Clinical Science (1992) 82, 127-
I3 I (Printed in Great Britain)
I27
Effect of haemodialysis on plasma levels of brain natriuretic peptide in patients with chronic renal failure Chim C. LANG', Anna Maria J. CHOY', lain S. HENDERSON*, Wendy J. COUTIE' and Allan D. STRUTHERS' 'Department of Clinical Pharmacology and 2Renal Unit, Ninewells Hospital and Medical School, Dundee, U.K. (Received
5 June/ I 9 July I99 I; accepted 6 September I99 I)
1. Plasma human brain natriuretic peptide-like immunoreactivity (hBNP-li) was measured in ten patients with chronic renal failure before and after 4 h of haemodialysis. 2. Plasma hBNP-li was elevated in all patients before dialysis (mean k SEM 2 1.0 k 3.8 pmol/l) compared with healthy control subjects (1.3 k 0.2 pmol/l, n = 1l ) , but showed considerable inter-patient variability. Before dialysis plasma hBNP-li bore no relationship to the serum creatinine level or to the mean blood pressure. 3. Plasma hBNP-li fell significantly ( P = 0.04) during 4 h of haemodialysis. The fall in plasma hBNP-li correlated significantly with the degree of postural blood pressure drop (9= 0.44, P = 0.05) and with the fall in body weight (9= 0.64, P < 0.01) after haemodialysis. In all patients, plasma hBNP-li at the end of treatment remained above that in healthy subjects. 4. There was no significant correlation between the fall in plasma hBNP-li and the fall in serum creatinine level, and between the fall in plasma hBNP-li and the fall in supine systolic or diastolic blood pressure, during haemodialysis. 5. We have shown that plasma hBNP-li is elevated in patients with chronic renal failure and is decreased during haemodialysis. The fact that the plasma hBNP-li was not reduced to normal by haemodialysis despite restoration to normovolaemia gives tentative support to the view that, in addition to hypervolaemia, another factor may also be responsible for the elevated plasma hBNP-li seen in these patients.
INTRODUCTION The discovery of atrial natriuretic peptide (ANP) by de Bold ef al. [l]in 1981 has led inevitably to a search for other peptides with natriuretic activity. Recently, Sudoh et al. [2]isolated a new natriuretic peptide from the pig brain and named it brain natriuretic peptide (BNP). BNP has striking similarity to ANP both in its amino acid sequence
and its biological actions [2-41. BNP has two forms with either 26 or 32 amino acid residues called porcine (p) BNP-26 and pBNP-32, respectively [2, 51. BNP, originally isolated from the brain, is also synthesized in and secreted from the heart in the pig [6]and rat [7]. Until recently, information on BNP in man has been scarce, mainly for lack of cross-reactivity of human BNP (hBNP) with antisera against pBNP or rat BNP (rBNP).However, the complementary DNA sequence encoding for the hBNP precursor, hBNP (1-108), was recently elucidated [8]. Since then hBNP has been isolated from the human atrium and found to comprise 32 amino acid residues, which are identical to the sequence [77-1081 of the hBNP precursor [9]. Furthermore, a hBNP-like immunoreactivity (hBNP-li) has been detected in plasma of healthy man at concentrations of 0.9 pmol/l and 1.5 pmol/l in two separate studies [lo, 111.Like ANP, plasma hBNP-li was found to be elevated in patients with congestive heart failure and chronic renal failure [ 111. However, there has been no information available on the release of BNP in man. To assess the effect of extracellular fluid expansion on circulating plasma hBNP-li, we have measured plasma hBNP-li in volume-overload chronic renal failure patients. We have also studied the effect of fluid removal by haemodialysis on plasma hBNPli in these patients.
SUBJECTSAND METHODS Ten patients (seven males) were studied. All had chronic renal failure and had been on conventional haemodialysis three times a week for at least 3 months at Ninewells Hospital and Medical School. They were aged 16-73 years (mean 47 years). The underlying diagnosis were glomerulonephritis (three), diabetes mellitus (two), polycystic renal disease (two), hydronephrosis (one), hypertension (one) and unspecified (one). All were taking phosphate binders; for hypertension two patients were receiving both /?-adrenoceptor blockers and calcium antagonists; one was receiving nifedipine alone and one
Key words: aldosterone, atrial natriuretic peptide, brain natriuretic peptide, chronic renal failure, haemodialysis. Abbreviations: ANP, atrial natriuretic peptide; ANP-Ii, atrial natriuretic peptide-like immunoreactivity; BNP, brain natriuretic peptide; hBNP, human brain natriuretic peptide; hBNP-li, human brain natriuretic peptide-like immunoreactivity; pBNP, porcine brain natriuretic peptide; rBNP, rat brain natriuretic peptide; TFA, trifluoroacetic acid. Correspondence: Dr C. C. Lang, Department of Clinical Pharmacology, Ninewells Hospital and Medical School, Dundee DD I 9SY, U.K.
I28
C.C.Lang et al.
was receiving methyldopa alone. No patient had clinical evidence of cardiac failure. Pre-dialysis blood pressures were 151 k 9 mmHg systolic and 85 f7 mmHg diastolic (mean fSEM). All patients gave informed consent to the study, which was approved by the hospital ethical committee. All patients were haemodialysed for 4 h. During haemodialysis, they remained in the semi-recumbent position and did not eat or drink. Haemodialysis was carried out by using a Gambro AK-10 module and Gambro GF-120H capillary dialysers. The dialysate contained: Na', 145 mmol/l; K', 1.3 mmol/l; Ca2+, 1.25 mmol/l; Mg2+, 0.25 mmol/l; C1-, 109.3 mmol/l. Each patient was weighed, their supine blood pressure measured and a sample of venous blood (25 ml) was removed before and immediately after each haemodialysis treatment. A 2 min standing blood pressure was also measured after each haemodialysis treatment. Venous blood was collected into chilled tubes (volume; preservative) for measurement of plasma hBNP-li [ l o ml; EDTA (potassium salt) and 4000 kallikrein inhibitory units of aprotonin (Bayer Ltd, Newbury, Berks, U.K.)], plasma ANP-like immunoreactivity (ANP-li) (as for plasma hBNP-li) and serum sodium and creatinine level (5 ml; no preservative). All samples were placed on ice (except for those for determination of electrolytes, which were allowed to clot), centrifuged at 4°C and the plasma/ serum was separated off and stored at -20°C (electrolytes, ANP-li, creatinine) or - 70°C (hBNP-li)until assay. Blood pressure and heart rate were measured by a semiautomatic sphygmomanometer (Dinamap vital signs monitor 1846; Critikon, Tampa, FL, U.S.A.). All samples were analysed together in one assay run and for any given variable within 1 month of the start of the study. Plasma hBNP-li was measured by ria., as previously described by our group [12], after plasma extraction, using a commercially available r.i.a. kit (Peninsula Laboratories Inc, Belmont, CA, U.S.A.). Plasma samples (1 ml) were acidified with 1 ml of 0.1% trifluoroacetic acid (TFA)and passed through Sep-Pak C, cartridges (Amersham International, Bucks, U.K.). The cartridges were preactivated with buffer B (4 ml) containing 60% acetonitrile/O.l% TFA and buffer A containing 0.1% TFA. After each aliquot the cartridges were washed with 0.1% TFA (10 ml) and the BNP was eluted with 3 ml of 60% acetonitrile/O.l% TFA solution. The eluate was evaporated under vacuum and the precipitate was resuspended in 250 pl of BNP assay buffer. The anti-hBNP-32 antibody used was generated against synthetic hBNP-32 (Peninsula Laboratories Inc, Belmont, CA, U.S.A.). The antibody showed the following cross-reactivity: hBNP-32, 100%; rBNP-32, 0.04%; rBNP-45, pBNP-26, 0%; a-human ANP (1-28), O%, angiotensin 11, endothelin-1 and vasopressin, 0%. A portion (100 pl) of reconstituted BNP extract was incubated in duplicate with 100 pl of anti-hBNP-32 antibody which had been reconstituted with 13 ml of r.i.a. buffer, at 4°C overnight. Then '251-hBNP-32 (between 10 000 and 20 000 c.p.m) was added, mixed and incubated at 4°C overnight. Standard curves were constructed
with standard hBNP-32 (Peninsula Laboratories) in r.i.a. buffer over a concentration range of 0.5-128 pg/tube. Precipitation was with 100 pl of diluted goat anti-rabbit IgG serum and 1000 pl of diluted normal rabbit serum. Precipitates were allowed to form for 2 h at room temperature before 0.5 ml of r i a . buffer was added. The pellets were separated by centrifugation at 1700 g for 20 min, supernatant was removed and the radioactivity in the pellets was counted in a LKB Clini Gamma counter. Plasma hBNP-li was estimated by direct comparison with the standard curve and was corrected for percentage extraction. The intra-subject coefficient of variation was 5.8% and the limit of detection for this assay was 0.5 pmol/l. Plasma ANP-11 was also measured by r i a . (Amersham International) after plasma extraction by the method of Richards et al. [13]. The intra-subject and inter-subject coefficients of variation were 4.9% and 7.8%, respectively. The serum creatinine level was measured by the modulated Jaffe method on an auto-analyser (Cobas Bio, Roche Diagnostica, Basle, Switzerland).Serum electrolyte levels were measured by an internal caesium standard flame emission photometer (model 943; Instrumental Laboratory, Milan, Italy). The significance of differences was analysed by the Wilcoxon signed-rank test. Linear regression lines were obtained by the least squares method. Results are expressed as means fSEM. A P value of < 0.05 was considered as being statistically significant.
RESULTS Nine out of ten patients lost weight during haemodialysis (Table 1).This was accompanied by a significant fall in both systolic and diastolic blood pressures (Table 1).Both serum creatinine and potassium levels fell significantly during haemodialysis ( P < 0.0 1 and P < 0.0 1, respectively) but the serum sodium level remained unchanged (Table 1). Both mean plasma hBNP-li (21.0k3.8 pmol/l) and ANP-li (104.7 f 11.4 pmol/l) were significantly higher in patients with chronic renal failure than that in 11 healthy male subjects aged 20-23 years (plasma hBNP-li, 1.3 f0.2 pmol/l; plasma ANP-li, 5.6 k 1.7 pmol/l). Our Table I. Changes in body weight, blood pressure and serum sodium, potassium and creatinine levels before and after haemodialysis in patients with chronic renal failure (n= 10). Values are means ~ S E M Abbreviation: . NS, not significant.
Body wt. (kg) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Serum Nat level (mmolll) Serum K t level (mmolll) Serum creatinine levels (mmolll)
Before dialysis (n= 10)
After dialysis ( n = 10)
P value
64.9t4.5 I5 I *9 85 f 7 139.5k1.1 5.6t0.2 1058t I14
62.0k4.4 134+8 77t5 140.5+ 1.2 3.4 t0.1 428+46
(0.05
I3 I (Printed in Great Britain)
I27
Effect of haemodialysis on plasma levels of brain natriuretic peptide in patients with chronic renal failure Chim C. LANG', Anna Maria J. CHOY', lain S. HENDERSON*, Wendy J. COUTIE' and Allan D. STRUTHERS' 'Department of Clinical Pharmacology and 2Renal Unit, Ninewells Hospital and Medical School, Dundee, U.K. (Received
5 June/ I 9 July I99 I; accepted 6 September I99 I)
1. Plasma human brain natriuretic peptide-like immunoreactivity (hBNP-li) was measured in ten patients with chronic renal failure before and after 4 h of haemodialysis. 2. Plasma hBNP-li was elevated in all patients before dialysis (mean k SEM 2 1.0 k 3.8 pmol/l) compared with healthy control subjects (1.3 k 0.2 pmol/l, n = 1l ) , but showed considerable inter-patient variability. Before dialysis plasma hBNP-li bore no relationship to the serum creatinine level or to the mean blood pressure. 3. Plasma hBNP-li fell significantly ( P = 0.04) during 4 h of haemodialysis. The fall in plasma hBNP-li correlated significantly with the degree of postural blood pressure drop (9= 0.44, P = 0.05) and with the fall in body weight (9= 0.64, P < 0.01) after haemodialysis. In all patients, plasma hBNP-li at the end of treatment remained above that in healthy subjects. 4. There was no significant correlation between the fall in plasma hBNP-li and the fall in serum creatinine level, and between the fall in plasma hBNP-li and the fall in supine systolic or diastolic blood pressure, during haemodialysis. 5. We have shown that plasma hBNP-li is elevated in patients with chronic renal failure and is decreased during haemodialysis. The fact that the plasma hBNP-li was not reduced to normal by haemodialysis despite restoration to normovolaemia gives tentative support to the view that, in addition to hypervolaemia, another factor may also be responsible for the elevated plasma hBNP-li seen in these patients.
INTRODUCTION The discovery of atrial natriuretic peptide (ANP) by de Bold ef al. [l]in 1981 has led inevitably to a search for other peptides with natriuretic activity. Recently, Sudoh et al. [2]isolated a new natriuretic peptide from the pig brain and named it brain natriuretic peptide (BNP). BNP has striking similarity to ANP both in its amino acid sequence
and its biological actions [2-41. BNP has two forms with either 26 or 32 amino acid residues called porcine (p) BNP-26 and pBNP-32, respectively [2, 51. BNP, originally isolated from the brain, is also synthesized in and secreted from the heart in the pig [6]and rat [7]. Until recently, information on BNP in man has been scarce, mainly for lack of cross-reactivity of human BNP (hBNP) with antisera against pBNP or rat BNP (rBNP).However, the complementary DNA sequence encoding for the hBNP precursor, hBNP (1-108), was recently elucidated [8]. Since then hBNP has been isolated from the human atrium and found to comprise 32 amino acid residues, which are identical to the sequence [77-1081 of the hBNP precursor [9]. Furthermore, a hBNP-like immunoreactivity (hBNP-li) has been detected in plasma of healthy man at concentrations of 0.9 pmol/l and 1.5 pmol/l in two separate studies [lo, 111.Like ANP, plasma hBNP-li was found to be elevated in patients with congestive heart failure and chronic renal failure [ 111. However, there has been no information available on the release of BNP in man. To assess the effect of extracellular fluid expansion on circulating plasma hBNP-li, we have measured plasma hBNP-li in volume-overload chronic renal failure patients. We have also studied the effect of fluid removal by haemodialysis on plasma hBNPli in these patients.
SUBJECTSAND METHODS Ten patients (seven males) were studied. All had chronic renal failure and had been on conventional haemodialysis three times a week for at least 3 months at Ninewells Hospital and Medical School. They were aged 16-73 years (mean 47 years). The underlying diagnosis were glomerulonephritis (three), diabetes mellitus (two), polycystic renal disease (two), hydronephrosis (one), hypertension (one) and unspecified (one). All were taking phosphate binders; for hypertension two patients were receiving both /?-adrenoceptor blockers and calcium antagonists; one was receiving nifedipine alone and one
Key words: aldosterone, atrial natriuretic peptide, brain natriuretic peptide, chronic renal failure, haemodialysis. Abbreviations: ANP, atrial natriuretic peptide; ANP-Ii, atrial natriuretic peptide-like immunoreactivity; BNP, brain natriuretic peptide; hBNP, human brain natriuretic peptide; hBNP-li, human brain natriuretic peptide-like immunoreactivity; pBNP, porcine brain natriuretic peptide; rBNP, rat brain natriuretic peptide; TFA, trifluoroacetic acid. Correspondence: Dr C. C. Lang, Department of Clinical Pharmacology, Ninewells Hospital and Medical School, Dundee DD I 9SY, U.K.
I28
C.C.Lang et al.
was receiving methyldopa alone. No patient had clinical evidence of cardiac failure. Pre-dialysis blood pressures were 151 k 9 mmHg systolic and 85 f7 mmHg diastolic (mean fSEM). All patients gave informed consent to the study, which was approved by the hospital ethical committee. All patients were haemodialysed for 4 h. During haemodialysis, they remained in the semi-recumbent position and did not eat or drink. Haemodialysis was carried out by using a Gambro AK-10 module and Gambro GF-120H capillary dialysers. The dialysate contained: Na', 145 mmol/l; K', 1.3 mmol/l; Ca2+, 1.25 mmol/l; Mg2+, 0.25 mmol/l; C1-, 109.3 mmol/l. Each patient was weighed, their supine blood pressure measured and a sample of venous blood (25 ml) was removed before and immediately after each haemodialysis treatment. A 2 min standing blood pressure was also measured after each haemodialysis treatment. Venous blood was collected into chilled tubes (volume; preservative) for measurement of plasma hBNP-li [ l o ml; EDTA (potassium salt) and 4000 kallikrein inhibitory units of aprotonin (Bayer Ltd, Newbury, Berks, U.K.)], plasma ANP-like immunoreactivity (ANP-li) (as for plasma hBNP-li) and serum sodium and creatinine level (5 ml; no preservative). All samples were placed on ice (except for those for determination of electrolytes, which were allowed to clot), centrifuged at 4°C and the plasma/ serum was separated off and stored at -20°C (electrolytes, ANP-li, creatinine) or - 70°C (hBNP-li)until assay. Blood pressure and heart rate were measured by a semiautomatic sphygmomanometer (Dinamap vital signs monitor 1846; Critikon, Tampa, FL, U.S.A.). All samples were analysed together in one assay run and for any given variable within 1 month of the start of the study. Plasma hBNP-li was measured by ria., as previously described by our group [12], after plasma extraction, using a commercially available r.i.a. kit (Peninsula Laboratories Inc, Belmont, CA, U.S.A.). Plasma samples (1 ml) were acidified with 1 ml of 0.1% trifluoroacetic acid (TFA)and passed through Sep-Pak C, cartridges (Amersham International, Bucks, U.K.). The cartridges were preactivated with buffer B (4 ml) containing 60% acetonitrile/O.l% TFA and buffer A containing 0.1% TFA. After each aliquot the cartridges were washed with 0.1% TFA (10 ml) and the BNP was eluted with 3 ml of 60% acetonitrile/O.l% TFA solution. The eluate was evaporated under vacuum and the precipitate was resuspended in 250 pl of BNP assay buffer. The anti-hBNP-32 antibody used was generated against synthetic hBNP-32 (Peninsula Laboratories Inc, Belmont, CA, U.S.A.). The antibody showed the following cross-reactivity: hBNP-32, 100%; rBNP-32, 0.04%; rBNP-45, pBNP-26, 0%; a-human ANP (1-28), O%, angiotensin 11, endothelin-1 and vasopressin, 0%. A portion (100 pl) of reconstituted BNP extract was incubated in duplicate with 100 pl of anti-hBNP-32 antibody which had been reconstituted with 13 ml of r.i.a. buffer, at 4°C overnight. Then '251-hBNP-32 (between 10 000 and 20 000 c.p.m) was added, mixed and incubated at 4°C overnight. Standard curves were constructed
with standard hBNP-32 (Peninsula Laboratories) in r.i.a. buffer over a concentration range of 0.5-128 pg/tube. Precipitation was with 100 pl of diluted goat anti-rabbit IgG serum and 1000 pl of diluted normal rabbit serum. Precipitates were allowed to form for 2 h at room temperature before 0.5 ml of r i a . buffer was added. The pellets were separated by centrifugation at 1700 g for 20 min, supernatant was removed and the radioactivity in the pellets was counted in a LKB Clini Gamma counter. Plasma hBNP-li was estimated by direct comparison with the standard curve and was corrected for percentage extraction. The intra-subject coefficient of variation was 5.8% and the limit of detection for this assay was 0.5 pmol/l. Plasma ANP-11 was also measured by r i a . (Amersham International) after plasma extraction by the method of Richards et al. [13]. The intra-subject and inter-subject coefficients of variation were 4.9% and 7.8%, respectively. The serum creatinine level was measured by the modulated Jaffe method on an auto-analyser (Cobas Bio, Roche Diagnostica, Basle, Switzerland).Serum electrolyte levels were measured by an internal caesium standard flame emission photometer (model 943; Instrumental Laboratory, Milan, Italy). The significance of differences was analysed by the Wilcoxon signed-rank test. Linear regression lines were obtained by the least squares method. Results are expressed as means fSEM. A P value of < 0.05 was considered as being statistically significant.
RESULTS Nine out of ten patients lost weight during haemodialysis (Table 1).This was accompanied by a significant fall in both systolic and diastolic blood pressures (Table 1).Both serum creatinine and potassium levels fell significantly during haemodialysis ( P < 0.0 1 and P < 0.0 1, respectively) but the serum sodium level remained unchanged (Table 1). Both mean plasma hBNP-li (21.0k3.8 pmol/l) and ANP-li (104.7 f 11.4 pmol/l) were significantly higher in patients with chronic renal failure than that in 11 healthy male subjects aged 20-23 years (plasma hBNP-li, 1.3 f0.2 pmol/l; plasma ANP-li, 5.6 k 1.7 pmol/l). Our Table I. Changes in body weight, blood pressure and serum sodium, potassium and creatinine levels before and after haemodialysis in patients with chronic renal failure (n= 10). Values are means ~ S E M Abbreviation: . NS, not significant.
Body wt. (kg) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Serum Nat level (mmolll) Serum K t level (mmolll) Serum creatinine levels (mmolll)
Before dialysis (n= 10)
After dialysis ( n = 10)
P value
64.9t4.5 I5 I *9 85 f 7 139.5k1.1 5.6t0.2 1058t I14
62.0k4.4 134+8 77t5 140.5+ 1.2 3.4 t0.1 428+46
(0.05
