Short-Term Oral Endothelin-Receptor Antagonist Therapy in Conventionally Treated Patients With Symptomatic Severe Chronic Heart Failure
Background—The vasoconstrictor peptide endothelin-1 (ET-1) is important for increased vascular tone in patients with chronic heart failure, but the effects of endothelin-receptor blockade in addition to conventional triple therapy are unknown.
Methods and Results—Thirty-six men (mean age±SD, 55±8 years) with symptomatic heart failure (NYHA class III; left ventricular ejection fraction, 22.4±4.5%) despite treatment with diuretics, digoxin, and ACE inhibitors received, in a double-blind and randomized fashion, either additional oral bosentan (1.0 g BID; n=24) or placebo (n=12) over 2 weeks. Hemodynamic and hormonal (plasma ET-1, norepinephrine, renin activity, and angiotensin II) measurements were obtained before and repeatedly for 24 hours after administration of bosentan on days 1 and 14. Bosentan was discontinued in 1 patient with symptomatic hypotension, and 2 patients (bosentan group) declined hemodynamic investigations on day 14. Compared with placebo, bosentan on day 1 significantly decreased mean arterial pressure (difference from baseline over 12 hours [95% CIs], −13.9% [−16.0% to −11.7%]), pulmonary artery mean (−12.9% [−17.4% to −8.3%]) and capillary wedge (−14.5% [−20.5% to −8.5%]) pressures, and right atrial pressure (−20.2% [−29.4% to −11.0%]). Cardiac output increased (15.1% [10.7% to 19.7%]), but heart rate was unchanged. Both systemic (−24.2% [−28.1% to −20.3%]) and pulmonary (−19.9% [−28.4% to −11.4%]) vascular resistance were reduced. After 2 weeks, cardiac output had further increased (by 15.2% [10.8% to 19.6%]) and systemic (−9.3% [−12.3% to −6.4%]) and pulmonary (−9.7% [−16.3% to −3.1%]) vascular resistances further decreased compared with day 1. Heart rate remained unchanged. Plasma ET-1 levels increased after bosentan, but baseline levels of the other hormones were unchanged.
Conclusions—Additional short-term oral endothelin-receptor antagonist therapy improved systemic and pulmonary hemodynamics in heart failure patients who were symptomatic with standard triple-drug therapy. Further investigations are warranted to characterize the effects of long-term endothelin-receptor antagonist therapy on symptoms, morbidity, and mortality in such patients.
There is both indirect and direct evidence that the vasoconstrictor peptide endothelin-1 (ET-1)1 participates in the regulation of vascular tone in patients with chronic heart failure. Thus, plasma levels of ET-1 are elevated, particularly in patients with severe heart failure,2 and intravenous infusion of an endothelin-receptor antagonist decreased both systemic and pulmonary vascular resistance acutely in such patients.3 Furthermore, plasma levels of its precursor big ET-1 are a strong and independent predictor of death in heart failure patients,4 and plasma ET-1 is also related to exercise capacity.5 Thus, therapeutic administration of an endothelin-receptor antagonist theoretically might improve the hemodynamic status of patients with heart failure and alleviate their symptoms. Indeed, venous, arterial, and pulmonary arterial vasodilation and an increase in stroke volume without sympathetic reflex activation were observed immediately after intravenous administration of the endothelin-receptor antagonist bosentan,3 which blocks both type A (ETA) and type B (ETB) receptors.6 However, these effects were seen after ACE inhibitors were discontinued, and no data exist in human heart failure with respect to more prolonged therapy. Because any new heart failure therapy is likely to be used in addition to ACE inhibition and animal experiments demonstrated additive hemodynamic effects of combined endothelin-receptor antagonist and ACE inhibitor therapy,7 we investigated hemodynamic and clinical effects of additional oral bosentan in patients who remained symptomatic despite conventional heart failure therapy, including ACE inhibitors.
All patients gave written informed consent to this prospective, double-blind, randomized, placebo-controlled clinical study, which was conducted in parallel at 3 centers in Switzerland and approved by the respective ethical committees.
Thirty-six patients (all male; age, 55.2±8.1 years; range, 34 to 68 years) with clinically stable chronic (>3 months) congestive heart failure and dyspnea in NYHA class III were studied.
Inclusion criteria were a left ventricular ejection fraction of <30%, a pulmonary capillary wedge pressure of ≥15 mm Hg, and/or a resting cardiac index of ≤2.5 L · min−1 · m−2. All patients were on stable doses of ACE inhibitors, diuretics, and digoxin except 1 patient with no digoxia. Oral anticoagulants (n=30) and antiarrhythmics (n=10, amiodarone) were continued, but long-acting nitrates and β-blockers were discontinued 2 and 5 days, respectively, before the baseline hemodynamic study to reduce variability due to differing background medication.
Hemodynamic measurements were obtained for 24 hours, with the patient supine, by standard techniques3 with a Swan-Ganz thermodilution catheter and by cannulation of the radial artery. Cardiac output was determined in duplicate by the thermodilution technique. Systemic and pulmonary vascular resistances were calculated according to standard formulas. Heart rate was derived from the continuously monitored ECG.
Blood for determination of plasma levels of ET-1, norepinephrine, renin activity, and angiotensin II (Ang II) was withdrawn from the pulmonary artery, separated, and stored at −70°C until assay.
ET-1 plasma concentrations were determined as previously described3 8 with polyclonal rabbit antiserum RAS 6901 (anti−ET-1). Cross-reactivity of antiserum with the precursor big ET-1, ET-1, and ET-3 is <10%. The range for ET-1 in normal subjects with this assay was 2.0 to 4.6 ng/L (mean, 3.3±0.66 ng/L).
Plasma norepinephrine was measured by high-performance liquid chromatography9 (normal range, 0.66 to 3.85 nmol/L; mean, 1.86±0.88 nmol/L).
Plasma renin activity (PRA) was measured by radioimmunoassay (normal range, 8.8 to 36 mL U/L) and plasma Ang II concentrations with a radioimmunoassay after solid-phase extraction with SepPak C1810 (normal range, 3.8 to 30 ng/L; mean, 12.1±5.7 ng/L).
Laboratory Analysis and Safety Data
Urinalysis, red and white blood cell counts, and multipanel blood chemistry were performed at screening and on days 1, 7, 14, and 21.
After screening, patients were admitted to the hospital on the morning of day 1 in the fasting state without their regular morning doses of medications. After placement of catheters and blood sampling, hemodynamic measurements were repeated until variation of cardiac output was <10%, and the last value was taken as baseline. Next, bosentan 1000 mg or matching placebo (2:1 randomization) was given orally, together with a standard hospital breakfast and each patient’s morning dose of diuretics and digoxin. The ACE inhibitor was administered 3 hours later because of safety concerns regarding hypotension during coadministration of 2 vasodilators. However, this allowed an assessment of bosentan effects at the time of peak plasma levels (V. Charlon, PhD, written communication, May 1998) against the background of chronic but without interference from acute ACE inhibition. Hemodynamics were measured repeatedly (at 1, 2, 3, 4, 6, 8, 12, 16, and 24 hours). Twelve hours after the first dose, the second dose of study medication was administered together with a meal, and standard evening medication and the ACE inhibitor 4 hours later. After the last measurements, catheters were removed. Vital signs were reassessed immediately before and 3 hours after the third dose of study medication, which was now given simultaneously with all other medications, including the ACE inhibitor during breakfast. Patients were then discharged and instructed to take the study medication twice daily together with food in addition to their standard medication. After 1 week, NYHA class, clinical status, and vital signs were assessed. After 2 weeks, patients were readmitted to the hospital, and all procedures were repeated in an identical way. Patients were then discharged without study medication and followed up clinically for at least 1 additional month by ambulatory visit or telephone interview.
All patients were included in the safety analysis. However, for hemodynamic evaluation, only paired data from days 1 and 14 were analyzed.
All calculations were performed with the StatView 4.5 (Abacus Inc) statistical program. Results are expressed as mean±SD. ANOVA was used to examine differences between the study groups at baseline. Because hemodynamics were assessed repeatedly, ANOVA for repeated measurements was performed for all changes from baseline over the 12-hour dosing interval to test for overall drug effects. In addition, an unpaired t test was used to calculate differences and 95% CIs between placebo- and bosentan-treated patients based on all changes for the first 3 hours (ie, without interference from acute ACE inhibition) and for the entire dosing interval of 12 hours (ie, including the effects of ACE inhibition from 3 to 12 hours). The trough effect was calculated as change from baseline obtained predrug in the morning of day 14. Hormone measurements were compared with their baseline values by paired t test and differences between groups by unpaired t test.
A value of P<0.05 (2-tailed) was considered to indicate a significant difference.
Baseline patient characteristics are shown in Tables 1 through 3⇓⇓⇓⇓. Placebo- and bosentan-treated patients were comparable in most regards, but patients assigned to bosentan had a somewhat higher ejection fraction, stroke volume index, and systolic blood pressure and significantly lower heart rate, suggesting somewhat greater hemodynamic impairment in the placebo group. Baseline plasma ET-1 levels were positively correlated with pulmonary capillary wedge (r=0.44, P<0.05), right atrial (r=0.54, P<0.01), and pulmonary artery mean (r=0.46, P<0.05) pressures and pulmonary vascular resistance (r=0.48, P<0.05) and negatively with stroke volume index (r=−0.51, P<0.05).
Safety and Clinical Outcome
One patient developed supine systolic blood pressure <90 mm Hg on the first day of bosentan treatment. Therapy was discontinued after 2 days because of persisting symptomatic hypotension. Two additional patients complained of dizziness; hypotension was noted on the second day of therapy, but symptoms subsided, and they continued bosentan therapy. Furthermore, 3 patients reported dizziness without detectable hypotension, and 2 of the 6 patients with dizziness reported mild headaches. Weight did not differ from baseline after 2 weeks (79.8±10.4 versus 79.4±9.6 kg in bosentan-treated and 79.2±11.1 versus 79.6±10.8 kg in placebo-treated patients). There were no specific changes of laboratory values. In particular, serum creatinine (93±25 versus 94±32 μmol/L) and liver function tests (ALAT, ASAT, γ-GT) remained unchanged during 2 weeks of bosentan and 1 week later. Two bosentan-treated patients declined to undergo repeat hemodynamic evaluation. In bosentan-treated patients, dyspnea improved by 1 NYHA class in 9, was unchanged in 11, and worsened in 4; in the placebo group, 1 patient improved, 8 were unchanged, and 3 became worse (P=0.18).
Hemodynamic measurements at baseline and 3 hours after administration of standard medication and study drug in the morning (ie, before intake of ACE inhibitor) are shown in Table 3⇑⇑. Although placebo did not change hemodynamics, bosentan significantly decreased arterial, pulmonary artery, and right atrial pressures and increased cardiac index within the first 3 hours on day 1 compared with placebo. Accordingly, both calculated systemic and pulmonary vascular resistances were significantly reduced. Heart rate did not change; thus, the increase in cardiac index was due to a significant increase in stroke volume index. As shown in Figures 1⇓ and 2⇓, these differences persisted after 3 hours when ACE inhibitors were added. Figure 3⇓ summarizes the differences between placebo- and bosentan-treated patients with respect to all measurements obtained over a period of 12 hours after study drug intake on both study days. As shown, the effect of bosentan persisted to a similar extent after 2 weeks with respect to arterial, pulmonary artery, and cardiac filling pressures, whereas cardiac (15.2% [10.8% to 19.6%]) and stroke volume (15.6% [8.8% to 20.1%]) indexes had increased further compared with bosentan effect on day 1 (both P<0.01); likewise, systemic (−9.3% [−12.3% to −6.4%]) and pulmonary (−9.7% [−16.3% to −3.1%]) vascular resistances had decreased further (both P<0.01) without change of heart rate.
At trough (before drug intake in the morning of day 14; Table 3⇑⇑ and Figure 2⇑), bosentan-treated patients had lower arterial pressure, higher cardiac and stroke volume indexes, and reduced systemic and pulmonary vascular resistances. Pulmonary arterial and left and right heart filling pressures were not significantly different from baseline.
Baseline neurohormones were not different in the two groups (Table 4⇓). After the morning dose of diuretic, digoxin, and study drug, plasma ET-1 increased by 134±70% after 3 hours in the bosentan-treated group (P<0.01) and was unchanged with placebo. PRA and Ang II levels increased similarly in the two groups, and plasma norepinephrine remained unchanged.
Compared with day 1, plasma ET-1 was elevated in bosentan-treated patients on the morning of day 14 before drug intake and unchanged in the placebo group. Plasma ET-1 increased 3 hours after bosentan, but the rise was less than that on day 1 (134±70% versus 37±38%, P<0.05). Neither baseline plasma ET-1 levels nor its changes during therapy predicted the clinical or hemodynamic response to bosentan.
Before drug intake, PRA and Ang II levels were similar on days 1 and 14, and the increases after diuretic administration were unchanged in the placebo group. In contrast, increases in PRA (175±226% versus 58±94%, P<0.05) and plasma Ang II (201±236% versus 89±116%, P<0.05) were attenuated in the bosentan group on day 14 compared with day 1. Norepinephrine remained unchanged during the study in both groups.
These data show that short-term oral endothelin-receptor blockade with bosentan resulted in improved hemodynamics without neurohormonal stimulation in patients who were symptomatic despite triple-drug heart failure therapy, including an ACE inhibitor. These data confirm that endogenously released ET-1 is important for the regulation of elevated vascular tone in these patients3 11 and suggest that its antagonism may be useful as adjunct therapy.
As expected, the combination of 2 vasodilators resulted in symptomatic hypotension in 3 patients and withdrawal from the study in 1. However, the fixed-dose drug regimen probably contributed to this problem. Moreover, the optimal dose of bosentan is not known, but in a small, open-label trial using a dose of 500 mg BID,12 we observed hemodynamic changes from day 1 to day 14 of a magnitude similar to those in this trial. This indicates that lower dosages of bosentan than the one used in this study also elicit hemodynamic effects.
Bosentan caused venous, pulmonary artery, and arterial vasodilation when acting alone, eg, in the first 3 hours after study drug administration, and when combined with ACE inhibitors for the remainder of the dosing interval. These changes were seen on top of effective long-term ACE inhibition, as hemodynamically reflected in almost normal baseline values of systemic vascular resistance. This hemodynamic constellation probably also explains why the absolute magnitude of the hemodynamic effects was not larger. In addition, pulmonary vascular resistance, which was elevated despite chronic ACE inhibition, was markedly reduced by bosentan. Elevated cardiac filling pressures and pulmonary hypertension are independent risk factors for survival in severe heart failure, and reversal of pulmonary hypertension by drug therapy improved prognosis in cardiac transplant candidates,13 rendering the present findings potentially important.
The effects seen after the first dose were not only maintained during prolonged therapy but were even more marked with regard to systemic and pulmonary artery vasodilation and increases in cardiac index after 2 weeks. The mechanism behind this finding is not clear. Possibly, reversing the long-term cardiovascular trophic effects of ET-1 is of greater importance than vasodilation due to endothelin-receptor blockade per se. This view would be compatible with animal experiments showing that endothelin-receptor antagonism improved survival in rats with heart failure after myocardial infarction,14 15 ie, the model that predicted the clinical effectiveness of ACE inhibitors in humans.16 Whether 2 weeks of therapy would be enough for such an effect is unknown.
Interestingly, the dilatory effect did not result in reflex tachycardia, a finding similar to that obtained after acute administration of bosentan in heart failure patients.3 The mechanism(s) behind this observation are not clear, but reduced baroreceptor sensitivity in heart failure patients17 probably contributes to this effect. Arterial pressure and systemic and pulmonary vascular resistance were still decreased 12 hours after the last dose of bosentan, whereas cardiac filling pressures, which reflect venous tone in the absence of weight and volume changes, were unchanged. Different binding characteristics of bosentan to venous and arterial endothelin receptors might be a possible explanation. Indeed, ETB-mediated vasoconstriction might predominate on the venous side.18 Because bosentan, a mixed ETA- and ETB-receptor antagonist,6 is somewhat more potent at the ETA receptor, this pharmacological property might explain the longer-lasting effect of bosentan on the arterial side.
Bosentan did not affect basal PRA or Ang II and catecholamine levels, indicating that basal activity of these systems is not influenced by endogenously released ET-1.19 After 2 weeks, however, bosentan attenuated the increases of PRA and Ang II levels after diuretic intake. We have no ready explanation for this effect, but interactions of reduced renal perfusion and natriuresis could render the system partially ET-1 dependent after stimulation with a diuretic.
Plasma ET-1 more than doubled within 3 hours after the first dose of bosentan. This might be explained by dissociation or prevention of binding of ET-1 from its receptors, in particular from the ETB receptor, which is believed to act as a clearance receptor.20 However, ET-1 levels increased 2- to 3-fold after specific ETA-receptor blockade in rabbits, which brings into question the clearance function of the ETB receptor.21 Plasma ET-1 was still significantly increased 12 hours after bosentan, which might indicate residual receptor occupancy at the end of the dosing interval. The response of plasma ET-1 after the morning dose of bosentan was attenuated on day 14, although hemodynamic effects were more pronounced compared with day 1. One might speculate that the hemodynamic improvement brought about by bosentan might have resulted in a relative decrease in ET-1 levels, as seen in heart failure patients treated with a β-blocker.22
As in previous studies,3 23 baseline plasma ET-1 levels were significantly correlated with several indexes of hemodynamic impairment. However, neither baseline ET-1 levels nor increases seen after bosentan correlated with the hemodynamic effects of endothelin-receptor blockade. A closer relationship might exist between the response to endothelin-receptor blockade and tissue levels of this paracrine/autocrine system.
Bosentan is a mixed ETA/ETB-receptor antagonist.6 Stimulation of endothelial ETB receptors leads to vasodilation.24 Accordingly, attenuation of endothelin-mediated vasodilator effects by bosentan cannot be excluded. Comparative studies will be necessary to show whether selective ETA-receptor blockade alone has even greater hemodynamic effects.
Placebo- and bosentan-treated patients were not ideally matched with regard to baseline hemodynamics and concomitant medication. Although inclusion of a placebo group and standardization of food intake served as a control for the stability of the population and for non–drug-related interventions (eg, food intake) throughout the 24-hour monitoring period, one obviously must be careful not to extend our findings to a population that might be sicker than the group that received bosentan in this trial. Also, nonstandardized use of ACE inhibitors might have influenced our results. However, the majority of patients in both groups received enalapril twice daily, and more patients in the bosentan group received short-acting captopril. Thus, it is unlikely that nonstandardized use of ACE inhibitors significantly influenced the hemodynamic measurements obtained during the first 3 hours in the morning.
We did not objectively quantify the degree of functional impairment of patients. Therefore, the severity of functional impairment might have been overestimated in some patients. This might be of importance because the target population for future administration of endothelin antagonists will be primarily NYHA class III to IV patients.
Pretreatment with long-acting nitrates and β-blockers was stopped before the trial. Potentially, patients maximally prevasodilated with long-acting nitrates and ACE inhibitors could have an attenuated response to bosentan. Also, endogenous nitric oxide inhibits ET-1 synthesis.25 Conceivably, exogenous nitric oxide donors may have similar effects. However, there was no difference in the hemodynamic response to bosentan in patients previously treated with nitrates.
Finally, we did not see any untoward effects other than hypotension in this trial using a fixed and, presumably, high dose of bosentan.12 However, reversible increases of serum ALAT and ASAT were found after 4 weeks of treatment with bosentan in some hypertensive patients.26 It is possible that the 2-week treatment period was too short to see such effects in our patients. It is obvious that the proper dose and the safety of bosentan will have to be established in long-term trials.
ET-1 plays an important role in the regulation of vascular tone in patients with severe heart failure. Counteracting vascular ET-1 effects by short-term administration of an orally active, mixed ETA/ETB-receptor antagonist improved hemodynamics in patients who were symptomatic with standard triple-drug therapy. These effects were observed without neurohormonal activation or volume retention. Further investigations are warranted to study the safety of long-term endothelin-receptor antagonist therapy and its effects on symptoms, morbidity, and mortality in such patients.
This study was supported by an educational grant from F. Hoffmann-LaRoche, Ltd, Basel, Switzerland, and from the Swiss National Science Foundation (grant 32-43612.95).
Guest Editor for this article was Wilson S. Colucci, MD, Boston Medical Center, Boston, Mass.
- Received June 8, 1998.
- Revision received July 14, 1998.
- Accepted July 21, 1998.
- Copyright © 1998 by American Heart Association
Yanagisawa M, Kurihara H, Kimura S, Tomobe Y, Kobayashi M, Mitsui Y, Yazaki Y, Goto K, Masaki T. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature. 1989;332:411–415.
Wei CM, Lerman A, Rodeheffer RJ, McGregor CG, Brandt RR, Wright S, Heublein DM, Kao PC, Edwards WD, Burnett JC. Endothelin in human congestive heart failure. Circulation. 1994;89:1580–1586.
Pacher R, Stanek B, Hülsmann M, Koller-Strametz J, Berger R, Schuller M, Hartter E, Ogris E, Frey B, Heinz G, Maurer G. Prognostic impact of big endothelin-1 plasma concentrations compared with invasive hemodynamic evaluation in severe heart failure. J Am Coll Cardiol. 1996;27:633–641.
Clozel M, Breu V, Gray GA, Kalina B, Löffler BM, Burri K, Cassal JM, Hirth G, Müller M, Neidhart W, Ramuz H. Pharmacological characterization of bosentan, a new potent orally active nonpeptide endothelin receptor antagonist. J Pharmacol Exp Ther. 1994;270:228–235.
Teerlink JR, Löffler BM, Hess P, Maire JP, Clozel M, Clozel JP. Role of endothelin in the maintenance of blood pressure in conscious rats with chronic heart failure: acute effects of the endothelin receptor antagonist Ro 47–0203 (bosentan). Circulation. 1994;90:2510–2518.
Love MP, Haynes WG, Gray GA, Webb DJ, McMurray JJV. Vasodilator effects of endothelin-converting enzyme inhibition and endothelin ETA receptor blockade in chronic heart failure patients treated with ACE inhibitors. Circulation. 1996;94:2131–2137.
Sütsch G, Bertel O, Kiowski W. Acute and short-term effects of the nonpeptide endothelin-1 receptor antagonist bosentan in humans. Cardiovasc Drugs Ther. 1996;10:717–725.
Mulder P, Richard V, Derumeaux G, Hogie M, Henry JP, Lallemand F, Compagnon P, Macé B, Comoy E, Letac B, Thuillez C. Role of endogenous endothelin in chronic heart failure: effect of a long-term treatment with an endothelin antagonist on survival, hemodynamics and cardiac remodeling. Circulation. 1997;96:1976–1982.
Pfeffer JM, Pfeffer MA, Braunwald E. Hemodynamic benefits and prolonged survival with long-term captopril therapy in rats with myocardial infarction and heart failure. Circulation. 1987;75(suppl I):I-149-I-155.
Ellenbogen KA, Mohanty PK, Szentpetery S, Thames MD. Arterial baroreflex abnormalities in heart failure: reversal after orthotopic cardiac transplantation. Circulation. 1989;79:51–58.
Seo B, Oemar BS, Siebenmann R, von Segesser L, Lüscher TF. Both ETA and ETB receptors mediate contraction to endothelin-1 in human blood vessels. Circulation. 1994;89:1203–1208.
Schricker K, Scholz H, Hamann M, Clozel M, Kramer BK, Kurtz A. Role of endogenous endothelins in the renin system of normal and two-kidney, one clip rats. Hypertension. 1995;25:1025–1029.
Spinale FG, Walker JD, Mukherjee R, Iannini JP, Keever AT, Gallagher KP. Concomitant endothelin receptor subtype-A blockade during the progression of pacing-induced congestive heart failure in rabbits: beneficial effects on left ventricular and myocyte function. Circulation. 1997;95:1918–1929.
Cody R, Haas GJ, Binkley PF, Capers Q, Kelley R. Plasma endothelin correlates with the extent of pulmonary hypertension in patients with chronic congestive heart failure. Circulation. 1992;85:504–509.
Boulanger C, Lüscher TF. Release of endothelin from the porcine aorta: inhibition by endothelium-derived nitric oxide. J Clin Invest. 1990;85:587–590.