortality. Patients with MMP-2 ≥ 492.9 ng/ml had a significant lower survival rate during 50 months of follow-up period, compared with those with MMP-2 b 492.9 ng/ml (p = 0.006). 4. Discussion The result of this study show that the HD patients who died were older, had a higher frequency of DM, a lower albumin concentration and a higher MMP-2 concentration. To the best of our knowledge, this is the first study to demonstrate that MMP-2 is a significant predictive marker for mortality in HD patients. Cardiovascular disease is the leading cause of morbidity and mortality in patients with CKD [26], including enhanced arterial calcification and activation of fibroblasts and cytokines, which eventually leads to vascular extracellular matrix remodeling [27,28]. MMP abnormalities have been reported to be involved in the vascular changes associated with kidney failure [17], and to be a mechanism for CVD complications in dialysis patients [10,29,30]. Interestingly, experimental evidence has shown that early upregulation of MMP-2 in areas of elastin degradation and smooth muscle cells change the phenotype during the course of CKD, and that this is associated with increased circulating MMP-2 concentrations [17,27,28]. These alterations have been reported to promote vascular medial layer calcification [17], and increases in MMP-2 concentrations have been found to correlate positively with vascular stiffness and phosphate concentration in CKD patients [31]. The clinical utility of MMP-2 in HD patients had been discussed previously. First, MMP-2 genes expressions have been associated with atherosclerotic CVD in patients with uremia [10,12,13]. Krystyna et al. reported significantly increased serum MMP-2 concentrations in patients with CVD, and that this was associated with the prevalence of
Table 3 Risk of mortality in the HD patients using stepwise multivariate Cox proportional hazards regression analysis.
Age Diabetes mellitus Albumin MMP-2
aHR
95% C.I.
p value
1.03 2.395 0.475 1.003
1.005–1.056 1.217–4.714 0.229–0.987 1.001–1.005
0.02⁎ 0.012⁎ 0.047⁎ 0.005⁎
MMP-2, matrix metalloproteinase-2. (Adjusted for age, diabetes mellitus, gender, mode of dialysis, serum creatinine, urea reduction rate, KT/V, intact parathyroid hormone, aspartate aminotransferase, alanine aminotransferase, albumin, myeloperoxidase, MMP-2, stromal cell-derived factor 1). ⁎ p b 0.05 indicates significance.
164
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
Fig. 1. Predictive accuracy of albumin and MMP-2 was estimated by using time-dependent ROC curve analysis. ROC curve analysis showed that albumin (AUC = 0.628, 95% CI: 0.542–0.709, p = 0.027) and MMP-2 (AUC = 0.643, 95% CI: 0.557–0.722, p = 0.004) could predict mortality in the HD patients, and that both have a similar ability (p = 0.76) for predicting mortality in the dialysis patients. The cut-off value of MMP-2 was ≥492.9 ng/ml (sensitivity = 68.89%, specificity = 56.57%) and albumin was ≤3.7 mg/dl (sensitivity = 40%, specificity = 93.94%). p b 0.05 indicates significance.
atherosclerosis and CVD [10]. Pasterkamp et al. reported elevated expressions of MMP-2 within plaques, and that this played a part in plaque rupture by weakening it [32]. In an animal study, MMP-2 was found to predispose the extracellular matrix to degradation and facilitate arterial calcification in rats with CKD [17]. Second, MMP-2 may be a clinically useful indicator of an inflammatory process in HD patients [33]. The initiation of HD therapy is associated with increased oxidative status, which may not only explain the relatively higher concentrations of serum MMP-2 [34], but also suggests that serum MMP-2 is associated with vascular calcification and CVD complications in dialysis patients [29,30]. Krystyna et al. revealed a possible association between elevated
oxidative stress and the MMP-2 in HD patients, which could represent one of the mechanisms involved in the accelerated atherosclerosis in this population [10]. Third, elastin degradation including MMP-2, was associated progressive aortic stiffening and all-cause mortality in predialysis CKD [35]. However, the predictive mortality role of MMP-2 in dialysis patients was still controversial. Marta et al. revealed that MMP-2 was not an independent mortality predictor in long-term HD patients (HR = 1.078, 95% CI = 0.921–1.262, p = 0.4) [36]. The MMP2 concentration in this study was 381 ± 123 ng/dl and was no significant correlation with all-cause mortality. In our patients, the concentration of serum MMP-2 was higher in the patients who died and it was an independent predictor for 5-year mortality. The cut-off value of the MMP-2 was ≥ 492.9 ng/ml, and the sensitivity and specificity for predicting mortality were 68.3% and 56.6% respectively. The MMP-2 has reliable sensitivity but limited specificity for predicting survival of HD patients. Further investigations are needed to determine a more accurate cut-off value and the clinical relevance of the MMP-2. Serum albumin concentration is a well-known and powerful predictive factor of mortality in HD patients [1,2]. The association between low serum albumin and higher mortality may be explained by concurrent inflammatory processes [37] or due to a poor nutritional status [2,6]. There were several studies indicated that serum albumin was a predictive marker in dialysis patients. Hirokazu et al. revealed serum albumin was a predictor of malnutrition, CVD, and mortality in patients with end stage renal disease [38]. The AUC value of albumin was 0.72 for total mortality and cut-off value was b 3.3 mg/dl with 26 months follow up. Another 5 y cohort study revealed that serum albumin correlated with all-cause mortality in HD patients (AUC value of albumin was 0.65) [39]. In our study, the AUC and cut-off value of albumin was 0.62 and ≤3.7 mg/dl respectively. This result was similar to the previous literature to demonstrate that serum albumin was powerful associated with mortality in patients undergoing dialysis. Based on our ROC curve analysis, MMP-2 was comparable to albumin in its capacity to predict mortality in HD patients. In other word, MMP-2 could be a significant predictive marker for mortality in HD patients. Matrix metalloproteinase-9 is mainly produced by blood monocytes, and chronic stimulation of these cells during extracorporeal circulation during HD sessions may be responsible for inducting its expression [40]. Progression of CKD has been shown to be related to a decrease in MMP-9 [14]. In addition, Rysz et al. reported a decreased concentration of MMP-9 during HD [41]. In the animal study, increased MMP-9 could
Fig. 2. Kaplan–Meier analysis curve in hemodialysis patients at risk for all-cause mortality. Patients with MMP-2 ≥ 492.9 ng/ml had a significant lower survival rate during the 50 months follow-up period, compared with those with MMP-2 b 492.9 ng/ml (p = 0.006).
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
lead to altered extracellular matrix and, altered arterial structure and cause arterial calcification in uremic rats [17]. However, Krystyna et al. reported no significant correlation between MMP-9 and the prevalence of CVD in HD patients [10]. Another prospective observational cohort study which included 261 long-term HD patients followed up for 5 y, MMP-9 was not a significant independent predictor of overall mortality in patients undergoing dialysis (HR = 0.953, 95% CI = 0.804–1.131, p = 0.6) [36]. This is consistent with our results in which there was no correlation between mortality and serum MMP-9 concentration in the dialysis patients. The process of HD can trigger inflammation as a result of exposing the blood to a bio-incompatible system which stimulates monocytes and macrophages, thereby inducing the production of proinflammatory cytokines and contributing to the maintenance of oxidative stress [42]. The production of interleukins leading to a release of MPO has been implicated in the development of CVD in dialysis patients [43]. In the previous study, increased concentrations of MPO have been closely linked to mortality in dialysis patients [44]. Kamyar et al. reported a total 356 HD patients with a 3 y cohort study, the median serum MPO concentration was 393 pg/ml with relative risk 1.14 (p = 0.01) for all-cause mortality. Each 272 pg/ml increase in serum MPO concentration, there was 8% to 15% increase in risk for death during the observed 3-y interval according to the concentration of multivariate adjustment [44]. However, the MPO concentration was not significantly increased (73.33 pg/ml) and that could be possibility no associated with mortality in our results. A soluble decoy receptor of the osteoclast activator, osteopontin, has been reported to act as an important regulatory molecule in vascular calcification independent of classical risk factors [22]. Increased osteopontin concentrations have also been associated with increased vascular stiffness [45], and higher osteopontin concentrations may be independent predictors of all-cause mortality in diabetic nephropathy patients [46]. However, the predictive mortality role of osteopontin in dialysis patients was still controversial. Serum osteopontin was not associated with cardiovascular events in a prospective observational study with 97 dialysis patients over 5-y observation [47]. Another 3-y cohort study with 602 dialysis patients, Scialla et al. reported osteopontin was not a significant risk factor for all-cause and cardiovascular mortality in patients undergoing dialysis [22]. Our study also revealed no association with osteopontin and mortality in HD patients. Stromal derived factor-1, also called CXC motif ligand 12 (CXCL12), is a complex chemokine with putative anti-inflammatory and antiatherogenic functions that has also been reported to play a role in recruiting endothelial cells in response to vascular injury [23]. Mehta et al. reported that in CKD patients, higher concentrations of plasma SDF-1 were associated with known cardiovascular risk factors and prevalent CVD, and also predicted incident myocardial infarction and mortality even after adjusting for traditional risk factors and kidney function [24]. However, there are currently no reports demonstrating an association between SDF-1 and mortality in dialysis patients, and SDF-1 was not a predictive factor for mortality in the current study. There are some limitations to this study. First, the number of patients was small. Second, since the concentrations of the serum biomarkers were measured only once. Third, this study was only a single-center nature. Large prospective studies are needed to confirm whether these biomarkers, and especially MMP-2, are useful for predicting mortality in HD patients. In conclusion, this study shows that concentrations of serum MMP-2 and albumin are both important predictive factors of mortality in HD patients, and that compared with albumin, serum MMP-2 is a noninferior marker for predicting survival of HD patients. Acknowledgments This study was supported by a grant from the National Science Council 101-2314-B-040-003. The authors wish to thank Mr. Nigel Daly for editing and reviewing the manuscript for English language.
165
References [1] A.A. Lopes, J.L. Bragg-Gresham, S.J. Elder, et al., Independent and joint associations of nutritional status indicators with mortality risk among chronic hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS), J. Ren. Nutr. 20 (2010) 224–234. [2] C. Combe, K.P. McCullough, Y. Asano, N. Ginsberg, B.J. Maroni, T.B. Pifer, Kidney Disease Outcomes Quality Initiative (K/DOQI) and the Dialysis Outcomes and Practice Patterns Study (DOPPS): nutrition guidelines, indicators, and practices, Am. J. Kidney Dis. 44 (2004) 39–46. [3] B.D. Bradbury, R.B. Fissell, J.M. Albert, et al., Predictors of early mortality among incident US hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS), Clin. J. Am. Soc. Nephrol. 2 (2007) 89–99. [4] A. Kato, T. Takita, M. Furuhashi, Y. Maruyama, A. Hishida, Comparison of serum albumin, C-reactive protein and carotid atherosclerosis as predictors of 10-year mortality in hemodialysis patients, Hemodial. Int. 14 (2010) 226–232. [5] M. Herselman, N. Esau, J.M. Kruger, D. Labadarios, M.R. Moosa, Relationship between serum protein and mortality in adults on long-term hemodialysis: exhaustive review and meta-analysis, Nutrition 26 (2010) 10–32. [6] B.A. Cooper, E.L. Penne, L.H. Bartlett, C.A. Pollock, Protein malnutrition and hypoalbuminemia as predictors of vascular events and mortality in ESRD, Am. J. Kidney Dis. 43 (2004) 61–66. [7] V. Menon, A. Gul, M.J. Sarnak, Cardiovascular risk factors in chronic kidney disease, Kidney Int. 68 (2005) 1413–1418. [8] R.N. Foley, P.S. Parfrey, J.D. Harnett, et al., Clinical and echocardiographic disease in patients starting end-stage renal disease therapy, Kidney Int. 47 (1995) 186–192. [9] R.J. Bisoendial, R.S. Birjmohun, F. Akdim, et al., C-reactive protein elicits white blood cell activation in humans, Am. J. Med. 122 (2009) 581–589. [10] K. Pawlak, D. Pawlak, M. Mysliwiec, Serum matrix metalloproteinase-2 and increased oxidative stress are associated with carotid atherosclerosis in hemodialyzed patients, Atherosclerosis 190 (2007) 199–204. [11] A. Page-McCaw, A.J. Ewald, Z. Werb, Matrix metalloproteinases and the regulation of tissue remodelling, Nat. Rev. Mol. Cell Biol. 8 (2007) 221–233. [12] M. Peiskerova, M. Kalousova, M. Kratochvilova, et al., Fibroblast growth factor 23 and matrix-metalloproteinases in patients with chronic kidney disease: are they associated with cardiovascular disease? Kidney Blood Press. Res. 32 (2009) 276–283. [13] M. Nagano, K. Fukami, S. Yamagishi, et al., Circulating matrix metalloproteinase-2 is an independent correlate of proteinuria in patients with chronic kidney disease, Am. J. Nephrol. 29 (2009) 109–115. [14] H.R. Chang, S.F. Yang, M.L. Li, C.C. Lin, Y.S. Hsieh, J.D. Lian, Relationships between circulating matrix metalloproteinase-2 and -9 and renal function in patients with chronic kidney disease, Clin. Chim. Acta 366 (2006) 243–248. [15] K. Pawlak, J. Tankiewicz, M. Mysliwiec, D. Pawlak, Systemic levels of MMP2/TIMP2 and cardiovascular risk in CAPD patients, Nephron Clin. Pract. 115 (2010) 251–258. [16] C. Bouvet, S. Moreau, J. Blanchette, D. de Blois, P. Moreau, Sequential activation of matrix metalloproteinase 9 and transforming growth factor beta in arterial elastocalcinosis, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 856–862. [17] N.X. Chen, K.D. O'Neill, X. Chen, K. Kiattisunthorn, V.H. Gattone, S.M. Moe, Activation of arterial matrix metalloproteinases leads to vascular calcification in chronic kidney disease, Am. J. Nephrol. 34 (2011) 211–219. [18] T. Naruko, M. Ueda, K. Haze, et al., Neutrophil infiltration of culprit lesions in acute coronary syndromes, Circulation 106 (2002) 2894–2900. [19] M.L. Brennan, M.S. Penn, F. Van Lente, et al., Prognostic value of myeloperoxidase in patients with chest pain, N. Engl. J. Med. 349 (2003) 1595–1604. [20] S. Baldus, C. Heeschen, T. Meinertz, et al., Myeloperoxidase serum levels predict risk in patients with acute coronary syndromes, Circulation 108 (2003) 1440–1445. [21] M. Scatena, L. Liaw, C.M. Giachelli, Osteopontin: a multifunctional molecule regulating chronic inflammation and vascular disease, Arterioscler. Thromb. Vasc. Biol. 27 (2007) 2302–2309. [22] J.J. Scialla, W.H. Kao, C. Crainiceanu, et al., Biomarkers of vascular calcification and mortality in patients with ESRD, Clin. J. Am. Soc. Nephrol. 9 (2014) 745–755. [23] S. Abi-Younes, A. Sauty, F. Mach, G.K. Sukhova, P. Libby, A.D. Luster, The stromal cellderived factor-1 chemokine is a potent platelet agonist highly expressed in atherosclerotic plaques, Circ. Res. 86 (2000) 131–138. [24] N.N. Mehta, G.J. Matthews, P. Krishnamoorthy, et al., Higher plasma CXCL12 levels predict incident myocardial infarction and death in chronic kidney disease: findings from the Chronic Renal Insufficiency Cohort study, Eur. Heart J. 35 (2014) 2115–2122. [25] D.E. Kleiner, W.G. Stetler-Stevenson, Quantitative zymography: detection of picogram quantities of gelatinases, Anal. Biochem. 218 (1994) 325–329. [26] S. Ardhanari, M.A. Alpert, K. Aggarwal, Cardiovascular disease in chronic kidney disease: risk factors, pathogenesis, and prevention, Adv. Perit. Dial. 30 (2014) 40–53. [27] A. Pai, E.M. Leaf, M. El-Abbadi, C.M. Giachelli, Elastin degradation and vascular smooth muscle cell phenotype change precede cell loss and arterial medial calcification in a uremic mouse model of chronic kidney disease, Am. J. Pathol. 178 (2011) 764–773. [28] C. Kumata, M. Mizobuchi, H. Ogata, et al., Involvement of matrix metalloproteinase2 in the development of medial layer vascular calcification in uremic rats, Ther. Apher. Dial. 15 (2011) 18–22. [29] K. Pawlak, M. Mysliwiec, D. Pawlak, Peripheral blood level alterations of MMP-2 and MMP-9 in patients with chronic kidney disease on conservative treatment and on hemodialysis, Clin. Biochem. 44 (2011) 838–843. [30] R.S. Friese, F. Rao, S. Khandrika, et al., Matrix metalloproteinases: discrete elevations in essential hypertension and hypertensive end-stage renal disease, Clin. Exp. Hypertens. 31 (2009) 521–533.
166
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
[31] A.W. Chung, H.H. Yang, M.K. Sigrist, et al., Matrix metalloproteinase-2 and -9 exacerbate arterial stiffening and angiogenesis in diabetes and chronic kidney disease, Cardiovasc. Res. 84 (2009) 494–504. [32] G. Pasterkamp, A.H. Schoneveld, D.J. Hijnen, et al., Atherosclerotic arterial remodeling and the localization of macrophages and matrix metalloproteases 1, 2 and 9 in the human coronary artery, Atherosclerosis 150 (2000) 245–253. [33] G.A. Preston, C.V. Barrett, D.A. Alcorta, et al., Serum matrix metalloproteinases MMP2 and MMP-3 levels in dialysis patients vary independently of CRP and IL-6 levels, Nephron 92 (2002) 817–823. [34] K. Pawlak, D. Pawlak, M. Mysliwiec, Impaired renal function and duration of dialysis therapy are associated with oxidative stress and proatherogenic cytokine levels in patients with end-stage renal disease, Clin. Biochem. 40 (2007) 81–85. [35] E.R. Smith, L.A. Tomlinson, M.L. Ford, L.P. McMahon, C. Rajkumar, S.G. Holt, Elastin degradation is associated with progressive aortic stiffening and all-cause mortality in predialysis chronic kidney disease, Hypertension 59 (2012) 973–978. [36] M. Kalousova, H. Benakova, A.A. Kubena, S. Dusilova-Sulkova, V. Tesar, T. Zima, Pregnancy-associated plasma protein A as an independent mortality predictor in long-term hemodialysis patients, Kidney Blood Press. Res. 35 (2012) 192–201. [37] R. de Mutsert, D.C. Grootendorst, F. Indemans, E.W. Boeschoten, R.T. Krediet, F.W. Dekker, Association between serum albumin and mortality in dialysis patients is partly explained by inflammation, and not by malnutrition, J. Ren. Nutr. 19 (2009) 127–135. [38] H. Honda, A.R. Qureshi, O. Heimburger, et al., Serum albumin, C-reactive protein, interleukin 6, and fetuin a as predictors of malnutrition, cardiovascular disease, and mortality in patients with ESRD, Am. J. Kidney Dis. 47 (2006) 139–148.
[39] S.C. Hung, T.W. Hsu, Y.P. Lin, D.C. Tarng, Decoy receptor 3, a novel inflammatory marker, and mortality in hemodialysis patients, Clin. J. Am. Soc. Nephrol. 7 (2012) 1257–1265. [40] A. Kalela, M. Ponnio, T.A. Koivu, et al., Association of serum sialic acid and MMP-9 with lipids and inflammatory markers, Eur. J. Clin. Investig. 30 (2000) 99–104. [41] J. Rysz, M. Banach, R.A. Stolarek, et al., Serum metalloproteinases MMP-2, MMP-9 and metalloproteinase tissue inhibitors TIMP-1 and TIMP-2 in patients on hemodialysis, Int. Urol. Nephrol. 43 (2011) 491–498. [42] C. Libetta, V. Sepe, P. Esposito, F. Galli, C.A. Dal, Oxidative stress and inflammation: implications in uremia and hemodialysis, Clin. Biochem. 44 (2011) 1189–1198. [43] G.A. Kaysen, N.W. Levin, W.E. Mitch, A.L. Chapman, L. Kubala, J.P. Eiserich, Evidence that C-reactive protein or IL-6 are not surrogates for all inflammatory cardiovascular risk factors in hemodialysis patients, Blood Purif. 24 (2006) 508–516. [44] K. Kalantar-Zadeh, M.L. Brennan, S.L. Hazen, Serum myeloperoxidase and mortality in maintenance hemodialysis patients, Am. J. Kidney Dis. 48 (2006) 59–68. [45] D.V. Barreto, F.C. Barreto, A.B. Carvalho, et al., Coronary calcification in hemodialysis patients: the contribution of traditional and uremia-related risk factors, Kidney Int. 67 (2005) 1576–1582. [46] D. Gordin, C. Forsblom, N.M. Panduru, et al., Osteopontin is a strong predictor of incipient diabetic nephropathy, cardiovascular disease, and all-cause mortality in patients with type 1 diabetes, Diabetes Care 37 (2014) 2593–2600. [47] J.E. Lee, H.J. Kim, S.J. Moon, et al., Serum osteoprotegerin is associated with vascular stiffness and the onset of new cardiovascular events in hemodialysis patients, Korean J. Intern. Med. 28 (2013) 668–677.
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
MMP-2 serum concentrations predict mortality in hemodialysis patients: a 5-year cohort study Kuang-Chih Hsiao a,b,1, Jen-Pi Tsai c,1, Shun-Fa Yang a, Wen-Chin Lee b, Jong-Yu Huang b, Shun-Chi Chang b, Chun-Shuo Hso b, Horng-Rong Chang a,d,e,⁎ a
Institute of Medicine, Chung Shan Medical University, Taichung, Taiwan Division of Nephrology, Department of Medicine, Show Chwan Memorial Hospital, Changhua, Taiwan Department of Nephrology, Buddhist Dalin Tzu Chi General Hospital, Chiayi, Taiwan d Division of Nephrology, Department of Medicine, Chung Shan Medical University Hospital, Taichung, Taiwan e School of Medicine, Chung Shan Medical University, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 11 September 2015 Received in revised form 17 November 2015 Accepted 19 November 2015 Available online 26 November 2015 Keywords: Albumin Matrix metalloproteinase Hemodialysis Survival
a b s t r a c t Background: We evaluated the ability of matrix metalloproteinase (MMP)-2, MMP-9, myeloperoxidase, osteopontin and stromal cell-derived factor 1 to predict mortality in hemodialysis (HD) patients. Methods: One hundred forty HD patients were enrolled and followed from December 2007 until December 2012. At the end of this 5-year period, data were compared between the patients who were alive and those who had died. Results: The patients who alive were younger (56 vs. 63 y), with lower frequency of diabetes mellitus (34.34% vs. 58.53%), higher concentrations of albumin (4.13 vs. 3.91 mg/dl) and lower concentrations of MMP-2 (430.76 vs. 521.59 ng/ml). Multivariate analysis showed that age (HR = 1.03, p = 0.02), diabetes mellitus (HR = 2.395, p = 0.012), albumin (HR = 0.475, p = 0.047) and MMP-2 (HR = 1.003, p = 0.005) were independent factors predicting mortality in HD patients. Receiver operating characteristic curve analysis showed that albumin (AUC = 0.628, p = 0.027) and MMP-2 (AUC = 0.643, p = 0.004) had a similar ability (p = 0.76) to predict survival of HD patients. Conclusions: Compared with albumin, serum MMP-2 is a non-inferior prognostic marker for predicting the survival of HD patients. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Serum albumin concentration is a powerful predictive marker of mortality in hemodialysis (HD) patients [1]. It is influenced by nutrition and other factors including inflammation and depression, and comorbid conditions including cardiac failure, ischemic heart disease, and cardiac mortality in HD patients [2]. The Dialysis Outcomes and Practice Patterns Study (DOPPS) reported that patients with a serum albumin concentration of b3.5 g/dl had a higher risk of mortality [3], and another 10-year cohort study reported an increased risk of mortality in patients with a serum albumin concentration below 3.8 g/dl [4]. A meta-analysis of 38 studies including a total of 265,330 patients showed an inverse relationship between albumin and all-cause and cardiovascular mortality [5]. Hypoalbuminemia predisposes dialysis patients to vascular events
⁎ Corresponding author at: Division of Nephrology, Department of Medicine, Chung Shan Medical University Hospital, No. 110, Sec. 1, Jianguo N Road, South District, Taichung 40201, Taiwan. E-mail address: [email protected] (H.-R. Chang). 1 Contributed equally to this article.
http://dx.doi.org/10.1016/j.cca.2015.11.019 0009-8981/© 2015 Elsevier B.V. All rights reserved.
(acute myocardial infection, hospitalization of heart failure), and this accounts for the increase in mortality [6]. Hemodialysis patients are at a significantly higher risk of cardiovascular events than the general population [7], and atherosclerotic cardiovascular disease (CVD) appears to increase mortality and morbidity in dialysis patients [8]. Long-term HD treatment leads to abnormalities in metabolic, humoral, inflammatory and hemodynamic factors (e.g., chronic volume overload, oxidative stress, hyperphosphatemia, upregulation of cytokines and C-reactive protein), which can promote the activation of matrix metalloproteinases (MMPs) [9,10]. MMPs are a large family of zinc-dependent proteases that regulate tissue remodeling, cell proliferation and angiogenesis by cleaving many components of the extracellular matrix [11]. MMP genes expressions are regulated by inflammatory cytokines and their excessive or inappropriate synthesis have been associated with atherosclerotic CVD in patients with uremia [10,12,13]. We previously reported increased MMP-2 and decreased MMP-9 concentrations in patients with chronic kidney disease (CKD) [14], and more recent studies have suggested that increased MMP-2 concentrations correlate with proteinuria, oxidative stress, and CVD in dialysis patients [10,15]. MMP-9 activation promotes transforming growth factor-beta signaling and a sequence of events that leads to
162
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
elastocalcinosis and vascular stiffness [16]. Upregulated MMP-9 disrupts elastin and affects normal vascular smooth muscle cells allowing the extracellular matrix to expand and calcify, eventually leading to overt hydroxyapatite deposition and vascular dysfunction [17]. Other biomarkers have been associated with atherosclerosis in dialysis patients in addition to MMP-2 and MMP-9. Myeloperoxidase (MPO) is an abundant hemoprotein, released mainly by activated neutrophils into the blood stream when neutrophils become activated [18]. Previous studies have shown that MPO concentrations in the blood predict the risk of clinical events in patients with acute coronary syndromes, demonstrating the potential usefulness of MPO to identify patients at high risk [19,20]. Osteopontin is a phosphoprotein expressed in mineral tissues which inhibits mineralization by blocking hydroxyapatite formation and activating osteoclast function [21]. Osteopontin has been associated with vascular calcification and may be a risk factor for all-cause and cardiovascular mortality in patients undergoing dialysis [22]. Furthermore, stromal cell-derived factor 1 (SDF-1) encodes an inflammatory chemokine expressed in cells of relevance to CVD [23]. Mehta et al. reported that a higher concentration of plasma SDF-1 was associated with CVD risk factors and prevalent CVD as well as the risk of incident myocardial infarction and mortality in CKD patients [24]. Taken together, we hypothesized that MMP-2, MMP-9, MPO, osteopontin and SDF-1may contribute to adverse outcomes in HD patients. Therefore, the aim of this study was to investigate the ability of MMP2, MMP-9, MPO, osteopontin and SDF-1 to predict long-term mortality in HD patients in a 5-y cohort study. 2. Materials and methods
loading buffer containing 0.01% SDS, but with no reducing agents or boiling of the samples before loading. After electrophoresis, the gel was washed twice with 100 ml distilled water containing 2% Triton X100 for 30 min at room temperature. The gel was then incubated in 100 ml reaction buffer (40 mmol/l Tris–HCl, pH 8.0, 10 mmol/l CaCl2, 0.02% NaN3) for 16 h at 37 °C, stained with Coomassie brilliant blue R250 and destained with methanol-acetic acid-water (50/75/875, v/v/v).
2.4. Measurement of plasma concentration of osteopontin, MPO and SDF-1 with Enzyme-Linked Immunosorbent Assay method (ELISA) The osteopontin, MPO and SDF-1 concentrations in the plasma samples were analyzed using human ELISA kits (DOST00, DMYE00B and DSA00 respectively; R&D Systems). Fifty microliters of each plasma sample was directly transferred to the microtest strip wells of the ELISA plate and then assayed according to the manufacturer's instructions. Absorbance was measured at 450 nm in a microtest plate spectrophotometer. Osteopontin, MPO and SDF-1 concentrations were quantified with a calibration curve using human osteopontin, MPO and SDF-1 as a standard. The detection concentrations of biomarkers are as follows: Osteopotin (0.024 ng/ml): MPO (0.618 ng/ml) and SDF-1 (47 pg/ml).
2.5. Outcome The study outcome was 5 y all-cause mortality. The causes of death were ascertained by review of the patients' medical records and death announcement registration.
2.1. Patients and laboratory measurements From December 2007 to December 2012, 140 HD patients were recruited. Approval for this study was granted by the Institutional Review Board of Chung Shan Medical University Hospital. Baseline characteristics were obtained from the database of Chung Shan Medical University Hospital, including age, gender, body mass index, diabetes mellitus (DM), hypertension, hepatitis B, hepatitis C, smoking, hemodialysis mode (low/high flux), urea reduction rate, Kt/V and ultrafiltration rate. Laboratory data were measured from fasting blood samples using an autoanalyzer (Roche Diagnostics GmbH, Cobas Integra 400). Laboratory measurements included pre-dialysis serum concentrations of albumin, urea nitrogen, serum creatinine, sodium, potassium, calcium, phosphate, uric acid, intact parathyroid hormone (iPTH), alanine aminotransferase, aspartate aminotransferase, white blood cells, hemoglobin, hematocrit, cholesterol, triglycerides and fasting blood glucose. Intact PTH chemiluminescence immunoassay (Roche) with a reported reference range of 15–88 pg/ml was used to assess iPTH concentrations. 2.2. Dialysis procedure All patients received regular HD for 4 to 5 h three times per week at a blood flow rate of 250 to 350 ml/min via arteriovenous fistulas or graft. A bicarbonate based dialysate was used at a flow rate of 500 to 700 ml/min in each patient. All HD sessions were performed using a high-flux or low-flux polysulfone membrane dialyzer (APS; Asahi Medical). The efficiency of dialysis was assessed based on the delivered dose of dialysis (Kt/V) calculated using a single-pool urea kinetic model. 2.3. Measurement of serum MMP-2 and MMP-9 concentrations Activities of MMP-2 and -9 were determined by gelatin zymography according to the protocol developed by Kleineret et al. [25]. Eight percent SDS polyacrylamide gels (acrylamide/bis-acrylamide = 30/1.2) containing 0.1% gelatin was used for electrophoresis. Plasma samples (containing 10 μg protein) were prepared with a standard SDS-gel-
2.6. Statistical analysis Continuous variables were expressed as mean ± SD, and categorical variables as number or percentage. To determine the differences between the patients who were still alive and those who had died, the Student's t-test or the Mann–Whitney U test was used for continuous variables, and the chi-square test for categorical variables. A Cox proportional hazards model was used to assess all variables and determine the significance of variables for predicting all-cause 5-y mortality. To determine the risk of mortality, hazard ratios (HRs) and 95% confidence intervals (CIs) were obtained using the Cox proportional hazards model. A univariate Cox proportional hazards regression model was used to calculate the risk of mortality, and stepwise multivariate Cox regression analysis was used to identify the risk factors for mortality in these patients. The positive predictive values for MMP-2 and albumin concentration were also analyzed using receiver operating characteristic (ROC) curve analysis by calculating the area under the ROC curve (AUC). A p b 0.05 was considered to be statistically significant. All data were analyzed using SPSS statistical software ver. 14.0.
3. Results 3.1. Patient characteristics Of the 140 HD patients, 99 were still alive and 41 had died (Table 1). The patients who were still alive were younger (56 ± 14 vs. 63 ± 13 y, p = 0.005), had a lower frequency of DM (34.34% vs. 58.53%, p = 0.014), more high flux dialysis (50.00% vs. 13.88%, p = 0.019), higher concentrations of serum creatinine (10.37 ± 2.37 vs. 8.99 ± 1.61 mg/dl, p = 0.001), albumin (4.13 ± 0.35 vs. 3.91 ± 0.48 mg/dl, p = 0.18), and iPTH (244.22 ± 305.15 vs. 114.39 ± 117.23 pg/ml, p = 0.045), but lower concentrations of MMP-2 (430.76 ± 181.92 vs. 521.59 ± 138.72 ng/ml, p = 0.008) and MPO (62.61 ± 23.36 vs. 73.33 ± 25.51 ng/ml, p = 0.018).
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166 Table 1 Demographic and clinical characteristics of the patients who survived and those who died.
Patient number Age (years) Gender (M/F) BMI Diabetes mellitus Hypertension HBV HCV Smoking Dialysis mode (low/high) BUN (mg/dl) Creatinine (mg/dl) URR Kt/V (Daugirdes) Kt/V (Gotch) Ultrafiltration rate Cholesterol (mg/dl) Triglyceride (mg/dl) Glucose [AC] (mg/dl) Albumin (mg/dl) Uric acid (mg/dl) Na (meq/l) K (meq/l) Ca(mg/dl) P (mg/dl) iPTH (pg/ml) ALT (IU/L) AST (IU/L) WBC (×1000/ul) Hemoglobin (g/dl) Hematocrit (%) MMP-2 (ng/ml) MMP-9 (ng/ml) MPO (ng/ml) Osteopontin (ng/ml) SDF-1 (pg/ml)
Alive
Dead
p value
99 56 ± 15 45/54 22.09 ± 3.37 34 (34.34%) 20 (20.20%) 11 (11.11%) 6 (6.06%) 16 (16.16%) 66/33 (50.00%) 72.3 ± 17.95 10.37 ± 2.37 0.74 ± 0.05 1.65 ± 0.26 1.37 ± 0.21 4.55 ± 1.70 163.88 ± 37.03 154.36 ± 127.93 112.16 ± 66.76 4.13 ± 0.35 7.18 ± 1.42 137.26 ± 2.71 4.61 ± 0.67 9.38 ± 0.84 4.68 ± 1.25 244.22 ± 305.15 16.24 ± 12.72 16.82 ± 8.52 6.37 ± 1.95 10.57 ± 1.24 32.98 ± 3.61 430.76 ± 181.92 331.9 ± 163.44 62.61 ± 23.36 20.54 ± 38.7 2884.3 ± 1330.72
41 63 ± 13 21/20 22.82 ± 3.41 24 (58.53%) 5 (12.19%) 7 (17.07%) 1 (2.43%) 9 (21.95%) 36/5 (13.88%) 68.35 ± 16.73 8.99 ± 1.61 0.73 ± 0.05 1.57 ± 0.22 1.32 ± 0.19 4.36 ± 1.57 158.85 ± 40.52 128.54 ± 62.89 117.46 ± 89.43 3.91 ± 0.48 6.95 ± 1.33 136.51 ± 3.61 4.74 ± 0.78 9.21 ± 0.63 4.47 ± 1.44 114.39 ± 117.23 18.39 ± 11.95 19.42 ± 11.67 6.16 ± 2.06 10.35 ± 1.26 33.45 ± 5.48 521.59 ± 138.72 344.76 ± 213.69 73.33 ± 25.51 26.31 ± 50.9 3306.96 ± 1236.85
0.005⁎ NS NS 0.014⁎ NS NS NS NS 0.019⁎ NS 0.001⁎ NS NS NS NS NS NS NS 0.018⁎ NS NS NS NS NS 0.045⁎ NS NS NS NS NS 0.008⁎ NS 0.018⁎ NS NS
HBV, hepatitis B; HCV, hepatitis C; URR, urea reduction ratio; iPTH, intact parathyroid hormone; MMP-2, matrix metalloproteinase-2; MMP-9, matrix metalloproteinase-9; MPO, myeloperoxidase; SDF-1, stromal cell-derived factor 1. ⁎ p b 0.05 indicates significance.
3.2. Univariate and stepwise multivariate Cox proportional hazards regression analysis to predict mortality in the HD patients In univariate Cox proportional hazards regression analysis, the significant factors predicting 5-year mortality of the HD patients included: age (HR = 1.032, 95% CI = 1.008–1.057, p = 0.009), DM (HR = 2.358, 95% CI = 1.27–4.387, p = 0.007), dialysis mode (HR = 0.319, 95% CI = 0.126–0.808, p = 0.017), serum creatinine (HR = 0.773, 95% CI = 0.661–0.905, p = 0.001), albumin (HR = 0.324, 95% CI = 0.167– 0.629, p = 0.001), iPTH (HR = 0.998, 95% CI = 0.996–1, p = 0.02), MMP-2 (HR = 1.003, 95% CI = 1.001–1.005, p = 0.004) and MPO (HR = 1.015, 95% CI = 1.003–1.027, p = 0.019) (Table 2).
Table 2 Univariate Cox proportional hazards regression analysis.
Age Diabetes mellitus Dialysis mode (low/high) Creatinine (mg/dl) Albumin (mg/dl) iPTH (pg/ml) MMP-2 (ng/ml) MPO (ng/ml)
HR
95% C.I.
p value
1.032 2.358 0.319 0.773 0.324 0.998 1.003 1.015
1.008–1.057 1.27–4.378 0.126–0.808 0.661–0.905 0.167–0.629 0.996–1 1.001–1.005 1.003–1.027
0.009⁎ 0.007⁎ 0.017⁎ 0.001⁎ 0.001⁎ 0.02⁎ 0.004⁎ 0.019⁎
iPTH, intact parathyroid hormone; MMP-2, matrix metalloproteinase-2; MPO, myeloperoxidase. ⁎ p b 0.05 indicates significance.
163
After conducting stepwise multivariate Cox proportional hazards regression analysis by adjusting for markers with a p value less than 0.1 in univariate analysis (age, DM, gender, mode of dialysis, serum creatinine, urea reduction rate, KT/V, iPTH, aspartate aminotransferase, alanine aminotransferase, albumin, MPO, MMP-2, and SDF-1), the independent factors predicting mortality in the HD patients were age (HR = 1.03, 95% CI = 1.005–1.056, p = 0.02), DM (HR = 2.395, 95% CI = 1.217–4.714, p = 0.012), albumin (HR = 0.475, 95% CI = 0.229–0.987, p = 0.047) and MMP-2 (HR = 1.003, 95% CI = 1.001–1.005, p = 0.005) (Table 3). 3.3. ROC and Kaplan–Meier curve analysis to predict mortality in the HD patients ROC curve analysis showed that albumin (AUC = 0.628, p = 0.027) and MMP-2 (AUC = 0.643, p = 0.004) could predict mortality in the HD patients, and that both had a similar ability (p = 0.76) to predict mortality in these patients (Fig. 1). The cut-off value of MMP-2 was ≥ 492.9 ng/ml (sensitivity = 68.9%, 95% CI = 51.9–81.9; specificity = 56.6%, 95% CI = 46.2–66.5) and albumin was ≤3.7 mg/dl (sensitivity = 40%, 95% CI = 24.9–57.7; specificity = 93.9%, 95% CI = 87.7–97.7). Fig. 2 was Kaplan–Meier analysis curve in HD patients at risk for allcause mortality. Patients with MMP-2 ≥ 492.9 ng/ml had a significant lower survival rate during 50 months of follow-up period, compared with those with MMP-2 b 492.9 ng/ml (p = 0.006). 4. Discussion The result of this study show that the HD patients who died were older, had a higher frequency of DM, a lower albumin concentration and a higher MMP-2 concentration. To the best of our knowledge, this is the first study to demonstrate that MMP-2 is a significant predictive marker for mortality in HD patients. Cardiovascular disease is the leading cause of morbidity and mortality in patients with CKD [26], including enhanced arterial calcification and activation of fibroblasts and cytokines, which eventually leads to vascular extracellular matrix remodeling [27,28]. MMP abnormalities have been reported to be involved in the vascular changes associated with kidney failure [17], and to be a mechanism for CVD complications in dialysis patients [10,29,30]. Interestingly, experimental evidence has shown that early upregulation of MMP-2 in areas of elastin degradation and smooth muscle cells change the phenotype during the course of CKD, and that this is associated with increased circulating MMP-2 concentrations [17,27,28]. These alterations have been reported to promote vascular medial layer calcification [17], and increases in MMP-2 concentrations have been found to correlate positively with vascular stiffness and phosphate concentration in CKD patients [31]. The clinical utility of MMP-2 in HD patients had been discussed previously. First, MMP-2 genes expressions have been associated with atherosclerotic CVD in patients with uremia [10,12,13]. Krystyna et al. reported significantly increased serum MMP-2 concentrations in patients with CVD, and that this was associated with the prevalence of
Table 3 Risk of mortality in the HD patients using stepwise multivariate Cox proportional hazards regression analysis.
Age Diabetes mellitus Albumin MMP-2
aHR
95% C.I.
p value
1.03 2.395 0.475 1.003
1.005–1.056 1.217–4.714 0.229–0.987 1.001–1.005
0.02⁎ 0.012⁎ 0.047⁎ 0.005⁎
MMP-2, matrix metalloproteinase-2. (Adjusted for age, diabetes mellitus, gender, mode of dialysis, serum creatinine, urea reduction rate, KT/V, intact parathyroid hormone, aspartate aminotransferase, alanine aminotransferase, albumin, myeloperoxidase, MMP-2, stromal cell-derived factor 1). ⁎ p b 0.05 indicates significance.
164
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
Fig. 1. Predictive accuracy of albumin and MMP-2 was estimated by using time-dependent ROC curve analysis. ROC curve analysis showed that albumin (AUC = 0.628, 95% CI: 0.542–0.709, p = 0.027) and MMP-2 (AUC = 0.643, 95% CI: 0.557–0.722, p = 0.004) could predict mortality in the HD patients, and that both have a similar ability (p = 0.76) for predicting mortality in the dialysis patients. The cut-off value of MMP-2 was ≥492.9 ng/ml (sensitivity = 68.89%, specificity = 56.57%) and albumin was ≤3.7 mg/dl (sensitivity = 40%, specificity = 93.94%). p b 0.05 indicates significance.
atherosclerosis and CVD [10]. Pasterkamp et al. reported elevated expressions of MMP-2 within plaques, and that this played a part in plaque rupture by weakening it [32]. In an animal study, MMP-2 was found to predispose the extracellular matrix to degradation and facilitate arterial calcification in rats with CKD [17]. Second, MMP-2 may be a clinically useful indicator of an inflammatory process in HD patients [33]. The initiation of HD therapy is associated with increased oxidative status, which may not only explain the relatively higher concentrations of serum MMP-2 [34], but also suggests that serum MMP-2 is associated with vascular calcification and CVD complications in dialysis patients [29,30]. Krystyna et al. revealed a possible association between elevated
oxidative stress and the MMP-2 in HD patients, which could represent one of the mechanisms involved in the accelerated atherosclerosis in this population [10]. Third, elastin degradation including MMP-2, was associated progressive aortic stiffening and all-cause mortality in predialysis CKD [35]. However, the predictive mortality role of MMP-2 in dialysis patients was still controversial. Marta et al. revealed that MMP-2 was not an independent mortality predictor in long-term HD patients (HR = 1.078, 95% CI = 0.921–1.262, p = 0.4) [36]. The MMP2 concentration in this study was 381 ± 123 ng/dl and was no significant correlation with all-cause mortality. In our patients, the concentration of serum MMP-2 was higher in the patients who died and it was an independent predictor for 5-year mortality. The cut-off value of the MMP-2 was ≥ 492.9 ng/ml, and the sensitivity and specificity for predicting mortality were 68.3% and 56.6% respectively. The MMP-2 has reliable sensitivity but limited specificity for predicting survival of HD patients. Further investigations are needed to determine a more accurate cut-off value and the clinical relevance of the MMP-2. Serum albumin concentration is a well-known and powerful predictive factor of mortality in HD patients [1,2]. The association between low serum albumin and higher mortality may be explained by concurrent inflammatory processes [37] or due to a poor nutritional status [2,6]. There were several studies indicated that serum albumin was a predictive marker in dialysis patients. Hirokazu et al. revealed serum albumin was a predictor of malnutrition, CVD, and mortality in patients with end stage renal disease [38]. The AUC value of albumin was 0.72 for total mortality and cut-off value was b 3.3 mg/dl with 26 months follow up. Another 5 y cohort study revealed that serum albumin correlated with all-cause mortality in HD patients (AUC value of albumin was 0.65) [39]. In our study, the AUC and cut-off value of albumin was 0.62 and ≤3.7 mg/dl respectively. This result was similar to the previous literature to demonstrate that serum albumin was powerful associated with mortality in patients undergoing dialysis. Based on our ROC curve analysis, MMP-2 was comparable to albumin in its capacity to predict mortality in HD patients. In other word, MMP-2 could be a significant predictive marker for mortality in HD patients. Matrix metalloproteinase-9 is mainly produced by blood monocytes, and chronic stimulation of these cells during extracorporeal circulation during HD sessions may be responsible for inducting its expression [40]. Progression of CKD has been shown to be related to a decrease in MMP-9 [14]. In addition, Rysz et al. reported a decreased concentration of MMP-9 during HD [41]. In the animal study, increased MMP-9 could
Fig. 2. Kaplan–Meier analysis curve in hemodialysis patients at risk for all-cause mortality. Patients with MMP-2 ≥ 492.9 ng/ml had a significant lower survival rate during the 50 months follow-up period, compared with those with MMP-2 b 492.9 ng/ml (p = 0.006).
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
lead to altered extracellular matrix and, altered arterial structure and cause arterial calcification in uremic rats [17]. However, Krystyna et al. reported no significant correlation between MMP-9 and the prevalence of CVD in HD patients [10]. Another prospective observational cohort study which included 261 long-term HD patients followed up for 5 y, MMP-9 was not a significant independent predictor of overall mortality in patients undergoing dialysis (HR = 0.953, 95% CI = 0.804–1.131, p = 0.6) [36]. This is consistent with our results in which there was no correlation between mortality and serum MMP-9 concentration in the dialysis patients. The process of HD can trigger inflammation as a result of exposing the blood to a bio-incompatible system which stimulates monocytes and macrophages, thereby inducing the production of proinflammatory cytokines and contributing to the maintenance of oxidative stress [42]. The production of interleukins leading to a release of MPO has been implicated in the development of CVD in dialysis patients [43]. In the previous study, increased concentrations of MPO have been closely linked to mortality in dialysis patients [44]. Kamyar et al. reported a total 356 HD patients with a 3 y cohort study, the median serum MPO concentration was 393 pg/ml with relative risk 1.14 (p = 0.01) for all-cause mortality. Each 272 pg/ml increase in serum MPO concentration, there was 8% to 15% increase in risk for death during the observed 3-y interval according to the concentration of multivariate adjustment [44]. However, the MPO concentration was not significantly increased (73.33 pg/ml) and that could be possibility no associated with mortality in our results. A soluble decoy receptor of the osteoclast activator, osteopontin, has been reported to act as an important regulatory molecule in vascular calcification independent of classical risk factors [22]. Increased osteopontin concentrations have also been associated with increased vascular stiffness [45], and higher osteopontin concentrations may be independent predictors of all-cause mortality in diabetic nephropathy patients [46]. However, the predictive mortality role of osteopontin in dialysis patients was still controversial. Serum osteopontin was not associated with cardiovascular events in a prospective observational study with 97 dialysis patients over 5-y observation [47]. Another 3-y cohort study with 602 dialysis patients, Scialla et al. reported osteopontin was not a significant risk factor for all-cause and cardiovascular mortality in patients undergoing dialysis [22]. Our study also revealed no association with osteopontin and mortality in HD patients. Stromal derived factor-1, also called CXC motif ligand 12 (CXCL12), is a complex chemokine with putative anti-inflammatory and antiatherogenic functions that has also been reported to play a role in recruiting endothelial cells in response to vascular injury [23]. Mehta et al. reported that in CKD patients, higher concentrations of plasma SDF-1 were associated with known cardiovascular risk factors and prevalent CVD, and also predicted incident myocardial infarction and mortality even after adjusting for traditional risk factors and kidney function [24]. However, there are currently no reports demonstrating an association between SDF-1 and mortality in dialysis patients, and SDF-1 was not a predictive factor for mortality in the current study. There are some limitations to this study. First, the number of patients was small. Second, since the concentrations of the serum biomarkers were measured only once. Third, this study was only a single-center nature. Large prospective studies are needed to confirm whether these biomarkers, and especially MMP-2, are useful for predicting mortality in HD patients. In conclusion, this study shows that concentrations of serum MMP-2 and albumin are both important predictive factors of mortality in HD patients, and that compared with albumin, serum MMP-2 is a noninferior marker for predicting survival of HD patients. Acknowledgments This study was supported by a grant from the National Science Council 101-2314-B-040-003. The authors wish to thank Mr. Nigel Daly for editing and reviewing the manuscript for English language.
165
References [1] A.A. Lopes, J.L. Bragg-Gresham, S.J. Elder, et al., Independent and joint associations of nutritional status indicators with mortality risk among chronic hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS), J. Ren. Nutr. 20 (2010) 224–234. [2] C. Combe, K.P. McCullough, Y. Asano, N. Ginsberg, B.J. Maroni, T.B. Pifer, Kidney Disease Outcomes Quality Initiative (K/DOQI) and the Dialysis Outcomes and Practice Patterns Study (DOPPS): nutrition guidelines, indicators, and practices, Am. J. Kidney Dis. 44 (2004) 39–46. [3] B.D. Bradbury, R.B. Fissell, J.M. Albert, et al., Predictors of early mortality among incident US hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS), Clin. J. Am. Soc. Nephrol. 2 (2007) 89–99. [4] A. Kato, T. Takita, M. Furuhashi, Y. Maruyama, A. Hishida, Comparison of serum albumin, C-reactive protein and carotid atherosclerosis as predictors of 10-year mortality in hemodialysis patients, Hemodial. Int. 14 (2010) 226–232. [5] M. Herselman, N. Esau, J.M. Kruger, D. Labadarios, M.R. Moosa, Relationship between serum protein and mortality in adults on long-term hemodialysis: exhaustive review and meta-analysis, Nutrition 26 (2010) 10–32. [6] B.A. Cooper, E.L. Penne, L.H. Bartlett, C.A. Pollock, Protein malnutrition and hypoalbuminemia as predictors of vascular events and mortality in ESRD, Am. J. Kidney Dis. 43 (2004) 61–66. [7] V. Menon, A. Gul, M.J. Sarnak, Cardiovascular risk factors in chronic kidney disease, Kidney Int. 68 (2005) 1413–1418. [8] R.N. Foley, P.S. Parfrey, J.D. Harnett, et al., Clinical and echocardiographic disease in patients starting end-stage renal disease therapy, Kidney Int. 47 (1995) 186–192. [9] R.J. Bisoendial, R.S. Birjmohun, F. Akdim, et al., C-reactive protein elicits white blood cell activation in humans, Am. J. Med. 122 (2009) 581–589. [10] K. Pawlak, D. Pawlak, M. Mysliwiec, Serum matrix metalloproteinase-2 and increased oxidative stress are associated with carotid atherosclerosis in hemodialyzed patients, Atherosclerosis 190 (2007) 199–204. [11] A. Page-McCaw, A.J. Ewald, Z. Werb, Matrix metalloproteinases and the regulation of tissue remodelling, Nat. Rev. Mol. Cell Biol. 8 (2007) 221–233. [12] M. Peiskerova, M. Kalousova, M. Kratochvilova, et al., Fibroblast growth factor 23 and matrix-metalloproteinases in patients with chronic kidney disease: are they associated with cardiovascular disease? Kidney Blood Press. Res. 32 (2009) 276–283. [13] M. Nagano, K. Fukami, S. Yamagishi, et al., Circulating matrix metalloproteinase-2 is an independent correlate of proteinuria in patients with chronic kidney disease, Am. J. Nephrol. 29 (2009) 109–115. [14] H.R. Chang, S.F. Yang, M.L. Li, C.C. Lin, Y.S. Hsieh, J.D. Lian, Relationships between circulating matrix metalloproteinase-2 and -9 and renal function in patients with chronic kidney disease, Clin. Chim. Acta 366 (2006) 243–248. [15] K. Pawlak, J. Tankiewicz, M. Mysliwiec, D. Pawlak, Systemic levels of MMP2/TIMP2 and cardiovascular risk in CAPD patients, Nephron Clin. Pract. 115 (2010) 251–258. [16] C. Bouvet, S. Moreau, J. Blanchette, D. de Blois, P. Moreau, Sequential activation of matrix metalloproteinase 9 and transforming growth factor beta in arterial elastocalcinosis, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 856–862. [17] N.X. Chen, K.D. O'Neill, X. Chen, K. Kiattisunthorn, V.H. Gattone, S.M. Moe, Activation of arterial matrix metalloproteinases leads to vascular calcification in chronic kidney disease, Am. J. Nephrol. 34 (2011) 211–219. [18] T. Naruko, M. Ueda, K. Haze, et al., Neutrophil infiltration of culprit lesions in acute coronary syndromes, Circulation 106 (2002) 2894–2900. [19] M.L. Brennan, M.S. Penn, F. Van Lente, et al., Prognostic value of myeloperoxidase in patients with chest pain, N. Engl. J. Med. 349 (2003) 1595–1604. [20] S. Baldus, C. Heeschen, T. Meinertz, et al., Myeloperoxidase serum levels predict risk in patients with acute coronary syndromes, Circulation 108 (2003) 1440–1445. [21] M. Scatena, L. Liaw, C.M. Giachelli, Osteopontin: a multifunctional molecule regulating chronic inflammation and vascular disease, Arterioscler. Thromb. Vasc. Biol. 27 (2007) 2302–2309. [22] J.J. Scialla, W.H. Kao, C. Crainiceanu, et al., Biomarkers of vascular calcification and mortality in patients with ESRD, Clin. J. Am. Soc. Nephrol. 9 (2014) 745–755. [23] S. Abi-Younes, A. Sauty, F. Mach, G.K. Sukhova, P. Libby, A.D. Luster, The stromal cellderived factor-1 chemokine is a potent platelet agonist highly expressed in atherosclerotic plaques, Circ. Res. 86 (2000) 131–138. [24] N.N. Mehta, G.J. Matthews, P. Krishnamoorthy, et al., Higher plasma CXCL12 levels predict incident myocardial infarction and death in chronic kidney disease: findings from the Chronic Renal Insufficiency Cohort study, Eur. Heart J. 35 (2014) 2115–2122. [25] D.E. Kleiner, W.G. Stetler-Stevenson, Quantitative zymography: detection of picogram quantities of gelatinases, Anal. Biochem. 218 (1994) 325–329. [26] S. Ardhanari, M.A. Alpert, K. Aggarwal, Cardiovascular disease in chronic kidney disease: risk factors, pathogenesis, and prevention, Adv. Perit. Dial. 30 (2014) 40–53. [27] A. Pai, E.M. Leaf, M. El-Abbadi, C.M. Giachelli, Elastin degradation and vascular smooth muscle cell phenotype change precede cell loss and arterial medial calcification in a uremic mouse model of chronic kidney disease, Am. J. Pathol. 178 (2011) 764–773. [28] C. Kumata, M. Mizobuchi, H. Ogata, et al., Involvement of matrix metalloproteinase2 in the development of medial layer vascular calcification in uremic rats, Ther. Apher. Dial. 15 (2011) 18–22. [29] K. Pawlak, M. Mysliwiec, D. Pawlak, Peripheral blood level alterations of MMP-2 and MMP-9 in patients with chronic kidney disease on conservative treatment and on hemodialysis, Clin. Biochem. 44 (2011) 838–843. [30] R.S. Friese, F. Rao, S. Khandrika, et al., Matrix metalloproteinases: discrete elevations in essential hypertension and hypertensive end-stage renal disease, Clin. Exp. Hypertens. 31 (2009) 521–533.
166
K.-C. Hsiao et al. / Clinica Chimica Acta 452 (2016) 161–166
[31] A.W. Chung, H.H. Yang, M.K. Sigrist, et al., Matrix metalloproteinase-2 and -9 exacerbate arterial stiffening and angiogenesis in diabetes and chronic kidney disease, Cardiovasc. Res. 84 (2009) 494–504. [32] G. Pasterkamp, A.H. Schoneveld, D.J. Hijnen, et al., Atherosclerotic arterial remodeling and the localization of macrophages and matrix metalloproteases 1, 2 and 9 in the human coronary artery, Atherosclerosis 150 (2000) 245–253. [33] G.A. Preston, C.V. Barrett, D.A. Alcorta, et al., Serum matrix metalloproteinases MMP2 and MMP-3 levels in dialysis patients vary independently of CRP and IL-6 levels, Nephron 92 (2002) 817–823. [34] K. Pawlak, D. Pawlak, M. Mysliwiec, Impaired renal function and duration of dialysis therapy are associated with oxidative stress and proatherogenic cytokine levels in patients with end-stage renal disease, Clin. Biochem. 40 (2007) 81–85. [35] E.R. Smith, L.A. Tomlinson, M.L. Ford, L.P. McMahon, C. Rajkumar, S.G. Holt, Elastin degradation is associated with progressive aortic stiffening and all-cause mortality in predialysis chronic kidney disease, Hypertension 59 (2012) 973–978. [36] M. Kalousova, H. Benakova, A.A. Kubena, S. Dusilova-Sulkova, V. Tesar, T. Zima, Pregnancy-associated plasma protein A as an independent mortality predictor in long-term hemodialysis patients, Kidney Blood Press. Res. 35 (2012) 192–201. [37] R. de Mutsert, D.C. Grootendorst, F. Indemans, E.W. Boeschoten, R.T. Krediet, F.W. Dekker, Association between serum albumin and mortality in dialysis patients is partly explained by inflammation, and not by malnutrition, J. Ren. Nutr. 19 (2009) 127–135. [38] H. Honda, A.R. Qureshi, O. Heimburger, et al., Serum albumin, C-reactive protein, interleukin 6, and fetuin a as predictors of malnutrition, cardiovascular disease, and mortality in patients with ESRD, Am. J. Kidney Dis. 47 (2006) 139–148.
[39] S.C. Hung, T.W. Hsu, Y.P. Lin, D.C. Tarng, Decoy receptor 3, a novel inflammatory marker, and mortality in hemodialysis patients, Clin. J. Am. Soc. Nephrol. 7 (2012) 1257–1265. [40] A. Kalela, M. Ponnio, T.A. Koivu, et al., Association of serum sialic acid and MMP-9 with lipids and inflammatory markers, Eur. J. Clin. Investig. 30 (2000) 99–104. [41] J. Rysz, M. Banach, R.A. Stolarek, et al., Serum metalloproteinases MMP-2, MMP-9 and metalloproteinase tissue inhibitors TIMP-1 and TIMP-2 in patients on hemodialysis, Int. Urol. Nephrol. 43 (2011) 491–498. [42] C. Libetta, V. Sepe, P. Esposito, F. Galli, C.A. Dal, Oxidative stress and inflammation: implications in uremia and hemodialysis, Clin. Biochem. 44 (2011) 1189–1198. [43] G.A. Kaysen, N.W. Levin, W.E. Mitch, A.L. Chapman, L. Kubala, J.P. Eiserich, Evidence that C-reactive protein or IL-6 are not surrogates for all inflammatory cardiovascular risk factors in hemodialysis patients, Blood Purif. 24 (2006) 508–516. [44] K. Kalantar-Zadeh, M.L. Brennan, S.L. Hazen, Serum myeloperoxidase and mortality in maintenance hemodialysis patients, Am. J. Kidney Dis. 48 (2006) 59–68. [45] D.V. Barreto, F.C. Barreto, A.B. Carvalho, et al., Coronary calcification in hemodialysis patients: the contribution of traditional and uremia-related risk factors, Kidney Int. 67 (2005) 1576–1582. [46] D. Gordin, C. Forsblom, N.M. Panduru, et al., Osteopontin is a strong predictor of incipient diabetic nephropathy, cardiovascular disease, and all-cause mortality in patients with type 1 diabetes, Diabetes Care 37 (2014) 2593–2600. [47] J.E. Lee, H.J. Kim, S.J. Moon, et al., Serum osteoprotegerin is associated with vascular stiffness and the onset of new cardiovascular events in hemodialysis patients, Korean J. Intern. Med. 28 (2013) 668–677.
