Brain Natriuretic Peptide and Its Relationship to Left Ventricular Hypertrophy in Patients on Peritoneal Dialysis or Hemodialysis Less Than 3 Years

Abstract

An increase of brain natriuretic peptide (BNP) levels is commonly observed in patients on dialysis. Increased circulating levels of BNP are related to future cardiac events and associated with shorter survival in patients on chronic hemodialysis (HD). During the first 1 or 2 years on dialysis, patients on peritoneal dialysis (PD) have been shown to have an improvement in left ventricular hypertrophy, blood pressure, and volume status. This study compares BNP levels and cardiac status of PD and HD patients without cardiovascular disease and on dialysis for less than 36 months. The correlation between plasma BNP concentration and findings of echocardiography before HD scans were examined and compared with findings of PD. Twenty-two HD patients (15 men, 7 women; mean age, 52.5 ± 13.9 years) and 19 PD patients (10 men, 9 women; mean age, 47.6 ± 11.3 years) were studied. There were no significant differences between HD and PD patients with regard to age, gender, duration of dialysis, left ventricular mass, left ventricular mass index (p > 0.05). Plasma BNP levels were markedly greater in HD patients (467.8 ± 466.5 pg/mL) than those of PD patients (143.1 ± 165.2 pg/mL). Urine output was significantly higher in PD patients compared with HD patients (p < 0.05). A positive correlation between systolic blood pressure, diastolic blood pressure, and plasma BNP in HD patients (r: 0.653, p: 0.001; r: 0.493, p: 0.023, respectively) was detected. Additional studies are needed to investigate whether lower BNP level in PD patients is an advantage.

Keywords:

• brain natriuretic peptide (BNP)
• hemodialysis (HD)
• peritoneal dialysis (PD)
• left ventricular hypertrophy (LVH)

INTRODUCTION

Today, we know that the heart secretes two cardiac natriuretic peptides with a homologous structure: atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP). Cardiac myocytes constitute the major source of natriuretic peptides in circulation. Recently, cardiac fibroblasts have also been shown to produce BNP. In contrast to ANPs, which originate mainly from atrial tissue, BNP-related peptides are produced mainly from ventricular myocytes. The main stimulus for BNP peptide synthesis and secretion is cardiac wall stress. The biological effects of BNP include diuresis, vasodilatation, inhibition of renin and aldosterone production, and inhibition of cardiac and vascular myocyte growth. Because increased cardiac wall stress is a common denominator of many cardiac diseases, it follows that circulating natriuretic peptides may serve as clinical biochemical markers of these states.[1],[2]

BNP gene expression can increase very rapidly in response to an appropriate stimulus.[3] The half life of BNP is 22 min, and prior studies have established that BNP can accurately reflect pulmonary capillary wedge pressure changes every 2 h.[4] A significant role in adaptation to pathophysiological processes is suggested by the discovery of elevated levels of the hormones in a range of conditions characterized by elevated filling pressures, increased afterload, and or myocardial hypertrophy, in addition to the now well-recognized associations with overt heart failure and subclinical ventricular dysfunction.[5],[6] Plasma BNP concentrations have been found to be significantly related to mortality and to complement standard risk stratification by providing additional prognostic information beyond left ventricular ejection fraction (LVEF).[7],[8]

An increase of BNP is commonly observed in patients on hemodialysis (HD). An earlier report by Ishizaka et al.[9] showed that changes in plasma BNP levels before and after HD correlated with changes in cardiac index and left atrial diameter determined by echocardiography in a small number of patients. Nitta et al. showed that plasma BNP levels inversely correlated with LVEF determined by M-mode echocardiography in HD patients.[10]

Plasma BNP levels in HD patients directly correlated with the degree of body fluid retention. The left ventricular hypertrophy (LVH) and endothelial dysfunction in severe chronic renal failure, systolic and diastolic left ventricular dysfunction, and the associated cardiac disease (usually ischemic) also contribute to this increase.[11],[12]

It is well recognized that dialysis patients are at an increased risk of cardiovascular death.[13] This is due to complications of both atherosclerotic coronary artery disease and LVH. LVH is most often related to volume overload and elevated blood pressure (BP). Dialysis patients with LVH, left ventricular dilatation, or systolic dysfunction have a threefold increased risk of heart failure and a decreased survival rate.[14] The hemodynamic alterations associated with HD (fluid/volume shifts, increased pulse pressure due to arteriovenous fistulas) predispose to the development of LVH, whereas the steady-state environment of peritoneal dialysis (PD) resembles that of normal health and is less likely to lead to the development of LVH.[15]

In English literature, there are few studies comparing BNP levels in PD and HD patients. All these studies' BNP levels were found lower in PD patients than in HD patients.[11],[16],[17] Nakatani et al. reported that both continuous ambulatory peritoneal dialysis (CAPD) and HD patients had significantly higher BNP levels than those with normal renal functions and a positive correlation with left ventricular mass index (LVMI) and a negative correlation with LVEF in both CAPD and HD patients.[17]

The peaks and valleys of solute concentrations and total body volume associated with intermittent HD may be avoided with daily therapies such as PD. It was shown that during the first 1 or 2 years on dialysis, patients on PD have been shown to have an improvement in LVH, BP, and volume status.[18] PD has more consistent volume removal than HD. This study compares BNP levels and cardiac status of PD and HD patients without cardiovascular disease and on dialysis for less than 36 months, examines the correlation between plasma BNP concentration and parameters of echocardiography just before HD seans, and compares the findings with those on PD.

MATERIAL AND METHODS

Twenty-two HD patients (15 men, 7 women; mean age, 52.5 ± 13.9 years) and 19 PD patients (10 men, 9 women; mean age, 47.6 ± 11.3 years) on dialysis for longer than 2 months were studied.

Patients with cardiovascular diseases such as myocardial infarction, angina pectoris, valvular heart disease, and hypertrophic and dilated cardiomyopathy were excluded in this study by history, physical examination, and echocardiography.

All HD patients received regular dialysis using hemophane hollow-fiber dialyzer three times per week in sessions lasting 4 hours. The dialysate was bicarbonate buffered and contained 140 mmol/L sodium. The dialysate flow rate was 500 mL/min, and blood flow ranged from 300 to 350 mL/min. All patients on CAPD received four exchanges per day using standard dialysis bags (8 L/day).

Information on patient's demographics and clinical parameters, including type and prescription of peritoneal dialysis, adequacy of dialysis parameters, and list of current and prior use of medications, was obtained from each center's medical recordings and also from patients themselves using a case report form.

Causes of renal failure in HD patients included chronic glomerulonephritis (n=4), diabetes mellitus (n =6), hypertensive nephrosclerosis (n=4), polycystic kidney disease (n= 2), and unknown (n=6). Causes of renal failure in PD patients were chronic glomerulonephritis (n =5), diabetes mellitus (n=3), hypertensive nephrosclerosis (n=2), polycystic kidney disease (n=1), and unknown (n=8).

A high sensitivity assay for C-reactive protein (CRP) was used. Serum CRP was measured by nephelometry on a BNAII (Dade Behring, Liederbach, Germany). After 20 to 30 min of quiet resting in semirecumbent position, samples for BNP were taken into chilled ethylenediaminetetraacetic acid Vacutainers, placed immediately on ice, and centrifuged within 30 min at −4°C. The plasma levels of BNP were determined using microparticle enzyme immunoassay (MEIA) on an Axsym system (Abbott Diagnostics, Chicago, IL, USA).

Echocardiograms were recorded in all patients before the HD session by an experienced cardiologist using a ATL HDI 5000 echocardiograph. Echocardiographic measurements were carried out by a cardiologist according to the recommendations of the American Society of Echocardiography.[19]

All M-mode studies of the left ventricular (LV) cavity were guided by two-dimensional echocardiography. Measurements of interventricular septal thickness, posterior wall thickness, and LV internal dimensions were made at end-diastole (LVDd) and end-systole (LVDs), in accordance with the recommendations of the American Society of Echocardiography.[19] LVDd, LVDs, interventricular septal thickness (IVST), and posterior wall thickness (PWT) were measured, and LVEF was calculated by standard techniques. Left ventricular mass (LVM) was calculated by the regression equation. Described by Devereux and Reichek,[20] LVM = 1.04 × [(IVST + PWT + LVDd)³ − LVDd³] − 13.6 was divided by body surface area to determine the LVMI, presented in g/m². LVH was defined as LVMI >114 g/m² for men and >106 g/m² for women, the criteria of LVH in a Japanese study.[6]

Blood pressure was determined using a standard mercury sphygmomanometer and cuffs adapted to arm circumference. The systolic blood pressure was taken as the point of appearance of Korotokoff sounds, and the diastolic blood pressure was taken as the point of disappearance of the sounds.

STATISTICAL ANALYSES

All results are expressed as mean ± SD and range, where appropriate. Associations between categorical variables and serum BNP levels were examined by using Pearson's chi-square tests. Associations with variables were examined with the use of two-tailed Fisher's exact tests and with Mann-Whitney rank-sum tests.

RESULTS

There were no significant differences between HD and PD patients with regard to age, gender, dialysis duration, body weight, BMI, CRP, homocysteine levels, systolic and diastolic blood pressure, ejection fraction, LVM, and LVMI (p > 0.05). Demographic and laboratory characteristics of all patients are shown in Table 1. Plasma BNP levels were markedly greater in HD patients (467.8 ± 466.5 pg/mL) than those of PD patients (143.1 ± 165.25 pg/mL). Urine output was significantly higher in PD compared with HD patients (p< 0.05). 54.5% of HD patients had no any urine output.

Table 1. Demographic and laboratory parameters of patients

	Hemodialysis (n:22)	Peritoneal dialysis (n:19)
	Mean ± std. deviation	Mean ± std. deviation	P
Gender (female/male)	7/15	9/10	p > 0.05
Age (years)	52.1 ± 13.9	47.6 ± 11.3	p > 0.05
Duration of dialysis (month)	20.6 ± 11.2	15.4 ± 11.6	p > 0.05
Weight (kg)	67.6 ± 14.4	67.6 ± 11.5	p > 0.05
BMI	25.2 ± 4.9	26.3 ± 4.9	p > 0.05
BNP (pg/mL)	462.8 ± 466.5	143.1 ± 165.2	0.000
Albumin (g/dL)	3.8 ± 0.6	3.5 ± 0.4	0.005
CRP (mg/L)	8.6 ± 13.2	13.3 ± 15.8	p > 0.05
Hb (g/dL)	10.6 ± 1.7	11.1 ± 1.5	p > 0.05
Homocysteine (umol/L)	24.2 ± 10.5	19.9 ± 8.9	p > 0.05
Ef	63.5 ± 7.1	64.1 ± 8.6	p > 0.05
SBP (mmHg)	129.6 ± 22.0	123.9 ± 15.8	p > 0.05
DBP (mmHg)	83.0 ± 10.4	76.6 ± 10.1	p > 0.05
LVM (g)	251.9 ± 90.82	243.0 ± 75.6	p > 0.05
LVMI (g/m^2)	98.6 ± 30.2	92.8 ± 26.1	p > 0.05

BMI, body mass index; CRP, C-reactive protein; Hb, hemoglobin; ef, ejection fraction; LVM, left ventricular mass; LVMI, left ventricular mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Our study also found a positive correlation between systolic blood pressure (SBP), diastolic blood pressure (DBP), and plasma BNP in HD patients (r: 0.653, p: 0.001; r: 0.493, p: 0.023, respectively). In PD, BNP levels were not correlated with SBP, and DBP (p > 0.05). BNP levels were not correlated with BMI, duration of PD, urine output, hemoglobin, serum albumin, CRP, ejection fraction, LVM, and LVMI in both groups (p > 0.05).

LVH was detected in 8 (36.4%) of 22 HD patients and in 4 (21.1%) of 19 PD patients, and the difference was not statistically significant (p > 0.05). Demographic and laboratory characteristics of PD patients with and without LVH were shown Table 2. Table 3 shows demographic and laboratory characteristics of HD patients with and without LVH (p < 0.05). In both HD and PD patients with LVH, SBP and DBP were significantly higher than those of the patients without LVH.

Table 2. Demographic and laboratory parameters of PD patients with and without LVH

	PD patients without LVH (n:15)	PD patients with LVH (n:4)
	Mean ± std. deviation	Mean ± std. deviation	p
Gender (female/male)	8/7	1/3
Age (years)	45.3 ± 11.1	56.5 ± 7.9	p > 0.05
Duration of dialysis (months)	13.4 ± 10.6	23.2 ± 13.6	p > 0.05
Weight (kg)	66.1 ± 11.2	69.2 ± 14.0	p > 0.05
BMI	25.4 ± 4.6	29.6 ± 4.7	p > 0.05
BNP (pg/mL)	110.4 ± 138.8	284.6 ± 228.1	p > 0.05
Albumin (g/dL)	3.5 ± 0.4	3.6 ± 0.1	p > 0.05
CRP (mg/L)	14.4 ± 17.4	9.2 ± 8.5	p > 0.05
Hb (g/dL)	11.2 ± 1.6	10.5 ± 0.8	p > 0.05
Homocysteine (umol/L)	19.5 ± 9.5	21.5 ± 6.7	p > 0.05
SBP (mmHg)	120.1 ± 15.7	137.5 ± 5.0	0.046
DBP (mmHg)	74.2 ± 9.9	85.0 ± 5.7	0.035
ef	63.8 ± 9.1	66.0 ± 2.8	p > 0.05
LVM (g)	216.3 ± 48.7	343.1 ± 79.2	0.001
LVMI (g/m^2)	82.8 ± 17.8	130.1 ± 15.8	0.002

LVH, left ventricular hypertrophy; BMI body mass index; CRP, C-reactive protein; Hb, hemoglobin; ef, ejection fraction; LVM, left ventricular mass; LVMI, left ventricular mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 3. Demographic and laboratory parameters of HD patients with and without LVH

	HD patients without LVH (n: 15)	HD patients with LVH (n: 8)
	Mean ± std. deviation	Mean ± std. deviation	P
Gender	3/11	4/4
Age (years)	50.3 ± 14.9	56.3 ± 11.8	p > 0.05
Duration of dialysis	21.1 ± 12.9	19.6 ± 8.0	p > 0.05
Weight (kg)	64.6 ± 14.0	72.9 ± 14.5	p > 0.05
BMI	23.8 ± 4.6	27.6 ± 4.8	p > 0.05
BNP (pg/mL)	307.1 ± 120.5	774.3 ± 724.0	p > 0.05
Albumin (g/dL)	3.8 ± 0.7	3.8 ± 0.3	p > 0.05
CRP (mg/L)	9.9 ± 15.8	6.6 ± 8.0	p > 0.05
Hb (g/dL)	10.7 ± 2.0	10.2 ± 1.0	p > 0.05
Homocysteine (umol/L)	21.1 ± 6.5	30.2 ± 14.8	p > 0.05
SBP (mm/Hg)	121.8 ± 18.4	143.1 ± 22.2	0.042
DBP (mm/Hg)	79.0 ± 9.8	90.0 ± 7.5	0.020
ef	66.1 ± 4.2	59.0 ± 9.2	0.040
LVM (g)	201.7 ± 53.1	339.8 ± 75.0	0.000
LVMI (g/m^2)	81.9 ± 20.9	127.9 ± 19.7	0.000

LVH, left ventricular hypertrophy; BMI, body mass index; CRP, C-reactive protein; Hb, hemoglobin; ef, ejection fraction; LVM, left ventricular mass; LVMI, left ventricular mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Fourteen (63.6%) of 22 HD patients were on antihypertensive treatment: 7(31.7%) on monotherapy with angiotensin-converting enzyme inhibitors, 3(13.6%) on calcium channel blockers, and 3(13.6%) on double or triple therapy with various combinations of these drugs. Six (31.6%) of 19 PD patients were on antihypertensive treatment: 4(21.1%) on monotherapy with angiotensin-converting enzyme inhibitors, and 2(10.5%) on calcium channel blockers. Antihypertensive drug usage was more common in HD patients (p < 0.05).

DISCUSSION

In our study, BNP levels of both HD and PD patients (467.8 ± 466.5 pq/mL; 143.1 ± 165.2 pq/mL, respectively) were in the upper limit of the normal range (<100 pg/mL). In individuals with normal renal function, congestive heart failure is usually absent at BNP levels less than 100 pg/mL, possible between 100 and 500 pg/mL, and probable at levels greater than 500 pg/mL. BNP levels between 100 and 500 pg/mL may also be seen in patients with left ventricular dysfunction, lung disease, renal failure, myocardial infarction, or pulmonary embolism.[21] Cardiac natriuretic peptide levels are frequently raised in dialysis patients. The high plasma concentration of BNP in dialysis patients is multifactorial and depends on extracellular volume expansion, concomitant heart disease, and abolished renal clearance.[22],[23]

Our study revealed that plasma BNP concentration was significantly lower in CAPD patients than in HD patients. We studied BNP levels in HD patients when they are on peak volume status, just before HD. There were no statistically significant difference between PD and HD patients with regard to duration of age, gender, duration of dialysis, DBP, SBP, hemoglobin, and homocysteine levels. Whereas 14 (63.6%) HD patients were on antihypertensive therapy, 6 (31.6%) PD patients were on antihypertensive therapy, and the difference was statistically significant (p < 0.05). We found a statistically significant correlation with BNP, SBP, and DBP in HD patients. Urine output was significantly higher in PD patients compared with HD patients. 54.5% of HD patients had no any urine output. It is known that in dialysis patients the main factor for hypertension is volume load. Expansion of extracellular volume causing myocardial stretching and increased left ventricular pressures may be the principal cause of increased BNP in HD patients.

In addition to glomerular filtration, BNP is eliminated from plasma mainly through natriuretic peptide receptors and degraded by neutral endopeptidases.[24] We did not measure levels of BNP in the dialysate and any residual urine. However, the removal of the metabolic or excretory function of the kidney would not expected to cause a significant rise in BNP concentration.[25]

A 68% increase in circulating BNP levels has been demonstrated over the 10 days after creation of an arteriovenous (AV) fistula for HD, which suggests that peripheral AV shunting itself has an effect on the cardiac production of BNP.[26]

Fagugli et al. investigated the relationship between plasma BNP and volume load by using bioimpedance. These authors reported an association between BNP and extracellular volume, and indicated that the increased synthesis and release of BNP from the LV in HD patients appeared to be mainly related to volume stress rather than to pressure load.[27]

Goto et al. demonstrated that increased circulating levels of BNP are associated with a high occurrence of cardiac events in patients with chronic hemodialysis without any other signs of cardiac dysfunction.[28] In their study, Kaplan-Meier analysis demonstrated the incidence of cardiac events with higher levels of BNP than in those with lower levels of each peptide.

We could not find any significant association with BNP and LVMI and LVM in patients on dialysis for less than 36 months. Ishii et al. did not find a significant association of BNP with either an overall or cardiac mortality within 2 years in patients on chronic dialysis.[29]

Nakatani et al. reported a positive correlation with LVMI and a negative correlation with LVEF in both CAPD and HD patients. Plasma BNP concentration was significantly lower in CAPD patients than in HD patients.[17] In this study, duration of dialysis was longer than those of patients in our study.

It is known that intracardiac distension is the main regulator of BNP.[1],[2] Thus, in chronic dialysis patients, intracardiac distension might be less associated with cardiac problems within 3 years. Higher urine output could play a role on lower BNP levels in PD versus HD.

As a conclusion, BNP levels of HD, just before HD, were significantly higher than those of PD patients. We could not find any relationship among LVM, LVMI, and BNP levels of patients on dialysis for less than 36 months. In HD patients, higher BNP levels were associated with higher BP. Limitations of our study include the single center nature of its design and small sample size, which was not adequately powered to find smaller, clinically meaningful changes in BNP. It was reported that increased circulating levels of ANP or BNP are related to future cardiac events in patients with chronic hemodialysis.[30] Zoccali et al. reported that raised plasma BNP levels are associated with shorter survival.[30] Additional studies are needed to test the influence of dialysis modality on plasma concentrations of BNP and to elucidate the interdependence of the production, release, and elimination of these peptides in dialysis treatment and relationship mortality.