Copeptin and Its Relation to Arteriovenous Fistula (AVF) Type and NYHA Class in Hemodialysis Patients

Abstract

Copeptin is cosynthesized with vasopressin, also known as anti-diuretic hormone, with similar plasma levels. In the past 2 years, copeptin has been studied as a diagnostic and prognostic marker in infections and other diseases. In patients with decompensated heart failure, copeptin was an accurate prognostic marker for mortality. Cardiovascular disease is a major contributor to the mortality and morbidity in chronic kidney disease. Creation of an arteriovenous fistula (AVF) might contribute to the development or worsening of congestive heart failure (CHF). The aim of the study was to assess associations between copeptin, New York Heart Association (NYHA) class, and the location of the AVF in hemodialysis (HD) patients. The cross-sectional study was performed on a cohort of 93 clinically stable HD patients. Patients with proximal AVF tend to be older, with decreased renal residual function and increased NYHA functional class. These patients were also highly anemic, had more acidosis, and had increased high-sensitivity C-reactive protein along with increased copeptin and NT-proBNP levels. These changes were also associated with significant changes in all intra-cardiac dimensions, including right ventricle, both atria, and intraventricular septum and increase in end-systolic and end-diastolic left ventricular intra-cardiac dimensions. In multiple logistic regression analysis, the only associate of copeptin was NYHA functional class. Copeptin level in HD patients depends on cardiac function and it might be involved in the pathophysiology of cardiovascular disease in these patients. Proximal AVF creation might contribute to the development or worsening of CHF in HD patients.

• copeptin
• hemodialysis
• AV fistula
• chronic heart failure

INTRODUCTION

Cardiovascular disease is a major contributor to the mortality and morbidity in patients with chronic renal failure.¹ Chronic kidney disease (CKD) is common in patients with congestive heart failure (CHF), reaching up to 34.3% in the study of Gotsman et al.² Its presence adversely influences patients’ survival.^3,4 Although the presence of an arteriovenous fistula (AVF) has an adverse effect on cardiac function, the exact role of the vascular access in contributing to cardiovascular morbidity is unclear. Creation of an AVF has significant cardiovascular effects, particularly increased blood flow and cardiac output.⁵ It should be stressed that CHF is common in patients with CKD.⁶ Creation of AVF might contribute to the development or worsening of CHF.⁷

Copeptin is cosynthesized with vasopressin, also known as anti-diuretic hormone, thereby directly mirroring vasopressin levels, although copeptin levels are more stable in plasma and serum.⁸ In the past 2 years, copeptin has been studied as both a diagnostic and a prognostic marker in different diseases.⁹

In patients with decompensated heart failure, copeptin is an accurate prognostic marker for mortality.^10,11 The aim of the present study was to assess the association between copeptin levels, New York Heart Association (NYHA) class, and location of AVF in chronic hemodialysis (HD) patients.

PATIENTS AND METHODS

The study was performed on 93 clinically stable chronic HD patients (age range 26–82 years, 47 males) from one center. All the patients underwent regular HD for 4–5 h, three times per week for a mean duration of 48 ± 35 months (range 8–179 months). Blood flow was usually 180–280 mL/min with a dialysate flow of 500 mL/min. All the patients were dialyzed using low-flux dialyzers with bicarbonate-buffered dialysate. Enoxaparin (Clexane, Aventis, Paris, France) was given as an anticoagulant during HD, as a single intravenous injection at the beginning of each dialysis session.

The causes of renal failure among HD patients included chronic glomerulonephritis (n = 28), diabetic nephropathy (n = 21), chronic interstitial nephritis (n = 12), polycystic kidney disease (n = 12), amyloidosis (n = 7), and others or unknown (n = 13).

Inclusion criteria were no thrombosis, C-reactive protein (CRP) within normal range (below 6 mg/L, using standard semiquantitative laboratory method), controlled hypertension, and less than twice of normal alanine aminotransferase and aspartate aminotransferase levels (upper range 45 IU/L).

Seventy-three patients were treated only with recombinant human erythropoietin alfa and 67 with antihypertensive drugs. Forty-four patients were anuric and 61 had residual renal function (median urine output of 1550 mL, range 500–3100 mL/day). All the patients were tested for the HBV and anti-HCV antibodies: 20 were anti-HCV positive, 10 were HBV positive, and 4 were positive for both HBV and anti-HCV.

In all HD patients, blood was drawn both in the morning between 8.00 and 9.00 am before the onset of midweek dialysis session (and before heparin administration) and after HD from the arterial line of HD system, immediately before discontinuation of the extracorporeal circulation (only for urea concentration necessary for Kt/V determination, used as a marker of adequacy of dialysis; mean Kt/V was 1.18 ± 0.19). Blood samples were aliquotted and serum was stored at −40°C before assay.

Hemoglobin, hematocrit, platelet count, fibrinogen, total protein, cholesterol, low-density lipoprotein (LDL), high-density lipoprotein, triglycerides, albumin concentration, bicarbonates, and CRP (for screening purposes) were measured by standard laboratory methods. Copeptin was assessed using commercially available kits from Phoenix Pharmaceuticals, Inc. (Burlingame, CA, USA) and high-sensitivity C-reactive protein (hsCRP) was assessed using commercially available kits from American Diagnostica (Greenwich, CT, USA).

Echocardiography b-mode was performed in each patient. All the patients were informed about the aim of the study and gave their consent. The study was approved by the local Medical University Ethic Committee.

The various clinical, laboratory, and cardiac parameters were compared between HD patients with distal forearm AVF (n = 72) and patients with proximal arm AVF (n = 21).

Statistical Analysis

Data were analyzed using Statistica 6.0 computer software (Tulsa, OK, USA). Data are expressed as mean ± SD or median (minimum and maximum values). The examination of the distribution normality of variables was done using Shapiro–Wilk W-test. Comparisons between groups were done by either t-test or Mann–Whitney test, as appropriate. Correlations between copeptin and other variables were evaluated by Pearson’s or Spearman’s test, as appropriate. Values of p < 0.05 were taken as statistically significant. Multiple regression analysis was used to determine independent factors affecting dependent variable. Factors showing linear correlation with copeptin (p < 0.1) were included in the multiple regression analysis.

Table 1. Clinical characteristics of hemodialysis patients with proximal and distal AVF.

	Proximal AVF	Distal AVF
Age (years)	63.04 ± 12.55	57.21 ± 13.61
Diabetes mellitus (%)	33	20
Ischemic heart disease (%)	49	26*
Hypertension (%)	76	75
SBP (mmHg)	131.6 ± 17.6	134.7 ± 17.4
DBP (mmHg)	88.9 ± 9.3	89.6 ± 9.8
RRF (mL)	320 (0.1100)	821 (0.3100)*
Time on HD (months)	25.56 ± 11.19	58.10 ± 45.54*
Kt/V	1.14 ± 0.24	1.22 ± 0.25
BMI (kg/m^2)	24.67 ± 3.82	25.03 ± 4.03
ACE inhibitors (%)	69	61
Beta-blockers (%)	31	34
Calcium channel blockers (%)	42	43
AT1 receptor blockers (%)	8	0*
Diuretics (%)	46	15*
Statins (%)	15	8
Nitrates (%)	35	23
Acetylsalicylic acid (%)	19	7

Notes: Data are given as percent, mean ± SD, *p < 0.05, or median and ranges, proximal versus distal AVF. SBP, systolic blood pressure; DBP, diastolic blood pressure; RRF, residual renal function; BMI, body mass index.

RESULTS

Clinical Characteristics

Basal clinical characteristics of the studied groups are presented in Table 1. The patients with proximal AVF tend to be older, although this difference was not statistically significant. The prevalence of diabetes mellitus and ischemic heart disease was higher in the patients with proximal AVF. There was no difference between both groups in both the prevalence and the control rates of hypertension. The patients with proximal AVF had lower residual renal function, although they had a significantly shorter duration of HD treatment. Dialysis adequacy was similar between both groups. BMI was also similar between both groups.

Table 2. Laboratory and echocardiographic parameters of hemodialysis patients with proximal and distal AVF.

	Proximal AVF	Distal AVF
Hemoglobin (g/dL)	11.47 ± 1.28	12.22 ± 1.18*
Hematocrit (%)	34.14 ± 3.10	35.97 ± 3.71*
Total cholesterol (mg/dL)	188.86 ± 42.30	169.52 ± 48.73
HDL (mg/dL)	47.28 ± 17.40	44.37 ± 11.86
LDL (mg/dL)	110.45 ± 38.22	101.24 ± 33.07
Triglycerides (mg/dL)	176.09 ± 71.66	148.37 ± 69.27
Total protein (g/dL)	6.45 ± 0.37	6.67 ± 0.84*
Albumin (g/dL)	3.97 ± 0.32	4.06 ±0.47
hsCRP (mg/L)	7.75 ± 5.45	3.91±3.15^**
pH	7.34 ± 0.06	7.37±0.06
Bicarbonate (mmol/L)	19.56 ± 4.02	21.01 ± 2.25*
Copeptin (ng/mL)	1.32 ± 0.43	1.11 ± 0.22*
NT-proBNP (pg/mL)	9608 ± 5145	7608 ± 2221^**
NYHA class	2.16 ± 1.02	1.81 ± 0.76*
EF (%)	58.20 ± 8.23	59.71 ± 7.82
LVIDd (mm)	52.10 ± 7.88	51.00± 4.80
LVISd (mm)	37.36 ± 1.86	35.00 ± 1.68*
RV (mm)	27.71 ± 4.70	24.29 ± 1.81*
LA (mm)	44.44 ± 5.62	40.52±6.06*
RA (mm)	51.50 ± 8.82	44.00 ± 5.28*
IVS (mm)	14.31 ± 1.91	12.71 ± 1.84*

Notes: Data are given as mean ± SD, *p < 0.05, ^**p < 0.01, proximal versus distal AVF. HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, high-sensitivity C-reactive protein; EF, ejection fraction; LVIDd, left ventricular internal end-diastolic dimension; LVISd, left ventricular internal end-systolic dimension; RV, right ventricle; LA, left atrium; RA, right atrium; IVS, interventricular septum diastole dimension.

Laboratory and Echocardiographic Parameters

The laboratory parameters of both studied groups are presented in Table 2. The patients with proximal AVF had significantly lower hemoglobins and hematocrit levels. The lipids profile and serum albumin were similar in both groups. The patients with proximal AVF had more significant metabolic acidosis compared to the patients with distal AVF. hsCRP levels were significantly higher in patients with proximal AVF. Copeptin and NT-proBNP were also significantly higher in those patients with proximal AVF. The NYHA functional class was significantly higher in patients with proximal AVF. This was not associated with a significant difference in the left ventricular ejection fraction of the patients in both groups; however, it was related to a significant elevation in all intra-cardiac structure dimensions, including the right ventricle, both left and right atria, and the internal end-systolic and end-diastolic dimensions of left ventricle and the interventricular septum diastole dimension.

Copeptin Correlations to Clinical, Laboratory, and Cardiac Parameters

Copeptin was related to the presence of CHF (r = −0.21, p < 0.05), NYHA class (r = −0.20, p < 0.05), total cholesterol (r = 0.28, p < 0.05), LDL (r = 0.30, p < 0.01), triglycerides (r = 0.28, p < 0.05), and NT-proBNP (r = 0.20, p < 0.05). NT-proBNP was related to LDL (r = 0.25, p < 0.05), left atrium dimension(r = −0.34, p < 0.01), and pH (r = −0.28, p < 0.01).

Copeptin was significantly higher in patients with NYHA class I/II (n = 64) compared to patients with NYHA class III/IV (n = 29) (Figure 1), whereas NT-proBNP was significantly lower in patients with NYHA class I/II than in patients with NYHA class III/IV(Figure 2). In multiple logistic regression analysis, the only associate of copeptin were NYHA class (β = −0.29, p = 0.22) and HR (0.39 (0.01–0.77), p < 0.05).

Figure 1. Copeptin in HD patients: copeptin in patients with NYHA class I/II in comparison to NYHA class III/IV.

Note: ^**p < 0.01.

Figure 2. NT-proBNP in HD patients: NT-proBNP in patients with NYHA class I/II versus NYHA class III/IV.

Note: ^**p < 0.01.

Copeptin did not differ significantly between female and male patients. Similarly, diabetic and hypertensive patients showed similar copeptin levels to their nondiabetic and normotensive counterparts in both groups of HD patients. Copeptin was similar in patients with and without residual renal function. In univariate analysis, copeptin was not related to residual renal function. Prevalence of history of myocardial infarction and stroke was similar in patients with proximal and distal AVF(8% vs. 3%, not significant (NS) and 15% vs. 13%, NS, respectively). In the group with distal AVF only one patient had persistent atrial fibrillation, and in the group with proximal AVF one patient had persistent atrial fibrillation and one had AV block I°.

We found that patients dialyzed on the polysulfone dialyzers (n = 61, F5 and F6, Fresenius Medical Care, Bad Homburg, Germany) had a statistically significant lower copeptin relative to the other types of dialyzers used (GFE 09, GFS plus, Althin Ultra Nova 140, AV160S, BL621, FB130T) (0.96 ± 0.27 vs. 1.16 ± 0.32 ng/mL, p < 0.05).

DISCUSSION

Creation of an AVF has significant effects on both systolic and diastolic functions. It is associated with increased blood flow, pulmonary hypertension, and increased cardiac output. Although usually these changes are hemodynamically insignificant, the increased cardiac output and blood flow may be so great that they could result in overt cardiac failure. Patients with severe heart failure with NYHA functional class IV, due to intrinsic heart disease, are at considerably much higher risk for worsening of the cardiac function after the creation of an AVF; in general they are not good candidates for proximal AVF. This risk is greater for a proximal arm fistula than with one located in the forearm.^7,8 However, it should be stressed that data on the possible relation between cardiovascular system and AVF are limited; often case reports are presented. Therefore, individualized approach should be employed toward every patient at high or very high risk; in certain cases permanent catheter could be an option. Close monitoring of cardiovascular status should also be considered in such patients.

Till date, only scarce data are available regarding the location of the AVF to NYHA functional class and other intra-cardiac structural parameters. In the present study, we demonstrate that presence of the proximal AVF is associated with clinical, laboratory, and cardiac structural changes. A proximal AVF was associated with increased NYHA functional class, decreased hemoglobin levels, more significant metabolic acidosis, along with increased hsCRP and elevated copeptin and NT-proBNP levels. The creation of a proximal AVF was also associated with intra-cardiac structural changes, probably leading to the increased NYHA functional class in those patients. However, without prospective studies we could not answer the questions whether copeptin could be an appropriate marker to those changes after an AVF creation and whether copeptin is better than NT-proBNP in HD patients.

In the present study, we have also demonstrated that copeptin levels are higher in HD patients, in correlation with the NYHA functional class. Our data on the possible relations between kidney function and increased NT-proBNP, and copeptin levels, together with higher NYHA class, could support previous reports on the relation between CKD and CHF.^4,12,13 In a recent study by Bhandari et al.¹⁴ that included 706 healthy volunteers, copeptin levels were significantly higher in males, together with higher NT-proBNP and serum creatinine. Only in males, copeptin was related in univariant analysis to eGFR (modification of diet in renal disease formula), whereas in females copeptin was not significantly related to eGFR (r = −0.097, p = 0.095). In multivariant analysis, male gender and eGFR were independent predictors of copeptin levels in healthy volunteers. Fenske et al.¹⁵ found that in the HD patients from 4D study, copeptin correlated with markers of residual renal function, for example, dialysis vintage and diuretic treatment. In our study, we assessed residual renal function and found that copeptin was unrelated to it, neither to dialysis vintage nor to diuretic treatment, despite the fact that patients with distal AVF had higher residual diuresis and lower use of diuretics. However, in the study of Fenske et al.,¹⁵ copeptin was not associated with the risk of death caused by CHF in HD patients with type 2 diabetes from the 4D study. It could be explained by the low incidence of events. Moreover, no associations were detected for copeptin and future myocardial infarction. In patients with stable heart failure, in survival analysis, both elevated copeptin and hs-cTnT concentrations were significant predictors of outcome (p < 0.001 for both).¹⁶ The combination of both markers showed a graded and highly significant association with impaired clinical outcome, which was independent of plasma NT-proBNP levels (adjusted hazard ratio 1.40, 95% CI 1.20–1.70; p < 0.001). Authors suggested that adding copeptin concentrations to a prediction model with NT-proBNP and hs-cTnT resulted in significant improvement in model performance (net reclassification improvement 0.208; p < 0.05). In the study by Alehagen et al.¹⁷ on older patients with chronic heart failure, increased concentration of copeptin was associated with increased risk of all-cause mortality and cardiovascular mortality. In addition, the combination of elevated NT-proBNP concentrations and elevated copeptin concentrations also was associated with increased risk of all-cause mortality. In the BACH trial on 1641 patients with dyspnea, where 557 patients with acute heart failure were included in the analysis, copeptin was found to predict 90-day mortality, re-admissions, and visits in emergency department, especially in those with hyponatremia.¹⁸

In our previous study on heart and kidney allograft recipients, we reported that copeptin was higher in CKD stage 4 than in stage 2 and similarly in NYHA class III versus I. There was no effect of gender, diabetes, or hypertension on copeptin levels in either group of transplanted patients. Among the heart transplant population, copeptin is independently associated with kidney and heart function, but not in kidney allograft recipients. It may also predict outcomes of orthotopic heart transplant patients. Copeptin was positively related with deceleration time and left atrial size. In our previous study on heart transplant recipients, copeptin was not related to any echocardiographic parameters, except the intraventricular septum, which remained an independent predictor together with cystatin C of copeptin levels in this population.¹⁹

Khan et al.²⁰ reported that in patients with acute myocardial infarction, copeptin was related to age and eGFR, and copeptin was higher in males. In our study, copeptin levels were not significantly different between males and females. Natriuretic peptides (BNP and NT-proBNP) are considered as established diagnostic and prognostic markers for CHF. Gegenhuber et al.¹¹ showed a prognostic utility of copeptin comparable to BNP in acute heart failure. In addition, Stoiser et al.¹² demonstrated that copeptin was an even superior predictor of composite end points in advanced heart failure over BNP. They also reported that in patients after acute myocardial infarction, copeptin was elevated in patients who died, compared with survivors. Copeptin was thereby a significant independent predictor of death or heart failure within 60 days.¹² Of course, any biomarker will always oversimplify the interpretation of important variables, and therefore biomarkers are meant to complement, rather than to supersede, the judgment of clinicians and/or validated clinical severity scores. In the case of copeptin, renal dysfunction and probably gender should be taken into account when interpreting copeptin levels.

One of the limitations of this study is that the cohort of the prevalent HD patients consisted of entirely Caucasians and the prevalence of diabetes was relatively small. Therefore, our findings should be extrapolated with caution to other ethnic groups. Another limitation of the study is its cross-sectional design, as well as lack of measurement of blood flow in the AVF by Doppler method (lack of technical possibility).

Creation of an AVF in patients with already established cardiovascular complications may enhance the risk of decompensation of CHF, pulmonary hypertension, inadequate low blood flow in the AVF, with subsequent inadequate dialysis and AVF thrombosis, fistula infection (with local and systemic symptoms), and micro-inflammation. According to our results, it should be suggested to create a distal AVF, if it is possible, in most patients, especially in patients with established CHF and maybe also in patients with a basally increased copeptin levels. Prospective studies are needed to assess the effects of AVF creation on the cardiovascular system in HD patients.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.