Interaction of intravenous atrial natriuretic factor with furosemide in patients with heart failure

Interaction of intravenous atrial natriuretic factor with furosemide in patients with heart failure Furosemide is frequently administered intravenously to patients with chronic heart failure. However, use of diuretics may cause neuroendocrine activation and by itself may not consistently afford diuresis. Atrial natriuretic factor (ANF) in pharmacologic doses is a vasodilator and has favorable neuroendocrinologic effects in patients with congestive heart failure. To examine whether exogenous ANF might enhance the effects of acute furosemide injection, we studied 14 patients with chronic stable heaR failure and measured the effects of the combination of ANF and furosemide on hemodynamics, neuroendocrine activation, and urine output. Eight patients were randomly assigned to receive placebo plus furosemide (1.3 mglkg intravenously). Six patients received ANF (2 rg/kg intravenously) plus furosemide at the same dose in a double-blind manner. The group receiving placebo plus furosemide exhibited a slight increase in mean arterial pressure (92 to 96 mm Hg; p < 0.03), systemic vascular resistance (1969 to 2271 dynes . set . cme5; p = 0.0007), and pulmonary capillary wedge pressure (22 to 24 mm Hg; p < 0.04) from baseline to 10 minutes. The group receiving ANF plus furosemide exhibited no change in mean arterial pressure and systemic vascular resistance from baseline to 10 minutes. Pulmonary capillary wedge pressure and mean pulmonary pressure were unchanged. In the group receiving placebo plus furosemide. Plasma renin activity and norepinephrine levels increased significantly from baseline to 10 minutes (plasma renin activity 12.3 t 5 to 31.0 + 10 ng/ml/hr; p < 0.01; plasma norepinephrine level 699 r 100 to 876 + 134 pg/ml; p < O.Oi), whereas in the group receiving ANF plus furosemide no significant increase was noted in neuroendocrinologic activity. However, ANF did not further augment the diuretic response to furosemide. We conclude that pretreatment with ANF may blunt the mild systemic vasoconstrictor response to intravenous furosemide. The ability of ANF to attenuate activation of the renin-angiotensin and sympathetic nervous systems in response to furosemide may explain the lack of vasoconstriction in the ANF-treated group. There appears to be no important augmentation of natriuresis or diuresis when bolus ANF and intravenous furosemide are administered concomitantly to patients with congestive heart failure, although more patients will need to be studied to more precisely define these effects. (AM HEART J 1994;127:392-99.)

Terence P. Connelly, MD, Gary S. Francis, MD, Karen J. Williams, RN, Alfred0 M. Beltran, MD, and Jay N. Cohn, MD Minneapolis, Minn.

Furosemide is commonly administered intravenously to induce diuresis in patients with congestive heart failure. We have previously shown that administration of intravenous furosemide may cause a slight in-

From the Cardiovascular Division, Minnesota Medical School. Supported Received

in part

by Scios,

for publication

Inc.,

Mountain

View,

Division, School,

Inc.

of

Calif. 7, 1993.

by Mosby-Year Book, + .lO 4/l/51124

accepted

University

Reprint requests: Gary S. Francis, MD, Cardiovascular ment of Medicine, University of Minnesota Medical UMHC, 420 Delaware St. SE, Minneapolis, MN 55455.

392

23, 1993;

of Medicine,

June

Copyright @ 1994 0002-8703/94/$1.00

April

Department

DepartBox 608

crease in mean arterial pressure and systemic vascular resistance and a decrease in cardiac output before the onset of diuresis in patients with chronic stable heart fai1ure.l These effects may be related to activation of the renin-angiotensin system and can be at least partially inhibited by the angiotensin-converting enzyme (ACE) inhibitor captopril.2 In addition, the plasma norepinephrine level increases acutely after the administration of intravenous fur0semide.l It is possible but not proved that furosemide-induced neuroendocrine activation may to some extent offset the diuretic and natriuretic responses in patients with heart failure. Atria1 natriuretic factor (ANF) levels are known to

Volume 127, Number 2 American Heart .iournal

Connelly et al.

393

Table I. Baseline characteristics Age (YT)

68 70 47 49 62 59 70 63 60 63 71 48 66 52 61 k 8

Heart

failure type

Ischemic Ischemic Ischemic Idiopathic Ischemic Ischemic Ischemic Ischemic Idiopathic Ischemic Idiopathic Ischemic Ischemic Ischemic

Drug

ANF + ANF + ANF + ANF + ANF + ANF + Placebo Placebo Placebo Placebo Placebo Placebo Placebo Placebo

Atria1 natriuretic factor; PCWP, New York Heart Association.

ANF,

PCWP

Furosemide Furosemide Furosemide Furosemide Furosemide Furosemide + Furosemide + Furosemide + Furosemide + Furosemide + Furosemide + Furosemide + Furosemide + Furosemide

17 24 12 23 23 23 28 30 23 25 15 21 15 20 21 t 5

NE

4.8 4.09 6.89 4.02 4.08 2.89 3.53 3.7 2.86 4.56 5.14 2.59 3.22 2.53 3.9 * 1.2

227 140 483 720 1822 415 380 673 1147 884 612 771 270 855 671 f 436

est diuretic and natriuretic properties,6l 7 although these effects are somewhat blunted in the setting of heart failure.8 Given the contrasting properties of furosemide and ANF on neuroendocrine activation

in the setting of heart failure,

the combination

of

agents could potentially be beneficial. ANF could reduce the mild acute vasoconstrictor and neuroendocrinologic properties of furosemide while augmenting the diuretic and natriuretic effects. Therefore the

aim of this study was to examine the acute interaction of ANF and intravenous furosemide on the he-

profile, neuroendocrine

activation,

and

urine output in patients with chronic stable heart

failure. METHODS Study population. After signing informed consent, 14 patients with mild-to-moderate (New York Heart Association classII to III) congestiveheart failure were studied in a double-blind, placebo-controlled parallel manner. Patients were all men between the agesof 47 and 71 years (mean age60) (Table I). Eleven patients had had a previ-

ANF

PRA

64.5 145.0 115.8 284.0 335.0 285.0 100.5 253.5 455.5 188.5 333.9 263.7 172.5 367.5 240 f 113


pulmonary capillary wedge pressure; CO, cardiac output; NE, norepinephrine;

be chronically elevated in patients with congestive heart failure3 and may decrease after treatment with diuretics.435 ANF in pharmacologic doses has been found to have favorable effects on the hemodynamic profile of patients with heart failure, producing a fall in systemic vascular resistance and a mild reduction in arterial blood pressure.6 The neuroendocrinologic alterations that occur after administration of ANF include a decrease in aldosterone levels and a mild decrease in plasma renin activity (PRA) and plasma norepinephrine levels. 6,7 In addition, ANF has mod-

modynamic

co

PRA,

NYHA

class

II II II II II III II III III III II II II II

plasma renin activity;

NYHA,

ous myocardial infarction with the subsequentdevelopment of chronic heart failure and were categorized ashaving “ischemic” cardiomyopathy. Three patients had no clinical evidence of coronary artery diseaseor alcoholism and wereconsideredto have “idiopathic” cardiomyopathy. All patients except onewerereceiving digoxin, and this was continued throughout the study. Each was being treated with diuretics and ACE inhibitors. Patients were excluded if they were unable to tolerate temporary withdrawal of ACE inhibitors, vasodilators, or diuretics. ACE inhibitors were withheld for 72 hours and diuretics and vasodilators for 24 hours. Additional exclusion criteria included pregnancy, myocardial infarction, angioplasty or coronary artery bypass operation within 3 months, serum creatinine greater than 2.5 mg/dl, total bilirubin greater than 2.0 mg/ dl, supinesystolic blood pressurelessthan 100mm Hg, serum potassiumlessthan 3.5 mEq/dl, hematocrit lessthan 30%) and a pulmonary capillary wedge pressure before drug administration of lessthan 12 mm Hg. A total of 20 patients were screenedand 14 were entered into the trial. Protocol. Patients were studied in the supineposition at least 12hours after eating. Right-heart catheterization was performed with a 110cm triple-lumen, flow-directed catheter from an antecubital vein. A 7.6 cm Teflon catheter was placed in the brachial artery to record systemic blood pressure.All waveforms were recorded on a Gould fourchannel recorder (model RS3400,Gould Inc., Valley View, Ohio) with zero reference level at the midaxillary line. Heart rate was measuredfrom a continuously recorded ECG lead. Cardiac output wasmeasuredin triplicate by the thermodilution technique utilizing cold saline solution. Cardiac index, pulmonary vascular resistance, and systemic vascular resistancewere calculated from standard formulas. Blood wasdrawn from the right atria1port of the Swan-Ganz catheter (Baxter Healthcare Corp., Santa Ana,

394

Connelly et al.

American

February 1994 Heart Journal

Protocol -

Lines

placed

- Baseline patients

hemodynamic and neuroendocrine asked to urinate - ANF

or placebo

administered

Furosemide administered urine collection begins

data obtained,

IV (2 mcg/kg) IV (1.3 mg/kg),

Ten minute hemodynamic measurement, blood drawn neuroendocrine levels 4” urine collection

for

ends 1

Fig. 1. Protocol for giving ANF or placebo plus furosemide in 14 patients with congestive heart failure.

Calif.), and all drugs were injected intravenously through the right atria1 port. Every attempt was made to ensure that the patients had an empty urinary bladder before study drugs were administered, although bladder catheterization was not performed. After the arterial and Swan-Ganz catheters were in place, the patients were asked to lie quietly for 30 minutes and baseline hemodynamic data were obtained (Fig. 1). At least three sets of hemodynamic data, including cardiac output, were obtained 15 minutes apart. If two sets had less than a 10 % variance, the last set obtained was used as the baseline measurement. Patients were randomly allocated to receive either synthetic human ANF (Scios, Mountain View, Calif.) or placebo. All patients were given 1.3 mg/kg of furosemide 5 minutes after the dose of ANF or placebo. The ANF was given as a dose of 2.0 pg/kg diluted with 10.5 ml of 10% mannitol solution given intravenously over 1 minute. Placebo (11 ml of 0.9% saline solution) was given in a similar manner. Six patients received ANF plus furosemide and eight received placebo plus furosemide. Hemodynamic measurements including cardiac output, pulmonary artery pressure, pulmonary capillary wedge pressure, right atria1 pressure, and systemic arterial pressure were obtained at baseline and at 10 minutes after the dose of furosemide. Blood was drawn at the same intervals for determination of PRA, arginine vasopressin, norepinephrine, epinephrine, aldosterone, ANF, and hematocrit. levels. Urine was collected throughout the study period (4 hours) and analyzed for sodium, potassium, creatinine, and volume measurements. Patients tolerated the study well with all patients entered completing the study.

Hormone assays. Blood samples, except for PRA tubes, were immediately placed on ice and spun in a refrigerated centrifuge within 30 minutes to separate the plasma. Plasma catecholamine levels were measured by a radioenzymatic method,g ANF by radioimmunoassay,1° PRA by radioimmunoassay,‘l arginine vasopressin by radioimmunoassay,i2 and aldosterone by radioimmunoassay.13 Hematocrit levels were obtained by the centrifuged capillary tube method. Urinary sodium and potassium levels were determined by a potentiometric method in which methylmonensin for sodium and valinomycin acid were used for potassium. An enzymatic creatinine reaction was used for urinary creatinine determination. Statistical analysis. Paired and unpaired t tests were used to assign p values for intergroup and intragroup analysis. p Values <0.05 were considered significant. Because PRA was not normally distributed, Wilcoxon and Spearman rank-order tests were used. RESULTS Hemodynamics. Hemodynamic measurements were obtained from all 14 patients at baseline and at 10 minutes after the furosemide dose. Ten minutes was chosen on the basis of previous experiencel, 2 to evaluate the maximal ANF and furosemide effects after bolus intravenous administration of the agents. The group given placebo plus furosemide exhibited a mild systemic vasoconstrictor effect with an increase in mean arterial pressure from 92 -t 5 (+SEM) to 96 t 5 mm Hg (p < 0.03), systemic vascular resis-

Volume 127, Number 2 American Heart Journal

Connelly et al.

Mean Arterial Pressure

Heart Rate

PI”5 -u- Placebo

IlJrosemlde

105

I

-e

ANF

pius

turosem~de

p c 0 03 p = NS I

p=NS

Cardiac Output

Pulmonary Capillary Wedge Pressure

Systemic Vascular Resistence +

Placebo

-o-

PiNF

plus plus

furosemide

f”msemilJe

Mean Pulmonary Pressure Placebo

p c 0 001

plus

furosemlde

p = NS

p = NS

T

Fig. 2. Hemodynamics at baseline and 10 minutes after furosemide dose. Open circles represent placebo plus furosemide group; closed circles with broken lines represent ANF group. p Values after legends in each graph represent difference between baseline and for that group. p Values to right of lo-minute points represent difference between placebo at 10 minutes.

with solid lines plus furosemide lo-minute values and ANF groups

395

396

Connelly

et al.

American

^^^ Plasma

Renin Activity

Plasma -e-

-o-Placebo plus furosemide ~~0.01 --t ANF plus furosemide p=NS

275

February 1994 Heart Journal

Norepinephrine

PlaceDo plus fUroSem!de PLO 01 ANF plus furosemide p=NS

-

c-

? 225 -

c

200..

5 .E

175.

/ / /

F

/

Baseline

10 min

1

p.NS 1

%0.

i

P

J Baseline

IO min

Fig. 3. PRA and plasma norepinephrine levels at baseline and 10 minutes after furosemide dose. Open circles with solid lines represent placebo plus furosemide group; closed circles with broken lines represent furosemide group. p Values after legend in each graph represent difference between baseline and lo-minute values for that group. p Values to the right of the lo-minute points represent difference between placebo

and ANF groups at 10 minutes.

tance from 1988 ? 157 to 2270 f 137 dynes . set . cmU5 (p = 0.0007), and heart rate from 80 + 7 to 85 + ‘7 beats/min 0, < 0.04). Cardiac output was unchanged from 3.5 -t 0.3 to 3.2 +- 0.3 L/min (p = NS). Stroke volume was slightly decreased from 47 + 6 to 40 +- 5 ml/beat (p < 0.04). Pulmonary capillary wedge pressure increased slightly from 22 +- 2 to 24 -t 2 mm Hg (p < 0.04), whereas mean pulmonary artery pressure was unchanged from 38 + 3 to 40 +- 3 mm Hg. These findings are consistent with our previous experience whereby intravenous furosemide induced a transient and mild pressor response in patients with chronic stable heart failure.r, 2 The group of patients given ANF in conjunction with furosemide exhibited no significant change from baseline to 10 minutes in mean arterial pressure (96 & 7 to 95 + 6 mm Hg), systemic vascular resis-

tance (1687 -t 114to 1763 + 129dynes . set . cmM5), or heart rate (69 + 5 to 72 _+ 7 beats/min; p = NS). Cardiac output, stroke volume, and pulmonary capillary wedge pressure did not change. Mean pulmonary artery pressure also did not change from baseline to 10 minutes in the patients receiving ANF plus furosemide (34 * 5 to 30 * 4 mm Hg; p = NS) (Fig. 2). Changes in mean arterial pressure for the group of 14 patients correlated with changes in PRA (r = 0.73; p < O.OOS), suggesting that activation of the reninangiotensin system is related to the pressor response produced by intravenous furosemide in patients with heart failure. Hormones. Blood was analyzed at baseline and 10 minutes after the furosemide dose. PRA increased in the placebo plus furosemide group from baseline to 10 minutes (12.3 f 5.0 to 31.0 +- 10.2 ng/ml/hr;

Volume 127, Number 2 American Heart J~umal

Connelly et al.

397

Total Urine Volume - Four Hours

Fig. 4. Total urine volume collected over 4-hour study period. p = NS representsno difference in urine volumes between groups after 4-hour collection period.

p < O.Ol), with seven of eight patients

exhibiting an increase (Fig. 3). Plasma norepinephrine levels increased in the placebo plus furosemide group from baseline to 10 minutes (699 +- 100 to 876 + 134 pg/ ml; p < 0.01) (Fig. 3). These data reconfirm our previous observations showing activation of the reninangiotensin and sympathetic nervous systems by intravenous furosemide in patients with chronic, stable congestive heart failure. PRA and plasma norepinephrine levels revealed no significant mean increase from baseline to 10 minutes in the group receiving ANF plus furosemide, although two patients with high baseline PRA exhibited pronounced increases (Fig. 3). The mean values did actually increase, however, and a type II beta error cannot be excluded. There was no significant change in serum levels of aldosterone, ANF, arginine vasopressin, epinephrine, or hematocrit from baseline to 10 minutes after the dose of furosemide between the patients receiving placebo plus furosemide and those receiving ANF plus furosemide. Urine. Urine was collected for 4 hours and analyzed at 2 and 4 hours after administration of study drugs for volume, sodium, potassium, and creatinine. Urine volumes (Fig. 4) were unchanged between the two groups. Four-hour urine volumes were 2184 +- 191 ml in the placebo plus furosemide group and 2410 ? 131 ml in the ANF plus furosemide group (p = NS).

Two-hour urine volumes were 1426 + 135 ml in the placebo plus furosemide group and 1677 +_ 133 ml in the ANF plus furosemide group (p = NS). Urinary sodium, potassium, and creatinine levels were not significantly different between the two groups. OfSCUSSlON

Congestive heart failure is a condition associated with hemodynamic and neuroendocrinologic abnormalities.14, l5 These disturbances include decreased cardiac output, increased systemic vascular resistance and pulmonary artery pressure, and increased PRA and plasma norepinephrine levels. Furosemide, a common and often-necessary agent used in the treatment of acute and chronic congestive heart failure, has been shown to cause neuroendocrine activation and to mildly increase systemic vascular resistance, mean arterial pressure, and heart rate when given intravenously to patients with chronic congestive heart fai1ure.l These acute responses in patients with chronic congestive heart failure appear to be at variance with the hemodynamic effects of furosemide reported in patients with acute myocardial infarction, who exhibited a reduction in cardiac filling pressures after administration of furosemide.16 ANF is known to be chronically elevated in patients with congestive heart failure.i7 It has been considered to be a counterregulatory hormone, which

398

Connelly

et al.

may offset neuroendocrine activation, peripheral vasoconstriction, and sodium retention.ls When ANF is administered by intravenous bolus to patients with heart failure, there is a reduction in systemic vascular resistance, pulmonary capillary wedge pressure, and plasma aldosterone levels.6 In most studies intravenous bolus ANF causes a decrease or no change in circulating PRA and plasma norepinephrine levels.6t lg The observed effects of ANF on hemodynamic measurements and neuroendocrine activation in patients with chronic heart failure led us to reason that the peptide might counteract the acute vasoconstrictor effects of intravenous furosemide. In this study we have reaffirmed the acute but mild vasoconstrictor and neuroendocrine activation properties of intravenous furosemide in patients with chronic heart failure. We also observed that the acute vasoconstrictor and neuroendocrinologic properties of furosemide were attenuated by pretreatment with ANF. Atria1 natriuretic factor given before furosemide administration was associated with a blunting of furosemide-induced increases in systemic vascular resistance and mean arterial pressure. The blunting of the furosemide-induced vasoconstriction appears to be mediated by inhibition of the reninangiotensin and sympathetic nervous systems, indicated by an attenuation of the increases in PRA and norepinephrine levels by pretreatment with ANF. The attenuation of the diuretic-induced increase in PRA by ANF was particularly evident in those patients with near-normal PRA at baseline. Such patients exhibited less of an increase in PRA after treatment with furosemide and ANF, and two patients exhibited a decrease in PRA. Alternatively two patients with increased PRA at baseline exhibited approximately fourfold increases in PRA in response to intravenous furosemide despite pretreatment with ANF. This suggests that ANF is able to block renin release in response to furosemide when plasma renin is near normal but not when the renin-angiotensin system is already activated. It is possible to postulate a dual mechanism of furosemide-induced renin release, one ANF sensitive and the second ANF insensitive. However, far more patients would have to be studied to adequately test this hypothesis. The design and scope of this study were rather limited, inasmuch as it was an exploratory effort to determine whether ANF might interact with furosemide in patients with chronic heart failure. Because the number of patients in each group was small, the results must be considered very preliminary. The interpretation of results is hampered by the lack of

American

February 1994 Heart Journal

ANF concentration differences between the two groups, and it is possible that ANF did not reach a steady-state distribution after 10 minutes in these patients. In conclusion, this study demonstrates that ANF may attenuate neuroendocrine activation and the mild peripheral vasoconstrictor responses to intravenous furosemide in selected patients with moderate-to-severe congestive heart failure. We found no important natriuretic or diuretic synergism between bolus ANF and intravenous furosemide at the doses tested. However, cautious interpretation is warranted because of the small number of patients studied. A longer observation period in which a smaller furosemide dose, a larger ANF dose, or both are used may reveal the presence of ANF enhancement. The blunting of furosemide-induced activation of the reninangiotensin and sympathetic nervous systems and subsequent mild vasoconstriction by ANF may offer a mechanism to offset a potentially deleterious effect of chronic diuretic use and should be explored in a larger study. Orally active neutral endopeptidase inhibitors that block the enzymatic breakdown of ANF and increase plasma ANF levelszO may represent a potential means of blocking chronic diuretic-induced neuroendocrine activation and vasoconstriction in patients with heart failure or hypertension. We thank Dr. Ada Simon, who performed the hormone analyses and Dr. Thomas Rector, who aided in the statistical analysis.

REFERENCES

1. Francis GS, Siegel R, Goldsmith SR, Olivari MT, Levine TB, Cohn JN. Acute vasoconstrictor response to furosemide in patients with chronic congestive heart failure. Ann Intern Med 1985;103:1-6. 2. Goldsmith SR. Francis GS. Cohn JN. Attenuation of the pressor response to intravenous furosemide by angiotensin converting enzyme inhibition in congestive heart failure. Am J Cardiol 1989;64:1382-5. 3. Rouleau JL, Kortas C, Bichet D, Champlain J. Neurohumoral and hemodynamic changes in congestive heart failure: lack of correlation and evidence of compensatory mechanisms. AM HEART J 1988;116:746-57. 4. Anderson JV, Woodruff PWR, Bloom SR. The effect of treatment of congestive heart failure on plasma atria1 natriuretic peptide concentration: a longitudinal study. Br Heart J 1988;59:207-11. 5. Singer DRJ. Shore AC. Markandu ND. Bucklev MG. MacGre>or GA. ‘Atria1 natriuretic peptide levels in “treated congestive heart failure. Lancet 1986;1:851. 6. Riegger GAJ, Kromer EP, Kochsiek K. Human atria1 natriuretic peptide: plasma levels, hemodynamic, hormonal and renal effects in patients with severe congestive heart failure. J Cardiovasc Pharmacol 1986;8:1107-12. 7. Saito H, Ogihara T, Nakamara M, Hara H, Higaki J, Rakugi H, Tateyama H, Minamuro T, Iinuma K, Kumahara Y. Hemodynamic, renal and hormonal responses to alpha-human

Volume 127, Number 2 American Heart Journal

8.

9. 10.

11. 12. 13. 14.

atria1 natriuretic peptide in patients with congestive heart failure. Clin Pharmacol Ther 1987;42:142-7. Cody RJ, Atlas SA, Laragh JH, Kubo SH, Covit AB, Ryman KS, Shaknovich A, Pondolfino K, Clark M, Camargo MJF, Scarborough RM, Lewicki JA. Atria1 natriuretic factor in normal subjects and heart failure patients. J Clin Invest 1986; 78~1362-74. Passon PG, Peuler JD. A simplified radiometric assay for norepinephrine and epinephrine. Anal Biochem 1973;51:61831. Burnett JC, Kao PC, Hu DC, Heser DW, Heublein D, Granger JP, Opgenorth TJ, Reeder GS. Atria1 natriuretic peptide elevation in congestive heart failure in the human. Science 1986; 231:1145-7. Sealey JE, G&en-Banes J, Laragh JH. The renin system: variations in man measured by radioimmunoassay or bioassay. Kidney Int 1972;1:240-53. Cowley AW, Skelton MM, Valasquez M. Sex differences in the endocrine predictors of essential hypertension. Hypertension 1985;79(suppl 1):1151-60. Mayes D, Furuyama S, Kern DC. A radioimmunoassay for plasma aldosterone. J Clin Endocrinol 1970;30:682-5. Francis GS, Benedict C, Johnstone DE, Kirlin PC, Nicklas J, Liang C, Kubo SH, Rudin-Toretisky E, Yusuf S. Comparison of neuroendocrine activation in patients with left ventricular

Connelly et al.

15.

16.

17. 18.

19.

20.

399

dysfunction with and without congestive heart failure: a subset of the studies of left ventricular dysfunction (SOLVD). Circulation 1990;82:1724-9. Cohn JN, Levine TB, Olivari MT, Garberg V, Lura D, Francis GS, Simon AB, Rector T. Plasma norepinephrine as a guide to prognosis in patients with congestive heart failure. N Engl J Med 1984;311:819-23. Dikshit K, Vyden JK, Forrester JS, Chatterjee K, Prakash R, Swan HJC. Renal and extrarenal hemodynamic effects of furosemide in congestive heart failure after myocardial infarction. N Engl J Med 1973;288:1087-90. Shenker Y, Sider RS, Ostafin EA, Grekin RJ. Plasma levels of immunoreactive atria1 natriuretic factor in healthy subjects and in patients with edema. J Clin Invest 1985;76:1684-98. Goy JJ, Waeber B, Nussberger J, Bidiville .I, Biollaz J, Nicod P, Mooser V, Kappenberger L, Brunner HR. Infusion of atria1 natriuretic peptide to patients with congestive heart failure. J Cardiovasc Pharmacol 1988;12:1562-70. Mtinzel T, Drexler H, Holtz J, Kurtz S, Just H. Mechanisms involved in the response to prolonged infusion of atria1 natriuretic factor in patients with chronic heart failure. Circulation 1991;83:191-201. Northridge DB, Alabaster CT, Connell JML. Effect of UK69578. A novel atriopeptidase inhibitor. Preliminary communication. Lancet 1989;2:591-3.