(Circulation. 2005;112:2411-2416.)
© 2005 American Heart Association, Inc.
Exercise Physiology |
From the Department of Medicine, Montreal Heart Institute, Montreal (A.D., A.D., P.K., M.J., F.M., L.A.M.); Department of Pediatric Cardiology, Quebec Heart Institute, Quebec (C.H.); and Department of Cardiology, Ottawa Heart Institute, Ottawa (K.L.C.), Canada.
Correspondence to Annie Dore, MD, Adult Congenital Heart Center, Montreal Heart Institute, 5000 Belanger St E, Montreal, Quebec, Canada H1T 1C8. E-mail annie.dore{at}icm-mhi.org
Received February 16, 2005; revision received June 22, 2005; accepted July 29, 2005.
| Abstract |
|---|
|
|
|---|
Methods and Results A multicenter, randomized, double-blind, placebo-controlled, crossover clinical trial was performed to assess the effects of losartan on exercise capacity and neurohormonal levels in patients with systemic RVs. Of 29 patients studied (age, 30.3±10.9 years), 21 had transposition of the great arteries with a Mustard baffle, and 8 had congenitally corrected transposition of the great arteries. Baseline values were as follows:
O2max, 29.8±5.6 mL · kg1 · min1 (73.5±12.9% predicted value); RV ejection fraction, 41.6±9.3%; N-terminal pro brain natriuretic peptide (NT-proBNP), 257.7±243.4 pg/mL (normal <125 pg/mL); and angiotensin II, 5.7±4.9 pg/mL (normal <5.0 pg/mL). Comparing losartan to placebo showed no differences in
O2max (29.9±5.4 versus 29.4±6.2 mL · kg1 · min1; P=0.43), exercise duration (632.3±123.0 versus 629.9±140.7 seconds; P=0.76), and NT-proBNP levels (201.2±267.8 versus 229.7±291.5 pg/mL; P=0.10), despite a trend toward increased angiotensin II levels (15.2±13.8 versus 8.8±12.5 pg/mL; P=0.08).
Conclusions In adults with systemic RVs, losartan did not improve exercise capacity or reduce NT-proBNP levels. Minimal baseline activation of the renin-angiotensin system may explain this lack of benefit and imply an alternative pathophysiological mechanism for the progressive ventricular dysfunction and impaired exercise capacity observed in such patients.
Key Words: angiotensin exercise transposition of great vessels
| Introduction |
|---|
|
|
|---|
In both symptomatic and asymptomatic patients with left ventricular (LV) dysfunction, pharmacological blockade of the renin-angiotensin system (RAS) improves LV filling pressures,810 cardiac index,8,9 exercise tolerance,812 and overall survival.1315 Results of such studies are often extrapolated to patients with systemic RVs, and therapy frequently is empirically initiated. However, beneficial effects of inhibiting the RAS in patients with systemic RVs have not yet been demonstrated. This study therefore was designed to assess the effects of losartan on exercise capacity and neurohormonal levels in adults with systemic RVs.
| Methods |
|---|
|
|
|---|
18 years of age and on a stable medical regimen with no change in therapy or hospitalization in the preceding 3 months. Patients were excluded if they had severe symptomatic heart failure (New York Heart Association [NYHA] functional class III or IV), were unable to exercise, were pregnant, had a fixed-rate pacemaker, had a creatinine level >250 mmol/L, or had a history of angioedema. Patients with any of the following hemodynamic criteria were likewise excluded: right-to-left shunting; substantial left-to-right shunting (Qp/Qs >1.5); severe mitral, aortic, or pulmonary regurgitation; systemic or pulmonary inflow obstruction with a peak velocity >1.5 m/s by transthoracic echocardiography; and severe outflow tract obstruction with a peak systolic gradient >80 mm Hg.
Study Design
A multicenter, randomized, double-blind, placebo-controlled, crossover trial was performed. Patients already receiving ACE inhibitors or angiotensin receptor blockers were required to discontinue therapy 4 weeks before their preliminary assessment. All other medications were continued at the same doses.
One week before randomization, all patients were preliminarily assessed with standard exercise treadmill tests using a RAMP protocol. On the day of randomization, baseline testing consisted of a transthoracic echocardiogram, maximal symptom-limited exercise treadmill test with spiroergometry, and laboratory testing that included creatinine, potassium, and neurohormonal levels. For a patient to be eligible for randomization, differences in exercise duration between preliminary and baseline tests could not exceed 15% or 60 seconds. After providing informed consent, qualifying patients were then randomized to therapy with losartan or placebo in a double-blinded fashion. Treatment was continued for 15 weeks (phase 1), after which crossover to the second treatment arm was performed for an additional 15 weeks (phase 2).
On initial randomization, patients received losartan 50 mg or placebo once a day for the first 4 weeks. At 14 days, potential side effects were assessed by telephone interviews. If therapy was well tolerated, the dose was doubled at 28 days (ie, losartan 50 mg or placebo twice daily). Once again, patients were contacted 14 days after the dosage increase to reassess side effects. At the conclusion of each phase, repeated testing was performed with a transthoracic echocardiogram, exercise treadmill spiroergometry, and blood tests, including neurohormonal levels. The study protocol was approved by all local institutional review boards and ethics committees.
Cardiopulmonary Testing
At baseline and after each treatment phase, patients underwent symptom-limited treadmill exercise testing using the same RAMP protocol and were encouraged to exercise to exhaustion. The RAMP protocol has been recommended by many authors16,17 because it allows the increase in work to be individualized. The speed and incline are increased at fixed intervals of 6 to 10 seconds, permitting a constant and continuous increase in external work. Gas exchange was assessed during each exercise test with a computerized metabolic cart (Oxycon Alpha, Jaegger). Gas was sampled through a Rudolph mask. Recorded variables included maximal oxygen uptake (
O2max), oxygen pulse, metabolic units, exercise duration, minute ventilation, heart rate, respiratory rate, and blood pressure. The primary end point,
O2max, is the product of cardiac output and AV difference in oxygen content and was defined as the highest value recorded during the last minute of exercise.
O2 at the gas exchange anaerobic threshold was determined by means of the V-slope method.
O2 at peak exercise was expressed both as milliliters per kilogram per minute and as a percentage of the
O2max predicted for age, sex, weight, and height according to previously published normograms.17
Transthoracic Echocardiography
A standard 2D Doppler transthoracic echocardiogram was performed at baseline and after each treatment phase. RV dimensions were qualitatively assessed as normal, mildly dilated, or moderately to severely dilated. RV systolic function was assessed both qualitatively (ie, normal, mild, or moderately to severely impaired) and by calculation of an RV ejection fraction using the simplified Simpson method in an apical 4-chamber view.18 Valvular stenosis and regurgitation were quantified by color and continuous-wave Doppler flow. Systemic and pulmonary venous inflows were evaluated by pulsed-wave Doppler. The presence of an intracardiac shunt was likewise systematically explored.
Neurohormone Levels
Levels of angiotensin II, epinephrine, norepinephrine, and N-terminal pro brain natriuretic peptide (NT-proBNP) were drawn at baseline and at the completion of each treatment phase. Peripheral venous blood samples were obtained in the morning after the subjects rested in a supine position in a dimly lit room for at least 30 minutes. All samples were centrifuged and frozen at 80°C until analysis. Analyses were performed in triplicate at the end of the study. Angiotensin II, epinephrine, and norepinephrine were measured by previously described techniques.1921 An angiotensin II level <5 pg/mL was considered normal.22 Epinephrine and norepinephrine values for normal control subjects in our laboratory (mean±SD) were 40.6±24.3 and 291±111 pg/mL, respectively.23 NT-proBNP levels were measured with the Elecsys NT-proBNP assay from Roche Diagnostics. A value of P<125 pg/mL was considered normal.24
Statistical Analysis
Sample size calculations were based on
O2max, the primary end point of interest. Assuming an SD of 7 mL · kg1 · min1 as reported by Musewe et al6 and a conservative intraclass correlation coefficient of 0.3, a total of 26 patients (ie, 13 per treatment arm) were required to obtain 80% power to detect a 4.65-mL · kg1 · min1 difference between losartan and placebo with a 2-sided
of 0.05.
Continuous normally distributed variables are expressed as mean±SD. Differences in outcome variables were compared by use of ANOVA models for repeated measures that accounted for assigned treatment arm (losartan versus placebo) and order of treatment phase (first 15 weeks versus second 15 weeks). Pearson correlation coefficients were used to assess bivariate associations. Two-tailed values of P<0.05 were considered statistically significant. Testing was performed with SAS software, version 8 (SAS Institute).
| Results |
|---|
|
|
|---|
|
Clinical Follow-Up
Mean treatment duration was 106.1±6.1 and 108.6±8.6 days for losartan and placebo, respectively (P=NS). Twenty-five patients (86%) reached the target dose of losartan 50 mg twice daily, whereas 4 remained at 50 mg once a day throughout the study. Mild side effects were equally distributed between the losartan and placebo groups and included dyspnea, fatigue, palpitations, dizziness, diarrhea, cough, and muscular pain. No patient experienced a change in functional class during the course of study.
Cardiopulmonary Testing
At baseline, the mean
O2max was 29.8±5.6 mL · kg1 · min1 (range, 18.5 to 38.6 mL · kg1 · min1), equivalent to 73.5±12.9% of the predicted value (Table 2). All subjects exercised to the point of exhaustion. No patient reached 100% of the predicted value. The maximum heart rate achieved was
85% predicted in 20 patients, between 80% and 84% in 3, and <80% in 6. Comparing losartan to placebo showed no differences in
O2max, total exercise duration, anaerobic threshold, heart rate, systolic blood pressure, and oxygen saturation.
|
Echocardiographic Evaluation
Baseline echocardiographic characteristics are described in Table 1. RV size was abnormal in 95% of the patients. RV systolic function was qualitatively normal in 28%, mildly impaired in 69%, or moderately to severely impaired in 3%. It was possible to quantify the RV ejection fraction in 27 patients (range, 21% to 60%), with no statistically significant improvement in patients on losartan versus placebo (43.1±12.2% versus 40.5±9.9%; P=0.24). The degree of tricuspid regurgitation remains unaltered regardless of treatment modality.
Neurohormonal Levels
The mean baseline level of angiotensin II was nearly normal at 5.7±4.9 pg/mL (range, 0.4 to 24.1 pg/mL), with 48% of patients having values below the 5-pg/mL cutoff for normal. No significant correlations were detected between levels of angiotensin II and the baseline RV ejection fraction (r=0.18, P=0.36) or
O2max (r=0.04, P=0.82). The mean baseline level of epinephrine was 32.6±26.7 pg/mL, with 82% of patients having values below the mean normal value of 41 pg/mL for healthy subjects. The mean level of norepinephrine was 204.3±57.7 pg/mL, with 89% of patients having values below the mean 291-pg/mL cutoff for normal subjects.
At baseline, the mean level of NT-proBNP was mildly increased at 257.7±243.4 pg/mL (range, 33.2 to 1048.0 pg/mL), with 44% of patients having lower than the 125-pg/mL cutoff. Patients with lower RV ejection fractions had significantly higher NT-proBNP levels (r=0.42, P=0.02), but no correlation was found between NT-proBNP levels and baseline
O2max (r=0.26, P=0.16). Comparing losartan and placebo showed no significant differences in levels of epinephrine, norepinephrine, and NT-proBNP, but an expected increase in angiotensin II, indicating inhibition of the AT1 receptor, was noted (Table 3).
|
| Discussion |
|---|
|
|
|---|
It is well known that LV dysfunction is characterized by increased circulating levels of neurohormones. BNP is secreted by the heart in response to pressure or volume overload and is produced as a prohormone, pro-BNP, that is enzymatically cleaved into BNP and an amino-terminal portion, NT-proBNP. Plasma BNP and NT-proBNP levels are clinically useful in the diagnosis, management, and risk stratification of adults with LV congestive heart failure.2528 Moreover, they are sensitive markers for asymptomatic LV dysfunction25 and correlate with impaired exercise capacity and lower LV ejection fraction.26,27,29 NT-proBNP levels have been shown to increase proportionally with NYHA functional class.29 In 105 patients with chronic heart failure in NYHA functional classes I, IIA, IIB, and III, NT-proBNP levels were 103.4±97.6, 622.2±472.7, 2241.0±1528.4, and 10 018.0±12 727.0 pg/mL, respectively.29 Comparatively, our patients had mildly higher levels of NT-proBNP (257±243 pg/mL) than NYHA class I patients but lower levels than NYHA class IIA patients. This finding may be explained in part by an objectively decreased exercise capacity (
O2max 73.5% predicted), despite 27 of 29 patients reporting being in NYHA class I. This impaired exercise capacity has been consistently demonstrated in patients with systemic RVs.6,7 In addition, discrepancies between NYHA class and cardiopulmonary exercise testing are not uncommon, with reports suggesting <50% concordance.30
The detrimental effects of an overactive RAS in LV dysfunction31 and benefits of pharmacological blockade on exercise tolerance have been repeatedly demonstrated over the past 20 years. In 1983, Kramer et al8 reported that
O2max increased from 12.9±2.3 to 15±1.8 mL · kg1 · min1 after 3 months of therapy with captopril in patients with NYHA class II to IV symptoms and an LV ejection fraction <25%. Sharpe et al9 also observed an increase in exercise duration in NYHA class II and III patients, from 9.3±5.7 to 17.6±5.6 minutes after 3 months of treatment with enalapril. When an angiotensin II receptor inhibitor, losartan 50 mg daily, was compared with enalapril 20 mg daily in a 12-week randomized parallel trial of 116 patients in NYHA class II to IV with an LV ejection fraction <45%, equivalent benefits in exercise duration, 6-minute walk test, and LV ejection fraction were found.32
Unfortunately, the situation is less clear when the RV is systemic because prior data were limited to nonrandomized studies involving few patients. In 7 patients >13 years of age with a systemic RV, Lester et al33 found a small but statistically significant increase in exercise duration after 8 weeks of losartan therapy. The results of our trial, however, are consistent with a subsequent nonrandomized and unblinded study of 8 young patients (mean age, 13.8 years) with D-TGA, a Mustard baffle, and qualitatively depressed RV function that found baseline reductions in
O2max, exercise duration, and cardiac index compared with control subjects with a systemic LV, but no improvements occurred after 12 months of enalapril.34
It is important to note that the lack of benefit observed in our patients occurred despite most having a baseline dilated RV, depressed RV systolic function, decreased exercise tolerance, and an increase angiotensin II level on losartan, indicating that the desired pharmacokinetic response was achieved. Intriguingly, the baseline circulating angiotensin II level was only mildly elevated, with nearly half the patients having normal levels. This provocative finding, combined with the observed lack of clinical benefit, raises the possibility that, unlike patients with LV dysfunction, activation of the RAS is not a dominant pathophysiological contributor to impaired exercise tolerance observed in patients with a systemic RV.
In fact, animal models suggest that activation of the RAS is not implicated in an RV hypertrophy response to pressure overload.35 When pulmonary arterybanded cats were compared with controls, similar levels of angiotensin II were found, prompting the authors to conclude that RV pressure-overload hypertrophy can occur without RAS activation. Moreover, pharmacological modulation of the RAS with losartan and captopril did not alter the extent of hemodynamic overload or hypertrophic response. A clinical study in 53 adults with a variety of congenital heart defects further supports these findings.36 Asymptomatic patients and control subjects had no evidence of RAS activation, unlike those with symptoms.
If RAS activation is not the primary mechanism implicated in the almost universal findings of decreased exercise capacity, RV dilatation, and RV systolic dysfunction in patients with intra-atrial baffle redirection surgery, other hypotheses merit consideration. Chronotropic incompetence, ischemia, and diastolic dysfunction have been previously proposed. As a result of extensive intra-atrial surgery, sinus node dysfunction is indeed prevalent in this patient population and may contribute to decreased exercise tolerance.1,6,37 Chronotropic incompetence, however, could not be confirmed in our study because 80% of patients achieved >80% of their maximum predicted heart rate, albeit with some help from implanted pacemakers. Second, reversible and fixed perfusion defects with concordant regional wall motion abnormalities have indeed been documented.38 Although this may play a role, if ischemic RV cardiomyopathy was primarily responsible for the impaired exercise capacity, some benefit may be expected from afterload reduction with losartan.
Finally, both systolic function and diastolic function have been previously assessed in patients with Mustard and Senning procedures. Reich et al39 studied 153 children with radionuclide ventriculography and found that, although RV and LV ejection fractions were usually normal 4.4 years postoperatively, the LV peak filling rate was significantly depressed with a marked increase in filling fraction resulting from atrial systole. LV diastolic abnormalities progressed in a subset of 99 patients studied 8.8 years later, despite unaltered LV and RV ejection fractions. Inversely, RV peak filling rate was higher than normal, with a prolonged right atrial contraction time at the expense of shorter RV rapid and slow filling times. The authors surmised that abnormal filling patterns resulted from decreased capacity of the systemic venous atrium and diminished ability of atrial contraction to boost LV filling. In 14 older patients, Derrick et al40 reported reductions in
O2max (77% predicted) despite appropriate responses of load-independent indexes of contraction and relaxation. This finding was attributed to a failure to augment RV filling rates during tachycardia, presumably resulting from AV transport impaired by abnormal intra-atrial pathways.
On the whole, these studies suggest that the main factor limiting an appropriate increase in stroke volume during exertion is decreased preload, which accompanies the exercise-induced tachycardia response. If this were borne out in further studies, ß-blockers may prove to be a sounder approach to this therapeutic conundrum than treatment aimed at inhibiting the RAS.
Study Limitations
Given the complexity of the patient population, some potential confounding variables such as technical surgical variations, type of palliative surgery before intra-atrial repair, perioperative support, and form of myocardial protection are difficult to quantify and adjust for in analyses. Nevertheless, the randomized and crossover nature of the study design offers some protection over confounding by immeasurable variables. All subjects who participated in this study were in NYHA functional class I or II. Therefore, results may not be applicable to more symptomatic patients. Finally, our study was designed to assess the superiority of losartan versus placebo and is likely underpowered to determine equivalency with placebo. Given that no statistical difference in
O2max between the losartan and placebo groups was noted, it is important to consider the risk of a type II error (ie, failing to reject the null hypothesis when a true difference actually exists). Because the observed SD (ie, 3.7 mL · kg1 · min1) was substantially lower than predicted (ie, 7 mL · kg1 · min1) and more patients were enrolled than sample size calculations required, >99% study power was obtained to detect the hypothesized difference of 4.65 mL · kg1 · min1, indicating a <1% risk of a type II error. Indeed, the study had >80% power to detect a difference in
O2max of as little as 2.0 mL · kg1 · min1. It is therefore highly improbable that the lack of improvement in exercise capacity from losartan therapy resulted from insufficient statistical power.
Conclusions
Despite impairments in baseline functional testing, adult patients with a systemic RV did not have improved exercise capacity with losartan therapy. Minimal baseline activation of the RAS may explain this lack of benefit. Caution is warranted when results from studies assessing the effects of inhibition of the RAS on the failing LV are extrapolated without independent validation on the systemic RV.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
F. Haddad, R. Doyle, D. J. Murphy, and S. A. Hunt Right Ventricular Function in Cardiovascular Disease, Part II: Pathophysiology, Clinical Importance, and Management of Right Ventricular Failure Circulation, April 1, 2008; 117(13): 1717 - 1731. [Full Text] [PDF] |
||||
![]() |
R. E. Shaddy, M. M. Boucek, D. T. Hsu, R. J. Boucek, C. E. Canter, L. Mahony, R. D. Ross, E. Pahl, E. D. Blume, D. A. Dodd, et al. Carvedilol for Children and Adolescents With Heart Failure: A Randomized Controlled Trial JAMA, September 12, 2007; 298(10): 1171 - 1179. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2005 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |