Pulmonary and Systemic Responses to Exogenous Endothelin-1 in Patients with Left Ventricular Dysfunction

Abstract

Plasma levels of immunoreactive endothelin-1 (ET-1) are elevated in chronic heart failure(CHF) and have been reported to correlate closely with pulmonary hemodynamic measurements. We investigated the effects of exogenous ET-1 on the pulmonary vasculature in patients with left ventricular systolic dysfunction (LVD), with or without overt heart failure. ET-1 was infused at 1, 5, and 15 pmol/min into a distal pulmonary artery of 10 patients with LVD. Hemodynamics were measured by a thermodilution catheter and arterial line. Intravascular Doppler and local pulmonary angiography were used to assess local pulmonary blood flow in the first four patients. Systemic hemodynamic changes occurred with ET-1 infusion in a dose-dependent fashion. Mean arterial pressure (100 ± 8-107± 11 mm Hg; p < 0.01) and systemic vascular resistance (1,699± 375-2,033 ± 427 dynes/s/cm⁻⁵; p < 0.001) rose, whereas the cardiac index fell from 2.43 ± 0.53 to 2.20 ± 0.491/min/m²(p < 0.002). However, mean pulmonary artery pressure (21 ± 7 mm Hg) and pulmonary vascular resistance (151 ± 43-147 ± 43 dynes/s/cm⁻⁵) did not change. Exogenous ET-1, when infused into patients with LVD, causes systemic but not pulmonary vasoconstriction.

The endothelins(ETs) are a family of potent vasoconstrictor peptides ⁽¹⁾. ET-1 is the predominant isoform expressed in the human vasculature. Plasma concentrations of ET-1 are elevated in patients with moderate or severe heart failure ^(2-4). Several authors have noted a positive relationship between plasma immunoreactive ET-1 (irET) and pulmonary hemodynamic measurements ^(4-6), particularly pulmonary vascular resistance (PVR) and the ratio of pulmonary to systemic vascular resistance (the resistance ratio).

It is not yet clear whether the relationship between plasma ET-1 and PVR is one of association or of cause and effect. To resolve this issue, interventional studies are required. Bosentan, a nonselective ET receptor antagonist, reduced both SVR and PVR in patients with chronic heart failure (CHF) when administered over a 2-week period ⁽⁷⁾. It is possible that the decrease in PVR observed during administration of bosentan was passive, secondary to an improvement in cardiac and systemic hemodynamics, rather than due to inhibition of a direct pulmonary vasoconstrictor effect of ET. An alternative intervention to attempt to resolve the role of ET-1 in the control of PVR is to infuse the agonist to achieve high local concentrations of ET-1 in the pulmonary vascular bed without increasing concentrations in the systemic vasculature. Accordingly, we infused ET-1 in a range of concentrations directly into the distal pulmonary artery of patients with left ventricular dysfunction with or without overt heart failure. We used intravascular Doppler ultrasound and local pulmonary angiography to study the local effects of ET-1 in the pulmonary resistance vessels.

MATERIALS AND METHODS

Patient selection

Patients with chronic left ventricular systolic dysfunction (LVD) were eligible for study (left ventricular ejection fraction of <40%). Patients with severe coronary disease, valvular heart disease, atrial fibrillation, insulin-dependent diabetes, uncontrolled hypertension, and chronic renal impairment (creatinine > 200 μmol/ml) were excluded.

Patient characteristics

Ten patients (nine male) aged 51-74 (mean 62) years took part in the study. The mean LVEF was 27 ± 6% (SD). One patient had a history of hypertension and one patient had noninsulin-dependent diabetes. Seven patients had LVD secondary to ischemic heart disease, and three patients had a dilated cardiomyopathy. Two patients were asymptomatic, and four patients were in NYHA classes II and III. Medications were as follows: angiotensin-converting enzyme inhibitors eight patients; diuretics five patients; digoxin one patient; β-blockers three patients; calcium antagonists three patients; aspirin five patients; and lipid-lowering therapy seven patients.

Study protocols

Studies were conducted with the approval of the local ethics committee and the written, informed consent of each patient. Cardiac medications were withheld for 24 h before the study. Patients were fasted for 4 h before the study. A 7F thermodilution catheter was positioned in a distal pulmonary artery percutaneously via a femoral vein under fluoroscopic control. In the first four patients, a 6F multipurpose catheter was also passed to the same pulmonary artery to allow intravascular Doppler studies. A 4F femoral arterial line was also placed to allow continuous intra-arterial blood pressure monitoring. Heparin 2,500 U was given as standard prophylaxis against thrombus formation.

After baseline values were established, sodium nitroprusside (SNP) was infused into the pulmonary artery under study at 0.56 and 1.12 μg/kg/min to assess vasodilator reserve. Five patients also received 1.68 μg/kg/min. After 5 min of each dose, a complete set of hemodynamic measurements was taken. Time was allowed for hemodynamic values to return to baseline (approximately 30 min). ET-1 (Clinalfa, Switzerland) was then infused at 1, 5, and 15 pmol/min into the same pulmonary artery. Two patients did not receive the 15 pmol/min infusion because of a decrease in cardiac output of >15% in one patient and a 20 mm Hg systolic BP rise in the other. Each dose was infused for 20 min, with hemodynamic measurements being made at 5 and 15 min. Further measurements were taken 5 and 15 min after the infusion was complete.

Intravascular Doppler studies

Intravascular Doppler studies were performed in the first four patients. A 0.018-inch Doppler guide wire (Flowire, Cardiometrics Inc.) was passed down the multipurpose catheter and positioned in a distal pulmonary artery under fluoroscopic control. Peak instantaneous velocities were analyzed, with the formula(average peak velocity)/2 used to calculate mean velocity in cm/s. The velocity signals for 10 consecutive sinus beats were averaged. This method has previously been validated ex vivo and demonstrates excellent linear correlation to volumetric flow, with r² values between 0.98 and 1.00 ⁽⁸⁾. The Doppler guidewire was positioned distal to the tip of the thermodilution catheter through which the SNP and ET-1 were infused. Recordings were made at baseline and at each dose of SNP and ET-1. Local pulmonary angiography was performed at baseline and at peak dose via the multipurpose catheter to allow estimation of the diameter of the pulmonary artery under study.

Statistical analysis

All values are reported as mean ± SD. The primary end points of the study were the changes in PVR and SVR from baseline to the maximal achieved dose of ET-1. Student's pairedt test (two-tailed) was used to compare baseline and peak hemodynamic measurements in the SNP and ET-1 infusion study.

RESULTS

Sodium nitroprusside infusion

Sodium nitroprusside (SNP) infusion led to systemic vasodilatation. Heart rate rose (75 ± 17 to 88 ± 20; p< 0.001), right atrial pressure remained unchanged (4.8 ± 1.4 to 4.5 ± 1.9), and mean arterial pressure (99 ± 9 to 76 ± 10; p< 0.001), mean pulmonary artery pressure (19 ± 5 to 11 ± 3; p< 0.001), pulmonary wedge pressure (11 ± 4 to 4 ± 2; p< 0.001), and systemic vascular resistance (1,575 ± 287 to 1,112 ± 272; p < 0.001) fell. Cardiac index rose (2.64 ± 0.58 to 2.84± 0.68; p < 0.05). The decrease in PVR did not reach statistical significance (124 ± 30 to 94 ± 40; p = 0.06).

ET-1 infusion

No patient noted any adverse effects during the infusion of ET-1.Figure 1 shows the response of SVR and PVR to ET-1. No hemodynamic change was observed with ET-1 at 1 pmol/min and only trends to increased SVR at 5 pmol/min (Fig. 1). Table 1 shows the peak hemodynamic effects of ET-1 infusion in these patients compared to the baseline taken after SNP infusion. At 15 pmol/min of ET-1 the heart rate remained unchanged, mean arterial pressure (MAP) rose by 7%, cardiac index fell by 9% and, consequently, systemic vascular resistance (SVR) rose by 20%. Mean pulmonary artery pressure (MPAP) and PVR were unchanged.

FIG. 1: Changes in systemic(SVR) (A) and pulmonary vascular resistance (PVR) (B) in response to graded infusion of endothelin-1 (ET-1). The increase in SVR was statistically significant at the highest concentrations of ET-1 (p < 0.001).

TABLE 1: Endothelin-1 hemodynamic data

Intravascular Doppler and local pulmonary angiography

Mean velocity did not change with infusion of SNP or ET-1: baseline 0.41 + 0.07 m/s, SNP 0.39 + 0.04 m/s, baseline 0.35 ± 0.04 m/s, ET-1 0.36 ± 0.06 m/s. No change could be observed in the diameter of the pulmonary conduit artery under study: baseline 0.36 ± 0.11 cm, SNP 0.38 ± 0.13 cm, ET-1 0.36 ± 0.12 cm.

DISCUSSION

This study shows that, in patients with LVD, infusions of ET-1 caused systemic vasoconstriction but with little or no change in mean pulmonary artery pressure or PVR. Intravascular Doppler and local pulmonary angiography were used to study the local effects of ET-1 in four patients. There was no noticeable change in pulmonary artery diameter or mean velocity, and therefore no change in local pulmonary blood flow, despite what must have been high local concentrations of ET-1. This lack of a direct effect of exogenous ET-1 infused into the lumen of a pulmonary artery casts doubt on a simple causal relationship between plasma concentrations of ET-1 and pulmonary vascular resistance.

The hemodynamic effects of ET-1 that we observed are broadly similar to those found in studies of healthy volunteers. We infused a maximum of 15 pmol/min, which is approximately 0.5 ng/kg/min assuming a 70-kg patient. Wagner et al. ⁽⁹⁾ infused ET-1 (∼1 ng/kg/min) into healthy volunteers to achieve a 4.5-fold increment in pulmonary arterial plasma concentrations, achieving levels compatible with those found in heart failure. MPAP fell but PVR did not change. Weitzberg et al. ⁽¹⁰⁾ infused approximately 10 ng/kg/min to achieve a 20-fold increase in pulmonary arterial plasma concentrations of ET-1, well above the pathophysiologic range, and reported a slight increase in MPAP (10.7 to 12.8 mm Hg), with a greater rise in PVR than SVR Pulmonary wedge pressure was again unchanged.

In healthy volunteers, ET_A-mediated vasoconstriction predominates in the systemic circulation. The increase in systemic vascular resistance observed in the present study is consistent with this. The lack of obvious pulmonary vasoconstriction in response to infused ET-1 in our study requires some explanation in view of the previously noted relationship between plasma ET-1 and pulmonary hemodynamic measurements and a considerable body of in vitro work. ET-1 is a potent constrictor of normal pulmonary arteries in vitro, but because vessels are bathed in ET-1, the peptide could be acting predominantly and directly on vascular smooth muscle. Tissue rather than plasma concentrations of ET-1 (in view of its abluminal secretion) may be a more important determinant of PVR so that, although plasma concentrations of ET-1 may reflect spill-over from the tissues, infusion of exogenous ET-1 may not accurately simulate the pathophysiologic state. Exogenous ET-1 in the plasma may have more prominent effects on the endothelial ET_B receptors than endogenously formed ET-1. Pulmonary endothelial ET_B receptor stimulation may lead to the release of nitric oxide and/or vasodilator prostaglandins that counteract ET-1's vasoconstrictor actions via smooth muscle ET_A and ET_B receptors.

An alternative explanation for the relationship between plasma ET-1 and PVR is that plasma concentrations of ET-1 may be merely a marker for endothelial dysfunction. There is considerable evidence for endothelial dysfunction in patients with heart failure, and in many respects this dysfunction reflects endothelial hyperactivity. Von Willebrand factor, a glycoprotein released from endothelial cells, is elevated and correlates directly not only with PVR ⁽¹¹⁾ but also with plasma concentrations of endothelin ⁽¹²⁾. Habib et al. ⁽¹³⁾ also demonstrated an enhanced response to L-NMMA in patients with heart failure, suggesting enhanced basal nitric oxide production in heart failure that presumably also reflects disordered endothelial function.

Increased plasma ET-1 might also reflect reduced clearance of ET-1 in CHF, because the pulmonary vasculature has been reported to be a major site of clearance of ET-1 in many studies^(9,10,14). A recent report has suggested that the pulmonary clearance of ET-1 in heart failure is inversely related to the severity of pulmonary hypertension ⁽¹⁵⁾.

Study limitations

After a period of rest to achieve stable baseline hemodynamics each study began with an infusion of SNP to assess pulmonary vascular responsiveness. Despite a 30-min wash-out period during which mean systemic and pulmonary arterial pressures were noted to have returned to normal, some rebound increase in PVR and SVR was noted. This could have obscured a small increase in PVR during infusion of ET-1. Even if this were the case, any rise in PVR was no greater than the increase in SVR, and a preferential vasoconstrictor effect in the pulmonary circulation can be excluded.

CONCLUSION

ET-1, when infused into patients with LVD, causes systemic vasoconstriction, with little or no effect on the pulmonary vasculature. Further studies of selective ET antagonists are required to determine the importance of endogenous ET-1 production in the control of pulmonary vascular tone.

Acknowledgments: Dr. Peter Cowburn and Dr. John Cleland are supported by grants from the British Heart Foundation. We wish to thank our patients and our nursing and technical staff for their help with this study.