Soluble CD146 and B-type natriuretic peptide dissect overhydration into functional components of prognostic relevance in haemodialysis patients

Soluble CD146 and B-type natriuretic peptide dissect overhydration into functional components of... Abstract Background Accurate volume status evaluation and differentiation of cardiac and non-cardiac components of overhydration (OH) are fundaments of optimal haemodialysis (HD) management. Methods This study, by combining bioimpedance measurements, cardiovascular biomarkers and echocardiography, aimed at dissecting OH into its major functional components, and prospectively tested the association between cardiac and non-cardiac components of OH with mortality. In the first part, we validated soluble CD146 (sCD146) as a non-cardiac biomarker of systemic congestion in a cohort of 30 HD patients. In the second part, we performed a prospective 1-year follow-up study in an independent cohort of 144 HD patients. Results sCD146 incrementally increased after the short and long intervals after HD (+53 ng/mL, P = 0.006 and  +91 ng/mL, P < 0.001), correlated with OH as determined by bioimpedance and well-diagnosed OH (area under the receiver operating characteristics curve 0.72, P = 0.005). The prevalence of OH was lower for low-sCD146 and low-BNP patients (B-type natriuretic peptide, 29%) compared with subjects with either one or both biomarkers elevated (65–74%, P < 0.001). Notably, most low-BNP but high-sCD146 subjects were overhydrated. Systolic dysfunction was 2- to 3-fold more prevalent among high-BNP compared with low-BNP patients (44–68% versus 21–23%, chi-square P < 0.001), regardless of sCD146. One-year all-cause mortality was markedly higher in patients with high-BNP (P = 0.001) but not with high-sCD146. In multivariate analysis, systolic dysfunction and BNP, but not OH, were associated with lower survival. Conclusions The combination of BNP and sCD146 dissects OH into functional components of prognostic value. OH in HD patients is associated with higher mortality only if resulting from cardiac dysfunction. BNP, haemodialysis, heart failure, overhydration, sCD146 INTRODUCTION Volume management is a critical element in the care of haemodialysis (HD) patients and requires accurate assessment of both the hydration state and the underlying pathophysiology. Overhydration (OH), a state of generalized oedema formation, is a frequent condition in HD patients and is associated with reduced survival [1–6]. In patients with end-stage renal disease (ESRD), OH is related either to renal salt and water retention and insufficient ultrafiltration during HD or concomitant congestive heart failure (CHF). Indeed, increased intracardiac and upstream vascular hydrostatic pressures (systemic congestion) may contribute to oedema formation in patients with ESRD and CHF [7, 8]. The dissection of OH into cardiac and non-cardiac components in the individual HD patient is challenging because ESRD and CHF often coexist and because rapid and reliable diagnostic tools are still lacking. However, the differentiation of the underlying mechanisms is of significant clinical relevance since it might implicate different treatments. Bioimpedance analysis has become a popular tool in the assessment of OH in HD patients [1]. Body composition measurement (BCM) by bioimpedance not only allows to quantify malnutrition and sarcopenia, but also to provide a quantitative estimation of fluid compartments. This general estimate of the volume status supports the clinician in HD volume management but does not distinguish the underlying mechanisms of OH. Cardiovascular biomarkers might provide complementary diagnostic information for this purpose. Natriuretic peptides—B-type natriuretic peptide (BNP) and the amino-terminal pro-BNP (NT-proBNP)—are released by cardiomyocytes in the presence of increased intracardiac pressures. Previous studies showed discordant results regarding the ability of natriuretic peptides to detect OH and guide volume management [9–15], but consistently displayed excellent negative predictive values for CHF, even in HD patients [16, 17]. Thus, natriuretic peptides are considered a biomarker of cardiac dysfunction rather than OH in HD patients. Notably, elevated plasma concentrations of natriuretic peptides correlated well with mortality in HD patients [15, 17–23]. Our group recently described soluble CD146 (sCD146) as a novel vascular biomarker, released by vessels in response to endothelial stretch [24–26]. The extra-cardiac origin of sCD146 provides complementary information to BNP, and its mode of release into circulation makes it a more promising direct marker of systemic congestion, regardless of cardiac function. In this study, after validating sCD146 as a non-cardiac biomarker of systemic congestion in HD patients, by measuring BNP and sCD146 we dissected OH into its major functional components and prospectively tested the association between cardiac and non-cardiac components of OH with mortality in HD patients. MATERIALS AND METHODS Study population The study population consisted of 174 clinically stable chronic outpatients on the maintenance HD. The first cohort of patients (‘Zurich’, n = 30) underwent a mechanistic study to validate sCD146 as a biomarker of systemic congestion in HD. This part of the study was performed from 1 October to 31 December 2016, at the University Hospital of Zurich, Switzerland. Inclusion criteria were age >18 years and HD since at least 1 month; exclusion criteria were acute illness or hospitalization in the last week before inclusion, and implanted cardiac device, as it precludes reliable bioelectrical impedance measurement. The second cohort of patients (‘London’, n = 144) was used for external validation of the ‘Zurich’ study and for the clinical outcome study, aimed at investigating the relationship between OH, cardiac dysfunction and outcome in stable HD patients. This part of the study was performed at the Royal Free Hospital of London, UK. Patients attending for their routine outpatient midweek HD session on 5 May and 6 May 2016 were included and followed up until the end of May 2017. Mechanistic ‘Zurich’ study The mechanistic ‘Zurich’ study consisted of four visits. The first visit was performed after a regular HD session, followed by a short interdialysis interval (duration: 2 days). The second visit was performed at the end of the short interval, before the HD session, and the third visit after the same HD session. The fourth visit was performed at the end of the subsequent prolonged interdialysis interval (‘long interval’, duration: 3 days). Briefly, the four visits allowed for study of a short and a long inter-dialysis interval to test different levels of OH. Total body weight was measured at each visit, to calculate the net body weight increase during the short and the long intervals, respectively. Furthermore, OH was quantitatively assessed at each study visit with a portable whole-body bioimpedance spectroscopy device (Fresenius Medical Care GmbH, Bad Homburg, Germany). BCM was performed with the patient in a supine position according to the manufacturer’s description. Based on a fluid assessment model using resistances to electrical currents of 50 discrete frequencies, the total body water (TBW), the extracellular water (ECW), the intracellular water (ICW) and the OH are calculated [27, 28]. Relevant OH was previously defined as an OH/ECW ratio >0.15 [1, 2]. Patient history and clinical evaluation were collected at each study visit. Comorbidities and HD-specific information were retrieved from electronic medical charts. Prospective clinical ‘London’ outcome study The clinical ‘London’ outcome study consisted of one visit, before a regular outpatient midweek HD session, followed by a prospective follow-up of 1 year for all-cause mortality. The quantitative assessment of OH was performed before the HD session using multifrequency bioimpedance (InBody 720, Seoul, South Korea), in a standardized protocol [29]. The TBW, ECW and ICW were calculated. Since the OH value was not determined in this cohort, relevant OH was defined in the presence of an ECW/TBW ratio >0.4 according to the literature [1]. All patients underwent transthoracic cardiac echocardiography for determination of the left-ventricular ejection fraction on a non-dialysis day. Cardiac systolic dysfunction was defined as left-ventricular ejection fraction <50%, according to the guidelines of the European Society of Cardiology [30]. Clinical data were retrieved from electronic medical charts. Biomarker testing Venous blood samples for measurement of cardiovascular biomarkers were drawn at each visit, centrifuged within 6 h and ethylenediaminetetraacetic acid (EDTA) plasma aliquots were stored at −80°C until analysed at Lariboisière University Hospital, Paris, France. The concentration of sCD146 was determined by ELISA (CY-QUANT ELISA sCD146©, Biocytex, France) with a detection limit of sCD146 of 10 ng/mL and coefficients of variation for both repeatability and reproducibility < 20% in the measured range. Measurement of concentration of B-type natriuretic peptide (BNP) was performed with the Architect i2000 platform (Abbott Diagnostics, Abbott Park, IL, USA). Ethical considerations The study was conducted according to the standards of the Declaration of Helsinki and approved by the local ethics committee of Zurich (2016-451) and London (13/Lo/0912). All patients provided written informed consent (ClinicalTrials.gov NCT02962635). Statistical analysis Continuous variables are expressed as median [interquartile range (IQR)], categorical variables as number (percentage). Independent samples were compared using the Mann–Whitney U test or chi-square test, as appropriate. Related samples were compared using the Wilcoxon signed rank test. Correlation analysis was performed using Spearman’s correlation coefficient (rho). The diagnostic performance of the biomarkers was assessed by receiver operating characteristic (ROC) analysis and expressed as area under the curve (AUC). Survival was plotted with the Kaplan–Meier curve and differences between groups were assessed by the log-rank test. Univariate and covariate-adjusted Cox proportional hazards regression models were used to estimate the association between circulating biomarkers and risk of death. The covariates included in the regression models were a priori selected among baseline variables with known associations with study outcomes: age, sex, cardiac systolic dysfunction, relevant OH, anuria, log-transformed BNP and log-transformed sCD146. The null hypothesis was rejected with an adjusted two-sided P-value <0.05. All analyses were performed with the use of IBM SPSS Statistics, Version 25.0 (IBM Corp, Armonk, NY, USA). RESULTS Characterization of the study population The study population consisted of 174 HD patients, predominantly middle-aged men with ESRD (Table 1). Dialysis access was mostly an arteriovenous fistula, and anuria was highly prevalent. Cardiovascular risk factors and vascular complications (coronary and peripheral artery disease) were highly prevalent. Table 1 Patient characteristics of the study population ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 Data are presented as median [quartiles] or number (percentage) as appropriate. Table 1 Patient characteristics of the study population ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 Data are presented as median [quartiles] or number (percentage) as appropriate. Mechanistic ‘Zurich’ study on vascular sCD146 as a biomarker of systemic congestion The median value of circulating sCD146 in all studied patients at all time points (120 measurements) was 530 ng/mL (range: 123–1730 ng/mL). The median intraindividual variability was 43% (range: 6–354%, Supplementary data, Figure S1). Dialysis did not alter circulating sCD146, as shown by similar sCD146 levels before and after HD [543 (384; 625) versus 492 ng/mL (393; 646), P = 0.52, Supplementary data, Figure S2]. Moreover, we measured changes in body weight and sCD146 concentrations after a short (2 days) and a long (3 days) interdialysis interval. We observed an incremental increase in both body weight [+0.9 kg (+0.3; +1.8), P < 0.001 and  + 1.8 kg (+0.1; +2.7), P < 0.001] and circulating sCD146 [+53 ng/mL (−16; +118), P = 0.006 and  +91 ng/mL (+20; +179), P < 0.001] after the short and long intervals, respectively (Figure 1). FIGURE 1 View largeDownload slide Median changes in body weight and sCD146 during the short and the long interval in the ‘Zurich’ cohort. Median values and quartiles are reported. P-values refer to related samples of Wilcoxon signed-rank test. N = 30. FIGURE 1 View largeDownload slide Median changes in body weight and sCD146 during the short and the long interval in the ‘Zurich’ cohort. Median values and quartiles are reported. P-values refer to related samples of Wilcoxon signed-rank test. N = 30. Notably, while anuric patients (n = 14) had higher baseline concentration of sCD146 compared with non-anuric patients [564 (463; 766) versus 435 (335; 499) ng/mL, P = 0.022], the median increase in circulating sCD146 during the short interval was more pronounced in non-anuric patients [median: +89 (+5; +124) ng/mL, P = 0.004] compared with anuric patients [median: +30 (−91; +88) ng/mL, P = 0.47]. Similar results were found during the long interval (data not shown). Furthermore, we assessed the correlation between sCD146 and OH determined by BCM at the end of both short and long intervals (n = 60; Figure 2). Circulating sCD146 positively correlated with OH (Spearman’s ρ 0.37, P = 0.004). Circulating sCD146 showed good properties for diagnosing patients with relevant OH [area under the receiver operating characteristics curve, AUROC: 0.72 (0.58; 0.86), P = 0.005]. FIGURE 2 View largeDownload slide Correlation between OH and circulating sCD146 after the short and long interval (A) and ROC curve of sCD146 (B) for diagnosing relevant OH (OH/ECV >0.15) in the ‘Zurich’ cohort. Dotted lines refer to the definition of relevant OH (OH/ECW >0.15). N = 60. FIGURE 2 View largeDownload slide Correlation between OH and circulating sCD146 after the short and long interval (A) and ROC curve of sCD146 (B) for diagnosing relevant OH (OH/ECV >0.15) in the ‘Zurich’ cohort. Dotted lines refer to the definition of relevant OH (OH/ECW >0.15). N = 60. Finally, we evaluated the diagnostic properties of circulating sCD146 to detect dyspnoea, a clinical parameter of systemic congestion. At the end of both short and long intervals (n = 60), 50% of patients reported dyspnoea, defined as breathlessness at rest or during stress (New York Heart Association class II–IV). Dyspneic patients displayed higher plasma concentrations of sCD146 [633 ng/mL (541; 770 ng/mL)] compared with patients without dyspnoea [523 ng/mL (364; 622 ng/mL), P = 0.006] (Figure 3A). Notably, BNP levels did not discriminate dyspneic and non-dyspneic patients [332 (191; 697) versus 147 pg/mL (63; 639), P = 0.08]. The AUROC of sCD146 for diagnosing dyspnoea was 0.71 (0.58; 0.84), P = 0.006 (Figure 3B). Plasma concentrations of sCD146 <461 ng/mL had a sensitivity  >90% of not being associated with dyspnoea. FIGURE 3 View largeDownload slide Circulating sCD146 according to the symptoms of dyspnoea (A) and ROC curve of sCD146 for diagnosing dyspnoea in the ‘Zurich’ cohort. (B) N = 60. CI, confidence interval. FIGURE 3 View largeDownload slide Circulating sCD146 according to the symptoms of dyspnoea (A) and ROC curve of sCD146 for diagnosing dyspnoea in the ‘Zurich’ cohort. (B) N = 60. CI, confidence interval. Validation of the ‘Zurich’ study in the ‘London’ cohort The data obtained in the ‘Zurich’ cohort were validated in the larger independent ‘London’ cohort of 144 stable HD outpatients. We confirmed that subjects with relevant OH (n = 84) had higher levels of sCD146 [598 (454; 744) versus 479 ng/mL (416; 597), P = 0.003] compared with non-OH subjects (n = 60). Circulating sCD146 levels positively correlated with OH (Spearman’s ρ 0.28, P = 0.001). Cardiac and non-cardiac components of OH In a further analysis in the ‘London’ cohort (n = 144), we found that the prevalence of relevant OH was much lower for patients with both low sCD146 and low BNP (29%) compared with patients with either an increase in one or both biomarkers (65–74%, chi-square P < 0.001) (Figure 4A). Notably, in the group of subjects with low-BNP but high-sCD146 (upper left quadrant), nearly three of four showed relevant OH, indicating that OH and systemic congestion correlated in most patients, but were not necessarily related to high BNP. FIGURE 4 View largeDownload slide Distribution of relevant OH (A) and cardiac systolic dysfunction (B) according to the circulating concentrations of sCD146 and BNP (n = 144) in the ‘London’ cohort. Values of BNP and sCD146 were log-transformed. Groups were divided according to the median values of BNP and sCD146 (represented by lines). Percentages refer to the prevalence of relevant OH (A) and cardiac systolic dysfunction (B). N = 144. FIGURE 4 View largeDownload slide Distribution of relevant OH (A) and cardiac systolic dysfunction (B) according to the circulating concentrations of sCD146 and BNP (n = 144) in the ‘London’ cohort. Values of BNP and sCD146 were log-transformed. Groups were divided according to the median values of BNP and sCD146 (represented by lines). Percentages refer to the prevalence of relevant OH (A) and cardiac systolic dysfunction (B). N = 144. Furthermore, patients with cardiac systolic dysfunction (n = 48) had higher BNP levels [1507 (504; 4245) versus 473 pg/mL (259; 1368), P < 0.001] compared with patients with preserved systolic function. However, sCD146 levels were similar in patients with or without cardiac systolic dysfunction [525 (387; 690) versus 559 ng/mL (436; 714), P = 0.47]. The prevalence of systolic dysfunction was ∼2- to 3-fold among patients with elevated BNP (44–68%) compared with patients with low-BNP (21–23%, chi-square P < 0.001), regardless of sCD146 (Figure 4B). These findings indicate that OH and systemic congestion not always reflect CHF in HD patients and that circulating sCD146 is a biomarker of OH/systemic congestion independent of CHF. Prospective clinical ‘London’ outcome study Based on the previous data, we investigated the prognostic impact of OH and its functional components on all-cause mortality in HD patients of the ‘London’ cohort. During a median follow-up of 365 days, a total of 27 deaths occurred. In univariate analysis (Table 2) age, relevant OH, cardiac systolic dysfunction and log-transformed BNP were shown to be associated with all-cause mortality, while sex, anuria and log-transformed sCD146 were not. In multivariate analysis, systolic dysfunction and log-transformed BNP were still strongly associated with all-cause mortality, while age and relevant OH were not. BNP shows excellent properties for predicting all-cause mortality during the follow-up [AUROC 0.81 (0.73; 0.90), P < 0.001], with specificity >90% for values exceeding 3000 pg/mL. Table 2 Univariate and multivariate analysis for prediction of long-term all-cause mortality (‘London’ cohort, n = 144) HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR, hazard ratio; CI, confidence interval. Table 2 Univariate and multivariate analysis for prediction of long-term all-cause mortality (‘London’ cohort, n = 144) HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR, hazard ratio; CI, confidence interval. Figure 5 illustrates the marked difference in survival of HD patients according to the presence of relevant OH (log-rank P = 0.041), systolic dysfunction (log-rank P < 0.001) and high BNP (log-rank P = 0.001) but not high sCD146 levels. FIGURE 5 View largeDownload slide Survival of HD patients according to the presence of relevant OH, cardiac systolic dysfunction and biomarker concentrations in the ‘London’ cohort. Relevant OH was defined as ECW/TBW >0.4. Low/high BNP was defined according to the median (631 pg/mL), low/high sCD146 was defined according to the median (542 ng/mL). N = 144. FIGURE 5 View largeDownload slide Survival of HD patients according to the presence of relevant OH, cardiac systolic dysfunction and biomarker concentrations in the ‘London’ cohort. Relevant OH was defined as ECW/TBW >0.4. Low/high BNP was defined according to the median (631 pg/mL), low/high sCD146 was defined according to the median (542 ng/mL). N = 144. DISCUSSION Optimal volume management in HD patients is of significant clinical relevance and relies on the accurate assessment of OH and its functional components. Indeed, since CHF is highly prevalent in ESRD patients, OH may have both cardiac and non-cardiac components. Our study, after validating the novel vascular biomarker sCD146 as a non-cardiac biomarker of systemic congestion in HD patients, by combining two cardiovascular biomarkers (BNP and sCD146) dissected OH into its major functional components. Furthermore, we prospectively showed that cardiac systolic dysfunction and not systemic congestion per se is associated with high all-cause mortality in HD patients. Vascular sCD146 is released from endothelial cells upon mechanical stress and is considered a biomarker of systemic congestion, independent form cardiac function [24–26, 31]. In the mechanistic part of our study, we confirmed for the first time, in two independent cohorts of stable HD patients (‘Zurich’ and ‘London’), that sCD146 is an accurate non-cardiac biomarker of systemic congestion, as previously shown in CHF [24–26]. Indeed, circulating sCD146 concentrations well correlated with clinical markers such as weight gain, dyspnoea and BCM-derived parameters of OH. Using sCD146 (reflecting systemic congestion, independently from the cardiac function) and combining it with BNP and echocardiographic data (reflecting cardiac function), we dissected OH into its major functional components (cardiac versus non-cardiac). Our data showed that OH (as determined by BCM), systemic congestion and cardiac dysfunction do not necessarily coexist in HD patients. Our data showed that relevant OH might be associated with three biomarker phenotypes: high-BNP/low-sCD146 (indicating structural cardiac impairment), low-BNP/high-sCD146 (indicating OH without cardiac involvement) and high-BNP/high-sCD146 (indicating OH of mixed origin). The application of this novel biomarker-based dissection of functional components of OH led us to the second main finding of this study: the high all-cause mortality associated with OH in HD patients is strongly related to cardiac systolic dysfunction. In light of these data, the previously reported increased mortality related to OH is probably indicative of a higher prevalence of prognostically relevant CHF among patients with BCM-determined OH, and not to systemic congestion per se. Clinical significance Our findings need to be validated in prospective larger studies but might contribute to a change in the management of HD patients according to the biomarker profile. High-BNP patients have a high likelihood of CHF and display high mortality independently from sCD146 concentration. High-BNP/low-sCD146 patients might require extensive cardiovascular evaluation and specific CHF therapy. High-BNP/high-sCD146 patients likely display OH of mixed origin and might benefit from a combination of specific CHF-therapy and intensive volume management (e.g. ultrafiltration) [32]. Low-BNP patients display good survival independently from sCD146 values. In these patients, sCD146 might help to guide volume management, in particular in the presence of symptoms such as dyspnoea, but in most patients, a less restrictive volume management to avoid hypotension, cardiac stunning and anuria might be appropriate. Further studies correlating sCD146 levels with blood pressure and the incidence of cardiovascular events will provide additional data of interest for an optimal volume management. Furthermore, the application of a biomarker approach might lead to a more precise cardiovascular stratification of patients to be included in interventional studies aiming at the improvement of cardiovascular mortality on HD. Limitations This study has several limitations. (i) Bioimpedance measurement cannot distinguish intravascular ECW from extravascular ECW and changes in overall ECW may not be equally distributed in the intravascular and extravascular components. Despite this significant limitation, bioimpedance measurements are still routinely used in clinical practice because of the lack of better alternatives. (ii) Bioimpedance measurements in the two cohorts were performed with different devices using different models. For practical reasons we did not perform repeated measurements, nor validate BCM measurements with InBody measurements and vice versa. Hence, comparison of bioimpedance-derived parameters across cohorts should be done with caution. (iii) A more precise characterization of the heart function might provide additional information of interest, e.g., the additional identification of patients with heart failure with preserved ejection fraction might be particularly informative, in consideration of the high prevalence of this condition in the HD population [33]. (iv) Including further parameters to assess OH and systemic congestion (e.g. by measuring the diameter if the inferior vena cava or by lung echocardiography) before and after dialysis might contribute to a more precise characterization of biomarker properties of sCD146 in this setting. Notably, the influence of other biological processes on circulating sCD146 (e.g. inflammation, angiogenesis) needs further investigation. (v) Larger multicentre studies, allowing a more comprehensive covariate adjustment and subgroup analysis, are needed to confirm our findings on long-term prognostic properties of BNP and sCD146 and to determine the absolute cut-off values of high- and low-BNP and sCD146 for a direct clinical application of this approach. This further step would also help us to understand if patients with high-BNP and low-sCD146 represent a population of patients with CHF without systemic congestion, or if this group is an artefact generated by the arbitrary definition of cut-offs. In summary, the combination of BNP and sCD146 dissects OH into functional components of prognostic value. SUPPLEMENTARY DATA Supplementary data are available at ndt online. FUNDING This study was supported by the Swiss Kidney Foundation and the Alfred and Erika Bär-Spycher Foundation. AUTHORS’ CONTRIBUTIONS M.A., P.E.C., S.V.M., S.S. and A.M. designed the study. S.V.M., K.G., M.S., K.T. and A.D. carried out measurements. M.A. and P.E.C. analysed the data and drafted the article. All authors revised and approved the final version of the article. CONFLICT OF INTEREST STATEMENT None declared. REFERENCES 1 Davies SJ , Davenport A. The role of bioimpedance and biomarkers in helping to aid clinical decision-making of volume assessments in dialysis patients . Kidney Int 2014 ; 86 : 489 – 496 Google Scholar CrossRef Search ADS PubMed 2 Wizemann V , Wabel P , Chamney P et al. The mortality risk of overhydration in haemodialysis patients . Nephrol Dial Transplant 2009 ; 24 : 1574 – 1579 Google Scholar CrossRef Search ADS PubMed 3 Agarwal R. Hypervolemia is associated with increased mortality among hemodialysis patients . Hypertension 2010 ; 56 : 512 – 517 Google Scholar CrossRef Search ADS PubMed 4 Onofriescu M , Siriopol D , Voroneanu L et al. Overhydration, cardiac function and survival in hemodialysis patients . PLoS One 2015 ; 10 : e0135691 Google Scholar CrossRef Search ADS PubMed 5 Caetano C , Valente A , Oliveira T et al. Body composition and mortality predictors in hemodialysis patients . J Ren Nutr 2016 ; 26 : 81 – 86 Google Scholar CrossRef Search ADS PubMed 6 Kim E-J , Choi M-J , Lee J-H et al. Extracellular fluid/intracellular fluid volume ratio as a novel risk indicator for all-cause mortality and cardiovascular disease in hemodialysis patients . PLoS One 2017 ; 12 : e0170272 Google Scholar CrossRef Search ADS PubMed 7 Arrigo M , Parissis JT , Akiyama E et al. Understanding acute heart failure: pathophysiology and diagnosis . Eur Heart J Suppl 2016 ; 18 : G11 – G18 Google Scholar CrossRef Search ADS 8 Picano E , Gargani L , Gheorghiade M. Why, when, and how to assess pulmonary congestion in heart failure: pathophysiological, clinical, and methodological implications . Heart Fail Rev 2010 ; 15 : 63 – 72 Google Scholar CrossRef Search ADS PubMed 9 Dastoor H , Bernieh B , Boobes Y et al. Plasma BNP in patients on maintenance haemodialysis: a guide to management? J Hypertens 2005 ; 23 : 23 – 28 Google Scholar CrossRef Search ADS PubMed 10 Morishita Y , Ando Y , Ishii E et al. Comparison of markers of circulating blood volume in hemodialysis patients . Clin Exp Nephrol 2005 ; 9 : 233 – 237 Google Scholar CrossRef Search ADS PubMed 11 Sheen V , Bhalla V , Tulua-Tata A et al. The use of B-type natriuretic peptide to assess volume status in patients with end-stage renal disease . Am Heart J 2007 ; 153 : 244.e1 – 244.e5 Google Scholar CrossRef Search ADS 12 Tapolyai M , Faludi M , Réti V et al. Volume estimation in dialysis patients: the concordance of brain-type natriuretic peptide measurements and bioimpedance values . Hemodial Int 2013 ; 17 : 406 – 412 Google Scholar CrossRef Search ADS PubMed 13 Basso F , Milan Manani S , Cruz DN et al. Comparison and reproducibility of techniques for fluid status assessment in chronic hemodialysis patients . Cardiorenal Med 2013 ; 3 : 104 – 112 Google Scholar CrossRef Search ADS PubMed 14 Donadio C , Bozzoli L , Colombini E et al. Effective and timely evaluation of pulmonary congestion . Medicine (Baltimore) 2015 ; 94 : e473 Google Scholar CrossRef Search ADS PubMed 15 Sivalingam M , Vilar E , Mathavakkannan S et al. The role of natriuretic peptides in volume assessment and mortality prediction in haemodialysis patients . BMC Nephrol 2015 ; 16 : 406 Google Scholar CrossRef Search ADS 16 Mallamaci F , Zoccali C , Tripepi G et al. Diagnostic potential of cardiac natriuretic peptides in dialysis patients . Kidney Int 2001 ; 59 : 1559 – 1566 Google Scholar CrossRef Search ADS PubMed 17 Zoccali C , Mallamaci F , Benedetto FA et al. Cardiac natriuretic peptides are related to left ventricular mass and function and predict mortality in dialysis patients . J Am Soc Nephrol 2001 ; 12 : 1508 – 1515 Google Scholar PubMed 18 Naganuma T , Sugimura K , Wada S et al. The prognostic role of brain natriuretic peptides in hemodialysis patients . Am J Nephrol 2002 ; 22 : 437 – 444 Google Scholar CrossRef Search ADS PubMed 19 Winkler K , Wanner C , Drechsler C et al. Change in N-terminal-pro-B-type-natriuretic-peptide and the risk of sudden death, stroke, myocardial infarction, and all-cause mortality in diabetic dialysis patients . Eur Heart J 2008 ; 29 : 2092 – 2099 Google Scholar CrossRef Search ADS PubMed 20 Paniagua R , Ventura MDJ , Avila-Diaz M et al. NT-proBNP, fluid volume overload and dialysis modality are independent predictors of mortality in ESRD patients . Nephrol Dial Transplant 2010 ; 25 : 551 – 557 Google Scholar CrossRef Search ADS PubMed 21 Breidthardt T , Kalbermatter S , Socrates T et al. Increasing B-type natriuretic peptide levels predict mortality in unselected haemodialysis patients . Eur J Heart Fail 2011 ; 13 : 860 – 867 Google Scholar CrossRef Search ADS PubMed 22 Artunc F , Nowak A , Müller C et al. Mortality prediction using modern peptide biomarkers in hemodialysis patients—a comparative analysis . Kidney Blood Press Res 2014 ; 39 : 563 – 572 Google Scholar CrossRef Search ADS PubMed 23 Ishii J , Takahashi H , Kitagawa F et al. Multimarker approach to risk stratification for long-term mortality in patients on chronic hemodialysis . Circ J 2015 ; 79 : 656 – 663 Google Scholar CrossRef Search ADS PubMed 24 Gayat E , Caillard A , Laribi S et al. Soluble CD146, a new endothelial biomarker of acutely decompensated heart failure . Int J Cardiol 2015 ; 199 : 241 – 247 Google Scholar CrossRef Search ADS PubMed 25 Arrigo M , Truong QA , Onat D et al. Soluble CD146 is a novel marker of systemic congestion in heart failure patients: an experimental mechanistic and transcardiac clinical study . Clin Chem 2017 ; 63 : 386 – 393 Google Scholar CrossRef Search ADS PubMed 26 Kubena P , Arrigo M , Parenica J et al. Plasma levels of soluble CD146 reflect the severity of pulmonary congestion better than brain natriuretic peptide in acute coronary syndrome . Ann Lab Med 2016 ; 36 : 300 Google Scholar CrossRef Search ADS PubMed 27 Moissl UM , Wabel P , Chamney PW et al. Body fluid volume determination via body composition spectroscopy in health and disease . Physiol Meas 2006 ; 27 : 921 – 933 Google Scholar CrossRef Search ADS PubMed 28 Chamney PW , Wabel P , Moissl UM et al. A whole-body model to distinguish excess fluid from the hydration of major body tissues . Am J Clin Nutr 2007 ; 85 : 80 – 89 Google Scholar CrossRef Search ADS PubMed 29 Kumar S , Khosravi M , Massart A et al. Are serum to dialysate sodium gradient and segmental bioimpedance volumes associated with the fall in blood pressure with hemodialysis? Int J Artif Organs 2014 ; 37 : 21 – 28 Google Scholar CrossRef Search ADS PubMed 30 Ponikowski P , Voors AA , Anker SD et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure . Eur Heart J 2016 ; 37 : 2129 – 2200 Google Scholar CrossRef Search ADS PubMed 31 Boneberg E-M , Illges H , Legler DF , Fürstenberger G. Soluble CD146 is generated by ectodomain shedding of membrane CD146 in a calcium-induced, matrix metalloprotease-dependent process . Microvasc Res 2009 ; 78 : 325 – 331 Google Scholar CrossRef Search ADS PubMed 32 Chazot C , Rozes M , Vo-Van C et al. Brain natriuretic peptide is a marker of fluid overload in incident hemodialysis patients . Cardiorenal Med 2017 ; 7 : 218 – 226 Google Scholar CrossRef Search ADS PubMed 33 Van Aelst LNL , Arrigo M , Placido R et al. Acutely decompensated heart failure with preserved and reduced ejection fraction present with comparable haemodynamic congestion . Eur J Heart Fail 2018 ; 20 : 738 – 747 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nephrology Dialysis Transplantation Oxford University Press

Soluble CD146 and B-type natriuretic peptide dissect overhydration into functional components of prognostic relevance in haemodialysis patients

Loading next page...
 
/lp/ou_press/soluble-cd146-and-b-type-natriuretic-peptide-dissect-overhydration-M87qQs2PJS
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
ISSN
0931-0509
eISSN
1460-2385
D.O.I.
10.1093/ndt/gfy113
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Accurate volume status evaluation and differentiation of cardiac and non-cardiac components of overhydration (OH) are fundaments of optimal haemodialysis (HD) management. Methods This study, by combining bioimpedance measurements, cardiovascular biomarkers and echocardiography, aimed at dissecting OH into its major functional components, and prospectively tested the association between cardiac and non-cardiac components of OH with mortality. In the first part, we validated soluble CD146 (sCD146) as a non-cardiac biomarker of systemic congestion in a cohort of 30 HD patients. In the second part, we performed a prospective 1-year follow-up study in an independent cohort of 144 HD patients. Results sCD146 incrementally increased after the short and long intervals after HD (+53 ng/mL, P = 0.006 and  +91 ng/mL, P < 0.001), correlated with OH as determined by bioimpedance and well-diagnosed OH (area under the receiver operating characteristics curve 0.72, P = 0.005). The prevalence of OH was lower for low-sCD146 and low-BNP patients (B-type natriuretic peptide, 29%) compared with subjects with either one or both biomarkers elevated (65–74%, P < 0.001). Notably, most low-BNP but high-sCD146 subjects were overhydrated. Systolic dysfunction was 2- to 3-fold more prevalent among high-BNP compared with low-BNP patients (44–68% versus 21–23%, chi-square P < 0.001), regardless of sCD146. One-year all-cause mortality was markedly higher in patients with high-BNP (P = 0.001) but not with high-sCD146. In multivariate analysis, systolic dysfunction and BNP, but not OH, were associated with lower survival. Conclusions The combination of BNP and sCD146 dissects OH into functional components of prognostic value. OH in HD patients is associated with higher mortality only if resulting from cardiac dysfunction. BNP, haemodialysis, heart failure, overhydration, sCD146 INTRODUCTION Volume management is a critical element in the care of haemodialysis (HD) patients and requires accurate assessment of both the hydration state and the underlying pathophysiology. Overhydration (OH), a state of generalized oedema formation, is a frequent condition in HD patients and is associated with reduced survival [1–6]. In patients with end-stage renal disease (ESRD), OH is related either to renal salt and water retention and insufficient ultrafiltration during HD or concomitant congestive heart failure (CHF). Indeed, increased intracardiac and upstream vascular hydrostatic pressures (systemic congestion) may contribute to oedema formation in patients with ESRD and CHF [7, 8]. The dissection of OH into cardiac and non-cardiac components in the individual HD patient is challenging because ESRD and CHF often coexist and because rapid and reliable diagnostic tools are still lacking. However, the differentiation of the underlying mechanisms is of significant clinical relevance since it might implicate different treatments. Bioimpedance analysis has become a popular tool in the assessment of OH in HD patients [1]. Body composition measurement (BCM) by bioimpedance not only allows to quantify malnutrition and sarcopenia, but also to provide a quantitative estimation of fluid compartments. This general estimate of the volume status supports the clinician in HD volume management but does not distinguish the underlying mechanisms of OH. Cardiovascular biomarkers might provide complementary diagnostic information for this purpose. Natriuretic peptides—B-type natriuretic peptide (BNP) and the amino-terminal pro-BNP (NT-proBNP)—are released by cardiomyocytes in the presence of increased intracardiac pressures. Previous studies showed discordant results regarding the ability of natriuretic peptides to detect OH and guide volume management [9–15], but consistently displayed excellent negative predictive values for CHF, even in HD patients [16, 17]. Thus, natriuretic peptides are considered a biomarker of cardiac dysfunction rather than OH in HD patients. Notably, elevated plasma concentrations of natriuretic peptides correlated well with mortality in HD patients [15, 17–23]. Our group recently described soluble CD146 (sCD146) as a novel vascular biomarker, released by vessels in response to endothelial stretch [24–26]. The extra-cardiac origin of sCD146 provides complementary information to BNP, and its mode of release into circulation makes it a more promising direct marker of systemic congestion, regardless of cardiac function. In this study, after validating sCD146 as a non-cardiac biomarker of systemic congestion in HD patients, by measuring BNP and sCD146 we dissected OH into its major functional components and prospectively tested the association between cardiac and non-cardiac components of OH with mortality in HD patients. MATERIALS AND METHODS Study population The study population consisted of 174 clinically stable chronic outpatients on the maintenance HD. The first cohort of patients (‘Zurich’, n = 30) underwent a mechanistic study to validate sCD146 as a biomarker of systemic congestion in HD. This part of the study was performed from 1 October to 31 December 2016, at the University Hospital of Zurich, Switzerland. Inclusion criteria were age >18 years and HD since at least 1 month; exclusion criteria were acute illness or hospitalization in the last week before inclusion, and implanted cardiac device, as it precludes reliable bioelectrical impedance measurement. The second cohort of patients (‘London’, n = 144) was used for external validation of the ‘Zurich’ study and for the clinical outcome study, aimed at investigating the relationship between OH, cardiac dysfunction and outcome in stable HD patients. This part of the study was performed at the Royal Free Hospital of London, UK. Patients attending for their routine outpatient midweek HD session on 5 May and 6 May 2016 were included and followed up until the end of May 2017. Mechanistic ‘Zurich’ study The mechanistic ‘Zurich’ study consisted of four visits. The first visit was performed after a regular HD session, followed by a short interdialysis interval (duration: 2 days). The second visit was performed at the end of the short interval, before the HD session, and the third visit after the same HD session. The fourth visit was performed at the end of the subsequent prolonged interdialysis interval (‘long interval’, duration: 3 days). Briefly, the four visits allowed for study of a short and a long inter-dialysis interval to test different levels of OH. Total body weight was measured at each visit, to calculate the net body weight increase during the short and the long intervals, respectively. Furthermore, OH was quantitatively assessed at each study visit with a portable whole-body bioimpedance spectroscopy device (Fresenius Medical Care GmbH, Bad Homburg, Germany). BCM was performed with the patient in a supine position according to the manufacturer’s description. Based on a fluid assessment model using resistances to electrical currents of 50 discrete frequencies, the total body water (TBW), the extracellular water (ECW), the intracellular water (ICW) and the OH are calculated [27, 28]. Relevant OH was previously defined as an OH/ECW ratio >0.15 [1, 2]. Patient history and clinical evaluation were collected at each study visit. Comorbidities and HD-specific information were retrieved from electronic medical charts. Prospective clinical ‘London’ outcome study The clinical ‘London’ outcome study consisted of one visit, before a regular outpatient midweek HD session, followed by a prospective follow-up of 1 year for all-cause mortality. The quantitative assessment of OH was performed before the HD session using multifrequency bioimpedance (InBody 720, Seoul, South Korea), in a standardized protocol [29]. The TBW, ECW and ICW were calculated. Since the OH value was not determined in this cohort, relevant OH was defined in the presence of an ECW/TBW ratio >0.4 according to the literature [1]. All patients underwent transthoracic cardiac echocardiography for determination of the left-ventricular ejection fraction on a non-dialysis day. Cardiac systolic dysfunction was defined as left-ventricular ejection fraction <50%, according to the guidelines of the European Society of Cardiology [30]. Clinical data were retrieved from electronic medical charts. Biomarker testing Venous blood samples for measurement of cardiovascular biomarkers were drawn at each visit, centrifuged within 6 h and ethylenediaminetetraacetic acid (EDTA) plasma aliquots were stored at −80°C until analysed at Lariboisière University Hospital, Paris, France. The concentration of sCD146 was determined by ELISA (CY-QUANT ELISA sCD146©, Biocytex, France) with a detection limit of sCD146 of 10 ng/mL and coefficients of variation for both repeatability and reproducibility < 20% in the measured range. Measurement of concentration of B-type natriuretic peptide (BNP) was performed with the Architect i2000 platform (Abbott Diagnostics, Abbott Park, IL, USA). Ethical considerations The study was conducted according to the standards of the Declaration of Helsinki and approved by the local ethics committee of Zurich (2016-451) and London (13/Lo/0912). All patients provided written informed consent (ClinicalTrials.gov NCT02962635). Statistical analysis Continuous variables are expressed as median [interquartile range (IQR)], categorical variables as number (percentage). Independent samples were compared using the Mann–Whitney U test or chi-square test, as appropriate. Related samples were compared using the Wilcoxon signed rank test. Correlation analysis was performed using Spearman’s correlation coefficient (rho). The diagnostic performance of the biomarkers was assessed by receiver operating characteristic (ROC) analysis and expressed as area under the curve (AUC). Survival was plotted with the Kaplan–Meier curve and differences between groups were assessed by the log-rank test. Univariate and covariate-adjusted Cox proportional hazards regression models were used to estimate the association between circulating biomarkers and risk of death. The covariates included in the regression models were a priori selected among baseline variables with known associations with study outcomes: age, sex, cardiac systolic dysfunction, relevant OH, anuria, log-transformed BNP and log-transformed sCD146. The null hypothesis was rejected with an adjusted two-sided P-value <0.05. All analyses were performed with the use of IBM SPSS Statistics, Version 25.0 (IBM Corp, Armonk, NY, USA). RESULTS Characterization of the study population The study population consisted of 174 HD patients, predominantly middle-aged men with ESRD (Table 1). Dialysis access was mostly an arteriovenous fistula, and anuria was highly prevalent. Cardiovascular risk factors and vascular complications (coronary and peripheral artery disease) were highly prevalent. Table 1 Patient characteristics of the study population ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 Data are presented as median [quartiles] or number (percentage) as appropriate. Table 1 Patient characteristics of the study population ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 ‘Zurich’ cohort ‘London’ cohort P-value (n = 30) (n = 144) Age (years) 67 [56; 78] 73 [59; 81] 0.15 Gender (male) 25 (83%) 88 (61%) 0.021 Height (cm) 172 [162; 180] 163 [157; 171] 0.001 Weight (kg) 76 [62; 86] 71 [63; 79] 0.12 Dialysis access 0.05  AV fistula/graft 29 (97%) 119 (83%)  Catheter 1 (3%) 25 (17%) Dialysis vintage (months) 25 [12; 42] 35 [19; 61] 0.043 Anuria (<500 mL/day) 14 (47%) 106 (74%) 0.008 Arterial hypertension 30 (100%) 138 (96%) 0.59 Diabetes 11 (37%) 75 (52%) 0.16 History of coronary artery disease 12 (40%) 53 (37%) 0.84 History of chronic heart failure 4 (13%) 36 (25%) 0.23 History of peripheral artery disease 5 (17%) 23 (16%) 1.00 Plasma sCD146 (ng/mL) (before HD) 543 [390; 615] 542 [424; 705] 0.55 Data are presented as median [quartiles] or number (percentage) as appropriate. Mechanistic ‘Zurich’ study on vascular sCD146 as a biomarker of systemic congestion The median value of circulating sCD146 in all studied patients at all time points (120 measurements) was 530 ng/mL (range: 123–1730 ng/mL). The median intraindividual variability was 43% (range: 6–354%, Supplementary data, Figure S1). Dialysis did not alter circulating sCD146, as shown by similar sCD146 levels before and after HD [543 (384; 625) versus 492 ng/mL (393; 646), P = 0.52, Supplementary data, Figure S2]. Moreover, we measured changes in body weight and sCD146 concentrations after a short (2 days) and a long (3 days) interdialysis interval. We observed an incremental increase in both body weight [+0.9 kg (+0.3; +1.8), P < 0.001 and  + 1.8 kg (+0.1; +2.7), P < 0.001] and circulating sCD146 [+53 ng/mL (−16; +118), P = 0.006 and  +91 ng/mL (+20; +179), P < 0.001] after the short and long intervals, respectively (Figure 1). FIGURE 1 View largeDownload slide Median changes in body weight and sCD146 during the short and the long interval in the ‘Zurich’ cohort. Median values and quartiles are reported. P-values refer to related samples of Wilcoxon signed-rank test. N = 30. FIGURE 1 View largeDownload slide Median changes in body weight and sCD146 during the short and the long interval in the ‘Zurich’ cohort. Median values and quartiles are reported. P-values refer to related samples of Wilcoxon signed-rank test. N = 30. Notably, while anuric patients (n = 14) had higher baseline concentration of sCD146 compared with non-anuric patients [564 (463; 766) versus 435 (335; 499) ng/mL, P = 0.022], the median increase in circulating sCD146 during the short interval was more pronounced in non-anuric patients [median: +89 (+5; +124) ng/mL, P = 0.004] compared with anuric patients [median: +30 (−91; +88) ng/mL, P = 0.47]. Similar results were found during the long interval (data not shown). Furthermore, we assessed the correlation between sCD146 and OH determined by BCM at the end of both short and long intervals (n = 60; Figure 2). Circulating sCD146 positively correlated with OH (Spearman’s ρ 0.37, P = 0.004). Circulating sCD146 showed good properties for diagnosing patients with relevant OH [area under the receiver operating characteristics curve, AUROC: 0.72 (0.58; 0.86), P = 0.005]. FIGURE 2 View largeDownload slide Correlation between OH and circulating sCD146 after the short and long interval (A) and ROC curve of sCD146 (B) for diagnosing relevant OH (OH/ECV >0.15) in the ‘Zurich’ cohort. Dotted lines refer to the definition of relevant OH (OH/ECW >0.15). N = 60. FIGURE 2 View largeDownload slide Correlation between OH and circulating sCD146 after the short and long interval (A) and ROC curve of sCD146 (B) for diagnosing relevant OH (OH/ECV >0.15) in the ‘Zurich’ cohort. Dotted lines refer to the definition of relevant OH (OH/ECW >0.15). N = 60. Finally, we evaluated the diagnostic properties of circulating sCD146 to detect dyspnoea, a clinical parameter of systemic congestion. At the end of both short and long intervals (n = 60), 50% of patients reported dyspnoea, defined as breathlessness at rest or during stress (New York Heart Association class II–IV). Dyspneic patients displayed higher plasma concentrations of sCD146 [633 ng/mL (541; 770 ng/mL)] compared with patients without dyspnoea [523 ng/mL (364; 622 ng/mL), P = 0.006] (Figure 3A). Notably, BNP levels did not discriminate dyspneic and non-dyspneic patients [332 (191; 697) versus 147 pg/mL (63; 639), P = 0.08]. The AUROC of sCD146 for diagnosing dyspnoea was 0.71 (0.58; 0.84), P = 0.006 (Figure 3B). Plasma concentrations of sCD146 <461 ng/mL had a sensitivity  >90% of not being associated with dyspnoea. FIGURE 3 View largeDownload slide Circulating sCD146 according to the symptoms of dyspnoea (A) and ROC curve of sCD146 for diagnosing dyspnoea in the ‘Zurich’ cohort. (B) N = 60. CI, confidence interval. FIGURE 3 View largeDownload slide Circulating sCD146 according to the symptoms of dyspnoea (A) and ROC curve of sCD146 for diagnosing dyspnoea in the ‘Zurich’ cohort. (B) N = 60. CI, confidence interval. Validation of the ‘Zurich’ study in the ‘London’ cohort The data obtained in the ‘Zurich’ cohort were validated in the larger independent ‘London’ cohort of 144 stable HD outpatients. We confirmed that subjects with relevant OH (n = 84) had higher levels of sCD146 [598 (454; 744) versus 479 ng/mL (416; 597), P = 0.003] compared with non-OH subjects (n = 60). Circulating sCD146 levels positively correlated with OH (Spearman’s ρ 0.28, P = 0.001). Cardiac and non-cardiac components of OH In a further analysis in the ‘London’ cohort (n = 144), we found that the prevalence of relevant OH was much lower for patients with both low sCD146 and low BNP (29%) compared with patients with either an increase in one or both biomarkers (65–74%, chi-square P < 0.001) (Figure 4A). Notably, in the group of subjects with low-BNP but high-sCD146 (upper left quadrant), nearly three of four showed relevant OH, indicating that OH and systemic congestion correlated in most patients, but were not necessarily related to high BNP. FIGURE 4 View largeDownload slide Distribution of relevant OH (A) and cardiac systolic dysfunction (B) according to the circulating concentrations of sCD146 and BNP (n = 144) in the ‘London’ cohort. Values of BNP and sCD146 were log-transformed. Groups were divided according to the median values of BNP and sCD146 (represented by lines). Percentages refer to the prevalence of relevant OH (A) and cardiac systolic dysfunction (B). N = 144. FIGURE 4 View largeDownload slide Distribution of relevant OH (A) and cardiac systolic dysfunction (B) according to the circulating concentrations of sCD146 and BNP (n = 144) in the ‘London’ cohort. Values of BNP and sCD146 were log-transformed. Groups were divided according to the median values of BNP and sCD146 (represented by lines). Percentages refer to the prevalence of relevant OH (A) and cardiac systolic dysfunction (B). N = 144. Furthermore, patients with cardiac systolic dysfunction (n = 48) had higher BNP levels [1507 (504; 4245) versus 473 pg/mL (259; 1368), P < 0.001] compared with patients with preserved systolic function. However, sCD146 levels were similar in patients with or without cardiac systolic dysfunction [525 (387; 690) versus 559 ng/mL (436; 714), P = 0.47]. The prevalence of systolic dysfunction was ∼2- to 3-fold among patients with elevated BNP (44–68%) compared with patients with low-BNP (21–23%, chi-square P < 0.001), regardless of sCD146 (Figure 4B). These findings indicate that OH and systemic congestion not always reflect CHF in HD patients and that circulating sCD146 is a biomarker of OH/systemic congestion independent of CHF. Prospective clinical ‘London’ outcome study Based on the previous data, we investigated the prognostic impact of OH and its functional components on all-cause mortality in HD patients of the ‘London’ cohort. During a median follow-up of 365 days, a total of 27 deaths occurred. In univariate analysis (Table 2) age, relevant OH, cardiac systolic dysfunction and log-transformed BNP were shown to be associated with all-cause mortality, while sex, anuria and log-transformed sCD146 were not. In multivariate analysis, systolic dysfunction and log-transformed BNP were still strongly associated with all-cause mortality, while age and relevant OH were not. BNP shows excellent properties for predicting all-cause mortality during the follow-up [AUROC 0.81 (0.73; 0.90), P < 0.001], with specificity >90% for values exceeding 3000 pg/mL. Table 2 Univariate and multivariate analysis for prediction of long-term all-cause mortality (‘London’ cohort, n = 144) HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR, hazard ratio; CI, confidence interval. Table 2 Univariate and multivariate analysis for prediction of long-term all-cause mortality (‘London’ cohort, n = 144) HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR CI P-value HR CI P-value Age 1.03 0.99–1.06 0.052 1.03 0.99–1.07 0.09 Sex 1.08 0.50–2.36 0.84 Cardiac systolic dysfunction 5.24 2.07–13.3 <0.001 3.43 1.23–9.09 0.018 Relevant OH 2.50 1.01–6.19 0.048 0.86 0.26–2.79 0.80 Anuria 0.80 0.35–1.82 0.59 log BNP 7.29 3.44–15.5 <0.001 5.19 1.78–15.2 0.003 log sCD146 1.10 0.14–8.64 0.93 HR, hazard ratio; CI, confidence interval. Figure 5 illustrates the marked difference in survival of HD patients according to the presence of relevant OH (log-rank P = 0.041), systolic dysfunction (log-rank P < 0.001) and high BNP (log-rank P = 0.001) but not high sCD146 levels. FIGURE 5 View largeDownload slide Survival of HD patients according to the presence of relevant OH, cardiac systolic dysfunction and biomarker concentrations in the ‘London’ cohort. Relevant OH was defined as ECW/TBW >0.4. Low/high BNP was defined according to the median (631 pg/mL), low/high sCD146 was defined according to the median (542 ng/mL). N = 144. FIGURE 5 View largeDownload slide Survival of HD patients according to the presence of relevant OH, cardiac systolic dysfunction and biomarker concentrations in the ‘London’ cohort. Relevant OH was defined as ECW/TBW >0.4. Low/high BNP was defined according to the median (631 pg/mL), low/high sCD146 was defined according to the median (542 ng/mL). N = 144. DISCUSSION Optimal volume management in HD patients is of significant clinical relevance and relies on the accurate assessment of OH and its functional components. Indeed, since CHF is highly prevalent in ESRD patients, OH may have both cardiac and non-cardiac components. Our study, after validating the novel vascular biomarker sCD146 as a non-cardiac biomarker of systemic congestion in HD patients, by combining two cardiovascular biomarkers (BNP and sCD146) dissected OH into its major functional components. Furthermore, we prospectively showed that cardiac systolic dysfunction and not systemic congestion per se is associated with high all-cause mortality in HD patients. Vascular sCD146 is released from endothelial cells upon mechanical stress and is considered a biomarker of systemic congestion, independent form cardiac function [24–26, 31]. In the mechanistic part of our study, we confirmed for the first time, in two independent cohorts of stable HD patients (‘Zurich’ and ‘London’), that sCD146 is an accurate non-cardiac biomarker of systemic congestion, as previously shown in CHF [24–26]. Indeed, circulating sCD146 concentrations well correlated with clinical markers such as weight gain, dyspnoea and BCM-derived parameters of OH. Using sCD146 (reflecting systemic congestion, independently from the cardiac function) and combining it with BNP and echocardiographic data (reflecting cardiac function), we dissected OH into its major functional components (cardiac versus non-cardiac). Our data showed that OH (as determined by BCM), systemic congestion and cardiac dysfunction do not necessarily coexist in HD patients. Our data showed that relevant OH might be associated with three biomarker phenotypes: high-BNP/low-sCD146 (indicating structural cardiac impairment), low-BNP/high-sCD146 (indicating OH without cardiac involvement) and high-BNP/high-sCD146 (indicating OH of mixed origin). The application of this novel biomarker-based dissection of functional components of OH led us to the second main finding of this study: the high all-cause mortality associated with OH in HD patients is strongly related to cardiac systolic dysfunction. In light of these data, the previously reported increased mortality related to OH is probably indicative of a higher prevalence of prognostically relevant CHF among patients with BCM-determined OH, and not to systemic congestion per se. Clinical significance Our findings need to be validated in prospective larger studies but might contribute to a change in the management of HD patients according to the biomarker profile. High-BNP patients have a high likelihood of CHF and display high mortality independently from sCD146 concentration. High-BNP/low-sCD146 patients might require extensive cardiovascular evaluation and specific CHF therapy. High-BNP/high-sCD146 patients likely display OH of mixed origin and might benefit from a combination of specific CHF-therapy and intensive volume management (e.g. ultrafiltration) [32]. Low-BNP patients display good survival independently from sCD146 values. In these patients, sCD146 might help to guide volume management, in particular in the presence of symptoms such as dyspnoea, but in most patients, a less restrictive volume management to avoid hypotension, cardiac stunning and anuria might be appropriate. Further studies correlating sCD146 levels with blood pressure and the incidence of cardiovascular events will provide additional data of interest for an optimal volume management. Furthermore, the application of a biomarker approach might lead to a more precise cardiovascular stratification of patients to be included in interventional studies aiming at the improvement of cardiovascular mortality on HD. Limitations This study has several limitations. (i) Bioimpedance measurement cannot distinguish intravascular ECW from extravascular ECW and changes in overall ECW may not be equally distributed in the intravascular and extravascular components. Despite this significant limitation, bioimpedance measurements are still routinely used in clinical practice because of the lack of better alternatives. (ii) Bioimpedance measurements in the two cohorts were performed with different devices using different models. For practical reasons we did not perform repeated measurements, nor validate BCM measurements with InBody measurements and vice versa. Hence, comparison of bioimpedance-derived parameters across cohorts should be done with caution. (iii) A more precise characterization of the heart function might provide additional information of interest, e.g., the additional identification of patients with heart failure with preserved ejection fraction might be particularly informative, in consideration of the high prevalence of this condition in the HD population [33]. (iv) Including further parameters to assess OH and systemic congestion (e.g. by measuring the diameter if the inferior vena cava or by lung echocardiography) before and after dialysis might contribute to a more precise characterization of biomarker properties of sCD146 in this setting. Notably, the influence of other biological processes on circulating sCD146 (e.g. inflammation, angiogenesis) needs further investigation. (v) Larger multicentre studies, allowing a more comprehensive covariate adjustment and subgroup analysis, are needed to confirm our findings on long-term prognostic properties of BNP and sCD146 and to determine the absolute cut-off values of high- and low-BNP and sCD146 for a direct clinical application of this approach. This further step would also help us to understand if patients with high-BNP and low-sCD146 represent a population of patients with CHF without systemic congestion, or if this group is an artefact generated by the arbitrary definition of cut-offs. In summary, the combination of BNP and sCD146 dissects OH into functional components of prognostic value. SUPPLEMENTARY DATA Supplementary data are available at ndt online. FUNDING This study was supported by the Swiss Kidney Foundation and the Alfred and Erika Bär-Spycher Foundation. AUTHORS’ CONTRIBUTIONS M.A., P.E.C., S.V.M., S.S. and A.M. designed the study. S.V.M., K.G., M.S., K.T. and A.D. carried out measurements. M.A. and P.E.C. analysed the data and drafted the article. All authors revised and approved the final version of the article. CONFLICT OF INTEREST STATEMENT None declared. REFERENCES 1 Davies SJ , Davenport A. The role of bioimpedance and biomarkers in helping to aid clinical decision-making of volume assessments in dialysis patients . Kidney Int 2014 ; 86 : 489 – 496 Google Scholar CrossRef Search ADS PubMed 2 Wizemann V , Wabel P , Chamney P et al. The mortality risk of overhydration in haemodialysis patients . Nephrol Dial Transplant 2009 ; 24 : 1574 – 1579 Google Scholar CrossRef Search ADS PubMed 3 Agarwal R. Hypervolemia is associated with increased mortality among hemodialysis patients . Hypertension 2010 ; 56 : 512 – 517 Google Scholar CrossRef Search ADS PubMed 4 Onofriescu M , Siriopol D , Voroneanu L et al. Overhydration, cardiac function and survival in hemodialysis patients . PLoS One 2015 ; 10 : e0135691 Google Scholar CrossRef Search ADS PubMed 5 Caetano C , Valente A , Oliveira T et al. Body composition and mortality predictors in hemodialysis patients . J Ren Nutr 2016 ; 26 : 81 – 86 Google Scholar CrossRef Search ADS PubMed 6 Kim E-J , Choi M-J , Lee J-H et al. Extracellular fluid/intracellular fluid volume ratio as a novel risk indicator for all-cause mortality and cardiovascular disease in hemodialysis patients . PLoS One 2017 ; 12 : e0170272 Google Scholar CrossRef Search ADS PubMed 7 Arrigo M , Parissis JT , Akiyama E et al. Understanding acute heart failure: pathophysiology and diagnosis . Eur Heart J Suppl 2016 ; 18 : G11 – G18 Google Scholar CrossRef Search ADS 8 Picano E , Gargani L , Gheorghiade M. Why, when, and how to assess pulmonary congestion in heart failure: pathophysiological, clinical, and methodological implications . Heart Fail Rev 2010 ; 15 : 63 – 72 Google Scholar CrossRef Search ADS PubMed 9 Dastoor H , Bernieh B , Boobes Y et al. Plasma BNP in patients on maintenance haemodialysis: a guide to management? J Hypertens 2005 ; 23 : 23 – 28 Google Scholar CrossRef Search ADS PubMed 10 Morishita Y , Ando Y , Ishii E et al. Comparison of markers of circulating blood volume in hemodialysis patients . Clin Exp Nephrol 2005 ; 9 : 233 – 237 Google Scholar CrossRef Search ADS PubMed 11 Sheen V , Bhalla V , Tulua-Tata A et al. The use of B-type natriuretic peptide to assess volume status in patients with end-stage renal disease . Am Heart J 2007 ; 153 : 244.e1 – 244.e5 Google Scholar CrossRef Search ADS 12 Tapolyai M , Faludi M , Réti V et al. Volume estimation in dialysis patients: the concordance of brain-type natriuretic peptide measurements and bioimpedance values . Hemodial Int 2013 ; 17 : 406 – 412 Google Scholar CrossRef Search ADS PubMed 13 Basso F , Milan Manani S , Cruz DN et al. Comparison and reproducibility of techniques for fluid status assessment in chronic hemodialysis patients . Cardiorenal Med 2013 ; 3 : 104 – 112 Google Scholar CrossRef Search ADS PubMed 14 Donadio C , Bozzoli L , Colombini E et al. Effective and timely evaluation of pulmonary congestion . Medicine (Baltimore) 2015 ; 94 : e473 Google Scholar CrossRef Search ADS PubMed 15 Sivalingam M , Vilar E , Mathavakkannan S et al. The role of natriuretic peptides in volume assessment and mortality prediction in haemodialysis patients . BMC Nephrol 2015 ; 16 : 406 Google Scholar CrossRef Search ADS 16 Mallamaci F , Zoccali C , Tripepi G et al. Diagnostic potential of cardiac natriuretic peptides in dialysis patients . Kidney Int 2001 ; 59 : 1559 – 1566 Google Scholar CrossRef Search ADS PubMed 17 Zoccali C , Mallamaci F , Benedetto FA et al. Cardiac natriuretic peptides are related to left ventricular mass and function and predict mortality in dialysis patients . J Am Soc Nephrol 2001 ; 12 : 1508 – 1515 Google Scholar PubMed 18 Naganuma T , Sugimura K , Wada S et al. The prognostic role of brain natriuretic peptides in hemodialysis patients . Am J Nephrol 2002 ; 22 : 437 – 444 Google Scholar CrossRef Search ADS PubMed 19 Winkler K , Wanner C , Drechsler C et al. Change in N-terminal-pro-B-type-natriuretic-peptide and the risk of sudden death, stroke, myocardial infarction, and all-cause mortality in diabetic dialysis patients . Eur Heart J 2008 ; 29 : 2092 – 2099 Google Scholar CrossRef Search ADS PubMed 20 Paniagua R , Ventura MDJ , Avila-Diaz M et al. NT-proBNP, fluid volume overload and dialysis modality are independent predictors of mortality in ESRD patients . Nephrol Dial Transplant 2010 ; 25 : 551 – 557 Google Scholar CrossRef Search ADS PubMed 21 Breidthardt T , Kalbermatter S , Socrates T et al. Increasing B-type natriuretic peptide levels predict mortality in unselected haemodialysis patients . Eur J Heart Fail 2011 ; 13 : 860 – 867 Google Scholar CrossRef Search ADS PubMed 22 Artunc F , Nowak A , Müller C et al. Mortality prediction using modern peptide biomarkers in hemodialysis patients—a comparative analysis . Kidney Blood Press Res 2014 ; 39 : 563 – 572 Google Scholar CrossRef Search ADS PubMed 23 Ishii J , Takahashi H , Kitagawa F et al. Multimarker approach to risk stratification for long-term mortality in patients on chronic hemodialysis . Circ J 2015 ; 79 : 656 – 663 Google Scholar CrossRef Search ADS PubMed 24 Gayat E , Caillard A , Laribi S et al. Soluble CD146, a new endothelial biomarker of acutely decompensated heart failure . Int J Cardiol 2015 ; 199 : 241 – 247 Google Scholar CrossRef Search ADS PubMed 25 Arrigo M , Truong QA , Onat D et al. Soluble CD146 is a novel marker of systemic congestion in heart failure patients: an experimental mechanistic and transcardiac clinical study . Clin Chem 2017 ; 63 : 386 – 393 Google Scholar CrossRef Search ADS PubMed 26 Kubena P , Arrigo M , Parenica J et al. Plasma levels of soluble CD146 reflect the severity of pulmonary congestion better than brain natriuretic peptide in acute coronary syndrome . Ann Lab Med 2016 ; 36 : 300 Google Scholar CrossRef Search ADS PubMed 27 Moissl UM , Wabel P , Chamney PW et al. Body fluid volume determination via body composition spectroscopy in health and disease . Physiol Meas 2006 ; 27 : 921 – 933 Google Scholar CrossRef Search ADS PubMed 28 Chamney PW , Wabel P , Moissl UM et al. A whole-body model to distinguish excess fluid from the hydration of major body tissues . Am J Clin Nutr 2007 ; 85 : 80 – 89 Google Scholar CrossRef Search ADS PubMed 29 Kumar S , Khosravi M , Massart A et al. Are serum to dialysate sodium gradient and segmental bioimpedance volumes associated with the fall in blood pressure with hemodialysis? Int J Artif Organs 2014 ; 37 : 21 – 28 Google Scholar CrossRef Search ADS PubMed 30 Ponikowski P , Voors AA , Anker SD et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure . Eur Heart J 2016 ; 37 : 2129 – 2200 Google Scholar CrossRef Search ADS PubMed 31 Boneberg E-M , Illges H , Legler DF , Fürstenberger G. Soluble CD146 is generated by ectodomain shedding of membrane CD146 in a calcium-induced, matrix metalloprotease-dependent process . Microvasc Res 2009 ; 78 : 325 – 331 Google Scholar CrossRef Search ADS PubMed 32 Chazot C , Rozes M , Vo-Van C et al. Brain natriuretic peptide is a marker of fluid overload in incident hemodialysis patients . Cardiorenal Med 2017 ; 7 : 218 – 226 Google Scholar CrossRef Search ADS PubMed 33 Van Aelst LNL , Arrigo M , Placido R et al. Acutely decompensated heart failure with preserved and reduced ejection fraction present with comparable haemodynamic congestion . Eur J Heart Fail 2018 ; 20 : 738 – 747 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Nephrology Dialysis TransplantationOxford University Press

Published: May 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off